o

n l

co *-. x co — co £ — w

BRARIES SMITHSONIAN INSTITUTION NOIlfUllSNI NWINOSHIIIAIS SSIHVHail LIBRARIES SMITHSONIAI

co 2 v CO 2 ^ 00 2 .... CO

2 «t AS .,2 < s <

1

t— 2 \c3®a^pv 2 V/MM# r—

— s Ay/Z'ji ~ xOauTc v — * Jytfflr ~

2 GO k 2 CO 2 CO _

UTlON NOIlfUllSNI NVIN0SH1IWS S3IUVUSn LIBRARIES SMITHSONIAN INSTITUTION NOIlfUllSNI NVIN0SH1IIN
co — co X co ~ co

co

co

q >\ x^oxixsgx O o u<-^ __ -x^uav^-' q

aan^LiBRAR i es^ Smithsonian-* institution NoiiruiiSNi“JN\/iNOSHiitNS S3iavyan-lLiBRARi es^smithsonia
2 z , ^ ^ !I » 2 t~ 2

».''cX m xfXxTx 2 sM ra m X 2 m xTH55x 2

—

m m vJvasv^x m ' \\y ^ \^vas»^/ m

CO _ CO — CO — co

UTlON NOIlfUllSNI NVIN0SH1HMS Sliavaail LIBRARIES SMITHSONIAN INSTITUTION NOIlfUllSNI NVINOSHim

CO 2 <£ Z w 2 CO 2

X IP#, o X 2 ^ =! 2 ^

co co x co

O £ X# 2

a VcXSSUS^V I— .V'VXX ~Z. . "’f /•* • \ ,.■*

' 2 ■' > ■ 5

CO 2 CO * 2 Jo ■• - 2 CO

'dail LIBRARIES SMITHSONIAN INSTITUTION NOIlfUllSNI_NVINOSHllWS S3iaVaail LIBRARIES SMITHSONIA
co — co \ co — co

CO

LjJ

CO

CO

my,

... < Mjrr- s

WuiHfJ “ . <v/<v 5

o “ O ~ o ^s» _ x«,'»,y O

2 _j 2 _J 2 _J 2

UTlON NOIlfUllSNI NVIN0SH1IIAIS SBIBVaail LIBRARIES SMITHSONIAN INSTITUTION NOIlfUllSNI NVIN0SH1IIA

2 r~ 2 r~ 2 i~

x ^ —
m X^Oi us^x m

co*-. X co X co X co

faan libraries Smithsonian institution NouruiiSNi nvinoshihns S3iavaan libraries Smithsons

5 , CO 2 ,v CO 2 W 2 . , w

, . 1 £ ,<s2£x § jfc> £ . . vi £.

X - = VX«J\X - y . y- /C^^tW^-VX 2

2 . ‘tf*

CO .y-i
°

5 x > X^iiu^ 2 2 >

UTlON NOIlfUllSNI NVIN0SHlllNSt°S3 I yva a 11 LIBRARIES SMITHSONIAN INSTITUTION NOIlfUliSNI NVINOSHIM
- ,v <o X co x ^ ^ co x

a- .X. V I. i ^\rZTT>>^ v<T7cr»Trv. it I ✓Xv-isOaTxs. Li 1 /Xicv

Y%w z| m ,

Q '*i\ 'X^UIJLV^ o X^-)V IJL^ j Q X^Oai pcjy XxOcjixSJjJ,/ Q

fa a 11 ^Ll B RAR I ES^SMITHSONIAN^INSTITUTION2 NOIlfUllSNI "JNVIN0SH1IWS:ZS3 I H V a 3 11 ^ LT B R A R I E S^ SMITHSONI/

“ 2 n , 2 r- z

x O m XT^q^oX o a. 2 X^AoX “ O

CO

CO X CO 5 CO X CO

'UTlON NOIlfUllSNI NVIN0SH1IWS S3IHVaOI1 LIBRARIES SMITHSONIAN INSTITUTION NOIlfUllSNI NVINOSHlIlf

CO 2 co . 2 , CjO 2 CO 2

< S M < -a. 2 xSstitgXv < ^ S <

vaan libraries smithsonian^institution NoiiniiiSNi_NViNOSHiiiNS ssiavaan^ libraries smithsoni/s

to lZ ■/ Jt- /^WFI^A lZ /o'jSfiKIA w r- APOT^IA w /MfftlA ZZ /

2 ? ,./#^ 5 (i# M S 2 (ffii 5 2 ? ,;#

« 'W S N^&S/ g S ' "V K n,

to _ co — co \ ± co Sr

TUTION NOlinillSNI NW1NQSH1SWS S3 IMVaa II LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1II

to z „ ^ « 5 .*•-■ !£ z ,v to z

S s .< 5 „ ,/ - s .<

X
CO

— § ’^Plfy

> W S ^ > XiOmvjX 5

LI BRAR I ES^SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHIIWS^S 3 I H V 0 8 1 1 2 L I B R A R I E S^SMITHSONI/

CO 5 ^ ^ CO — CO — CO

w .co u co . J&XX ^ z^-'Tx w w

c WKSw «; J3?l c

O x\ ~ '^OiCLSjX O Xg*~pC

LI BRAR I ES^SMITHSONIAnZiNSTITUTION NOlinillSN
z

NVINOSHill

%<

TUTION NOlinillSNI NVIN0SH1IIMS saiavaan

Z r~ , Z i

O ><S™C//X ~ * O -XjsrTriX

m I #% S %% i #

J® P tOf 2 W P (?(

oft/ /n Vx .*/ m >>>1^ ^ x*As*iz m XL*os*£Z «; m

^ co — co _ co

vaail LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IWS S3I0V00I1 LIBRARIES SMITHSONI,

z <o Z » CO Z co _ z (O

5 -SW > ^ 5'* ' ' I 5

CO 2 CO ■•*“ Z CO z to

TUTION NOIinillSNI_ NVINOSHillAJS S3I0V0ai1 LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI

CO \ co _ “ co — CO

K'AsV^Z

NVIN0SH1I

LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IWS S3iaV00!1

Z r- Z r* ;z

LIBRARIES SMITHSONI

<i\&$y m ^ m Ny^*£S^// ^ m 00

co ± co £ — co \ 2 — to

TUTION NOlinillSNI NV1N0SH1IINS S3iaV0 8n LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHil

co Z _ CO 2 ..... CO Z , CO z

2 CO * Z — >

Vaan^LIBRARIES SMITHSONIAN INSTITUTION" NOlinillSNI NVlNOSHillMS^Saiyvyen^LIBRARIES^SMITHSONI

CO CO — to — CO

CO

o

TUTION NOlinillSNI """nVINOSHIIIAIS S3 I 0V0 8 11 L1BRARI ES 2 SMITHSON IAN-* INSTITUTION ~ NOlinillSNI "^NVINOSHil I
z r- , z r- 2 r- z r~

2 5 XS^lri7X “ zTXa CX O O

_ m • ^ vTO1'>r m ^ m xc~osv*iz ^ x^va^x m

vaa n~LI BRAR I ES SMITHSONIAN INSTITUTION 'NOlinillSNI _ NVINOSHil WS S3 I ava 0 11 ~LI B RAR I ES SMITHSONI

5 X W Z V CO z CO Z , CO .

< s < -x S < /®\ s A < g

A ^ — s? 0 .-fX-Vx' ,'r; . U'

K UJ XCvnrS^s. rn 'Sg. UJ ^ .

z

CO

z

CO

H

5

CO

L52X

NH

Number 32:4
10 My 1980

M

NTRIBUTIONS IN SCIENCE

NATURAL HISTORY MUSEUM OF LOS ANGELES COUNTY

A NEW GENUS AND SPECIES OF CHELID TURTLE
FROM QUEENSLAND, AUSTRALIA

By John M. Legler and John Can.n

• • •

I Published by the NATURAL HISTORY MUSEUM OF LOS ANGELES COUNTY • 900 EXPOSITION BOULEVARD - LOS ANGELES, CALIFORNIA 90007

mi ‘

1

viliilHBHp

ft sS livJa

’4|*

SERIAL PUBLICATIONS

OF THE NATURAL HISTORY MUSEUM OF LOS ANGELES COUNTY

The scientific publications of the Natural History Museum of Los Angeles County have been issued
at irregular intervals in three major series; the articles in each series are numbered individually, and
numbers run consecutively, regardless of subject matter:

• Contributions in Science, a miscellaneous series of technical papers describing original
research in the life and earth sciences.

• Science Bulletins, a miscellaneous series of monographs describing original research in the
life and earth sciences. This series was discontinued in 1978 with the issue of Numbers 29
and 30; monographs are now published by the Museum in the Contributions in Science

series.

• Science Series, long articles on natural history topics, generally written for the layman.
Copies of the publications in these series are available on an exchange basis to institutions and
individual researchers. Copies are also sold through the Museum Bookshop.

jeurv*.

1

A NEW GENUS AND SPECIES OF CHELID TURTLE
FROM QUEENSLAND, AUSTRALIA1 2

By John M. Legler1 2 3 and John Cann4

Abstract. The Fitzroy Tortoise, Rheodytes leukops, new genus and species (Family Chelidae), is de-
scribed from a large series of adults and juveniles from the Fitzroy drainage of Queensland, Australia (type
locality 23°09'S, 149°55'E). No previous mention of this taxon appears under any name in the scientific
literature. Each of the following characteristics alone will distinguish R. leukops from all other chelids:
interlateral seam contacts on posterior parts of marginals 6 and 8; rib tips of costals 2-4 forming gomphoses
with centers of peripherals 4-6; splenial bone lacking. Rheodytes leukops occurs in microsympatry with
Elseya dentata and Emydura kreffti. Chelodina longicollis, C. expansa, and Elseya latisternum occur
nearby in the same drainage. Rheodytes leukops is completely carnivorous (feeding chiefly on aquatic insect
larvae), occurs in fast, clear water, and seems to be specialized for bottom probing and scraping. Annual
reproductive potential is 46-59 eggs in 3-5 clutches. The eggs are small and have a mean incubation time of
47 days at 30°C.

Rheodytes leukops is seemingly most closely related to the eastern Australian short-necked chelids (gen-
era Elseya and Emydura ), but the phylogeny of Rheodytes is not yet understood. The following related
generic groups of short-necked Australian chelids are defined on the basis of shell and scute proportions,
form, skull and shell osteology, integumentary topography, color, and reproductive and dietary habits: (1)
Rheodytes; (2) Elseya dentata group; (3) Elseya latisternum group; and (4) Emydura. Pseudemydura
constitutes the fifth short-necked genus and seems not to be closely related to any other group in Australia.

Cann received a moribund specimen of an unusual turtle from
a reptile park in 1973; the specimen had been collected on the
Fitzroy River “near Rockhampton” and is now cataloged as
AM R41274. Another specimen (AM R41794) was received
by the Australian Museum in April 1974. We examined these
two specimens in July 1974 and quickly recognized them as a
strikingly new and different taxon. Seemingly, no other speci-
mens of the Fitzroy Tortoise were in existence at that time. We
obtained the specimens upon which this account is based in Oc-
tober 1976. An illustrated popular account of this expedition
and the specimens collected appeared in Cann (1978).

METHODS AND MATERIALS

Abbreviations used for museums are as follows: AM, Aus-
tralian Museum, Sydney; LACM, Los Angeles County
Museum of Natural History; NMV, National Museum of Vic-
toria, Melbourne; QM, Queensland Museum, Brisbane; UU,
University of Utah, Salt Lake City; WAM, Western Australian
Museum, Perth. Other frequently used abbreviations: CL, car-
apace length; CW, carapace width; CBL, condylobasilar
length; M, marginal; P, peripheral.

Measurements of the shell and terminology of shell elements,
unless specifically explained, are as outlined by Carr (1952).
Lengths of plastral scutes are interlaminal (i.e., an average of
right and left scutes as measured on their common midventral
seam).

Skull measurements that are not self-evident from their ter-
minology are: Condylobasilar Length — from posteriormost

point of occipital condyle to anteriormost point of premaxillary
region; Squamosonasal Length — from a line joining posterior
tips of squamosals to the anteriormost projection of the nasals;
Height of Snout — from top of nasal bones to tomial edge of
maxilla, in a line perpendicular to basicranial-palatal plane;
Maxillary Breadth — width of skull across posteriormost edges
of maxillary tomium; Length of Dentary Symphysis — mea-
sured with mandible oriented on a plane surface and parallel to
that plane surface (i.e., not maximal); Greatest Length of
Mandible — on midlongitudinal axis from line joining posterior-
most points of rami to anterior tip of dentary symphysis; Great-
est Width of Mandible — maximum outside breadth — the base
of the mandibular triangle.

Tinkle (1962) demonstrated the utility of expressing the
points at which the five interlateral seams of the carapace inter-
sect the marginal scutes. The terminology used here is as fol-

1 Review Committee for this contribution:

Ernest E. Williams
John W. Wright
George Zug

2Send reprint requests to Legler at Utah address.

department of Biology, University of Utah, Salt Lake City, Utah
84112, and Department of Zoology, University of New England, Ar-
midale, NSW 2351 Australia; Research Associate in Herpetology of
the Natural History Museum of Los Angeles County.

426 Yarra Road, Phillip Bay, NSW 2036 Australia.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 324: 1-18
ISSN: 0459-8113

2

Legler and Cann: New Australian Chelid

lows: P, contact on posterior half of marginal; M, contact at
midpoint; A, contact on anterior half (this terminology is a sim-
plification of that used by Tinkle).

The type locality and most of the other localities mentioned
in this paper can be found on the following maps. Australia
1:250,000, Series R 502, Ed. 1-DNM: Duaringa SF 55-16;
Rockhampton SF 56-13. World Aeronautical Chart ICAO
1:1,000,000: Rockhampton, 3235 5th Ed.; Operational Naviga-
tion Chart 1:1,000,000: ONC-Q15: ONC-P15; The Reader’s
Digest Complete Atlas of Australia, 1968, Reader’s Digest
Assn. Pty. Ftd. Sydney, 183 pp (ca. 1:1,265,000): Clermont, pp
84-85; Brisbane-Rockhampton, pp 60-61. All the aforemen-
tioned maps have longitude and latitude and permit precise
reckoning of localities. The Reader’s Digest Atlas and the
1:1,000,000 aeronautical sheets are especially recommended.
Coverage of the latter is worldwide.

SYSTEMATICS

Rheodytes, new genus
Figures 1 through 7

TYPE SPECIES. Rheodytes leukops, new species, by
monotypy.

VERNACULAR NAME. Fitzroy Tortoise.

DIAGNOSIS. A short-necked Australian chelid dis-
tinguished from all other members of the Chelidae by the fol-
lowing characters (each character marked with an asterisk (*)
is alone diagnostic among chelids): (1)* interlateral seam con-
tacts on the posterior parts of the sixth and eighth marginal
scutes; (2)* rib tips of costals 2-4 forming gomphoses with the
centers of peripherals 4-6; (3) a narrow, unridged maxillary
triturating surface that becomes even narrower in the premaxil-
lary region; (4)* splenial bone lacking; (5) a long completely
coossified dentary symphysis; (6) a maxillary tomial edge that
is straight in profile; (7) a white ring around the iris; (8) rela-
tively small eggs and short incubation period; (9) huge cloacal
bursae.

ETYMOLOGY. The generic name is derived from the Greek
roots rheos (current or stream) and dytes (diver) and alludes to
the speed and agility of these animals in fast currents. The spe-
cific name is derived from the Greek leukos (white) and ops
(eye) and refers to the distinctive white ring around the iris.

RELATIONSHIPS. Superficially similar to and probably
most closely related to the genera Emydura and Elseya but dis-
tinguished from them by the above phenotypic characters (in
combination or by each of the first four diagnostic characters
alone). Of these, Elseya dent at a seems most closely to resemble
R. leukops (Table 1). Rheodytes leukops occurs microsym-
patrically with Elseya dentata and Emydura krejfti.

Rheodytes leukops, new species
Figures 1 through 7

HOLOTYPE. QM J31701, whole adult female with car-
apace length of 253 mm, collected 7-8 October 1976 by J.M.
Legler and J. Cann: Fitzroy River, 63 km N and 25 km E of
Duaringa, Queensland, Australia, elevation 40 m (23°09'S,
149°55'E). Bearing also the numbered tags “UU Field JML
8217” and “UU 17111” (Fig. 1, Table 2).

ALLOTYPE. QM J31702, whole adult male with carapace

length of 235 mm, same data as holotype (JML 8215 and UU
17109).

PARATOPOTYPES. LACM 127779, UU 17131, AM
R44651, whole adult males; UU 17103 male skeleton; UU
17104-7 males, dry shell with soft parts and viscera in liquid
(hereinafter “S & P” specimens); LACM 127778, UU 17110,
17113, 17122, AM R44650, whole adult females; UU 17114-5
female skeletons; UU 171 16-21 females, S & P specimens. UU
16805-820, 17150-86, 17197-226, QM J31704-7, AM R44652,
hatchlings representing six clutches of eggs from paratopotypic
females.

OTHER PARATYPES. Fitzroy River, Glenroy Crossing, 60
km N and 23.5 km E of Duaringa, Queensland, elevation 40 m
(23° ITS, 149°56'E), 9 October 1976, Legler and Cann: UU
17124-6 females, S & P specimens. UU 17132-49, 17227-45,
LACM 127780, NMV 50435, UU 17248-66, hatchlings repre-
senting three clutches of eggs from paratypic females. Dawson
River, 2 km N of Gainesford, Queensland, elevation 64 m
(23°47'S, 149°46'E), 10 October 1976, Legler and Cann: UU
17127 male, 17128 female, QM J3 1 703 female, all prepared as
S & P specimens. Windah Creek, near Gogango, Queensland,
elevation 80 m (23°37'S, 150°02'E), 15 September 1976, AM
R44650 whole adult female. “Mackenzie River, Dawson Valley,
Queensland” 3 April 1974, R. Stokes: AM R41794 whole adult
male. “Fitzroy River, Queensland” November 1973, R. Ohl:
AM R41274 whole adult female.

DIAGNOSIS. See diagnosis of genus Rheodytes.

GENERAL DESCRIPTION OF SPECIES (based on hypo-
digm). Dorsal silhouette of adult carapace a tapered ellipse
lacking any distinct, straight, lateral edge. Widest point at M7
or M7-8. Juvenile carapace nearly round. Plastron never visible
in dorsal view of either whole animal or dry shell (Figs. 1, 5,
and 7).

Edge of carapace smooth in adults, extremely serrate in juve-
niles up to 95 mm long and 2-Vi years old (older juvenile stages
unknown at this writing); serrations consist of single projections
on individual marginal scutes as follows: Ml unmodified; M2-4
slightly pointed; M4-6 sharply pointed with spines directed pos-
teriorly; spines on individual marginals 7 through 12 become
progressively broad and blunt (Fig. 3).

Optical cross section low (Table 3 and Fig. 5) either peaked
or slightly flattened middorsally; marginal index (vertical
height of margin expressed as a percentage of total height),
0.45.

Plastral lobes relatively narrow, tapered, and straight-sided.
Forelobe foreshortened and tapered to a blunt point (rather
than truncated). Anal notch semicircular, never angular (Figs.
1 and 7).

Head narrow and high. Orbits small and (because of narrow
rostrum and interorbital region) appear to be directed more an-
teriorly than in other short-necked chelids. Maximum breadth
of head at midposterior border of tympanum (at quadrato-
squamosal suture); optical cross section of head suggests a high
rectangle with sides angled slightly inward dorsally. In profile,
there is a continuous even slope formed by dorsum of rostrum
and head (no supraorbital or rostral bulges). In general, the
head appears relatively small and the neck relatively large, the
combination suggesting a powerful instrument for digging or
prying (Figs. 3, 4, and 6).

Posterior edge of maxillary sheath (near angulis oris) forms
angle of about 45 degrees with main axis of maxillary tomial

Contrib. Sci. Natur. Hist. Mas. Los Angeles County. 1980. 324: 1-18

Legler and Cann: New Australian Chelid

3

edge. From that angle to premaxillary region, the tomial edge is
almost perfectly straight in profile (no concavity, no convexity).
In anterior view, the sheath is flat or slightly concave in pre-
maxillary region. Maxillary tomial edge lacking distinct
notches or denticulation of any sort. Tomial edge of mandibular
sheath concave in profile, gently curved from angulis oris to tip;
tip bluntly rounded, neither a hook nor a crushing device. In
general, the tomial edges of the jaw sheaths appear to be worn
in adults; in the largest adults, the anterior parts of the tomial
apparatus barely occlude.

Precentral scute present in 27 (93 percent) adults. Modal
carapacal seam contacts 2M 5A 6P 8P 1 1 A (see “Methods and
Materials”). Modal formula for three longest scutes on plas-
tron Fern » An > Ab. Length of femoral scute approximately
20 percent of carapace length (Table 3). A slight plastral con-
cavity along midline in adults of both sexes.

Epidermal laminae relatively thin. Intercostal sutures visible
through lateral scutes in adults except in most heavily pig-
mented specimens. Plastral and interperipheral sutures easily
discernible through scutes. Epidermal laminae thickened and
weltlike over interosseous sutures.

Surface texture of carapace (under low magnification) con-
sisting of a primary series of sharp parallel ridges (longitudinal
or diagonal, depending on scute) separated by narrow sulci;

smaller secondary cross ridges create a series of pits that ap-
pears as a reticulation to the unaided eye. Plastron smooth,
lacking a distinctive texture. Growth zones weakly evident in
captive-reared juveniles but not at all evident in adults.

Feet and hands large and fully webbed. Claws 5-4. An exten-
sive ulnar fringe of enlarged scales on foreleg. Tibial edge of leg
lacking a distinct series of enlarged scales (Figs. 3 and 4). Two
musk gland orifices — one below posterior part of M3 and one
below anterior part of M7.

Triturating surfaces of maxillary sheath relatively narrow
and at right angles to plane of tomial edge, about as wide as
tomial edge is deep. A slight expansion in width of triturating
surface just anterior to internal nares at which point the sheath
is about 1 mm thick. From that point anteriad, triturating sur-
face narrows quickly and significantly to about 25 percent of its
narrowest posterior width. The “U” formed by the two triturat-
ing rami anteriorly bounds a thick pad of buccal mucosa that
covers nearly the entire premaxillary region of the palate. This
anterior extension of palatal mucosa is rounded and tumescent-
appearing in live specimens but is dented or collapsed in most
preserved specimens. The tumescence consists of a thick (16 to
20 cell layers) pad of nonkeratinized epithelium underlain by
loose connective tissue and a large vascular sinus.

Choanal papillae lacking; mucosa of buccal and pharyngeal

Figure 1. Rheodytes leukops, new genus and species: Holotype QM J31701, adult9, carapace length, 253 mm.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 324: 1-18

4

Legler and Cann: New Australian Chelid

Table 1. Comparison of selected characters in four major groups of Australian short-necked chelids, excluding Pseudemydura (see also “Discus-
sion”). Remarks under Emydura kreffti apply to all known members of the genus in Australia and New Guinea.

Rheodytes leukops

Elseya dentata

Elseya latisternum

Emydura kreffti

Cross section relatively low,
slightly peaked, relatively
steep-sided; marginal index,
0.45 (see description).

Anterior lobe of plastron not
extending anterior to edge of
carapace, not visible in dorsal
view.

Cross section evenly rounded
from margin to margin, never
peaked or depressed mid-
dorsally; marginal index, 0.44

Anterior lobe of plastron with
anterior projection to or
beyond anterior edge of
carapace, usually visible in
dorsal view.

Cross section evenly rounded from
margin to margin; often flat-
topped or with a midlongitudinal
depression; marginal index, 0.53

Cross section rounded above (never
peaked or depressed) and steep-
sided; marginal index, 0.40

Anterior lobe of plastron not extending anterior to edge of carapace, not
visible in dorsal view.

Lateral and posterior marginal serration extreme in hatchlings but
completely lost in adults.

Marginal serration moderate to No marginal serration at any

extreme posteriorly in young and stage,

adults.

Precentral scute usually present
(93%).

Precentral scute usually absent
(94%).

Precentral scute geographically
variable; usually absent in far
northern populations (Cape York,
93%) usually present and well
developed in populations south of
29 °S (98%).

Precentral scute usually present
(99%).

Interlateral seams C and D forming contacts with marginals 7 and 9, usually medially or anteriorly, contacts
with M6 and M8 occurring only rarely or as anomalies.

Interlateral seams (C and D
after classification of Tinkle
1962) contacting posterior
parts of marginals 6 and 8
(Fig. 7).

Modal plastral formula
Femoral *■ Pectoral * Anal
(41%); Femoral always
longest (100%).

Modal plastral formula Femoral
*■ Pectoral * Abdominal
(50%); Femoral (79%) or
Pectoral (21%) longest.

Modal plastral formula Anal »
Pectoral » Intergular (24%); Anal
(73%), Femoral (10%), or Pectoral
(10%) longest.

Modal plastral formula Pectoral *•
Femoral » Abdominal (48%);
Pectoral (78%) or Femoral
(21%) longest.

Maxillary tomial edge nearly Maxillary tomial edge curved in profile from premaxillary region to angulis oris (a “smile”); premaxillary
straight in profile from region depressed at least slightly,

premaxillary region to
posterior end (not curved, not
indented) where it forms a
45° angle with straight
posterior edge of sheath
extending to angulis oris.

No median alveolar ridge.

Diet carnivorous, insectivorous;
no known malacophagous
tendencies.

Always one pair of distinct,
conical, small to medium
barbels; little variation.

No rostral pores.

A well defined median ridge on
alveolar surfaces of maxilla
and dentary, evident on both
osseous and keratinous parts of
jaw apparatus (evident in all
populations of E. dentata but
less well developed in
northwestern Australian E.
dentata and in E.
novaeguineae).

Diet herbivorous (Legler 1976).

One, two, or three barbels; often
a single median barbel in
northwestern Australian
populations; a tendency toward
thick, fleshy, variable barbels.

Rostral pores rare.

No median alveolar ridge.

Diet chiefly carnivorous with
insectivorous tendencies
(vegetation rare in guts
examined); no malacophagous
tendencies.

Number of barbels variable but
usually two in a balanced pair,
never single and median; usually
well developed and tending to be
long.

No median alveolar ridge.

Diet omnivorous, opportunistic,
geographically variable; aquatic
vegetation common in guts;
malacophagous modifications in
several populations.

Barbels poorly developed or absent,
rounded tubercles at most;
seldom conical, never single and
median; often indistinguishable
from other scales on skin.

Rostral pores evident and well developed (Winokur and Legler 1974).

Head shield fully keratinized and extensive at all ages; gnarled in older individuals. Head shield usually absent; weakly

developed only in largest adults
(usually females); never gnarled.

Continued

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 324:1-18

Legler and Cann: New Australian Chelid

5

Table 1 continued

Rheodytes leukops

Elseya dentata

Elseya latisternum

Emydura kreffti

Inguinal buttress nearly as Inguinal buttress much narrower than axillary buttress in transverse section. Posterior part of sternal cavity

broad transversely as axillary much less commodious than anterior part,
buttress; posterior part of
sternal cavity almost as
commodious as anterior part.

Rib tips of costals 2 through 4 Rib tips of costals 2 to 4 form articulations with the interperipheral sutures,

articulate with gomphoses
formed in the centers of
peripherals 4 through 6,
respectively.

Dentary symphysis coossified in adults, never visible. Dentary symphysis an unmodified Dentary symphysis coossified in

suture in most adults; coossified adults; never visible.

(but still visible) in a few large
individuals.

Splenial bone lacking, probably Splenial bone present as a distinct separate element of lower jaw.

fused with prearticular.

A well-developed retroarticular Mandibular retroarticular process small or lacking,

process on mandible.

Ventral processes of prefrontals Ventral processes of prefrontals not closely approximated, not constricted, permitting virtually unobstructed view
closely approximated at posteriorly through fissura ethmoidalis.

midpoint of fissura
ethmoidalis to create
keyholelike aperture in fossa
nasalis.

Ventral ridges of frontal bones vertical, not turned inward, not
closely approximated.

Ventral ridges of frontal bones Ventral ridges of frontals as in R

turned sharply inward and closely leukops.
approximated, nearly closing
sulcus olfactorius at midpoint.

Prefrontals narrowly exposed Prefrontals broadly exposed on anterior part of dorsal orbital rim.

dorsally, forming only a small
part of dorsal orbital rim.

At least a narrow maxillary- Prefrontals and nasals in contact, excluding a maxillary-frontal contact on dorsal exposure of skull,

frontal contact preventing
prefrontal-nasal contact on
dorsal exposure of skull.

Iris ivory in adults, silvery in
young.

Iris dark (brown to black) even
in young specimens.

Iris pale, usually yellowish, never ivory, silver or dark

roof more or less smooth. Tongue fleshy, unspecialized, bearing
transverse or oblique folds that are devoid of complex, mac-
roscopically visible papillae. Nictitating membrane lacking;
lower eyelid semitransparent (juvenile) to translucent (adult),
permitting view of eye even when closed.

One pair of distinct uniformly pale barbels; in form of flat-
tened cones with rounded tips, basal diameter 60 to 80 percent
of length. Other slightly pointed tubercular scales (quite dis-
tinct from the barbels) along medial edge of each mandibular
ramus just behind barbels. Barbels tend to point slightly for-
ward in living specimens (Figs. 3, 4, and 6). Surface topogra-
phy of barbels (under magnification) consists of shallow
anastomosing grooves on distal half and a series of juxtaposed
scalelike structures on basal half.

No sign of rostral pores (Winokur and Legler 1974) in any
specimen examined; soft skin of snout absolutely smooth be-
tween natural boundaries of maxillary sheath and rostral exten-
sion of head shield.

Entire epidermis pitted in adults, degree of pitting seemingly

directly correlated with age; cause of pitting unknown but
probably pathologic; most evident on head shield and plastron;
reflected secondarily on underlying dermal bone.

Large pointed conical tubercles on dorsum of neck, a few of
which may have flattened apices but none of which were found
to have specialized follicular centers (Legler and Winokur
1979). Each tubercle set on a small hillock, which bears also a
number of smaller granular scales; the larger the neck tubercle,
the larger the hillock. Rows of large tubercles seemingly trans-
verse in nape region and longitudinal elsewhere — five discern-
ible rows on posterior part of neck, seven in region of nape.
Neck tubercles of adults upright in most cases; those of juve-
niles in shape of flattened triangles or flattened cones, de-
pressed, soft, and lying down upon the skin — not rigid and
upright as in adults.

Ventral surface of neck bearing a series of very low conical or
hemispherical tubercles, each on a hillock that is surrounded by
many very low granular scales. Granular scales have a rough
feel, which is probably imparted by micropustules.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 324:1-18

6

Legler and Cann: New Australian Chelid

In juveniles, the primary topography of the neck skin seems
to be that of rather low plaquelike juxtaposed scales, which are
in turn covered with discernible (under low magnification) mi-
cropustules. The shape of the scales and their pustulate second-
ary topography gives the skin the appearance of crystallized
sugar. This is less distinct in specialized areas such as the dorsal

neck, where tubercles occur, but the tubercles regardless of size
also have the micropustulate secondary topography.

Head shield well developed and fully keratinized, smooth to
slightly pitted and rugose in females, extremely gnarled in
older males. Head shield extends posteriorly from rostrum (an-
terior edges of nasal bones) and corresponds thence to the gen-

Figure 2. Rheodytes leukops, new genus and species: Paratype UU 171 149, condylobasilar length 41.6 mm; three views of entire cranium, three
views of mandible, and a view directly into the narial aperture (note keyhole-shaped fissura ethmoidalis formed by descending processes of prefrontal
bones).

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 324:1-18

Figure 3. Rheodytes ieukops juveniles approximately 16 months old and 60 mm carapace length (UU 17219-22). Top: Three juve-
niles swimming rapidly apart after clinging together in a tight group; one individual is inconspicuous at bottom; note shell serration and
relative proportions of juveniles. Bottom: Three of above individuals in gregarious clinging postures; note development of mandibular

symphysis.

Contrib. Sci. Natitr. Hist. Mus. Los Angeles County. 1980. 324: 1-18

8

Legler and Cann: New Australian Chelid

Table 2. Selected measurements (in mm) of the holotype (QM J317019) and nine topotypic paratypes of Rheodytes leu-
kops , new genus and species. Widths of plastral forelobe and hindlobe were measured at humeropectoral seam and at
midfemoral scute, respectively.

Plastral Plastral

Collection and Carapace Carapace Plastron Forelobe Hindlobe Shell Head

Catalog No. Length Width Length Width Width Height Width

Females

UU 17110

257

201

211

75

81

88

33

QM J31701

253

203

205

73

80

70

31

UU 17112

245

204

208

73

76

80

33

UU 17113

241

187

198

72

73

73

31

UU 17114

253

198

216

71

78

78

33

Mean

249.8

198.6

207.6

72.8

77.6

77.8

32.2

Males
UU 17104

251

200

208

69

77

72

36

UU 17105

262

207

212

72

80

76

38

UU 17106

246

200

198

68

77

78

37

UU 17107

238

188

195

63

70

77

38

QM J31702

235

195

188

67

71

64

33

Mean

246.4

198.0

200.2

67.8

75.0

73.4

36.4

eral dorsal exposure of dermal skull roofing elements, closely
following free upper edge of orbit, having a slight ventral exten-
sion at origin of postorbital bar, corresponding to lateral edges
of parietals, and ending more or less evenly with the posterior
edges of parietals. A slight posteroventral extension along post-
temporal bar. Occasionally a blunt extension near base of su-
praoccipital spine. Posterior edge of head shield usually broken
up partly or completely into scalelike divisions. In general, all
extensions, ridges, and other surface topography of head shield
are intensified in older males to produce an overall gnarled ap-
pearance. Anterior vertical extension of maxillary shield (ex-
tending between eye and snout) also higher in old males than in
females (Fig. 6).

In profile, tympanum and head shield delimit a distinct re-
gion, in the shape of an inverted parenthesis, beginning at an-
gulis oris and extending to posterior temporal region — about 8
mm wide. This is an area of soft skin in which there are rather
widely spread low, pointed, conical tubercles or very low
rounded tubercles, with a few soft longitudinal ridges. Tym-
panum covered by a series of soft longitudinal skin ridges, none
of which is ossified. Tympanum bordered by tubercular scales
but lacking such scales on its surface. Large scales of temporal
region separated by smaller granular scales and small longitu-
dinal ridges.

KARYOTYPE. One juvenile was karyotyped (see Bull and
Legler, 1980, for account of karyotypic relationships in
pleurodires). Diploid number 50; fundamental number 72; 11
pairs of biarmed macrochromosomes of which the first three
are metacentric, submetacentric, and subtelocentric, respec-
tively. The remaining 14 pairs are microchromosomes. The
karyotype is virtually indistinguishable from those of other
short-necked Australian chelids.

OSTEOLOGY. The following account is based on 3 com-
plete and 16 partial skeletons (lacking skull). Phalangeal for-
mula 2- 3- 3- 3-3 on hand and foot. Cervical central articulations
typically chelid (Williams 1950) — 5 and 8 biconvex, 7 amphi-
coelous, 6 procoelous, the others opisthocoelous. Sclerotic ossi-
cles 13.

Shell thin and in general lightly built. Plastron typically
chelid, having four pairs of osseous elements plus an ento-

plastron. Hyo- and hypoplastral bones greatly thickened along
median interosseous sutures, the anteroposterior extent of the
thickening corresponding almost perfectly to anteroposterior
limits of bridge. Inguinal buttresses almost as broad trans-
versely as axillary buttresses; posterior part of sternal cavity
almost as deep (commodious) as anterior part. Neural bones
lacking. Rib tips of costals 2-4 forming gomphotic articulations
in the centers of peripherals 4-6, respectively (not in the inter-
peripheral sutures; Fig. 7).

Carapacoplastral articulation sutural but containing enough
soft connective tissue to permit some kinesis. Dried shells pre-
pared in caustic solutions show considerable gaps between car-
apace and buttresses.

Skull definitely prognathous in profile (Figs. 2, 4, and 6) due
to anterior projection and overhang of premaxillary region; su-
praoccipital process relatively short; trigeminal foramen not
visible in lateral view. In dorsal aspect, maxillary bone contacts
frontal and separates prefrontal from nasal; prefrontals very
narrowly exposed dorsally, forming only a small anterior part of
dorsal orbital rim. Ventral ridges on frontal bones (forming sul-
cus olfactorius [Gaffney 1972]) straight, not turned inward,
not closely approximated. Descending processes of prefrontals
having transverse plates forming posterior wall to nasal cavity
and bounding a keyhole-shaped fissura ethmoidalis (Gaffney
1972). Occipital region of skull bearing massive facets for at-
tachment of nuchal muscles (see Shah 1963) as follows: deep
facets on posterior faces of squamosal and opisthotic bones; a
large concave facet and four heavily developed posterior pro-
cesses on ventral surface of basioccipital (Fig. 2).

Internal nares large. Anteriormost part of palate thin and
forming a distinct semicircular depression bounded anteriorly
and laterally by low ridges for attachment of maxillary sheath.
Prepalatine foramina bordered by premaxillaries anteromedial-
ly and by maxillaries posterolaterally. Palate and maxillary re-
gion relatively narrow. Base of skull especially narrow at
pterygoid “waist” just posterior to trochlear processes of
pterygoids.

Mandibular ramus straight and lightly built; in dorsal as-
pect, lines drawn from tip of symphysis to centers of articular
facets lie almost entirely within the ramus (and form a mean

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 324:1-18

Legler and Cann: New Australian Chelid

9

angle of 46.5 degrees; Table 4). Intermandibular articulation
long and coossified without trace of suture or symphysis.
Splenial bone absent as individual bone in lower jaw. Meckelian
canal and fossa broadly exposed below coronoid on media! face
of mandibular ramus; coronoid-prearticular contact lacking in
two specimens, a narrow contact in one specimen. Angular
long, extending well past anterior extent of coronoid; tip of cor-
onoid (1.5 mm) clearly visible in lateral view. Mandible bearing
well developed retroarticular processes formed chiefly by the
angular bones.

COLOR IN LIFE. The following color description is based
on live, fresh-caught material. Adults were examined indoors in
diffuse summer daylight. All parts described were viewed when
slightly moist (after wiping with a damp cloth). Descriptions of
hatchlings are based on numerous close observations in an
aquarium lighted by a fluorescent bulb (Vita light). See Cann
(1978) for color photos.

ADULT FEMALES (UU 17110, 17113, QM J31701,
LACM 127778). Overall ground color of carapace medium to
dark brown with a suggestion of olive in palest specimens; car-
apace dull (probably due to complex surface texture); slightly
paler over interosseous sutures; no regular markings on car-
apace; a few specimens showing one to three small (5 to 15 mm
diam.) elliptical black spots irregularly placed on carapace.
Plastron unicolored in general aspect, ground color dark neu-
tral straw to horn, somewhat darker near bridges; no distinct
dark markings; no distinct darkening of interlaminal seams;
areas where interosseous sutures are clearly visible through

scutes are suffused with pink; inframarginal surfaces slightly
darker than ventral plane of plastron, interperipheral sutures
showing through as distinct pale lines. Soft parts (other than
head) generally pale to medium slatey olive-grey, not at all
bright or distinctive. Interdigital webbing paler than rest of
foot; enlarged scales on fore and hind limbs darker than ground
color; ventral surface of thigh paler than dorsal surface — a
slightly brighter pale ground color than plastron; skin of ingui-
nal pocket dull yellowish cream. Dorsal aspect of head and
neck uniform brownish (same shade as carapace), olive in pal-
est specimens; ventral ground color pale yellowish-orange, the
division between dark and pale colors along a line extending
posteriorly from angulis oris; orange color intensified in a verti-
cal bar on anterior edge of tympanum and just posterior to
mandibular symphysis where it is nearly pink; barbels paler at
tips than bases. Jaw sheaths pale neutral olive where underlain
by bone, straw at free edges; maxillary sheath streaked with
slatey gray; mandibular sheath uniform. Iris a narrow ivory-
colored band surrounded by rich brown. Tongue and inside of
mouth pale salmon pink.

HATCHLINGS (approximately 150 from 9 clutches).
Carapace tan to pale brown; marked with blackish-brown
flecks; each lateral and central scute bearing a keel on which
there is an intensification of dark color; this results in a series of
five vague black dots along middorsal line and four such dots on
each side; dots not necessarily more evident than general fleck-
ing. Plastral ground color pale neutral slate; yellowish suffusion
produced by translucency of plastron and underlying yolk mass;

Figure 4. Rheodytes leukops, one of the individuals shown in Figure 3; note heavy development of neck and limbs and prognathous
appearance of head.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 324:1-18

Table 3. Comparisons of scute, head, and shell proportions for adults of four species of short-necked chelids (means and extremes for each sex). All
specimens are from the Burnett, Fitzroy, and Raglan Creek drainage systems (latitudinal range 23° IT to 26° 19'S). Note average size of adults (E.
dentata largest, E. latisternum smallest) and degree of sexual dimorphism in size (least in R. leukops).

Greatest
Diameter of

Carapace
Length (mm)

Head
Width
h- CL

Least Inter-
orbital Breadth
-h Head Width

Orbit +
Length of
Mandibular
Symphysis

Carapace
Width -h
CL

Shell

Height

CL

Femoral
Scute
Length
h- CL

Anal
Scute
Length
h- CL

Rheodytes leukops

Males, N = 9

243

0.15

0.20

0.73

0.81

0.29

0.21

0.12

235-262

0.13-0.16

0.18-0.21

0.55-1.02

0.79-0.83

0.27-0.30

0.20-0.22

0.11-0.13

Females, N = 19

249

0.13

0.19

0.82

0.80

0.31

0.22

0.12

231-259

0.12-0.14

0.17-0.21

0.75-1.02

0.76-0.85

0.30-0.34

0.19-0.24

0.11-0.14

Elseya dentata

Males, N = 9

263

0.17

0.27

0.91

0.79

0.33

0.22

0.10

249-269

0.15-0.19

0.24-0.29

0.81-1.17

0.77-0.85

0.31-0.36

0.19-0.23

0.09-0.12

Females, N = 9

343

0.17

0.26

0.85

0.79

0.35

0.21

0.10

309-374

0.15-0.19

0.23-0.30

0.64-0.99

0.75-0.83

0.31-0.40

0.18-0.24

0.09-0.12

Elseya latisternum

Males, N = 6

164

148-172

0.19

0.24

0.22-0.26

0.97

0.91-1.09

0.80

0.79-0.83

0.30

0.29-0.31

0.15

0.13-0.17

0.18

0.17-0.19

Females, N = 3

232

0.20

0.24

0.80

0.80

0.31

0.14

0.18

172-271

0.20-0.21

0.23-0.26

0.70-0.98

0.78-0.81

0.30-0.32

0.13-0.15

0.16-0.19

Emydura kreffti

Males, N = 10

221

0.15

0.25

0.90

0.73

0.36

0.19

0.1 1

187-256

0.13-0.17

0.22-0.28

0.75-1.07

0.75-0.79

0.34-0.38

0.17-0.21

0.10-0.13

Females, N = 16

272

0.17

0.26

0.73

0.76

0.38

0.18

0.12

234-277

0.15-0.19

0.22-0.30

0.59-0.95

0.71-0.82

0.36-0.41

0.15-0.20

0.09-0.13

Table 4. Skull proportions in adults of four species of short-necked chelids from eastern Australia. Means and extremes are given for each sex. See
“Methods and Materials” for explanation of measurements.

Condylo-

basilar

Length

(mm)

Squamoso-
nasal
Length
- CBL

Height
of Snout
- CBL

Mandible:
Greatest
Width -
Greatest
Length

Length
of Dentary
Symphysis
Greatest Width
of Mandible

Maxillary
Breadth
- CBL

Mandibular

Angle

(degrees)

Rheodytes leukops

Males, N = 1

40.0

0.94

0.16

0.85

0.33

0.49

46.3

Females, N = 2

42.3

41.6-42.9

0.92

0.91-0.94

0.15

0.13-0.17

0.85

0.84-0.87

0.35

0.34-0.35

0.50

46.4

46.3-47.0

Elseya dentata
(Burnett and N.
Johnstone drainages)

Males, N = 2

49.3

48.5-50.1

1.05

1.01-1.08

0.19

0.16-0.21

0.95

0.24

0.23-0.25

0.53

0.52-0.55

51.0

50.9-51.1

Females, N = 3

63.9

56.7-69.0

1.03

1.03-1.07

0.20

0.20-0.21

0.96

0.90-1.03

0.26

0.21-0.32

0.54

0.53-0.55

51.2

48.8-54.2

Elseya latisternum
(Cape York Peninsula)

Males, N = 2

40.3

38.9-41.8

0.98

0.95-1.01

0.22

0.21-0.23

0.95

0.95-0.96

0.28

0.27-0.29

0.57

0.55-0.59

51.1

50.8-51.3

Females, N = 4

48.4

46.6-49.2

1.01

0.99-1.03

0.21

0.20-0.22

0.95

0.92-0.98

0.31

0.29-0.32

0.59

0.57-0.61

50.9

49.7-52.3

Emydura kreffti
(Burdekin drainage)

Males, N = 3

39.3

39.0-39.4

0.99

0.97-1.02

0.21

0.19-0.23

0.90

0.86-0.93

0.25

0.23-0.27

0.54

0.53-0.55

48.8

46.5-49.7

Females, N = 4

43.9

41.5-46.5

0.99

0.98-0.99

0.19

0.18-0.19

0.95

0.91-0.98

0.27

0.26-0.27

0.59

0.57-0.61

50.7

49.0-52.5

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 324:1-18

Legler and Cann: New Australian Chelid

11

irregularly flecked with melanin, flecks tending to be concen-
trated on anterior lobe and bridge (posterior lobe immaculate
in some individuals). Iris a metallic silvery-blue, not the clear
ivory of adults. Hatchlings placed under water on fine brownish
river sand were well camouflaged.

ONTOGENETIC COLOR CHANGE. There seems to be
a gradual paling of colors in females as they grow older, but this
is not at all distinct. There is a rather distinct change that oc-
curs in males with age, culminating in a series of bright con-
trasting colors on the head and neck. This seems to be a process
of progressive melanin loss and its replacement by orangish-
yellow or rosy pigment in some areas. In general, these areas
are the upper eyelids, the region around the tympanum, and to
a lesser extent, the snout and throat. The oldest and brightest
males we have examined (not necessarily the largest) present a
rather handsome, striking, tricolored appearance with an im-
maculate greenish-olive maxillary sheath, an orange-ringed
eye, and to a lesser extent, an orange-ringed tympanum (giving
the animal a clown-faced appearance; Fig. 6). Although the
color of the iris does not change in older males, it is set in a
more contrasting situation (the pale iris, the surrounding dark
part of the eyeball, the surrounding orange ring on the eyelids,
and the very dark brown ground color).

ECOLOGY

HABITAT. In the region of the type locality, the Fitzroy
River flows northwestward; large deep pools alternate with fast
shallow riffles. The type locality is a riffle 15 to 27 m wide and
400 m long, grading from a depth of 0.3 to 0.6 m at its up-
stream limit (where the stream is fordable) to 2.5 to 3.0 m at
its downstream limit (Fig. 8). Above the riffle, there is a pool
about 1 km long with high, earthen, wooded banks. The riffle
flows into a similar downstream pool, which is 2 to 3 km long.
Current in the large pools is undetectable, and depths are in
excess of 10 m; the bottom is rocky, with a heavy layer of silt,
especially near the banks. Current in the riffle varies from very
fast at the upper end (difficult to wade without support) to slow
at the downstream end (scarcely noticeable when diving).

A total of 35 adults of Rheodytes leukops was obtained; 28
of these were retained for study, and 7 were released at point of
capture. Most of the specimens were captured by diving with a
face mask and snorkle (one was caught in a gill net). We there-
fore had the opportunity to observe all resident turtle species
under water. We dived once in the lower part of the large up-
stream pool and repeatedly in the entire length of the riffle. Gill
nets were set in the downstream pool, but we did not dive there
because of the presence of large crocodiles (Crocodylus por-
osus, 3 to 3.6 m total length). Underwater visibility in the riffle
was 1.2 to 2.4 m; water temperature was 25.5 °C.

Rheodytes leukops definitely prefers fast water. None were
seen in the upstream pool; one was taken in the downstream
pool in a gill net (no other turtles of any kind were captured in
the net). The remaining turtles at the type locality were taken
in the riffle — some from the fastest water, most from water of
medium depth and medium current, and none in a still backwa-
ter. About half of these were on clean sand or gravel bottom
facing upstream. Others were on the downstream sides of rocks
or actually under such rocks. In slower water, the turtles were
in a greater variety of places (near dead wood, etc.).

About half of the R. leukops broke cover when we were 1 to
2 m away and swam off about as fast as we could pursue on a
straight chase. However, typical escape behavior was a long ris-
ing curve and then an abrupt return to the bottom; if at this
point a diver were directly over the turtle, it would reverse its
course 180 degrees and swim away on the bottom. We consider
R. leukops to be faster than any other Australian chelid, but
Elseya dentata is almost as fast. The terminal reverse in escape
behavior is virtually identical in all short-necked chelids
(Legler has also observed it in kinosternids, emydids, and
trionychids).

The short-necked species Elseya dentata and Emydura
kreffti were also common in the river. Both occurred in the rif-
fle, but R. leukops was the commonest turtle there. Emydura
kreffti was most likely to be found in association with dead
wood or undercut banks and was never in the fastest water.
Elseya dentata tended to be in slower water but was seen or
caught in the entire range of microhabitat described. Chelodina
longicollis occurred in quiet, clear backwater situations where
there were no R. leukops but where a few of the other short-
necks were observed. We saw no Elseya latisternum in large
rivers (Fitzroy and Dawson) at this latitude but found them to
be common (with Emydura kreffti) in smaller tributary
streams where there were no R. leukops.

Rheodytes leukops was also taken or seen on the Fitzroy
River at Glenroy Crossing (23 ° 1 l'S, 149° 56'E), 3 to 4 km up-
stream from the type locality, and on the Dawson River, 2 km
N Gainesford, Queensland (23°47'S, 149°46'E). Habitat at
Glenroy Crossing was virtually identical to the riffle at the type
locality. The Dawson River consisted of deep pools connected
by shallower stretches in which there was no noticeable current;
R. leukops, Elseya dentata, and Emydura kreffti were all
found, in small numbers, in association with dead wood.

Figure 5. Optical cross sections and silhouettes of four short-necked
chelids from eastern Australia (prepared from photographs); length
and width of carapace (in millimeters) are given for each specimen. In
order from top to bottom rows: Elseva latisternum, UU 170749, CL
271, CW 219; Elseya dentata. UU 170989, CL 309, CW 243;
Emydura kreffti, UU 169099, CL 262, CW 200; Rheodytes leukops,
paratype, UU 171189, CL 238, CW 196. All specimens obtained at or
near the type locality of R. leukops.

Contrib. Sci. Natur. Hist. Mas. Los Angeles County. 1980. 324:1-18

12

Legler and Cann: New Australian Chelid

Figure 6. Heads of live short-necked Australian chelids, from left to right. Top row: Elseya dentata, UU 148629, Malanda, Queensland — note
piebald coloration characteristic of old females; Elseya latisternum, UU 151319, Bloomsbury, Queensland. Middle row: Pseudemydura umbrina, A.
Burbidge field number 46cf, Twin Swamps Reserve; Emydura krefftiQ from the series UU 15881-8, Eton, Queensland. Bottom row: Rheodytes
leukops, UU 171 129, paratype; Rheodytes leukops UU 171070”, paratype — note texture and coloration, which is typical of old males.

Contrib. Sci. Natur. Hist. Mits. Los Angeles County. 1980. 324: 1-18

Table 5. Eggs of five species of chelids occurring on the Fitzroy River drainage system of Queensland. Mean and standard deviation are given above
the extremes (in parentheses). Data for Elseya dentata are from other drainage systems. Annual reproductive potential is the sum of corpora lutea
and potentially preovulatory ovarian follicles (minimum considers all follicles greater than 10 mm; maximum considers all follicles greater than 5
mm). Eggs from nest on Fitzroy R. are assumed, by elimination, to be those of R. leukops.

Mean and Std. Deviation

Sample (N)

Annual Reproductive
Potential

Incubation Time

Species

Weight (g)

Length (mm)

Width (mm)

Eggs

Clutches

at 30.0°C

Minimum

Maximum

Rheodytes leukops

7.50±0.582

(4.70-9.84)

29.67 ±0.969
(23.2-33.1)

2 1 . 1 7 ± 0.595
(19.0-23.8)

188

10

46.7

(44-50)

45.9±6.25

(34-53)

58.6±8.76

(44-70)

Nest on Fitzroy R.

—

30.7

(30.0-31.2)

21.8

(21.4-22.0)

5

1

Emydura kreffti

9.75±0.374

(8.74-11.35)

36.45±0.936

(32.9-39.6)

2 1 . 1 3 ± 0.27 3
(20.0-22.4)

82

5

47.25

(46-48)

5 1 .0 ± 15.79
(29-66)

56. 2± 16.9
(35-75)

Elseya latisternum

1 2.05±0.980
(11.22-15.70)

35.85 ± 1.320
(35.0-40.8)

23.98±0.535

(23.2-25.7)

17

1

60.5

(60-61)

46

53

Chelodina longicollis

7.51 ±0.384
(6.73-8.09)

30.81 ±0.829
(29.0-32.0)

20.40±0.272

(19.9-20.7)

12

1

67.0

(66-68)

24

33

Elseya dentata

15.7

48.6

(48.1-50.2)

27.7

(27.2-28.1)

5

309

1

19

160

(160-161)

Insufficient data

DIET. Stomachs

of eight fresh-caught live

adults were

mm.

The nests

were on a flat

sand and gravel

bar approx-

flushed (Legler 1977) at the type locality on 7 October 1976.
Stomach contents consisted purely of animal material and were
entirely of insects except for fragments of freshwater sponge.
Of these, the commonest were Trichoptera (larvae and pupae)
of the families Helicopsychidae, Hydropsychidae, Hydrop-
tilidae, and Leptoceridae, the first family being of commonest
occurrence. By estimate, a single stomach might contain at
least 100 larvae and their cases (most of which were un-
crushed). Large aquatic lepidopteran pupae (Pyralidae?; 10 to
20 mm) were found in most of the stomachs and made up the
bulk of the food in some stomachs. Lepidopteran larval cases
were constructed partly of leaves, which were passed un-
digested in the feces. Dipteran and ephemeropteran larvae oc-
curred as occasional items as did adult terrestrial coleopterans.
Plant materials found in stomachs (bark, algal filaments, pieces
of fibrous stem) were either closely associated with insect lar-
vae or actually incorporated in their cases and were therefore
assumed to have been ingested incidentally.

REPRODUCTION. A female obtained in mid-September
was gravid, and 1 1 of the 18 females (61 percent) collected on
7-9 October were gravid. Eggs were obtained by hormonal in-
duction of oviposition (Ewert and Legler 1978) or by dissection
and were incubated under controlled laboratory conditions. All
of the dissected females bore corpora lutea, indicating that they
had produced at least two clutches of eggs by early October,
and enlarged ovarian follicles, suggesting that they would pro-
duce at least one more clutch in the same season. Multiple
clutches are, in fact, the rule for all species of chelids at this
latitude (Legler, personal observation). The reproductive char-
acteristics of five species of chelids occurring in the Fitzroy
drainage are given in Table 5. Reproduction in Australian
chelids will be the subject of a separate paper by Legler.

Two nests were found at the type locality; they had been con-
structed and robbed within the preceding 48-hour period (prob-
ably the previous night). Five eggs in one of the nests identified
the nest (by elimination) as that of a Rheodytes leukops. One
nest contained a slanting cavity with a maximum depth of 170

imately 9 m from the fast water of a riffle and slightly closer to
some stagnant water. They were less than 1 m above the level of
the flowing water.

BEHAVIOR OF YOUNG. Of the eggs incubated and
hatched in Armidale, 26 juveniles (representing two clutches)
were transported to the United States on 5 January 1977 and
have been under observation in aquaria at the University of
Utah since then. Several hatchlings from each of the same
clutches were retained by Cann for observation in Sydney. An
attempt was made to maintain captives in an environment like
that (temperature and physical characteristics) of the original
habitat.

Juveniles show a distinctive gregarious behavior, especially
when no individual refugia are available (e.g., a bare aquarium
or plastic bucket): They cling together, the claws of one grasp-
ing the shell of another — at various angles — to form a roughly
spherical mass of individuals (Fig. 3). On some occasions, an
entire clutch of hatchlings was involved in this clinging
behavior.

Small juveniles spend much of their time under cover. They
can conceal themselves effectively; in a 5-gallon (19-liter) tank
containing 10 to 13 individuals, few can be seen. Rock caves
and crevices are favorite retreats. When such retreats are not
available, the turtles are quite capable of burrowing into the
sand beneath a rock or piece of dead wood. Hatchlings were
also observed burrowing into the sand without the aid of lever-
age from a rock; as they dig into the substrate, sand is thrown
up and settles on the carapace. This behavior is reminiscent of
sand burrowing in Trionyx and Carettochelys (personal obser-
vation) but not nearly as efficient.

The young are violently aggressive toward one another at
times. This behavior could have been territorial, and it was
nearly always observed on a feeding day (when most of the ani-
mals were out of their refugia). The behavior consisted of a
visually oriented chase (begun in some instances from a dis-
tance of 15 cm) and violent biting of the head, neck, and fore-
quarters of one juvenile by the other. The bitten animal did not

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 324:1-18

Figure 7. Rheodytes leukops, new genus and species: Paratype UU 171 149, carapace length, 247 mm. Dorsal and ventral views of scuteless shell.
Dots on left indicate where rib tips form gomphoses with peripherals. Small crosses indicate position of dorsal tips of plastral buttresses. The crescent-
shaped break on left ninth marginal probably resulted from a Crocodylus porosus bite.

Figure 8. Type locality of Rheodytes leukops (Fitzroy River, 23°09'S, 149°55'E) looking downstream from the riffle. Water in left
foreground is about 1 m deep and very fast; water near dead wood in right background is much deeper and slower. Two nests of R. leukops
were found on the sandy area just to the right of this picture.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 324:1-18

Legler and Cann: New Australian Chelid

15

retaliate and was seldom chased very far. The dominant animal
seemed not to return to any set place. Aggressive behavior was
commonest just after feeding, occurred occasionally before
feeding (and at other times) but never occurred during feeding
(the animals were not fighting for food). Fighting after feeding
lasted from 2 to 5 minutes, after which the turtles disappeared
into refugia. Some refugia were shared by as many as three or
four turtles.

When the small captives were active, they engaged in a lot of
probing at the substrate with the snout. This probing seemed to
involve both olfactory and visual cues. Although there was little
or nothing for them to eat between feedings, they often stopped
and bit at objects on rock faces (presumably algae). This be-
havior is tentatively interpreted as foraging. The hatchlings are
easily able to move underwater objects at least ten times their
own weight (average weight of hatchlings in April 1977, ca. 5.0
g); to do so, they first wedge the head and forelimbs under an
object, then apply force with all four limbs. They moved the
small half flower pots (50 g) that were supplied as caves about
the aquarium in this way and were easily able to burrow com-
pletely into coarse gravel (individual stones 6 to 10 g).

The ability of young to secrete themselves in sand, gravel,
and crevices probably explains our failure to find any immature
stages at all in the field. This behavior also suggests some of the
methods that the turtles use in foraging for insect larvae.

DISCUSSION

RELATIONSHIPS. The “short-necked” Australian chelids
comprise the genera Elseya, Emydura, Rheodytes, and
Pseudemydura (Figs. 6 and 9). Although the necks of these
taxa are quite obviously relatively shorter than those of the
“long-necks” (genus Chelodina), necks are quite difficult to
measure accurately in live or preserved specimens. The groups
are better categorized as follows:

Frontal bones fused; no parietosquamosal articulation; no
posterior temporal arch; claws 4-4; intergular scute iso-
lated from anterior free margin of plastron by anterior
union of gular scutes; diploid chromosome number 54
(Bull and Legler 1980) Chelodina

Frontal bones separate; a parietosquamosal articulation
forming a posterior temporal arch (or contributing to more
extensive skull roofing in Pseudemydura)-, claws 5-4; inter-
gular scute completely separating gular scutes (partly sep-
arating humerals) and contributing to anterior free edge of

plastron; diploid chromosome number 50

Short-necked taxa

Of the short-necked taxa, Pseudemydura is rare (probably
nearing extinction), isolated, and the least understood of the
group (Burbidge 1967; Burbidge et al. 1974). Legler’s studies
of Pseudemydura (based upon all available specimens and to be
reported elsewhere) suggest that it is not especially closely re-
lated to other short-necked Australian chelids. The genera cur-
rently recognized as Emydura, Elseya, and Rheodytes bear
enough in common to be regarded tentatively as having com-
mon ancestry. In no case, however, is this similarity sufficient
(in Legler’s opinion) to warrant the lumping of these generic
groups (Table 1).

Within the genus Elseya, there seem to be two major groups
as follows; (1) Elseya dentata group — a series of populations
(probably discontinuous) from the Burnett R. drainage, in the
coastal lowlands and tablelands, northward and westward at
least to the Victoria drainage of the Northern Territory; at least
two Australian species; closely related to Elseya novaeguineae
in New Guinea; (2) Elseya latisternum group — Elseya latister-
num plus undescribed taxa in the Murray-Darling headwaters
and eastern coastal drainages from approximately 32 'S north-
ward to the tip of Cape York and westward to approximately
140°E; possible relationships to undescribed taxa in New
Guinea.

In the comparisons made in Table 1, the species names
Elseya dentata, Elseya latisternum, and Emydura kreffti are
used, but the characters discussed or compared are shared by
the entire genus or proposed generic group.

MANDIBLE. Gaffney (1976) stated that the splenial bone
was a common feature of chelids but occurred in no other ex-
tant group of chelonians. We have personally confirmed the oc-
currence of the splenial in all Australian chelids except R.
leukops. Lack of a splenial would seem to be a derived charac-
ter within the Family Chelidae. The shape of the prearticular in
Rheodytes suggests that the splenial has become fused with
that element (Fig. 2).

SEAM AND SUTURE CONTACTS. A midperipheral
gomphosis for the rib tips of costals 2-4 occurs elsewhere in the
Chelidae only in Chelodina; it is the usual condition in cryp-
todires. The same rib tips are interperipheral in all other
chelids and in pelomedusids. Modal seam contacts on the pos-
terior parts of M6 (75 percent of sample) and M8 (96 percent)
are unique to Rheodytes among chelids; in other chelids, these
contacts occur only as occasional variations, and they are rare
among other living chelonians (Tinkle 1962; illustrations in
Boulenger 1889).

We regard these two unusual situations in Rheodytes as re-
lated, derived characters. The relationships of lateral scutes to
costal bones and of marginal scutes to peripheral bones remain
fairly constant in all chelonians — the scutes tend to straddle the
sutures in a way that seems to create maximal structural ef-
ficiency. However, either of these series may (and often does)
shift upon the other. In Rheodytes, it is quite evident that the
marginal-peripheral series has shifted anteriorly (or conversely
the lateral-costal series has shifted posteriorly) for a distance
equal to one-half the length of a peripheral bone in the bridge
region. It is not clear what this shift has accomplished func-
tionally or structurally.

CLOACAL BURSAE. One of our vivid early impressions of
Rheodytes was that adults of both sexes swam with a widely
gaping cloacal orifice (up to 30 mm in diameter). The orifice
remains open when individuals are out of water. We first be-
came aware of the large cloacal bursae when a female was ex-
amined in bright sunlight; the carapace transmitted enough
light to illuminate the coelomic cavity and to produce a spec-
tacular view internally for at least 100 mm, via the cloaca, re-
vealing a large sac fined with a vascular, villose mucosa. We at
first thought this was an oviduct, but dissection revealed it to be
a large and unusual cloacal bursa. The structure and function
of these organs is under investigation and is summarized else-
where (Legler 1979). Water is pumped in and out of the bursae
of captives and experimental animals at rates of 15 to 60 times
per minute. Captives seldom breathe air, and we saw no heads

Contrib. Sci. Natur. Hist. Mas. Los Angeles County. 1980. 324: 1-18

16

Legler and Cann: New Australian Chelid

of Rheodytes at the surface under natural conditions. These
facts suggest a respiratory function for the cloacal bursae.
Elseya dentata also displays a gaping cloaca while swimming
but has smaller cloacal bursae than R. leukops.

EYELIDS. The lack of a nictitating membrane is charac-
teristic of all chelids (Legler, personal observation) and is un-
mentioned thus far in the literature. Early but brief mention of
the translucent lower eyelid in Chelodina was made by Gadow
(1909) and Walls (1942; “tertiary spectacle”). The anatomy of
eyelids and the distribution of nictitating membranes in living
chelonians is presently under study in Legler’s laboratory.

ADAPTATIONS. Rheodytes leukops has relatively massive
cervical vertebrae and a relatively small head in comparison to
Elseya dentata. If the condylobasilar length of the skull is di-

vided by the sum of the lengths of the centra of cervical ver-
tebrae 2-8, the following figures are obtained (mean, extremes,
number of specimens): Rheodytes leukops — 0.393 (0.38-0.40),
N = 3; Elseya dentata — 0.495 (0.47-0.52), N = 2.

The described structure of the skull and cervical vertebrae,
the observations on behavior, and the examination of fresh
stomach contents all strongly indicate that Rheodytes has be-
come adapted to bottom probing and possibly the scraping of
rocks and dead wood for insect larvae. The function of the pad
of erectile tissue on the anterior part of the palate is unknown.

GEOGRAPHIC DISTRIBUTION. Localities for two of the
paratypes in the Australian Museum collection are uncertain,
and attempts to ascertain them more specifically have failed.
Concerning AM R41794 (“Mackenzie River”), if this part of

Figure 9. Skulls and mandibles of three species of short-necked Australian chelids, in vertical rows. Left: Elseya dentata, UU 147999, Auvergne,
Northern Territory, condylobasilar length 64.5 mm. Middle: Elseya latisternum, UU 150379, northern New South Wales, CBL 46.3: Right:
Emydura kreffti, UU 156509, Cooktown, Queensland, CBL 44.8.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 324:1-18

Legler and Cann: New Australian Chelid

17

the locality is correct, it establishes the presence of the genus in
another major tributary of the Fitzroy River. The Mackenzie
River begins at the confluence of the Nogoa and Comet Rivers,
flows northeastward to its confluence with the Connors River
and then southeastward to its confluence with the Fitzroy River
(24°24'S, 149°53'E). AM R41274 was probably taken in the
general region of the type locality (i.e., downstream from the
confluence of all major tributaries). Except for the collection
sites of these two specimens, all localities used in this descrip-
tion may be regarded as exact.

The specimens listed in this paper define the entire known
range of R. leukops (Fig. 10). Little collecting has been done
elsewhere on the Fitzroy drainage. Emydura kreffti was taken
in traps (UU 15847-61) from the Nogoa River near Emerald,
Queensland (23°31'S, 148° 1 l'E), in July 1973, and a specimen
of Elseya dentata (QM J28449) is known from the same lo-
cality. The river was murky and without current in 1973 when
Legler observed it (due possibly to the first filling of Fairbairn
Dam immediately upstream). It seems likely that, if Elseya
dentata could survive in this kind of microhabitat, Rheodytes

could also. Some of the tributaries of the Connors River (ca.
22°00'S, 148°53'E) contain permanent, clear running water
and may provide suitable habitat for Rheodytes. It seems vir-
tually certain that Rheodytes occurs farther upstream in the
Dawson River than here reported (say, southward at least to
Taroom [25°39'S, 150° 1 l'E], where Elseya dentata is known
to occur [Legler, personal observation]).

It seems likely that R. leukops is confined to the Fitzroy
drainage. We have observed, dived, and trapped in the Burnett
drainage and short drainages between the Burnett and Fitzroy
(e.g., Raglan Creek) to the south and in the Burdekin drainage
and short drainages between the Burdekin and the Fitzroy to
the north. Thus far, there is no evidence that Rheodytes occurs
in any of these places.

COMPARATIVE MATERIALS EXAMINED

Legler’s work on Australian chelids since 1972 has produced a
collection of approximately 3,000 specimens (currently housed
at the University of Utah but to be divided among that institu-

I i 8° 130° 142°

Figure 10. Known geographic distribution of Rheodytes leukops, new genus and species, in Australia. Localities from which specimens
examined were taken are represented by four solid dots on the Fitzroy River and its tributaries. Arrows show mouths of Burdekin and
Burnett Rivers. Circles showing MacKay and Brisbane are for orientation only. Drainage map of Australia adapted from Vari 1978.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 324:1-18

18

Legler and Cann: New Australian Chelid

tion and various Australian collections eventually). Data have
been taken on approximately 1,500 other specimens in Aus-
tralian and American museums. This large bank of specimens
and data was consulted ad libitum in the preparation of this
paper. Specimens used especially for the comparisons made in
this paper are mentioned specifically in the text and legends
and are listed below by major drainage system and museum
number.

Elseya dentata. NORTHERN TERRITORY. Victoria R.:
UU 14777, 14793-800; NMV 10386-90, 10397-99, 10403-6,
10828-30, 10832-35, 10847-48, 10850, 10859-60, 10870-74,
10885. Daly R.: UU 14809-844. Finniss R.: UU 14776. Ade-
laide R.: UU 14772-5. South Alligator R.: UU 14784-92.
Roper R.: UU 14778-83. QUEENSLAND. Gregory R.: UU
14801-08. North Johnstone R.: UU 14845-71. Fitzroy R.: UU
17093-102. Burnett R.: UU 14872, 17085-92.

Elseya latisternum. QUEENSLAND. Cape York Peninsula:
UU 14873-929. Mitchell R.: UU 14930-79. Endeavour R.: UU
14980-991. North Johnstone R.: UU 14994-15008. Burdekin
R.: UU 15079. Andromache R.: UU 15124-51. Pioneer R.: UU
15081-123. Fitzroy R.: UU 17049-64, 17074-76. Raglan Cr.:
UU 17070-73. Burnett R.: UU 15199-201, 17065-69. Brisbane
R.: UU 15152-169. Nerang R.: UU 15170-71. Tallebudgera
Cr.: UU 15172-98. NEW SOUTH WALES. Tweed R.: UU
15026-78, 17042-48. Richmond R.: UU 15009-25, 17077-80.

Emydura kreffti. QUEENSLAND. Normanby R.: UU
15674-722. Endeavour R : UU 15619-673. Burdekin R.: UU
15723-32. Pioneer R.: UU 15867-98. Fitzroy R.: UU
15847-61, 16892-904, 16906-59, 16978-17029. Raglan Cr.:
UU 16883-91. Burnett R.: UU 15834-46, 16866-882, 16905,
16960-77, 17066.

Emydura macquarii. NEW SOUTH WALES. Murray-
Darling drainages: UU 15954-16085, 16863.

Emydura australis. NORTHERN TERRITORY. Victoria
R.: UU 15462-95. Finniss R.: UU 15437-57. QUEENSLAND.
Wen lock R.: UU 15319-69. Mitchell R.: UU 15371-405.

Pseudemydura umbrina. All specimens from Twin Swamps
and Ellen Brook Reserves, ca. 29 km NE Perth, Western Aus-
tralia: WAM 1 1092-3, 11386, 13385, 13744-5, 21559-64,
21859, 29320, 29337-40, 29342-3, 29345, 29348, 29350,
29376, 36159, 36179, 36338, 37495, 37977, 39040, 39956,
40535.

ACKNOWLEDGMENTS

We are grateful to the Queensland Fisheries Service and the
Australian National Parks and Wildlife Service for permission
to collect and export specimens, to A. Burbidge, H.G. Cogger,
J. Covacevich, and G. Storr for permission to utilize collections
in their care, to R. Ohl for assistance with field work, and to
Steve Johnson for identification of insects. A.F. Legler assisted
substantially in all phases of the work in Australia and pre-
pared the line drawings of the skull and shell. H.G. Cogger,
E.E. Williams, J.W. Wright, and G. Zug read the manuscript
and made useful comments on it. Special thanks are due the
University of New England for permitting the use of laboratory
facilities in Armidale, New South Wales. This work resulted
from projects supported partly by the National Geographic So-
ciety and the University of Utah Research Committee.

LITERATURE CITED

Boulenger, G.A. 1889. Catalogue of the chelonians.
rhynchocephalians and crocodiles in the British Museum
( Natural History ). London.

Bull, J.J., and J.M. Legler. 1980. Karyotypes of side-necked
turtles (Testudines: Pleurodira). Canadian Jour. Zool.
58:828-41

Burbidge, A. A. 1967. The biology of south-western Australian
tortoises. Ph.D. Thesis, Univ. Western Australia,
Nedlands.

Burbidge, A. A., J.A.W. Kirsch, and A.R. Main. 1974. Re-
lationships within the chelidae (Testudines: Pleurodira) of
Australia and New Guinea. Copeia 1974(2):392-409.

Cann, J. 1978. Tortoises of Australia. Sydney: Angus &
Robertson (92 color photographs).

Carr, A. 1952. Handbook of turtles. New York: Comstock
Publ. Assn., Cornel! Univ.

Ewert, M.A., and J.M. Legler. 1978. Hormonal induction of
oviposition in turtles. Herpetologica 34( 3): 3 1 4— 1 8 .

Gadow, H. 1909. Amphibia and reptiles. London: MacMillan.

Gaffney, E.S. 1972. An illustrated glossary of turtle skull no-
menclature. Amer. Mus. Novitates 2486.

. 1976. Cranial morphology of the European Jurassic

turtles Portlandemys and Plesiochelys. Bull. Amer. Mus.
Nat. Hist. 1 57(6):487-544.

Legler, J.M. 1976. Feeding habits of some Australian short-
necked tortoises. Victorian Nat. 93(2):40-43.

. 1977. Stomach flushing: A technique for chelonian di-
etary studies. Herpetologica 33:281-84.

. 1979. Cloacal gills in Australian chelid turtles. 15 min.

video tape and abstract. University of Utah Educational
Media Service, Salt Lake City.

Legler, J.M., and R.M. Winokur. 1979. Unusual neck tu-
bercles in an Australian turtle, Elseya latisternum (Test-
udines: Chelidae). Herpetologica 35(4): 325-29.

Shah, R.V. 1963. The neck musculature of a Cryptodire (De-
irochelys) and a Pleurodire (Chelodina) compared. Bull.
Mus. Comp. Zool. 1 29(6):343— 68.

Tinkle, D.W. 1962. Variation in shell morphology of North
American turtles, 1: The carapacial seam arrangements.
Tulane Stud. Zool. 9(5):33 1-49.

Vari, R.P. 1978. The Terapon perches (Percoidei, Teraponid-
ae): A cladistic analysis and taxonomic revision. Bull.
Amer. Mus. Nat. Hist. 159(5): 175-340.

Walls, G.L. 1942. The vertebrate eye and its adaptive radia-
tion. Bull. Cranbrook Inst. Sci. 19.

Williams, E.E. 1950. Variation and selection in the cervical
central articulations of living turtles. Bull. Amer. Mus.
Nat. Hist. 94(9):505-562.

Winokur, R.M., and J.M. Legler. 1974. Rostral pores in
turtles. Jour. Morph. 143:107-119.

Submitted 21 September 1978. Accepted for publication 18

May 1979.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 324: 1-18

t

SERIAL PUBLICATIONS

OF THE NATURAL HISTORY MUSEUM OF LOS ANGELES COUNTY

The scientific publications of the Natural History Museum of Los Angeles County have been issued
at irregular intervals in three major series; the articles in each series are numbered individually, and
numbers run consecutively, regardless of subject matter:

• Contributions in Science, a miscellaneous series of technical papers describing original
research in the life and earth sciences.

■ Science Bulletins, a miscellaneous series of monographs describing original research in the
life and earth sciences. This series was discontinued in 1978 with the issue of Numbers 29
and 30; monographs are now published by the Museum in the Contributions in Science

series.

• Science Series, long articles on natural history topics, generally written for the layman.
Copies of the publications in these series are available on an exchange basis to institutions and
individual researchers. Copies are also sold through the Museum Bookshop.

A NEW TURTLE (GENUS KINOSTERNON) FROM
NORTHWESTERN MEXICO1

By James F. Berry 2>3 and John M. Legler3

Abstract: The Alamos Mud Turtle, Kinosternon alamosae, new species, is described from southern
Sonora and northern Sinaloa, Mexico. The species is a member of the Kinosternon scorpioides complex
(males lacking clasping organs) and differs from other Kinosternon in its rounded, noncarinate carapace,
widely separated axillary and inguinal scutes, narrow first central scute, and reduced chin barbels. Kinoster-
non alamosae is known to occur in the Pacific coastal region from Guaymas, Sonora, southward at least to
Guasave, Sinaloa, at elevations from sea level to 1,000 m. It is at least partly sympatric with K. integrum.
Kinosternon alamosae seems to inhabit temporary aquatic situations; all existing specimens were taken in
the wet season (July to September). Follicular development and oviposition coincide with the wet season. A
dichotomous key is presented for the identification of the adults of Kinosternon species occurring in Sonora,
Sinaloa, and Chihuahua, Mexico (K. alamosae. K. flavescens, K. hirtipes, K. integrum, and K. sonoriense).

Three species of Kinosternon are currently recognized from
Sonora, Mexico. These are K. flavescens (Agassiz) 1857, K.
sonoriense LeConte 1854, and K. integrum LeConte 1854
(Bogert and Oliver 1945; Langebartel and Smith 1954; Zweifel
and Norris 1955; Conant and Berry 1978; Iverson 1978, 1979).
Kinosternon hirtipes Wagler 1830 is not presently known from
Sonora but may occur there considering its known distribution
in the Rio Papigochic (Rio Yaqui drainage) of extreme western
Chihuahua (Legler and Webb 1970; Van Devender and Lowe
1977; Iverson 1978). Of these four species, only K. integrum
was thought to occupy the Pacific slopes of the Sierra Madre
Occidental and the coastal plain from Guaymas southward
(Conant and Berry 1978; Iverson 1978, 1979).

Heringhi (1969) reported both K. integrum and K. hirtipes
from the vicinity of Alamos, Sonora. We examined his and
other specimens from that region and found that most speci-
mens of K. integrum have been correctly identified but that
most specimens of “K. hirtipes ” (total of 32) represent a pre-
viously undescribed species. In the course of preparing this pa-
per and in preparing much more extensive analyses of
Kinosternon, we have seen no specimens of K. hirtipes from
southeastern Sonora.

METHODS AND MATERIALS

Shells were measured using the method outlined by Carr
(1952) and described in detail by Berry (1978). Plastral scute
lengths are interlaminal (i.e., average length of right and left
scutes, as measured along their common midventral seam).
Condylobasilar length was measured in a straight line from the
posteriormost point of the occipital condyle to the anteriormost
point of the premaxillary region.

The contact of nuchal and first neural bones is easy to ob-
serve on any whole preserved specimen by gently loosening the
tapered posterior edge of the first central scute with a blunt
blade and curling it forward. If the condition of the first neural
is not immediately evident, a gentle scraping of the soft tissue
over the bone will usually reveal it. In most cases, the reflected
scute can be laid back in place without any damage to the
specimen.

Tinkle (1962) demonstrated the utility of expressing the
points at which the five lateral seams of the carapace intersect
the marginal scutes. Our terminology is as follows: R contact
on posterior half of marginal; M, contact at midpoint; A, con-
tact on anterior half; or, e.g., 9-10, on the seam between M9
and M10. This terminology is a simplification of that used by
Tinkle.

Materials examined are in the collections of the American
Museum of Natural History (AMNH), Arizona State Univer-
sity (ASU), John B. Iverson, Earlham College, Richmond, Ind.
(JBI), University of Kansas (KU), Natural History Museum of
Los Angeles County (LACM), Museum of Vertebrate Zoology,
University of California, Berkeley (MVZ), University of Ari-
zona (UAZ), University of Colorado Museum (UC), Univer-
sity of Illinois Museum of Natural History (UIMNH),

1 Review Committee for this contribution:

Carl H. Ernst
John B. Iverson
John W. W right

'Present address: Department of Biology, Elmhurst College, Elmhurst,
Illinois 60126, U.S.A.

'Department of Biology, University of Utah, Salt Lake City, Utah
84112, U.S.A.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 325:1-12
ISSN: 0459-8 II 3

2

Berry and Legler: A New Mexican Kinosternon

Table 1. Comparisons of scute and shell proportions for adults of the three species of Kinosternon of the scorpioides group in Mexico. The specimens
of K. integrum are from Sonora and Chihuahua and those of K. scorpioides are all from Tehuantepec, Oaxaca, and represent the subspecies K
scorpioides cruentatum. Values are means and extremes.

K.

alamosae

K.

integrum

K. scorpioides

Males

Females

Males

Females

Males

Females

(N = 15)

(N = 16)

(N = 24)

(N = 36)

(N = 22)

(N = 46)

Carapace length (CL; mm)

118

113

150

140

123

124

90-135

89-126

110-183

109-161

109-133

105-138

Shell height

0.562

0.604

0.579

0.584

0.633

0.667

-h carapace width

0.496-0.653

0.523-0.644

0.516-0.613

0.521-0.633

0.557-0.727

0.615-0.717

Length of ant.

0.290

0.298

0.316

0.326

0.319

0.317

plastral lobe -s- CL

0.273-0.313

0.282-0.318

0.296-0.351

0.287-0.353

0.274-0.339

0.295-0.362

Length of gular

0.557

0.564

0.489

0.488

0.465

0.495

scute -5- length of
ant. plastral lobe

0.468-0.643

0.482-0.592

0.400-0.556

0.435-0.634

0.395-0.512

0.400-0.615

Length of pectoral

0.109

0.106

0.235

0.201

0.074

0.065

scute -r length of post,
plastral lobe

0.028-0.180

0.057-0.156

0.100-0.364

0.105-0.366

0.000-0.167

0.000-0.167

Length of anal

0.726

0.746

0.741

0.800

0.835

0.897

scute -5- length of post,
plastral lobe

0.623-0.838

0.686-0.815

0.682-0.800

0.731-0.878

0.738-0.905

0.795-0.976

First central scute,

0.880

0.969

0.991

0.987

1.020

1.077

width -s- length

0.726-1 . 1 19

0.751-1.255

0.903- 1161

0.833-1.167

0.857-1.167

0.923-1.360

CW -5- CL

0.633

0.676

0.630

0.658

0.688

0.731

0.562-0.673

0.614-0.766

0.592-0.700

0.619-0.702

0.634-0.721

0.684-0.772

Bridge -s- CL

0.270

0.301

0.243

0.256

0.299

0.321

0.255-0.290

0.276-0.327

0.221-0.273

0.212-0.279

0.271-0.315

0.283-0.346

University of Michigan Museum of Zoology (UMMZ), Uni-
versity of New Mexico (UNM), National Museum of Natural
History (USNM), and University of Utah (UU).

SYSTEMATICS

Kinosternon alamosae new species
Figures 1 through 3

VERNACULAR NAME. Alamos Mud Turtle.

HOLOTYPE. LACM 127639, adult female, preserved
whole; obtained at Rancho Carrizal, 7.2 km north and 1 1.5 km
west of Alamos, Sonora, Mexico [27°05'N, 109°03'W] on 9
July 1966 by Harold L. Heringhi. Formerly ASU 6383 and
bearing a small tag with that number.

ALLOTYPE. LACM 127640, adult male, preserved whole;
same date and locality as holotype. Formerly ASU 6390 and
bearing a small tag with that number.

PARATOPOTYPES (total of 6). UU 142809; UU14281?
skeleton; ASU 6385o", 63869, 6387o", 63899.

OTHER PARATYPES (total of 17 from Sonora). MVZ
509079, 509089, 509099 im, 509100"; AMNH 641639,
64164o', 64165-6899; ASU 67819: Alamos [27°01'N,
108°56'W], UU 142790": 0.5 mi. W Alamos. LACM 105403a",
1054049: La Esmeralda Ranch, 1.2 mi. N Alamos. ASU
6547o": 8 mi. S Alamos. UU 1 1853o" skeleton: La Casa de la

Huerta, Sierras de Alamos [26°59'N, 109°00'W], UAZ
39891a": 4.5 mi. W Minas Nuevas (by road).

DIAGNOSIS. Kinosternon alamosae is a medium-sized spe-
cies (males to 135 mm, carapace length; females to 126 mm).
It is a member of the K. scorpioides complex (with K. scor-
pioides and K. integrum ) in which adult males lack clasping
organs (sensu Legler 1965) on the posterior thigh and leg.
Kinosternon alamosae is most similar to K. integrum but differs
from it and all other Kinosternon by the following unique com-
bination of characteristics in adults: (1) carapace of both sexes
broadly rounded or flat-topped in cross section; noncarinate; (2)
movable plastral lobes extensive in area, closing or nearly clos-
ing anterior and posterior orifices of shell (almost completely
concealing head, limbs, and tail); (3) anal notch small or lack-
ing; (4) bridge relatively long, 26 to 33 percent of carapace
length (Table 1); (5) axillary and inguinal scutes widely sepa-
rated, the inguinal narrowly in contact with sixth marginal
(M6) but never in contact with M5; (6) rear margin of car-
apace straight or evenly curved in profile, never recurved or
flared outward; (7) first central scute usually not in contact
with M2; (8) clasping organs lacking in both sexes; (9) adult
tail terminating in horny spine in both sexes.

GENERAL DESCRIPTION OF SPECIES (based on hypo-
digm). Carapace relatively narrow (Table 1), oval in dorsal as-
pect, evenly rounded or flat-topped in cross section (large adult
females may have slight concavity in middorsal region); no evi-
dence of carapacal keeling. Growth zones evident on all plastral

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 325:1-12

Berry and Legler: A New Mexican Kinosternon

3

and carapaca! scutes in most medium-sized individuals and on
some of largest adults. Scutes imbricate. First central scute
narrow, not in contact with second marginal scute (M2) or just
reaching the Ml -M2 interlamina! seam. Modal formula for
carapacal scute contacts: IP, 5M, 7M, 9A, I0P. Marginal
scutes 1 through 9 of approximately same height, their dorsal
margins forming an even line; M10 abruptly higher than M9
but only slightly higher than Mil. Carapace slightly flared lat-
erally (viewed in cross section) in region of M8-10, but extreme
posterior edge (the pygal region viewed in profile) straight,
often vertical, not flared, not recurved; suprapygal region
abruptly rounded where horizontal dorsal edge of profile meets
posterior vertical edge.

Plastral lobes extensive, nearly closing orifices of shell, al-
most completely concealing soft parts when closed in live ani-
mals. Plastral lobes evenly rounded; anal notch small or absent.
Plastron deeply concave in males, fiat or slightly convex in
females; posterior plastral lobe of females extending nearly to
extreme posterior edge of carapace (shorter in males; Fig. 1).
Plastral scutes ranked from longest to shortest: abdominal,
anal, gular, humeral, femoral, pectoral.

Rostrum short and broad in dorsal aspect; rostral pores well
developed. Premaxillary hook of upper jaw weakly to moder-
ately developed, occasionally lacking. Keratinized head shield
extensive, covering entire area underlain by frontal and pre-
frontal bones, abutting upon superiormost part of maxillary
sheath and upon soft skin of snout; posterior edge of head shield
straight, bulged posteriorly, or with blunt extensions onto
postorbital arches and onto parietal region, creating a trilobate
configuration; posterior edge of head shield never concave or V-
shaped.

One pair of small chin barbels, occasionally pointed (Fig. 2)
but more often low, flat, tubercular or wartlike, almost never
longer than basal width. Two rows of low, weak, indistinct pa-
pillae on each side of neck. Anterior part of tongue papillose.

Hands and feet small and fully webbed; digital claws well
developed. Clasping organs lacking on posterior thigh and op-
posed leg in both sexes. Spadelike falciform scales on ante-
brachium and heel typically kinosternine, variably keratinized.
A horny spine at tip of tail, that of male broader and more
massive than that of female; tail of male elongate and prehen-
sile, longer than one-half the length of posterior plastral lobe;
tail of female much shorter than one-half the length of pos-
terior lobe.

COLORATION OF PRESERVED SPECIMENS. Ground
color of carapace neutral tan, pale brown, or pale brownish
olive with interlaminal seams distinctly darkened. Plastral
ground color yellowish, golden, or yellowish brown, unmarked
except for distinctly darkened interlaminal seams. Interosseous
sutures often visible beneath translucent carapacal scutes.

Skin of soft parts medium to pale slatey gray or brown,
darker dorsally than ventrally, lacking a distinct pattern. Skin
of head and neck grayish to cream below, pale neutral brownish
above, pale and dark fields blending gradually on side of head
and neck without sharp contrast. Dark ground color of head
unmarked (most preserved specimens) or with a pattern of pale
vermiculations. Individual pale marks fused to form vague lon-
gitudinal temporal stripes in some specimens. Pale vermicula-
tions coalesce into what is a pale field marked with remnants
(dots) of the original darker ground color in largest specimens.
Jaw sheaths bearing vertical dark streaks in males, uniformly

straw colored in females. Head pattern never as bright, bold, or
contrasting as in K. integrum.

COLORATION OF LIVE SPECIMENS. The following de-
scriptions are based on observation of two live adults of K. al-
amosae (JBI 858c$ Fig. 2, and JBI 859$) made available
through the courtesy of John B. Iverson. All observations were
made in aquaria with clear water under white fluorescent il-
lumination. The shells of these specimens were free of dirt and
epizoic algae.

Coloration of shell and skin not greatly different from that of
preserved specimens. Carapace in both sexes olive to brown
with dark brown to black interlaminal seams; paler at margin.
Plastron yellowish with dark brown seams; growth rings con-
taining melanin and distinctly visible.

Skin of head pale gray above with numerous dark spots; sides
of head mottled pale gray and yellowish cream with fewer dark
spots. An indistinct pale stripe extending from posteroventral
edge of orbit, above maxillary sheath to jaw articulation.
Horny jaw sheaths yellowish cream to pale gray, male with nu-
merous faint brown vertical streaks. Dorsal surface of neck pale
gray grading into immaculate yellowish cream of ventral sur-
face of head and neck. Skin of limbs overall pale gray, slightly
darker above.

Iris of female dark orange with four evenly spaced con-
centrations of darker pigment immediately peripheral to the
pupil, forming a stellate figure. Iris of male darker orange than
that of female but bearing the same stellate arrangement of
dark pigment.

OSTEOLOGY. The following description is based on two
complete adult skeletons of Kinosternon alamosae (UU 1 1853cf
and UU 14281$). Comparisons are with K. integrum and are
based on 13 complete skeletons (UU 7680, 7696, 7827,
7850-51, 11678dcf; UU 7737, 7750, 7833, 7872, 7875,

1 1 855$$; AMNH 64161$$). When the skeletal characters of
K. integrum are judged to differ significantly from those of K.
alamosae, those of K. integrum appear in brackets. See Table 2
for a comparison of skull proportions in K. alamosae and K.
integrum.

The appendicular skeleton of K. alamosae is typically
kinosternine and does not differ significantly from that of K.
integrum. The phalangeal formula is 2- 3- 3- 3 -3 on hands and
feet.

Skull (Fig. 3) in general relatively broad, low, robust, and
compactly built. Dorsal and ventral aspects: skull tapers
abruptly from posterior edge of orbit to snout [a gradual taper-
ing from anterior edge of tympanum to snout]; temporal bars
nearly parallel or bowed outward [temporal bars angled for-
ward towards snout, not bowed outward]; snout short, broad,
and prominent [relatively longer and narrower]. Lateral aspect:
dorsal edge of profile at least slightly convex from rostrum to
parietal-supraoccipital suture [slightly concave]; supraoccipital
spine relatively short and low, its dorsal edge concave posterior
to parietal-supraoccipital suture [spine longer and higher, dor-
sal edge highest just posterior to suture then angling postero-
ventrad as a straight or slightly convex blade of bone]; orbit
relatively large; posterior edge of postorbital bar lies anterior to
posterior edge of foramen interorbitale (Gaffney 1972; the de-
scending process of the parietal and the ascending process of
the palatine), permitting one to see through a properly oriented
skull [postorbital bar posterior to foramen interorbitale— no
such view possible]; jugal contributing extensively to dorsal free

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 325:1-12

Figure 1. Dorsal, ventral, and lateral head views of Kinosternon alamosae. new species. Left column — UU 14279a' para-
type, 0.8 km W of Alamos, Sonora, Mexico; CL, 126 mm. Right column — LACM 1276399 holotype; Rancho Carrizal, 7.2
km N and 1 1.5 km W of Alamos, Sonora, Mexico; CL 122 mm.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 325:1-12

Berry and Legler: A New Mexican Kinosternon

5

edge of temporal arch where it is joined by postorbital arch
[excluded from free edge or making a small contribution]; pos-
terior tips of squamosals do not project posterior to occipital
condyle [project far posterior to condyle].

Surangular bone high and with straight-sided dorsal edge,
concealing prearticular in lateral view [emarginate above, pre-
articular exposed in lateral view], Dentary crushing surface
well developed, equal to or wider than ventral part of mandibu-
lar ramus or mandibular symphysis [not as well developed, nar-
rower than ventral part of ramus or symphysis]. Mandibular
hook well developed but blunt [sharper]. Maxillary crushing
surfaces broad and well developed. Premaxillary beak moder-
ately to strongly developed.

Caudal vertebrae 18 (male) to 20 (female), terminal ver-
tebrae (2 in female, 4 in male) fused to form skeleton of termi-
nal spine. Second cervical vertebra opisthocoelous, third
biconvex, all others procoelous; sixth and seventh doubled pos-
teriorly, seventh and eighth doubled anteriorly.

Male with six hexagonal neural bones, posteriormost round
and lying between sixth costals; female having five hexagonal
neurals, the hindmost narrowly contacting sixth costals. First
neural separated from nuchal by suture between first costals,
the suture greater than one-half the length of first neural. All
hexagonal neurals long-sided and tapered anteriorly. Pygal
bone nearly square, slightly notched posteriorly. Suprapygal
five-sided, partially fused to eighth costals in female. Eight
pairs of costals; pairs 1, 6, 7, and 8 in middorsal contact, pairs 2
through 5 separated by neurals. First peripherals in narrow
contact with first costals, narrowly separating nuchal from sec-
ond peripherals.

DISCUSSION

RELATIONSHIPS. The absence of clasping organs on the
hind limbs of males of Kinosternon alamosae seems clearly to
ally it with members of the K. scorpioides complex. In Mexico,
the other members of the complex are K. integrum and K. scor-
pioides (Figs. 4 and 5). The relationships among all members
of the K. scorpioides complex have been considered in a sepa-
rate report (Berry 1978). Kinosternon alamosae bears a super-
ficial resemblance (general coloration and shell shape) to K.
flavescens, but we ascribe this to convergent adaptation to tem-
porary aquatic habitats. Tables 1 and 3 summarize the dif-
ferences and similarities of the three species of the K.
scorpioides complex in Mexico (see also the taxonomic key).
We regard the significant morphological differences between K.
integrum and K. alamosae and the sympatric occurrence of
these populations to be sufficient evidence for their distinction
as species.

Berry (1978) used multivariate techniques in the analysis of
relationships within the K. scorpioides species complex and
demonstrated that K. alamosae is phenetically more similar to
K. integrum than to any other member of the complex but that
the two species are nonetheless phenetically distinct. Berry hy-
pothesized that K. integrum had reached the Mexican Plateau
via the Rio Balsas drainage and had then descended the Rio
Grande de Santiago drainage and spread northward along the
northern Pacific coast of Mexico. It is not clear which species
first colonized the region of Alamos or how K. alamosae may
have gotten there. The two best possibilities are the route sug-
gested by Berry for K. integrum, or downstream dispersal from

Figure 2. Kinosternon alamosae, new species, live adult male; JBI 858, 14.8 mi. W, 1 .8 mi. S Alamos, Sonora, Mexico. Width of head
21 mm.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 325:1-12

Figure 3. Lateral, dorsal, and ventral views of Kinosternon skulls: Left column — K. alamosae, new species, UU 1428l9paratopotype;
condylobasilar length 26.2 mm. Right column — K. integrum UU 77509, Laguna Rio Viejo, 2 km N of Eldorado, Sinaloa, Mexico; con-
dylobasilar length 35.3 mm.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 325:1-12

Berry and Legler: A New Mexican Kinosternon

7

Table 2. Skull proportions in Kinosternon alamosae (two specimens)
and K. integrum (ten specimens from Sinaloa and Guerrero). Selected
measurements are expressed (means and extremes) as proportions of
condylobasilar length (see Figure 3).

K. integrum

K. alamosae

Males
(N = 5)

Females
(N = 5)

Males
(N = 1)

Females
(N = 1)

Greatest width

0.80

0.77-0.83

0.80

0.77-0.83

0.78

0.82

Width, maxillary
crushing surface

0.15

0.15-0.16

0.15

0.14-0.17

0.17

0.16

Vertical diameter
of orbit

0.19

0.19-0.20

0.20

0.19-0.23

0.2!

0.23

Width of rostrum

0.25

0.23-0.25

0.24

0.21-0.27

0.27

0.28

Greatest height

0.53

0.50-0.56

0.51

0.47-0.54

0.49

0.50

the headwaters of the Rio Yaqui, Rio Mayo, or Rio Fuerte as
suggested for fishes (Meek 1904) and for other freshwater tur-
tles (Legler and Webb 1970). Since no member of the K. scor-
pioides complex presently occurs in the Rio Yaqui, Rio Mayo,
or Rio Fuerte on the Mexican Plateau, the former explanation
seems more plausible.

We regard K. alamosae as a population that has been iso-
lated in the Alamos region for a long period. Wider (un-
detected) geographic distribution of K. alamosae is unlikely
since Berry has examined most existing specimens of Kinoster-
non from within the range of K. integrum.

TYPE LOCALITY. Data associated with the holotype and
paratopotypes are “Mexico, Sonora, Rancho Carrizal, 7 mi.
[11.3 km] W Alamos, 9 July 1966, Heringhi.” Heringhi ( 1969)
states the following concerning Rancho Carrizal: “27°05'N,
109°13'W, 360 m. A ranch 6 mi. [9.7 km] W Alamos on the
Alamos-Navojoa Road; Shorttree Forest.” The longitude and
latitude stated by Heringhi would place the locality more
nearly 29 to 32 km NNW of Alamos. We conclude therefore
that Heringhi’s coordinates were in error and that he used
odometer distances. The NIS Gazetteer for Mexico (1956) lists
a village (“popl.”) named Carrizal at 27°05'N, 109°03'W, and
we regard this as the type locality as it lies approximately 1 1
km NNW of Alamos by road. The locality is just north of the
point where an intermittent tributary of the Rio Mayo crosses
the main Alamos-Navojoa road (“Agua Marin” fide Heringhi
1969). Our statement of the type locality in metric distance
from Alamos (7.2 km N and 1 1.5 km W) is based on all of the
above information.

GEOGRAPHIC DISTRIBUTION. Kinosternon alamosae
is now known from seasonal aquatic habitats on the Pacific
Coastal Lowlands of Mexico from the vicinity of Guaymas,
Sonora, southward at least to Guasave, Sinaloa (Fig. 6; pos-
sibly as far south as Culiacan — see “Other Specimens Exam-
ined”) at elevations of sea level to slightly higher than 1000 m
(in the Sierras de Alamos). The northern limit of the range of
K. alamosae corresponds to the limit of thorn scrub forest and

to the northern limit of many tropical vertebrates (Stuart
1964).

ECOLOGY. Kinosternon alamosae occurs partly within the
northern part of the known range of K. integrum. The two spe-
cies are known from the same localities in some cases (e.g.,
“Alamos,” and 8 mi. S of Alamos, Heringhi 1969). It is not
known whether the two species ever occur microsympatrically.
Legler has worked near Alamos on three occasions. On 21 -24
January 1959 and 1 5-19 June 1961, baited traps set in the Rio
Cuchujaqui (10 km SE of Alamos, 26°57'N, 108°53'W)
caught many specimens of K. integrum but no K. alamosae. On
19-26 May 1978, conditions were extremely dry near Alamos.
The Rio Cuchujaqui was barely flowing, and the water was
clear enough (visibility 1 to 2 m) for diving. No turtles of any
kind were seen at the aforementioned locality nor at a locality
farther downstream (26°58'N, 108°51'W). During the 1978
visit, we did locate large concentrations (up to two individuals
per square meter of surface area) of K. integrum near Alamos
in a small (10 by 10 m) spring-fed impoundment in thorn for-
est, and in a large-mouthed deep well on a ranch. No K. al-
amosae were present in the approximately 100 specimens we
examined from these two localities (we were aware of K. al-
amosae at that time and were seeking it). Exploration of the
deep mud bottoms (with hands, feet, and seine) of two drying
cattle reservoirs produced no turtles.

Heringhi (1969) stated that the two species of Kinosternon
he collected near Alamos (his "K. hirtipes " = K. alamosae)
“do not separate geographically" but was otherwise vague on
the ecological differences (if any) of the two species. In one
instance, he obtained Kinosternon from the bottom mud of a
drying odorous cattle reservoir containing decaying vegetation.
Wiewandt et al. (1972) show a photograph of a temporary, wet
season pool in thorn forest near Alamos in which "Kinosternon
sonoriense" were seen mating in July 1969 (there is no other
evidence that K. sonoriense occurs in that region — the turtles
could have been K. alamosae).

July, August, and September are the wettest months of the
year in the coastal region from Guaymas to Culiacan. All of
the specimens of K alamosae known to us were obtained dur-
ing these months. From these data and from our field observa-
tions, it seems likely that K alamosae is active only or mainly
in the wet season, occurs chiefly in temporary aquatic habitats,
and probably occurs in microsympatry with K. integrum only
where the latter enters temporary aquatic situations in the wet
season.

REPRODUCTION. A female paratype (AMNH 64168)
collected “27 Aug. -2 Sept. 1942” contained five oviducal eggs;
dimensions of two of these were 25.7 by 16.5 and 27.6 by 16.4
mm. The ovaries bore fresh (but no old) corpora lutea plus five
follicles 6 to 1 I mm in diameter (mean, 9.4 mm) that were
regarded as potentially ovulatory in the same season. The ova-
ries of another female (UU 14281, paratopotype) collected 9
July 1966 bore five follicles 12 to 13 mm in diameter (mean,
12.2), five follicles 8 mm in diameter, and an additional three
follicles of 5 mm diameter. No corpora lutea were visible.

The eggs of K. alamosae would rank among the smaller
Kinosternon eggs listed by Moll and Legler (1971). We hypoth-
esize that follicular development begins when the turtles be-
come active in the early wet season, that at least some females
lay more than one clutch per season, and that laying is proba-
bly completed by the end of the wet season (Oct to Nov). This

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 325:1-12

8

Berry and Legler: A New Mexican Kinosternon

Figure 4. Dorsal, ventral, and lateral head views of Kinosternon integrum from Rio San Lorenzo, 1 7 mi. km ENE of Eldorado,
Sinaloa, Mexico. Left column— UU 77780"; CL 176 mm. Right column— UU 77839, CL 163 mm.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 325:1-12

Berry and Legler: A New Mexican Kinosternon

9

Figure 5. Dorsal, ventral, and lateral head views of Kinosternon scorpioides eruentatum from vicinity of Tonala, Chiapas,
Mexico. Left column — UU 763 1 cf; CL 1 16 mm. Right column — UU 76339; CL 103 mm.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 325:1 — 12

10

Berry and Legler: A New Mexican Kinosternon

Table 3. Comparison of selected characters in Kinosternon alamosae, K. integrum, and K. scorpioides cruentatum. Values in parentheses are percent-
ages of total specimens examined with the given character. See also Table 1 for proportions, and “Other Specimens Examined” for further information
on samples.

K. alamosae

K. integrum

K. scorpioides

Intermediate adult size; carapace
length to 1 35 mm inc/c/ and 1 26
mm in 99 (Table 1 ).

Large adult size, 183 mm c/d', 164
mm 99.

Intermediate adult size; 151
mm c/d' and 1 33 mm 99.

Carapace noncarinate, broad, and
evenly rounded or flat-topped in
cross section.

Carapace tricarinate but keels may
be obscured in older individuals.
Not broad and evenly rounded in
cross section.

Carapace strongly tricarinate, not
evenly rounded in cross section.

Axillary and inguinal scutes never in
contact.

Axillary and inguinal scutes usually
in contact on one (16%) or both
(50%) sides.

Axillary and inguinal scutes rarely
(2.5%) in contact.

First central scute excluded from
(80%) or barely in contact with
second marginal (20%).

Cl in substantial contact with M2
on one (11%) or both (66%)
sides.

Cl in substantial contact with M2
on one (25%) or both (45%)
sides.

One pair of small chin barbels;
barbels absent in 13% of sample.

Two (35%) or three (60%) pairs of
chin barbels in most specimens;
barbels never lacking.

Two (30%) or three (70%) pairs of
chin barbels; barbels never
lacking.

Posterior plastral hinge curved
posteriorly (Fig. 1).

Posterior plastral hinge curved
posteriorly (Fig. 4).

Posterior plastral hinge nearly
straight (Fig. 5).

Anal scute relatively short when
expressed as a percentage of
posterior lobe length (Table 1 ).

Anal scute relatively long.

Anal scute relatively long.

Interpectoral seam relatively short;
humerals never completely
separating pectorals.

Interpectoral seam relatively long;
pectorals always in contact.

Interpectoral seam relatively short;
humerals completely separating
pectorals in 21% of sample.

Head dull, neither contrastingly
marked nor including red and
orange.

Head bright and contrastingly
marked but pale colors do not
include red and orange.

Head brightly marked, pale colors
are usually yellow, orange or red.

reproductive cycle does not differ substantially from that we
have observed in a few K. integrum (UU 7808, Guasave, Sin-
aloa, 20 July 1965 — gravid; ASU 6077 and UU 7677, Alamos,
Sonora, 16 June 1966 and 17 June 1961, respectively — devel-
oping follicles in 5 to 10 mm class).

REMARKS. Several collections of Kinosternon from Sonora
identified as “K. integrum" include specimens of K. alamosae.
Of the specimens listed by Bogert and Oliver (1945) and
Zweifel and Norris (1955), AMNH 64163-68 and MVZ
50907-10 (Alamos) are here identified as K. alamosae, while
AMNH 64161-62 (Alamos) and AMNH 63755-58 and MVZ
50889-902 (Guirocoba) are K. integrum. A shell lacking soft
parts from 23.7 km S of Empalme, Sonora (UIMNH 24456),
identified as K. integrum by Langebartel and Smith (1954) is
tentatively identified as K. alamosae based on scute
proportions.

OTHER SPECIMENS EXAMINED

Kinosternon alamosae (10). SONORA: JB1 859? live, 3.5
mi. W Alamos; JBI 858a* live, 14.8 mi. W, 1.8 mi. S Alamos
[JB1 858-9 will ultimately be deposited in the Florida State
Museum, Gainesville]; UAZ 31741 cT, ca. 23 mi. E (by road)
Navojoa; UAZ 39889cf, 7.6 mi. NE (by road to Tezapaco) Es-

peranza (ca. 27°40'N, 109°55'W); UC 35122c/ im, San Carlos
Bay [27 ° 56'N, 1 1 1 °04'W]; UIMNH 24456c/, 14.7 mi. S Em-
palme; UC 16145c/, 28 mi. S Navojoa. SINALOA: UAZ
27956, 7.4 mi. S Guasave (by road) [25°34'N, 108°27'W];
LACM 105396-97, “Culiacan” (locality incorrect fide R.L.
Bezy, personal communication).

Kinosternon integrum (292). SONORA: AMNH 64161-62,
ASU 6075, UU 7677, 1 1855, Alamos [27‘01'N, 108°56'W];
ASU 6545, 0.8 mi. S Alamos; ASU 6077, 6107, 4 mi. SE Al-
amos; KU 47589-90, 47592-94, 9 mi. SE Alamos, Rio Alamos;
JBI 862, 1 km E Alamos; LACM 105406, 8 mi. SSE Alamos,
Rio Cuchujaqui; UAZ 39892, ca. 2 mi. NE Alamos Church,
Alamos; UU 1 1678, 1 1852, 1 1854, La Casa de la Huerta, Sier-
ras de Alamos; UAZ 36480, N slope Sierras de Alamos; MVZ
28937, 50889-902, AMNH 63755-57, Guirocoba [26°53'N,
108°41'W]; LACM 75350, Barranca del Cobre, 2 km E
Guirocoba; LACM 105402, UAZ 38189-90, La Aduana; UAZ
38705-6, mine, 14 mi. NW La Aduana [27°03'N, 1Q9°01'W];
UNM 5787, UAZ 28015, 28019, 28022-23, ]ti mi. N La
Aduana; UNM 14492, abandoned mine nr. La Aduana;
UMMZ 79514, nr. Pilares Mine; UAZ 38864, 20 mi. by rd. to
Yecora, NE Nuri [28°02'N, 109°22'W]; UAZ 39890, 12 mi.
(by rd.) SW Santa Ana de Yecora; UAZ 28008-9, 10.5 mi. W
(by rd.) Rosario [27°59'N, 109'20'W], SINALOA: KU
63637-46, 3 mi. NE San Miguel [25°56'N, 109°03'W], Rio

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 325:1-12

Berry and Legler: A New Mexican Kinosternon

11

del Fuerte; KU 80772-73, 13 km NNE Vaca; AMNH 82142,
Guasave, Rio Sinaloa; UU 7786-7818, 12 mi. NE Guasave,
trib. Rio Sinaloa; UMMZ 122242, trib. Rio Bacaburito, Rio
Sinaloa ca. 13 mi. NE Palmar; LACM 75351, La Huerta, Rio
Mocorito; UMMZ 122235-39, 0.9 mi. N San Benito, trib. Rio
Mocorito; UAZ 36511, Rancho los Pocitos, 14.2 mi. WNW
Pericos Jet.; AMNH 82143, LACM 105394-95, Culiacan
[24°48'N, 107°24'W]; LACM 105393, 5.5 mi. N Culiacan;
UMMZ 121922, 7.6 mi. N Culiacan; UU 3799-3800, 11 mi.
NW Culiacan; KU 45398, Eldorado [24"17'N, 107°21'W[;
UU 7678-7774, 13007-8, 2 km N Eldorado, Laguna Rio Viejo;
UU 7775-85, 17 mi. ENE Eldorado, Rio San Lorenzo; KU
63633-36, 1 mi. SE Camino. CHIHUAHUA: UAZ 28238,
Milpillas [27 ° 1 3'N, 108’38'W]; UAZ 28141-74, vie. Milpillas;
UAZ 39893-94, 1.5 mi. (by rd. to San Antonio) SW Milpillas.

Kinosternon scorpioides cruentatum (78). OAXACA:
UMMZ 1 18633-34, USNM 109106-23, Tehuantepec

[16°20'N, 95‘14'W]; UMMZ 82226-35, vie. Tehuan-
tepec; UMMZ 82183-224, Tehuantepec, Tehuantepec R.;
UMMZ 82225, 12 km NW Tehuantepec; UMMZ 82238, Nisa
Pipi, 8 km NW Tehuantepec; UMMZ 82236-37, “5 leagues”
[ca. 20 km] S Tehuantepec, Rancheria Lamanga; USNM
113278, San Mateo del Mar [16‘12'N, 95°00'W]; UU 7950,
12 mi. NE Juchitan [16°26'N, 95°01'W].

KFY TO THF SPFCIFS OF
KINOSTERNON OF SONORA, SINALOA, AND
CHIHUAHUA, MEXICO

The following key is applicable only to adults or individuals
longer than 90 mm.

1. Marginal scute 9 distinctly higher than M8, about same

height as M10, forming a distinct peak where it meets seam
between laterals 3 and 4; head shield consisting of crescent over
each orbit, joined or not joined to shield on rostrum

K. flavescens

Marginal scute 9 approximately same height as M8, much
lower than M10, not distinctly peaked where it meets seam be-
tween laterals 3 and 4; head shield V-shaped, rhomboidal, or
with trilobate posterior margin; all parts of head shield inter-
connected 2

2. Adult males with discrete clasping organs on posterior thigh
and leg (opposed patches of horny scales); bridge relatively
short in both sexes (less than 25 percent of carapace length in
93 percent of sample); known distribution at higher elevations
east of continental divide or north of Guaymas on coastal plain.

3

Adult males lacking discrete clasping organs on posterior leg
and thigh; bridge relatively long in both sexes (greater than 25
percent of carapace length in 96 percent of sample); known dis-
tribution at lower elevations west of continental divide from

Guaymas southward 4

3. Head shield deeply notched posteriorly (V-shaped); usually

three pairs of relatively short chin barbels (longest barbel less
than one-half the vertical diameter of orbit); male plastron rel-
atively narrow (midfemoral width less than 60 percent of car-
apace width at same level, anterior hinge width less than 65
percent of carapace width at same level); first neural not in

contact with nuchal (96 percent of sample, see “Methods and

Materials”) K. hirtipes

Head shield triangular or rhomboidal, straight-sided or with
a single lobe posteriorly; usually 3 to 4 pairs of relatively long
chin barbels (length of at least one pair greater than one-half
the vertical diameter of orbit); male plastron relatively exten-
sive (midfemoral width greater than 60 percent of carapace
width at same level, anterior hinge width greater than 65 per-
cent of carapace width at same level); first neural usually in
contact with nuchal (73 percent of sample, see “Methods and
Materials”) K. sonoriense

4. Inguinal and axillary scutes in contact on at least one side
(66 percent of sample); inguinals extending anteriorly beyond
seam between marginals 5 and 6; first central scute wide, in
contact with marginal 2 on one or both sides (80 percent of
sample); at least two pairs of well-developed chin barbels ....

K. integrum

Inguinal and axillary scutes never in contact; inguinals not
extending anteriorly beyond seam between marginals 5 and 6;
first central scute narrow, never in contact with marginal 2 (in
contact with seam between Ml and M2 in 20 percent of sam-
ple); barbels lacking or wartlike and inconspicuous

K. alamosae

ACKNOWLEDGMENTS

We thank the following persons for permission to examine spec-
imens in their care: Robert L. Bezy and John W. Wright,
Natural History Museum of Los Angeles County; Martin J.
Fouquette, Arizona State University; T. Paul Maslin and Shi-
Kuei Wu, University of Colorado Museum; Charles H. Lowe,
University of Arizona; William G. Degenhardt and James S.
Jacob, University of New Mexico; William E. Buellman, Uni-
versity of Kansas; Donald F. Hoffmeister, University of Illinois

Figure 6. Geographic ranges of Kinosternon alamosae (dots) and K.
integrum (triangles) in northwestern Mexico. Star is type locality of K.
alamosae; half circle is locality at which sympatry occurs.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 325:1-12

12

Berry and Legler: A New Mexican Kinosternon

Museum of Natural History; Arnold J. Kluge, University of
Michigan Museum of Zoology; and George R. Zug, National
Museum of Natural History. We are especially grateful to John
B. Iverson, who loaned us live specimens of Kinosternon al-
amosae, and to John K. Cross, who facilitated a large loan and
examined a number of University of Arizona specimens for us.
John B. Iverson, Carl H. Ernst, John W. Wright, and Roger
Conant made useful comments on the manuscript. We also
thank Sr. Ignacio Ibarrola Bejar and the Direction General de
la Fauna Silvestre, Mexico DF, for issuing Scientific Collecting
Permit No. 68-78/160 for our work in May 1978.

The large collections of specimens and field data at the Uni-
versity of Utah were made with the substantial aid of the fol-
lowing grants to Fegler for studies of Mexican and Central
American turtles: National Science Foundation GB 4859, GB
2608, G 17659; American Philosophical Society, Grant 341,
Johnson Fund.

RESUMEN

Una especie nueva de casquito, Kinosternon alamosae, es de-
scribida del sur de Sonora y del norte de Sinaloa, Mexico. Fa
especie es miembra del complejo K. seorpioides (machos fal-
tando organos abrazadores) y es en parte simpatrica con otros
Kinosternon la cual se diferencea por su carapacho redondo, ax-
ila y escudos inguinal anchamente separados, primera escuda
central angosta, y barbilla reducida. Kinosternon alamosae se
conoce en la region pacifica costal de Guaymas, Sonora hacia el
sur por lo menos hasta Guasave, Sinaloa, de elevaciones de
nivel de mar hasta 1000 metros. Fa especie parece occurrir en
habitates acuaticas temporales; todos los ejemplares existentes
fueron obtenidos durante la temporada de lluvia en los meses de
julio, agosto, y septiembre. Desarrollamiento folicular y puesta
de huevos coincide con la temporada de lluvia. Una Have di-
cotomica se presenta para la identification de adultos de
Kinosternon en Sonora, Sinaloa, y Chihuahua, Mexico (K. al-
amosae, K. flavescens, K. hirtipes, K. integrum, y K. sonoriense).

LITERATURE CITED

Agassiz, F. 1857. Contributions to the natural history of the
United States of America, vol. 1 and 2. Boston, Mass: Fit-
tie, Brown and Co.

Berry, J.F. 1978. Variation and systematics in the Kinosternon
seorpioides and K. leucostomum complexes (Reptilia: Test-
udines: Kinosternidae) of Mexico and Central America.
Ph.D. dissertation, Univ. of Utah. [See Dissertation Ab-
stracts 39(7), No. 7824683].

Bogert, C.M., and J.A. Oliver. 1945. A preliminary analysis
of the herpetofauna of Sonora. Bull. Amer. Mus. Nat.
Hist. 8 3 (6) : 30 1 — 425 .

Carr, A. 1952. Handbook of turtles. New York: Comstock
Publ. Assoc., Cornell Univ.

Conant, R., and J.F. Berry. 1978. Turtles of the family
Kinosternidae in the southwestern United States and adja-
cent Mexico; identification and distribution. Amer. Mus.
Novitates 2642.

Gaffney, E.S. 1972. An illustrated glossary of turtle skull no-
menclature. Amer. Mus. Novitates 2486.

Heringhi, H.F. 1969. An ecological survey of the herpeto-
fauna of Alamos, Sonora, Mexico. Master’s thesis, Ari-
zona State Univ.

Iverson, J.B. 1978. Distributional problems of the genus
Kinosternon in the American southwest and adjacent Mex-
ico. Copeia 1 978( 3):476— 90.

. 1979. A taxonomic reappraisal of the yellow mud tur-
tle, Kinosternon flavescens (Testudines: Kinosternidae).
Copeia 1979(2):2 12-25.

Fangebartel, D.A., and H.M. Smith. 1954. Summary of the
Norris collection of reptiles and amphibians from Sonora,
Mexico. Herpetologica 10(2): 125-1 36.

FeConte, J. 1854. Description of four new species of Kinoster-
num. Proc. Acad. Nat. Sci., Philadelphia. 7:180-90.

Fegler, J.M. 1965. A new species of Kinosternon from Central
America. Univ. Kans. Publ. Mus. Nat. Hist. 1 5(1 3):61 5—
25.

Fegler, J.M., and R.G. Webb. 1970. A new slider turtle
(Pseudemys scripta) from Sonora, Mexico. Herpetologica
26(2): 1 57—68.

Meek, S.E. 1 904. The fresh-water fishes of Mexico north of the
Isthmus of Tehuantepec. Field Columbian Mus. Publ. 93,
Zool. Ser. 5.

Moll, E.O., and J.M. Fegler. 1971 . The life history of a Neo-
tropical slider turtle, Pseudemys scripta (Schoepff), in
Panama. Bull. Los Angeles Co. Mus. (Nat. Hist.) 11.

N1S Gazetteer, Mexico. 1956. Official standard names ap-
proved by the U.S. Board on Geographic Names. Office of
Geography, Dept, of Interior, Washington, D.C.

Stuart, F.C. 1964. Fauna of Middle America. In Handbook
of Middle American Indians, vol. 1 : Natural environment
and early cultures, ed. R. Wauchope and R.C. West, pp.
316-62. Austin: Univ. of Texas Press.

Ti nkle, D.W. 1962. Variation in shell morphology of North
American turtles, 1: The carapacial seam arrangements.
Tulane Stud. Zool. Bot. 9:331-49.

Van Devender, T.R., and C.H. Fowe, Jr. 1977. Amphibians
and reptiles of Yepomera, Chihuahua, Mexico. J. Her-
petology 11(1 ):41— 50.

Wagler, J.G. 1830. Natiirliches System der Amphibien, mit
vorangehender Classification der Saugethiere und Vogel.
Ein Beitrag zur vergleichender Zoologie. Munich.

Wiewandt, T.A., C.H. Fowe, and M.W. Farson. 1972. Oc-
currence of Hypopachus variolosus (Cope) in the short
tree forest of southern Sonora, Mexico. Herpetologica
28(2): 1 62—64.

Zweifel, R.G., and K.S. Norris. 1955. Contribution to the
herpetology of Sonora: Descriptions of new subspecies of
snakes (Micrurus euryxanthus and Lampropeltis getulus)
and miscellaneous collecting notes. Amer. Midi. Natur.
54(1 ):203— 49.

Submitted 10 August 1978; accepted for publication 18 May
1979.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 325:1-12

NH

iis'w.

ft

e';'

S»’:-

tftli -

Ip;

,

i

iffiC

lit!

m

PP

i i

m

i-

SPECIES LIMITS IN THE PEROMYSCUS MEXICANUS GROUP
(MAMMALIA: RODENTIA: MUROIDEA)

By David George Huckaby

laHKE

I:

giiil

mm

If!

2S

iliSfl

Published by the NATURAL HISTORY MUSEUM OF LOS ANGELES COUNTY • 300 EXPOSITION BOULEVARD * LOS ANGELES, CALIFORNIA 90007

iillMli::

SERIAL PUBLICATIONS

OF THE NATURAL HISTORY MUSEUM OF LOS ANGELES COUNTY

The scientific! publications of the Natural History Museum of Los Angeles County have been issued
at irregular intervals' in three major series; the articles in each series are numbered individually, and
numbers run consecutively, regardless of subject matter;

« Contributions in Science, a miscellaneous series of technical papers describing original
research in the life and earth sciences.

. Science Bulletins, a miscellaneous series of monographs describing original research in the
life and earth sciences. This series was discontinued in 1978 with the issue ol Numbers 29
and 30; monographs are now published by the Museum in the Contributions in Science

series.

. Science Series, long articles on natural history topics, generally written for the layman.
Copies of the publications in these series are available on an exchange basis to institutions and
individual researchers. Copies are also sold through the Museum Bookshop.

SPECIES LIMITS IN THE PEROMYSCUS MEXICANUS GROUP
(MAMMALIA: RODENTIA: MUROIDEA)1

By David George Huckaby2

Abstract. To estimate species limits within the Peromyscus mexicanus group of Hooper ( 1968) as modi-
fied by Musser (1969, 1971), 1 compared over 4000 specimens by conventional and multivariate techniques.
I based the comparison on a suite of characters chosen from the skull, male reproductive system, and exter-
nal morphology of preserved specimens.

Hooper (1968) recognized 14 nominal forms (P. allophylus, furvus, grandis, guatemalensis, gymnotis,
latirostris, megalops, melanocarpus, mexicanus, nudipes, ochraventer, stirtoni, yucatanicus, and zarhy fi-
chus). 1 follow Hall (1971) in synonymizing P. latirostris with furvus and Musser (1971) in synonymizing
P. allophylus with gymnotis. I recognize all the rest as full species except P. nudipes, which 1 synonymize
with P. mexicanus. I also recognize P. melanurus (formerly viewed as a subspecies of P. megalops) as a full
species.

Osgood (1909) grouped the species of the genus Peromyscus
into six subgenera and further subdivided the nominate sub-
genus into eight species groups. Three of these groups (the P.
lepturus group with lepturus, lophurus, simulatus, nudipes, fur-
vus, guatemalensis, and altilaneus; the P mexicanus group with
mexicanus, allophylus, banderanus, and yucatanicus; and the P.
megalops group with megalops, melanocarpus, and zarhyn-
chus ) contained species herein considered.

Hooper (1958), using characters derived from the glans pe-
nis of some of these species, suggested that many aspects of
Osgood’s arrangement did not fit with these new data. Hooper
and Musser (1964), again using data from the glans penis but
from many more species of the genus, submerged the P. mega-
lops group of Osgood into the P. mexicanus group. They then
raised the lepturus group to subgeneric rank with the new name
Habromys but removed P. nudipes, furvus, guatemalensis, and
altilaneus from the subgenus and added them to their P. mex-
icanus group. They removed P. banderanus from the P. mex-
icanus group and made it the type of their new subgenus
Osgoodomys. Finally, they added to the P. mexicanus group
seven species described since Osgood’s study (1909) (P. stirtoni
Dickey 1928, P. grandis Goodwin 1932, P. hondurensis Good-
win 1941, P. latirostris Dalquest 1950, P. ochraventer Baker
1951, P. sloeops Goodwin 1955, and P. angustirostris Hall and
Alvarez 1961). Their revision resulted in the P. mexicanus
group totaling 17 species.

Musser (1964) synonymized P. angustirostris with furvus.
Hooper (1968) transferred P. hondurensis to his boy lii group
and suggested synonymizing P. sloeops with mexicanus. Mus-
ser (1969) synonymized P. hondurensis with oaxacensis of the
boy Hi group and P. sloeops with mexicanus. Hall (1971) syn-
onymized P. latirostris with furvus. Finally, Musser (1971) res-
urrected P. gymnotis as a full species from its previous subspe-
cific status under P. mexicanus and synonymized P. allophylus
with it.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 326:1-24
ISSN: 0459-81 13

I started with the 12 resulting nominal species now recog-
nized as members of the P. mexicanus group (P. furvus,
grandis, guatemalensis, gymnotis, megalops, melanocarpus,
mexicanus, nudipes, ochraventer, stirtoni. yucatanicus, and
zarhynchus). 1 attempted to arrive at a better delimitation of
the species by an analysis of most of the specimens of these
species in museums in the United States.

MATERIALS

The data for the study came primarily from standardly pre-
pared skins and skulls. 1 examined museum specimens repre-
senting every named form suspected of belonging to the P.
mexicanus group, as well as representatives of most other well-
defined species of Peromyscus. I also examined the holotypes of
all named forms in the group except those of Hesperomys mex-
icanus Saussure, Peromyscus gymnotis Thomas, and Per-
omyscus cacabatus Bangs.

Lists of the skins and skulls examined, in alphabetical order
by species, country, and province, state or department follow
each species account. I abbreviated the institutions as follows:
The University of Michigan, Museum of Zoology (UM); The
Museum, Michigan State University (MSU); American Mu-
seum of Natural History (AM); National Museum of Natural
History (NM); Field Museum of Natural History (FM); Uni-
versity of Kansas, Museum of Natural History (KU); Loui-
siana State University, Museum of Zoology (LSU); Texas

' Review committee for this contribution:

Michael D. Carleton
Guy G. Musser
Donald R. Patten

2 Department of Biology, California State University, Long Beach, Cal-
ifornia 90840.

2

Huckaby: Species of Peromyscus mexicanus Group

Cooperative Wildlife Collection, Texas Agricultural and Me-
chanical University (TCWC); Museum of Vertebrate Zoology,
University of California, Berkeley (MVZ); California Acad-
emy of Sciences (CAS); Natural History Museum of Los An-
geles County (LACM); University of California, Los Angeles
(LJCLA); California State University, Long Beach (CSULB).

I obtained data on the male reproductive system from fluid-
preserved specimens consisting of either whole animals or spe-
cially preserved tracts. Hooper (1958) described the techniques
used here for preparing the glans penis for observation. I list
here the samples of glandes penis, all contained in The Univer-
sity of Michigan, Museum of Zoology.

P.furvus: 13 mi NE Metepec, 13; 2-7.3 mi SE Huauchinan-
go, 6; Xilitla, 1 .

P. grandis: Finca Concepcion, 4.

P. guatemalensis: Cerro Mozotal, 5; Finca El Injerto, 6;
Barillas, 1; Yayquich, 2; Cumbre Maria Tucum, 3.

P gymnotis: 13 km N Huixtla, 10; 1 mi E Escuintla, 2; Finca
El Rosario, 1.

P. megalops: near Puerto Chico, 1 6; Campemento Rio Mo-
lino, 13; 4 mi S Jalatengo, 3; 23 mi N San Gabriel Mixtepec, 8;
vicinity Santa Rosa, 6.

P melanocarpus: 116 km SW Tuxtepec, 3; 23 mi S Valle
Nacional, 2.

P. mexicanus: 5 mi N Berriozabal, 1 1 ; vicinity of Tuxtla
Gutierrez, 15; 3 km E Riso de Oro, 2; Volcan Agua, 1; Santa
Maria de Ostuma, 1; 2 mi N Teocelo, 5; Barranca Texcolo, 2;
vicinity of Tenochtitlan, 5.

P. nudipes: Volcan Irazu, 1; Moravia de Chirripo, 7; 3 mi SE
Turrialba, 3; Tapanti, 1; Cerro de la Muerte, 3; Monte Verde, 1.

P. ochraventer: El Carrizo, 4; Gomez Farias, 5.

P. yucatanicus: Chichen Itza, 1; vicinity of Escarcega, 9.

P. zarhynchus: Yerba Buena, 10; Cerro Tzontehuitz, 6.

The following list contains the examples of male accessory
glands, all in The University of Michigan, Museum of Zoology.

P.furvus: Zacapoaxtla, 1.

P. guatemalensis: Cerro Mozotal, 1; El Injerto, 2.

P. grandis: Finca Concepcion, 2.

P. megalops: Campemento Rio Molino, 2; vie. Santa Rosa, 2.

P. melanocarpus: 23 mi S Valle Nacional, 1.

P. mexicanus: 5 mi N Berriozabal, 2; Solusuchiapa, 2.

P. ochraventer: Rancho del Cielo, 2.

P. yucatanicus: Chichen Itza, 2.

P. zarhynchus: Cerro Tzontehuitz, 1.

ANALYSIS OF CHARACTERS

Method of Selection

In selecting characters, I accepted local population samples of
like kinds as the operational taxonomic units. I pooled local
population samples with their geographic neighbors until I ob-
served a discontinuity in character states. If any apparent dis-
continuities in character states, not obviously related to age or
sex within a single sample, suggested the presence of two or
more species in the sample, I then treated the respective sub-
samples as separate samples for the remainder of the study.

As the selection of qualitative and quantitative characters
may involve different methodologies, I shall discuss each sepa-
rately. To select qualitative characters, I visually compared se-
ries of specimens representing the various populations and
noted similarities and differences. Desirable characters re-

mained fairly constant within samples but varied between sam-
ples. This follows the principle of conservancy (Farris 1966). To
qualify as qualitative, a character had to have relatively dis-
crete states. Otherwise, I considered the character quantitative.

Students have used at least two methods in selecting mor-
phometric characters. One method consists of measuring many
dimensions of the specimen, running correlations between each
pair of characters within samples, and discarding all highly cor-
related characters as redundant. This technique may involve
the expenditure of a large amount of effort and could result in a
much smaller list of acceptable measurements than represented
on the original list. In addition, if the method is not used in
combination with a considerable amount of visual comparison
of samples, there is a risk of overlooking interesting differences
involving the relation of various anatomical parts. For these
reasons, I obtained my list of morphometric characters by the
same method I used to determine qualitative characters,
namely, by direct visual comparison. In observing series of
skulls representing different populations, I noted whether one
sample appeared to have a broader brain case or a narrower
interorbital area, for example, than another. By this procedure,
I obtained a list of 1 1 skull measurements that singly or in com-
bination seemed to separate samples. I then measured each of
the 1 1 characters on all of the skulls in all of my samples.

In addition to the character differences that I discovered
more or less on my own, some previous morphological work
done on the group (Hooper 1958; Hooper and Musser 1964;
Linzey and Layne 1969; and Carleton 1973) suggested other
differences. In considering all except the last, however, I ampli-
fied the published results with my own observations; I accepted
Carleton’s results as published.

Skull Measurements

I made 1 1 cranial measurements with the aid of a measuring
microscope (Anderson 1968). The skull rested ventrum down
on the microscope stage for measurements one to five (see be-
low), and dorsum down for seven to eleven. The suture between
the nasals, frontals, and parietals or maxillaries paralleled the
vertical line of the eyepiece crosshair for all lengths and the
horizontal line of the crosshair for all widths. For measurement
1 1, the skull rested on a piece of plasticene with ventrum up
and the longitudinal axis of the tooth parallel to the horizontal
line of the crosshair.

These measurements resulted:

1. Length of skull, from the anterior tip of the nasals to the
posterior tip of the supraoccipital.

2. Rostral length, from the anterior tip of the nasals to a
plane through the anterior face of the zygomatic plate.

3. Length of brain case, from the posterior edge of the trans-
verse depression in the frontal bone, which indicates the antero-
dorsal limits of the frontal lobes of the brain, posteriorly to the
tip of the supraoccipital.

4. Interorbital width, least interorbital width of the frontals.

5. Width of brain case, greatest width of the brain case dor-
sal to the root of the zygomatic arch.

6. Length of incisive foramen, greatest length of the left in-
cisive foramen.

7. Molar row, greatest crown length of the left maxillary mo-
lar row.

8. Length of interpterygoid fossa, from the posterior limit of
the hard palate to the posterior tip of the left alisphenoid.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County 1980. 326:1-24

Huckaby: Species of Peromyscus mexicanus Group

3

D E F

Figure 1. Dorsal view of Peromyscus skulls to illustrate distally expanded nasals (A), distally
unexpanded nasals (B,C,D,E,F); hourglass interorbital (A); unbeaded supraorbital ridge (C,D);
beaded supraorbital ridge (B,E); and partially beaded supraorbital ridge (F). A, Peromyscus fur-
vus, CAS 13791, Veracruz; B, P. melanurus, CAS 16412, Oaxaca; C, P. mexicanus, LACM 14263,
Chiapas; D, P gymnotis, LACM 18892, Chiapas; E, P. megalops. CAS 16409, Oaxaca; F, P.
melanocarpus, CSULB 8900, Oaxaca.

9. Intermolar width, greatest distance between the lingual-
most projection of the first upper molars.

10. Width of interpterygoid fossa, greatest width of the soft
palate.

1 1 . Molar width, greatest width of the crown of the first up-
per molar.

Skull Morphology

I discerned three mutually exclusive forms of skulls with re-
spect to the architecture of the supraorbital region (Fig. 1): no
ridge or bead, exhibiting a smooth hourglass shape; a ridge
often extending well out into the supraorbital area and occa-
sionally forming a slight bead at the suture between the frontal
and parietal bones; a strongly beaded shelf extending the length
of the interorbital area. These differences appear constant at
the population level.

The distal part of the rostrum remains unexpanded in most
populations, but old animals in some samples exhibit expanded
nasals (Fig. 1 ).

Dental Morphology

In assessing variation in dental patterns in the P. mexicanus
group, I recorded the presence or absence of the mesoloph, ec-
tolophid, and mesolophid in most of the populations at hand
using Hooper’s (1957) terminology. The presence of styles and
lophs appears much too variable for use, except in conjunction
with other details.

The dental pattern of those samples referable to P. furvus
(Fig. 2) and P ochraventer differs from that of the other sam-
ples and requires further discussion. These samples have
strongly developed lophs that nearly always extend, uninter-
rupted, from the mure to the appropriate style. In P. fun’us

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 326:1-24

4

Huckaby: Species of Peromyscus mexicanus Group

(less well-developed in P. ochraventer), the style in the minor
fold of the first lower molar resembles the two halves of the
anterior cingulum in size and so closely joins the labial half as
to appear an extension of it. The anterior cingulum of the first
upper molar in P. furvus has a smaller lingual and a larger la-
bial cusp and frequently has a small style anterior to the cleft
between the two cusps. The other species have less strongly de-
veloped and frequently interrupted lophs with the style in the
minor fold of the anterior cingulum of the first upper molar
usually single (corresponding to the labial half of the other
type, but occasionally with a small lingual cusp). These differ-
ent types of cingulum structure correlate strongly with com-
plexity in the accessory lophs. Accordingly, I differentiate
between the relatively complex teeth of P. furvus and P ochra-
venter and the relatively simple teeth of the remaining species.

External Measurements

I relied mainly on three external measurements taken from
data recorded by the collector on the specimen tag: head and
body length, tail length, and hind feet length. I judged the ear
lengths recorded by collectors as too variable for use.

Pelage Color

dorsal coloration. With large samples of properly prepared
material comparable in age and season of collection, color vari-
ation would provide one or two more characters for understand-
ing relationships. I made use, however, of subjective impres-
sions of both hue and intensity of the overall coat color in
assessing the samples.

The adults of P. ochraventer exhibit an ochraceous wash over
the venter, whereas no individuals in any other species exhibit
such a wash. A concentration of pectoral buff occurs in many
samples of other species and ranges from a faint suggestion to a
patch covering the chest. Only some populations from Oaxaca
(P. megalops melanurus) completely lack a pectoral patch.

Some individuals from various geographic areas exhibit dark
hairs, instead of the usual whitish hairs, on the carpus and tar-
sus. Darkness of carpus and tarsus correlates well with dark-
ness of dorsal pelage.

Tails consistently exhibit dark dorsal coloration. Ventral
hairs on a given tail range from all pale to all dark. Ventral
scales also range from pale to dark, the areas appearing as light
and dark. All populations, with one or two exceptions, resemble
one another closely. Specimens from low elevations tend to have
short, sparse hairs, and the basic color of the tail, which derives
almost wholly from that of the scales, usually appears mono-
colored dark or dark with pale ventral blotches. Specimens
from higher elevations possess hairier tails that usually appear
more bicolored to the unaided eye.

Dorsal coloration varies considerably between populations, but
many factors render this variation practically useless for this
study. Age variation, with at least the four discernible pelages
of juvenile, subadult, young adult, and old adult, greatly re-
duces the sample size of comparable material. The addition of
the clearly darker pelages of specimens collected in the rainy
season results in only a few samples having truly comparable
pelages. Because of this ontogenetic and seasonal variation and
because the variation induced by the unstandardized methods
of stuffing the skins precludes the use of a reflecting meter for a
more precise hue, 1 found no satisfactory, objective measure of

Figure 2. Diagram of the crown patterns of upper left molar teeth to
illustrate complex (A) and simple (B) conditions. 1, anterocone; 2,
mesoloph and style; 3, anterior; and 4, lingual.

Mammary Glands

Mammae occur either as one pair axillary and two pairs ingui-
nal or as two pairs inguinal. Number of mammae remains
constant within populations.

Stomach Morphology

Carleton (1973) recognized three types of stomach in the P.
mexicanus group. Most forms have the glandular portion of the
stomach evaginated into a pouch, others lack the pouch, and a
third type exhibits a partial pouch.

Glans Penis

Figure 3 illustrates the extreme states for all but the last char-
acter of the glans penis. Dorsal lappets occur either as two
distinct elongate structures in most forms (Fig. 3A) or as a se-
ries of fingerlike projections from the scalloped edge subtending
the protractile tip of the glans in P. yucatanicus (Fig. 3B).

Most species have a relatively narrow baculum with an at-
tenuate tip (Fig. 3B and C); P. furvus has a relatively broad
baculum with a slightly enlarged and upturned distal tip (Fig.
3A). The shape of the base of the baculum varies considerably
within samples.

The cartilaginous tip varies from a moderate cone in most
species (Fig. 3A and C) to a tiny, barely discernible cap pro-
jecting into the protractile tip in P. yucatanicus (Fig. 3B). I
found no correlation between the relative length of the protrac-
tile tip and the length of the cartilaginous tip.

Most species have a relatively long glans (Fig. 3E and F); P.
ochraventer has a short one (Fig. 3D). The total length of the

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 326:1-24

Huckaby: Species of Peromyscus mexicanus Group

5

Figure 3. Glandes penis to illustrate: 1, dorsal lappet; 2, baculum; 3, cartilaginous tip; 4, length of glans;
5, length of protractile tip; and 6, length of spinous portion of glans. Spines not illustrated. A, Peromyscus
furvus, UM 1 12940; B, P. yucatanicus, UM 122434; C, P grandis, UM 1 17949; D, P. ochravenier, UM
122427; E, P melanurus, UM 1 18072; F, P. megalops, UM 1 17299.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 326:1 24

6

Huckaby: Species of Peromyscus mexicanus Group

short glans corresponds only to the combined lengths of the
spiny portion and the protractile tip of the long glans.

The length of the protractile tip varies from less than one-
fourth to one-third the length of the glans. Since the apparent
length of the tip varies with the amount of protrusion, I used
only fully protruded specimens. Most species have a tip one-
third the length of the glans (Fig. 3F); P. megalops melanurus
has a tip less than one-fourth the length of the glans (Fig. 3E).
In general, the longer protractile tips appear more attenuate,
with a smaller bulb at the end, than do the shorter tips.

Most species have relatively large spines; P. ochraventer ex-
hibits relatively small ones (see Hooper and Musser 1964).

Male Accessory Glands

Linzey and Layne (1969) describe differences in the male re-
productive tracts exclusive of the glans penis for specimens of

A A

A

II

Figure 4. A projection of 30 samples on the first (I) and second (II)
principal components. The data consist of means of 1 1 skull measure-
ments. Component I accounts for 97.3% of the variance; component II
for 1.0%. Table 1 lists the correlations of the characters with the com-
ponents. Size correlates negatively with component I. Squares repre-
sent Peromyscus gymnotis; open circles, P. mexicanus; closed circles,
P. guatemalensis; closed triangles, P. zarhynchus; and open triangle, P.
grandis.

Table 1 . Correlations of Peromyscus skull characters with the
principal components for the analysis illustrated by Figure 4.

Components

Characters

I

II

Length of skull

-0.99

0.02

Rostral length

-0.98

0.16

Length of braincase

-0.98

-0.11

Interorbital width

-0.89

-0.17

Width of braincase

-0.95

-0.26

Length of incisive foramen

-0.97

0.09

Molar row

-0.96

-0.11

Length of interpterygoid fossa

-0.92

-0.22

Intermolar width

-0.87

-0.11

Width of interpterygoid fossa

-0.88

-0.01

Molar width

-0.94

-0.11

the P. mexicanus group. I examined the specimens that they
used (deposited either in the United States National Museum
or the Museum of Zoology, The University of Michigan) and
agree with all identifications except three of the four specimens
they assigned to P. mexicanus. I regard the specimen from
Chiapas, Mexico, as P. mexicanus, the two from Baja Verapaz,
Guatemala, as P. oaxacensis, and the one from Sacatepequez,
Guatemala, as P. boylii.

These three misidentified specimens contributed most of the
variation observed by Linzey and Layne in the P. mexicanus
group. Layne* informed me that the specimen from Chiapas
had two pairs of ventral prostates. Thus, since the number of
ducts for the prostates tends to vary as much individually as it
does between populations, only the bifurcation of the penis bulb
tends to vary along species lines in this group. My examinations
lead me to consider the assignment of “bifurcated” versus
“nonbifurcated” to a specimen as subjective. In addition, the
degree of bifurcation of the bulb appears to correlate with the
relative state of enlargement of the testes (related to season-
ality of sperm production?), but the few specimens preclude a
more definitive correlation. For these reasons, I used no charac-
ters of the male accessory glands in my estimates of relation-
ships. In addition, species that I examined but that Linzey and
Layne did not (P. gymnotis, melanocarpus, ochraventer, and
yucatanicus) conform to the general pattern they reported for
members of the P mexicanus group, as well as to the pattern of
much of the rest of the subgenus Peromyscus.

Statistical Analysis

I analyzed the morphometric data with the principal compo-
nents program in the MIDAS package of computer programs
of The University of Michigan. Morrison (1967) discusses the
methodology.

ANALYSIS OF POPULATIONS

In arriving at my estimates of species limits, I pooled local sam-
ples until I discerned a discontinuity in character states. The
species that I recognize consist of populations that I cannot dis-
tinguish among on any basis other than slight average dif-
ferences in size and/or color. The ranges of most of the
resulting species do not overlap. For most cases of allopatry, two
or more distinct qualitative differences not due to age, sex, or
season suggest that the forms in question probably would not
interbreed if they ever did contact one another. These tech-
niques allowed me to easily separate seven species (P. ochraven-
ter, stirtoni. yucatanicus, furvus, megalops, melanocarpus, and
melanurus ) both from one another and from the remainder of
the populations.

No really distinct qualitative characters other than size and
color distinguish the remaining populations. Yet their range of
variation greatly transcends the amount usually found within
species of Peromyscus. Figure 4 and Table 1 illustrate the out-
come of the analysis of the remaining populations that occur in
Chiapas and Guatemala. The five species that I recognize form
a clearly graded series from small (P. gymnotis) to large (P.
grandis ), yet their respective geographic distributions only par-
tially correspond to the size differences (refer to species ac-
counts). The three largest forms (P. grandis, guatemalensis,
and zarhynchus) do occur at generally higher elevations than

*J.N. Layne, Archbold Biological Station, Lake Placid, Florida, per-
sonal communication.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 326:1-24

Huckaby: Species of Peromyscus mexicanus Group

7

do the smaller two (P. mexicanus and gymnotis ), although the
largest one ( P . grandis) occurs at elevations commonly oc-
cupied by P. mexicanus in other areas in Guatemala.

The allopatric ranges of P. zarhynchus, guatemalensis, and
grandis suggest the fragmentation of the range of a single spe-
cies once occurring over their combined areas, and 1, therefore,
earlier synonymized them as subspecies of P. zarhynchus
(Huckaby 1973). On the basis of size (Fig. 4 and Appendix)
and color (see species accounts below), P. guatemalensis differs
more from P. zarhynchus and grandis than the latter two do
from one another, yet P. guatemalensis occurs between them.
This suggests that P. guatemalensis arose independently of the
other two and now occupies an area once inhabited by popula-
tions that intergraded between the present P. zarhynchus and
grandis. 1 cannot choose between these possibilities with the
data at hand. 1 also cannot discount the possibility that all
three arose independently of one another by separate invasions
of their montane habitats from one or more lowland forms and
owe much of their similarities to convergence. Since I can dis-
tinguish the three, albeit mostly on size and color, I now con-
sider them full species.

Size alone (Fig. 4) easily separates P. mexicanus from gran-
dis and zarhynchus, and P. mexicanus occurs with zarhynchus
at Tumbala. Figure 4 demonstrates, however, some overlap in
size between populations assigned to P. mexicanus and those
assigned to P. guatemalensis. The population of P. mexicanus
(Finca San Rafael, Guatemala) that falls within the size range
of P. guatemalensis appears, at first glance, to bridge the gap
between the two supposed species. Thus, P. guatemalensis may
grade into P. mexicanus in the eastern part of its range but not
elsewhere. I suggest, however, that the populations from the
southeastern volcanoes of Guatemala (slopes of Acatenango-
Fuego and Agua) and adjacent highlands, here assigned to P.
mexicanus , resemble true P. guatemalensis from further west
(highlands around Lake Atitlan) simply because they occur at
higher elevations in habitats commonly occupied by P.
guatemalensis and have responded to selection pressures similar
to those that produced P. guatemalensis in the first place. Al-
though these populations of large dark P. mexicanus resemble
P. guatemalensis closely, they generally lack the fairly bi-
colored tails and gray pelage of P. guatemalensis. They closely
resemble P. mexicanus from the highlands of Honduras and
Nicaragua in color (dark brown to nearly black, rather than
the gray to black of P. guatemalensis). More specimens from
the critical area in southern Guatemala could force the altera-
tion of these conclusions.

Peromyscus gymnotis resembles a diminutive, dark P. mex-
icanus, and Osgood (1909) arranged it as a race of the latter
while recognizing a second small dark species ( P . allophylus)
from the same area. Musser (1971) demonstrated that only one
species, P. gymnotis, occurs on the coastal plain of Chiapas and
Guatemala. Peromyscus mexicanus and gymnotis occur para-
patrically to one another in the area south of Guatemala City
and both occur on the south slopes of Volcan Agua (P. mex-
icanus above gymnotis) with no suggestion of intergradation.
Further collecting in Chiapas between Tonala (where P. mex-
icanus occurs) and Pijijiapan (where P. gymnotis occurs) may
demonstrate a similar distribution there.

From the time of its description, all previous workers have
considered P. nudipes as a species separate from P. mexicanus,
although frequently commenting on the similarity of the two
forms. Peromyscus nudipes supposedly consists of larger, darker

animals than P. mexicanus does. I compared the skull measure-
ments of samples of the two forms by principal component
analysis (Fig. 5 and Table 2). The five samples of P nudipes did
not cluster together on either of the first two components and
four P. mexicanus samples had larger overall skulls than any of
the P. nudipes. I can detect no qualitative characters that will
separate P nudipes from mexicanus. I, therefore, consider the
populations formerly assigned to P. nudipes as the southern-
most extension of P. mexicanus.

SPECIES ACCOUNTS

Peromyscus ochraventer

Brown-bellied Mouse

SYNONYMY. Peromyscus ochraventer Baker 1951:213.

HOLOTYPE. An old adult female, skin and skull, in good
condition, KU 36958, collected 12 January 1950, 70 km S of

• • o • •

• • •

• •

o •

o

II

Figure 5. A projection of 38 samples on the first (I) and second (II)
principal components. The data consist of means of 1 1 skull measure-
ments. Component I accounts for 75.7% of the variance; component II
for 9.5%. Table 2 lists the correlations of the characters with the com-
ponents. Size correlates negatively with component I. Closed circles in-
dicate Peromyscus mexicanus; open circles, “P. nudipes.”

Table 2. Correlations of Peromyscus skull characters with the
principal components for the analysis illustrated by Figure 5.

Components

Character

I

II

Length of skull

-0.99

0.10

Rostral length

-0.90

0.28

Length of braincase

— 0.81

-0.20

Interorbital width

-0.38

-0.48

Width of braincase

-0.45

-0.82

Length of incisive foramen

-0.67

-0.41

Molar row

-0.58

-0.53

Length of interpterygoid fossa

-0.76

0.04

Intermolar width

-0.63

-0.22

Width of interpterygoid fossa

-0.49

-0.01

Molar width

-0.50

-0.30

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 326:1-24

8

Huckaby: Species of Peromvscus mexicanus Group

Ciudad Victoria and 6 km W of Interamerican Highway at El
Carrizo, Tamaulipas, Mexico, 2800 ft.

DIAGNOSIS. A medium-sized species of the subgenus Per-
omyscus with unexpanded nasals in adults; a smoothly
hourglass-shaped interorbital area; relatively complex teeth
with divided anterocone; ochraceous ventral color in adults; a
pair of pectoral mammae; a discoglandular stomach; a rela-
tively short glans penis with small spines, a relatively short pro-
tractile tip, and undivided dorsal lappets; and a cylindrical
baculum with a slight distal enlargement and a small car-
tilaginous tip.

DISTRIBUTION. The moist temperate and adjacent humid
tropical forests along the eastern slopes of the Sierra Madre
Oriental from the Rancho del Cielo area in Tamaulipas south
to the vicinity of Platanito in San Luis Potosi (Fig. 6). Known
range probably conforms closely to actual range.

VARIATION. Dalquest (1953:152) allocated eight speci-

mens from 10 km E of Platanito to P. mexicanus. I have exam-
ined the specimens (LSU 5782-9) and consider them P. ochra-
venter. They differ in no essential respect from specimens of P.
ochraventer from Tamaulipas and differ from the specimens of
P. mexicanus from 3 km N Tamazunchale, San Luis Potosi, by
their more hourglass-shaped interorbital region, more complex
teeth, and ochraceous ventral color in adults. I can detect no
important geographic variation within the small known range
of this species.

IDENTIFICATION. The Appendix provides descriptive sta-
tistics for morphometric variables useful in identifying P.
ochraventer. Table 3 summarizes the major qualitative dif-
ferences between the species herein considered. The moderately
large size, ochraceous ventral color, hourglass-shaped interorbi-
tal areas, and heavy complex teeth easily separate it from all
the other species of Peromyscus known to inhabit the same area
{P. boylii, pectoral is, and leucopus). The ochraceous ventral

Contrib. Sci. Natur. Hist. Mus. Los Angeles County 1980. 326:1-24

Huckaby: Species of Peromyscus mexicanus Group

9

color and hourglass-interorbital areas together with heavy com-
plex teeth separate it from geographically adjacent and similar
sized species (e.g., P. mexicanus).

SPECIMENS EXAMINED (114): MEXICO. San Luis
Potosi: 8 mi W El Naranjo, 1 (MSU); 10 km E Platanito, 8
(LSU). Tamaulipas: El Carrizo, 15 (KU); 3 mi W El Carrizo,
1500 ft, 8 (UM); Gomez Farias, 7 (UM); 5 mi NW Gomez
Farias, 4 (UM); Rancho del Cielo, 3 (UCLA), 62 (UM), 6
(AM).

Peromyscus stirtoni

Stirton’s Mouse

SYNONYMY. Peromyscus stirtoni Dickey 1928:5.

HOLOTYPE. An old, adult female, skin and skull, in good
condition, UCLA 10634, collected 29 October 1925, 13°30N
on the Rio Goascoran, La Union, El Salvador, 100 ft.

DIAGNOSIS. A small species of the subgenus Peromyscus
with strongly beaded supraorbital ridges; relatively simple
teeth; a hairy, bicolored tail; and no pectoral mammae (one
female from Honduras). No data on glans penis or stomach.

DISTRIBUTION. Dry to semiarid valleys from south-
eastern Guatemala to southern Honduras (Fig. 7). Limits of
range unknown.

IDENTIFICATION. Only Peromyscus mexicanus, boylii,
oaxacensis, and lophurus have known ranges that broadly over-
lap the known range of P. stirtoni. Of these, only P. mexicanus
probably occurs in the same habitat as P. stirtoni. The small
size of P. stirtoni (Appendix) together with its beaded supraor-
bital ridges and hairy, bicolored tail (Table 3) will easily dis-
tinguish it from any of the above species.

SPECIMENS EXAMINED (31): EL SALVADOR. La
Union: Rio Goascoran, 2 (MVZ), 1 (UCLA); Pine Peaks, 3 mi
W Volcan de Conchagua, 3200 ft, 4 (MVZ). Morazan: 1 mi SE
Divisadero, 850 ft, 1 (MVZ). San Miguel: Lake Olomega, 200
ft, 1 (MVZ). Santa Ana: Lake Guija, 1450 ft, 1 (MVZ).
GUATEMALA. Jutiapa: Santa Catarina Mita, 2450 ft, 1
(FM). Zacapa: Trujillo, 1 mi E Progresso-Zacapa border, 700
ft, 1 (NM). HONDURAS. Francisco Morazon: 3 mi S La
Venta, 1 (TCWC); La Piedra de Jesus, Sabana Grande, 19
(AM).

Peromyscus yucatanicus

Yucatan Mouse

SYNONYMY. Peromyscus yucatanicus J.A. Allen and
Chapman 1897a:8. Peromyscus yucatanicus badius Osgood
1904:70. Apazote, Campeche, Mexico.

HOLOTYPE. An adult male skin and skull, in good condi-
tion, AM 12001/10434, collected 17 March 1896, Chichen
Itza, Yucatan, Mexico.

DIAGNOSIS. A small species of the subgenus Peromyscus
with unexpanded nasals; a moderately developed supraorbital
shelf without beading; moderately complex teeth usually with
single anterocone; no ochraceous ventral color; no pectoral
mammae; a discoglandular stomach; a relatively long glans pe-
nis with large spines, a relatively long protractile tip, divided
dorsal lappets; and a cylindrical baculum with no great distal
enlargement and a small cartilaginous tip.

DISTRIBUTION. The semideciduous to semievergreen for-
ests of the Yucatan Peninsula of Mexico (Fig. 7). Southern lim-
its of range unknown. Apparently no species of Peromyscus
occurs in the southern part of the peninsula including Belice
and El Peten, Guatemala, making this the largest area in Mid-

dle America north of Panama without any Peromyscus.

VARIATION. Lawlor (1965) discussed variation in P. yuca-
tanicus and concluded that, although certain trends existed,
they did not warrant subspecific recognition. As Lawlor
(1965:430-31) noted, the mice from the northern part of the
range have brighter, buffier coats as adults than do those from
southern Campeche. I have observed live animals in the labora-
tory obtained from near Escarcega, Campeche, by J.A. Lackey
and from near Chichen Itza, Yucatan. Those from Yucatan
start life with a paler coat of gray than those from Campeche
and develop bright buff over the dorsum as they mature. In
contrast, animals from Campeche over a year old show only
slight traces of buff, and at all ages appear much darker and
grayer than specimens from further north. Osgood (1904) seg-
regated these dark southern animals off as a separate sub-
species P. y. badius. Lawlor (1965) concluded that, since the
variation in size did not correlate with the variation in color, he
could not recognize distinct subspecies. For the present, I agree
with Lawlor. Those wishing to do so can, however, easily sepa-
rate the specimens from southern Campeche from those of
more northern localities on the basis of color and apply the sub-
specific epithet P. y. badius to them.

IDENTIFICATION. Of all Peromyscus. apparently only P.
leucopus occurs sympatrically with P. yucatanicus. Peromyscus

Figure 7. Geographic distributions of Peromyscus yucatanicus (cir-
cles) and P. stirtoni (squares). Vertical hatching indicates known range
of P. mexicanus.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 326:1-24

10

Huckaby: Species of Peromyscus mexicanus Group

Table 3. Selected characters for the species of the Peromyscus mexicanus group.

CHARACTERS OF THE GLANS PENIS

Species

Length

of

Glans

Protrac-
tile Tip Spines

Dorsal

Lappets

Tip of
Baculum

Cartila-

ginous

Tip

P. ochraventer

Short

Short

Small

Undivided

Slightly expanded

Small

P. stirtoni

Unknown

Unknown Unknown

Unknown

Unknown

Unknown

P. yucatanicus

Long

Long

Large

Divided

Unexpanded

Small

P. furvus

Long

Short

Large

Undivided

Expanded

Medium

P. megalops

Long

Long

Large

Undivided

Unexpanded

Large

P. melanocarpus

Long

Long

Large

Undivided

Unexpanded

Large

P. melanurus

Long

Short

Large

Undivided

Unexpanded

Medium

P. zarhynchus

Long

Long

Large

Undivided

Unexpanded

Large

P. guatemalensis

Long

Long

Large

Undivided

Unexpanded

Large

P. grand is

Long

Long

Large

Undivided

Unexpanded

Large

P. gymnotis

Long

Long

Large

Undivided

Unexpanded

Large

P. mexicanus

Long

Long

Large

Undivided

Unexpanded

Large

OTHER CHARACTERS

Supra-

Relative

Tip of

orbital

Complexity

Pectoral

Species

Nasals

Area

of Teeth

Mammae

Stomach

P. ochraventer

Unexpanded

Hourglass

Complex

Present

Discoglandular

P. stirtoni

Unexpanded

Beaded shelf

Simple

Absent?

Unknown

P. yucatanicus

Unexpanded

Nonbeaded shelf

Simple

Absent

Discoglandular

P. furvus

Expanded

Hourglass

Complex

Present

Discoglandular

P. megalops

Unexpanded

Beaded shelf

Simple

Absent

Pouched

P. melanocarpus

Unexpanded

Beaded shelf

Simple

Absent

Half-pouched

P. melanurus

Unexpanded

Beaded shelf

Simple

Absent

Half-pouched

P. zarhynchus

Unexpanded

Nonbeaded shelf

Simple

Absent

Pouched

P. guatemalensis

Unexpanded

Nonbeaded shelf

Simple

Absent

Pouched

P. grandis

Unexpanded

Nonbeaded shelf

Simple

Absent

Pouched

P. gymnotis

Unexpanded

Nonbeaded shelf

Simple

Absent

Pouched

P. mexicanus

Unexpanded

Nonbeaded shelf

Simple

Absent

Pouched

leucopus is on the average, smaller in most dimensions (Appen-
dix), lacks supraorbital ridging, and has three pairs of mammae
and undivided dorsal lappets (Table 3).

SPECIMENS EXAMINED (159). MEXICO. Campeche:
La Tuxpena, 1 5 (NM); Apazote, 21 (NM); San Juan, 5 (FM);
6 km S Chompoton, 2 (UM); 3 km W Escarcega, 1 (UM); 7
km W Escarcega, 11 (UM); 7.5 km W Escarcega, 4 (UM); 7
km W, 2 km N Escarcega, 1 (UM); 7 km W, 1 km N Escar-
cega, 1 (UM); 9.8 km W, 3.3 km N Escarcega, 1 (UM). Quin-
tana Roo: La Vega, 29 (NM); 20 km S Peto, Santa Rosa, 19
(UM); 45 km S Peto, Esmeralda, 9 (UM). Yucatan: Chichen
Itza, 24 (NM), 1 1 (UM); Calcehtok, 5 (UM).

Peromyscus furvus

Blackish Mouse

SYNONYMY. Peromyscus furvus J.A. Allen and Chapman
1897b:201. Peromyscus latirostris Dalquest 1950:8. Apetsco,
San Luis Potosi, Mexico. Peromyscus angustirostris Hall and
Alvarez 1961:203. 3 km W Zacualpan, Veracruz, Mexico.

HOLOTYPE. An adult male, skin and skull, in good condi-
tion, AM 1 2450a / 1 0769, collected 2 April 1897, 1.5 mi E Ja-
lapa, Veracruz, Mexico, 4400 ft.

DIAGNOSIS. A large species of the subgenus Peromyscus
with distally expanded nasals in old adults; a smoothly
hourglass-shaped interorbital area; relatively complex teeth
with divided anterocone; no ochraceous ventral color; a pair of

pectoral mammae; a discoglandular stomach; a relatively long
glans penis with large spines, a relatively short protractile tip,
and undivided dorsal lappets; and a laterally flattened baculum
with a bulbous distal end and a moderately elongate cartilagi-
nous tip.

DISTRIBUTION. The cool, humid forests between 1200
and 2200 m along the eastern slopes of the Sierra Madre Ori-
ental from southeastern San Luis Potosi to northern Oaxaca
(Fig. 8). Southward dispersal probably limited by the deep can-
yon of the Rio Santo Domingo-Quiotepec through the Sistema
Montanoso in Oaxaca. The known range probably corresponds
to the actual range.

VARIATION. Goodwin (1969:192) allocated four subadult
specimens (AM 207440-3) from 20 mi E Teotitlan, Oaxaca, to
P. melanocarpus. On the basis of their complicated teeth and
lack of supraorbital shelving, I allocate them to P. furvus. In
discussing variation in P. furvus. Hall (1971) concluded that
the small amount attributable to geography did not warrant the
recognition of subspecies. I fully concur but do not discount the
possibility of slight size differences between some of the scat-
tered localities. In particular, specimens from San Luis Potosi
have the distally expanded nasals carried to the greatest
degree.

IDENTIFICATION. Its large size (Appendix) easily sepa-
rates P. furvus from most other Peromyscus known to inhabit
its restricted range (P. boylii, aztecus, simulatus, leucopus, and

Contrib. Sci. Natur. Hist. Mus. Los Angeles County 1980. 326:1-24

Huckaby: Species of Peromyscus mexicanus Group

1 1

pectoralis). Peromyscus mexicanus, which occurs lower on the
eastern slopes than P. furvus, has unswollen nasals, usually
some supraorbital ridging, relatively simple teeth usually with
single anterocone, no pectoral mammae, and a round bacular
shaft unexpanded distally (Table 3). Peromyscus thomasi oc-
curs sympatrically with P. furvus and differs by having larger
dimensions throughout, a larger skull with supraorbital ridging,
and a large thick glans penis with a base much broader than
tip. Peromyscus difficilis occurs in drier habitat to the west of
P. furvus and differs in its relatively more inflated auditory
bullae, unexpanded nasals, and simpler teeth with less tendency
for a divided anterocone.

SPECIMENS EXAMINED (187). MEXICO. Hidalgo: 13
mi NE Metepec, 6600 ft, 29 (UM). Oaxaca: 20 mi E Teotitlan,
4 (AM). Puebla: Huauchinango, 1 (UM), 2 (NM); 2-2.5 mi
SW Huauchinango, 5500-7000 ft, 10 (UM); 5.7 mi SW
Huauchinango, 6600 ft, 6 (UM); 7.3 mi SW Huauchinango,
6800 ft, 7 (UM); Honey, 1 (UM); 2 mi NW Zacapoaxtla, 1520
m, 1 (UM). Queretaro: 6 mi W Ahuacatlan, 5800 and 5600 ft,
2 (LSU). San Luis Potosi: 3.5 mi SW Xilitla, 740 m, 1 (UM);
Llano de Conejo, 6000 ft, 5 (LSU); Lower Llano, 2 (LSU);
Xilitla, 1 (LSU); Apetsco, 8 (LSU); Cerro Miramar, 6 (LSU);
Rancho Miramar Grande, 6000 ft, 7 (LSU); Cerro San An-
tonio, 3 (LSU). Veracruz: Jalapa, 13 (AM), 4 (NM); 5 km N
Jalapa, 1 (UM), 5 (KU); 5 km S Jalapa, 1 (UM); 0.5 mi Ja-
lapa, 4500 ft, 1 (UM); 2 mi SE Huayacocotla, 6500 ft, 4
(UM); Xico, 2 (NM); 1 mi W Xico, 1340 m, 15 (UM); 2 km
W Jico, 4200 ft, 26 (KU); Zacualpan, 6000 ft, 7 (KU); 3 km
W Zacualpan, 6000 ft, 16 (KU); Puente San Bernardo, 1 mi ?
Cacahualco, 2 (CAS).

Peromyscus megalops

Broad-faced Mouse

SYNONYMY. Peromyscus megalops Merriam 1898:119.
Peromyscus auritus Merriam 1898:119. 15 mi W Oaxaca,
Oaxaca, Mexico. Peromyscus comptus Merriam 1898:120.
“Mts” W Chilpancingo, Guerrero, Mexico.

HOLOTYPE: An old adult male, skin and skull, in good con-
dition, NM 71592, collected 26 March 1895, at a ranch called
La Cieneguilla at 10,000 feet in the Sierra Madre del Sur, near
the village of Santa Maria Ozolotepec, Oaxaca, Mexico.

DIAGNOSIS. A large species of the subgenus Peromyscus
with unexpanded nasals; strongly beaded supraorbital ridges;
moderately complex teeth usually with undivided anterocone;
no ochraceous ventral color; no pectoral mammae; a fully
pouched stomach; a relatively long glans penis with a long pro-
tractile tip, large spines, and undivided dorsal lappets; and a
cylindrical baculum with no great distal enlargement and a
large cartilaginous tip.

DISTRIBUTION. Geographically disjunct populations dis-
tributed in the cool, moist forests of conifers and angiosperms
at elevations from 1800 to 3000 m in the Sierra Madre del Sur
of Guerrero and Oaxaca (Lig. 8). Limits of range probably
known. Taken with P. melanurus near Juquila.

VARIATION. As pointed out by Musser (1964), sampled
populations of P. megalops resemble one another closely. Aver-
age darker color of specimens from near the type locality of P.
megalops warrants mention but not formal subspecific designa-
tion. Except for color, differences ascribed to P. auritus and
comptus fall within the range of individual variation seen in
specimens from near the type locality of P. megalops. The iso-

lated nature of the populations of P megalops suggests that
future studies utilizing very large samples might demonstrate
slight average differences between them in spite of the seem-
ingly very similar habitat.

IDENTILICATION. The combination of large size of most
dimensions (Appendix) and beaded supraorbital ridges (Table
3) easily separates this species from most species possibly sym-
patric with it (P. boylii, evides, oaxacensis, maniculatus, leuco-
pus, truei, difficilis, melanophrys, mexicanus, and thomasi). Of
these, only P. thomasi averages larger in most dimensions, and
none have beaded supraorbital shelves (although P. thomasi.
melanophrys, mexicanus, evides, and oaxacensis often have
nonbeaded supraorbital shelves). Three species occur close to
the known range of P. megalops and share its beaded supraor-
bital ridges. Peromyscus melanurus occurs on the mountain
slopes below P. megalops, averages slightly smaller than mega-
tops in most dimensions, and has slightly shorter, lighter pelage.
The supraorbital ridges of P. melanurus show a much greater
tendency to curve than those of P. megalops (Lig. 1). The rela-
tively long protractile tip of the glans penis of P. megalops also
contrasts with the relatively shorter tip of melanurus (Lig. 3).
Peromyscus melanocarpus occurs in the mountains to the north
of the known range of P. megalops. It averages slightly smaller
than megalops, has a much darker pelage usually with dark
hairs on the dorsal surface of the carpus and tarsus, and has
somewhat less well-developed beads on its supraorbital shelves
(Lig. 1). Peromyscus banderanus occurs on the coastal plain
below P. megalops at least in Guerrero, its strongly beaded su-
praorbital ridges show a tendency to curve, and its very elon-
gate skull with oval braincase contrasts with the broader one of
megalops. The tiny awl-shaped glans penis of P. banderanus

Figure 8. Geographic distributions of Peromyscus furvus (triangles),
P melanocarpus (open circles), P. megalops (closed circles), and P.
melanurus (squares). Vertical hatching indicates known range of P
mexicanus.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 326:1-24

12

Huckaby: Species of Peromyscus mexicanus Group

(Hooper 1958, Plate X) also contrasts strongly with that of P.
megalops.

SPECIMENS EXAMINED (244). MEXICO. Guerrero:
mts. near Chilpancingo, 18 (NM), 1 (AM); Omilteme, 24
(UM), 13 (NM), 23 (KU); 18 km SSW Chichihualco, 2500 m,
18 (KU); Cuapongo, 9 (FM); near Puerto Chico, 46 (UM).
Oaxaca: 15 mi W Oaxaca, 6 (NM); Santa Maria Ozolotepec, 4
(NM); Santo Tomas Teipan, 3 (AM); 3 km S San Miguel
Suchixtepec, Rio Molino, 9 (AM), 35 (UM); Cerro Madrena,
7000 ft, 4 (AM); Santo Domingo Chicahuaxtla, 2 (AM); San
Andres Chicahuaxtla, 1 (AM); 2 km NE San Andres Chica-
huaxtla, 2300 m, 15 (UM); 19 mi SW Cuquila, 7800 ft, 3
(MSU); Juquila Rd., 10 mi from Hwy. 131, 4000 ft, 6 (CAS);
Juquila Rd., between Yolotepec and Juquila, 7000 ft, 4 (CAS).

Peromyscus melcinoearpus

Black-wristed Mouse

SYNONYMY. Peromyscus melanocarpus Osgood 1904:73.

HOLOTYPE. A young adult female, skin and skull, in good
condition, NM 68610, collected 8 July 1894, Mt. Zempoaltep-
ec, Oaxaca, Mexico. Goldman (1951:209) describes the locality
as somewhere between the elevations of 7700 and 10,500 feet
on the western slopes of the mountain above the Indian village
of Yacochi.

DIAGNOSIS. A large species of the subgenus Peromyscus
with unexpanded nasals; weakly beaded supraorbital ridges;
moderately complex teeth usually with undivided anterocone;
no ochraceous ventral color; no pectoral mammae; a partially
pouched stomach; a relatively long glans penis with large
spines, a long protractile tip, and undivided dorsal lappets; and
a cylindrical baculum with no great distal enlargement and a
large cartilaginous tip.

DISTRIBUTION. The “cloud forest,” at 1500 to well over
2000 m, on the northern slopes of the Sistema Montanoso
Poblano Oaxaqueno southeast of the gorge of the Rio
Quiotepec-Santo Domingo (Fig. 8). The gorge of the Rio Ca-
jonos splits the range into a segment on the Sierra de Juarez
and another on the slopes of Mt. Zempoaltepec. Limits of
range probably known.

VARIATION. I had available only 31 specimens of this spe-
cies and could detect no significant geographical variation. I
would not expect much in a species with such a small geo-
graphic range extending over apparently similar habitat. Mus-
ser (1969), however, demonstrated significant differences
between the two populations of P. lepturus occurring on the
same two mountain ranges that constitute the known range of
P. melanocarpus. Larger samples might provide evidence for
similar differences in the latter.

IDENTIFICATION. The combination of large size (Appen-
dix), weakly beaded supraorbital ridges, and the lack of pec-
toral mammae (Table 3) will distinguish P. melanocarpus from
most other Peromyscus that may share its habitat {P. boylii,
lepturus, oaxacensis, and chinanteco). Peromyscus thomasi oc-
curs in much the same habitat and averages larger in most di-
mensions, lacks beading on its supraorbital ridges, and has
heavier more complicated teeth, a glans penis much thicker
proximally than distally, and pectoral mammae. Peromyscus
mexicanus occurs at lower elevations on the same mountain
slopes and lacks beading on its poorly developed supraorbital
ridges (Fig. 1); in addition, its white tarsus and carpus contrast
with the dark ones of P. melanocarpus.

SPECIMENS EXAMINED (31). MEXICO. Oaxaca: To-

tontepec, 6 (NM); 104 km SW Tuxtepec, 1620 m, 2 (UM);

1 16 km SW Tuxtepec, 2000 m, 12 (UM); San Isidro, 5 (AM);
10 mi N Ixtlan de Juarez, 9300 ft, 2 (MSU); 23 mi S Valle
Nacional, 5600 ft, 2 (LACM), 2 (UM), 2 (CSULB).

Peromyscus melanurus

Black-tailed Mouse

SYNONYMY. Peromyscus megalops melanurus Osgood
1909:215. Peromyscus mexicanus putlaensis Goodwin 1964:5.
San Vicente, Oaxaca, Mexico.

HOLOTYPE. A moderately old adult male, skin and skull,
in good condition, NM 71385, collected 20 March 1895, below
Pluma, Oaxaca, Mexico, 3000 ft.

DIAGNOSIS. A large species of the subgenus Peromyscus
with unexpanded nasals; strongly beaded supraorbital ridges;
moderately complex teeth usually with undivided anterocone;
no ochraceous ventral color or buffy pectoral spot; no pectoral
mammae; a partially pouched stomach; a relatively long glans
penis with a short protractile tip, large spines, and undivided
dorsal lappets; and a cylindrical baculum with no great distal
enlargement and a moderate cartilaginous tip.

DISTRIBUTION. The humid “coffee belt” at medium ele-
vations (700 to 1900 m) on the Pacific slopes of the Sierra
Madre del Sur de Oaxaca (Fig. 8). May occur in similar hab-
itat in coastal Guerrero (perhaps on the slopes below the range
of P. megalops in the mountains west of Chilpancingo). Other-
wise, the known distribution probably conforms closely to the
actual distribution. Taken with P. mexicanus at Pluma Hidalgo
and Kilometer 212 on the Oaxaca-Puerto Escondido Road and
with P. megalops near Juquila.

VARIATION. The seven specimens from San Vicente, Oa-
xaca, that provide the basis for the name P. mexicanus putlaen-
sis Goodwin 1964 differ in no essential characters from other
specimens of P. melanurus. As seen in Goodwin’s (1964) photo-
graphs, the type specimen clearly shows the beaded supraorbi-
tal ridges characteristic of P. melanurus but not of P. mexi-
canus. I noted no striking individual or geographic variation in
this species. The small geographic range and relatively uniform
habitat suggest that little differentiation has occurred.

IDENTIFICATION. The beaded supraorbital ridges to-
gether with the lack of a buffy pectoral spot, lack of pectoral
mammae, and a long glans with short protractile tip (Table 3)
will separate P. melanurus from some other species of Per-
omyscus with which it may occur (P. boylii, evides, and melan-
ophrys). Of the others, only P. banderanus and megalops have
beaded supraorbital ridges, and both species lack pectoral
mammae. Peromyscus banderanus has a much narrower skull
with oval braincase, and its small glans penis (Hooper 1968,
Plate X) contrasts strongly with that of P. melanurus. Per-
omyscus megalops averages larger (Appendix) and has longer,
darker pelage frequently with a buffy pectoral spot, very
straight as opposed to slightly curved supraorbital ridges (Fig.
1), and a longer protractile tip on its glans penis (Fig. 3). Per-
omyscus mexicanus also lacks pectoral mammae but almost al-
ways possesses a buffy pectoral spot, has distinct but nonbeaded
supraorbital ridges, and has a relatively long protractile tip on
its glans penis.

SPECIMENS EXAMINED (237). MEXICO. Oaxaca: San
Vicente Putla, 9 (AM); 7 mi S Chicahuaxtla, 4700 ft, 1
(MSU); km 123 Tlaxiaco-Putla Rd., 4350 ft. 6 (CAS); La-
chao, 9 (AM); 20 mi S and 4 mi E Sola de Vega, 4800 ft,
12 (KU); Santa Rosa, 7 (UM); 5 mi NW Santa Rosa,

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 326:1-24

Huckaby: Species of Peromyscus mexicanus Group

13

14 (UM); 10 mi NW Santa Rosa, 7 (UM); 10 km E Nopala,
7200 ft, 16 (CAS); 9 mi W San Gabriel Mixtepec, 2 (CAS); 23
mi N San Gabriel, 6100 ft, 1 (MSU); 8 mi SSW Juchatengo,
6300 ft, 4 (MSU); 9 mi S Juchatengo, 5900 ft, 4 (MSU); 10 mi
S Juchatengo, 5350 ft, 11 (KU); km 178 Oaxaca-Puerto Es-
condido Rd., 6200 ft, 1 (CAS); km 183 Oaxaca-Puerto Escon-
dido Rd., 6000 ft, 35 (CAS); km 184.5 Oaxaca-Puerto
Escondido Rd., 6000 ft, 9 (CAS); 5759 ft, 3 (CAS);
km 187 Oaxaca-Puerto Escondido Rd., 5400 ft, 8 (CAS); km
193 Oaxaca-Puerto Escondido Rd., 4200 ft, 29 (CAS); km 195
Oaxaca-Puerto Escondido Rd., 3475 ft, 10 (CAS); km 212
Oaxaca-Puerto Escondido Rd., 2400 ft, 3 (CAS); Pluma
Hidalgo, 18 (NM); 4 mi S Jalatengo, 6 (UM); 22.5 mi N Can-
delaria, 1630 m, 3 (KU); Juquila Rd., 10 mi from Hwy. 131,
7000 ft, 9 (CAS).

Peromyscus zarhynchus

Long-nosed Mouse

SYNONYMY. Peromyscus zarhynchus Merriam 1898:117.
Peromyscus zarhynchus cristobalensis Merriam 1898:1 17. San
Cristobal de las Casas, Chiapas, Mexico.

HOLOTYPE. An adult female, skin and skull, in good con-
dition, NM 76117, collected 26 October 1895, on mountain
slopes above village of Tumbala, Chiapas, Mexico, 5500 ft.

DIAGNOSIS. A very large species of the subgenus Per-
omyscus with unexpanded nasals; nonbeaded supraorbital
ridges; moderately complex teeth usually with undivided
anterocone; basically brown dorsal pelage with no ochraceous
ventral color; no pectoral mammae; a fully pouched stomach; a
relatively long glans penis with a long protractile tip, large
spines, and undivided dorsal lappets; and a cylindrical baculum
with no great distal enlargement and a large cartilaginous tip.

DISTRIBUTION. “Cloud forests” of higher elevations of
mountains of north central Chiapas (Fig. 9). Limits of range
probably known. Collections made on the ridges north and east

Figure 9. Geographic distributions of Peromyscus zarhynchus (tri-
angles), P. guatemalensis (closed circles), P. grandis (open circles),
and P. gymnotis (squares). Vertical hatching indicates known range of
P. mexicanus.

of Comitan lack P. zarhynchus. and these ridges probably do
not have the appropriate habitat. Taken with P. mexicanus at
Tumbala.

VARIATION. Peromyscus zarhynchus apparently occurs
only as several disjunct populations in northern Chiapas. Speci-
mens from the type locality are, on the average, darker in color
than those from elsewhere, perhaps reflecting the wetter en-
vironment near Tumbala. I can appreciate little other geo-
graphic variation in the specimens at hand and do not recom-
mend the recognition of subspecies.

IDENTIFICATION. The large size (Appendix) of P.
zarhynchus readily separates it from other Peromyscus known
to inhabit its restricted range ( P . boylii, oaxacensis, lophurus,
and mexicanus). Peromyscus zarhynchus differs from the geo-
graphically adjacent P. guatemalensis in its slightly larger over-
all size and brown as opposed to grayish dorsal pelage.

SPECIMENS EXAMINED (190). MEXICO. Chiapas:
San Cristobal, 9500 ft, 21 (NM), 1 (AM); 4 mi W San
Cristobal, 1 (AM); 8 mi SE San Cristobal, 2 (UM); 10 mi SE
San Cristobal, 2300 m, 2 (UM); Cerro Tzontehuitz, 2900 m,
20 (KU), 41 (UM); Tumbala, 15 (NM), I (AM); vicinity of
Pueblo Nuevo, 50 (AM); 1 mi N Pueblo Nuevo, 5500 ft, 11
(UM); 5 mi N Pueblo Nuevo, 3 (UM); 11.6 mi N Pueblo
Nuevo, 3 (UM); 4 mi SE Rayon, 19 (MSU).

Peromyscus guatemalensis

Guatemalan Mouse

SYNONYMY. Peromyscus guatemalensis Merriam
1898:1 18. Peromyscus altilaneus Osgood 1904:74. Todos San-
tos, Huehuetenango, Guatemala.

HOLOTYPE. An adult male, skin and skull, in good condi-
tion, NM 76861, collected 31 December 1895, on slopes above
village of Todos Santos, Huehuetenango, Guatemala, 10,000 ft.

DIAGNOSIS. A large species of the subgenus Peromyscus
with unexpanded nasals; nonbeaded supraorbital ridges; moder-
ately complex teeth usually with undivided anterocone; basi-
cally gray dorsal pelage with no ochraceous ventral color; no
pectoral mammae; a fully pouched stomach; a relatively long
glans penis with a long protractile tip, large spines, and un-
divided dorsal lappets; and a cylindrical baculum with no great
distal enlargement and a large cartilaginous tip.

DISTRIBUTION. Humid forests from 1300 to over 3000 m
in the mountains of southern Chiapas and southwestern Guate-
mala (Fig. 9). Limits of range probably known. Taken with P.
gymnotis at Finca Helvetia, Guatemala, and within 5 miles of
P. mexicanus at El Injerto, Guatemala.

VARIATION. Osgood (1904) named Peromyscus altilaneus
based on one specimen from Todos Santos, Guatemala, the
same locality from which Merriam (1898) had named P.
guatemalensis. In his description, Osgood noted that P. al-
tilaneus differed from P. guatemalensis mainly in its smaller
size. The holotype of P. altilaneus has dark, subadult pelage on
the dorsum of its head and relatively unworn teeth. Carleton
and Huckaby (1975) speculated that the type of P. altilaneus
might represent a mismatched skin and skull. Due to the diffi-
culty of proving this for forms that differ as slightly as these, I
now think that synonymization of P. altilaneus represents the
best solution. After direct comparison of the two types to one
another and with the 1 1 other specimens taken with them, I
believe the type of P. altilaneus Osgood 1904 represents a sub-
adult individual of the same species as the type of P. guatema-

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 326:1-24

14

Huckaby: Species of Peromyscus mexicanus Group

lensis Merriam 1895. The other specimens adequately bridge
the size gap between the two holotypes.

1 can appreciate no geographic variation within the small
known range of P. guatemalensis. The disjunct nature of the
populations suggests the possibility of such, particularly since
P. guatemalensis contacts the much smaller P. gymnotis in part
of its range and the only slightly smaller P. mexicanus
elsewhere.

IDENTIFICATION. The larger size of P guatemalensis
(Appendix) readily separates it from some species of Per-
omyscus known to inhabit its range ( P . boylii, oaxacensis,
lophurus, and gymnotis). At the western edge of the range of P.
guatemalensis, P. mexicanus averages smaller in most dimen-
sions and has a brownish dorsal pelage. The populations here
assigned to P mexicanus from the departments of
Chimaltenango, Escuintla, and Sacatepequez, Guatemala, re-
semble P. guatemalensis closely in size and to a lesser extent in
color. 1 can distinguish between them on average differences
only. Specimens assigned to P. guatemalensis have a decided
grayish cast to the pelage and have a strong tendency toward a
bicolored tail. Specimens assigned to P mexicanus have dark
brown pelage and a monocolored or ventrally blotched tail with
no tendency toward bicoloration. In contrast to the geograph-
ically adjacent P. zarhynchus, P. guatemalensis exhibits a gray
rather than brown dorsal pelage. Peromyscus guatemalensis
differs from the geographically adjacent P grandis in its
smaller dimensions and gray rather than brown dorsal pelage.

SPECIMENS EXAMINED (502). GUATEMALA.
Chimaltenango : 5 mi N Tecpan, 4 (UM). El Quiche: Cotzal, 3
(UM). Huehuetenango: Barillas, 25 (UM); San Mateo, 4
(AM); Yayquich, 2900 m, 17 (UM); 2 mi S San Juan Ixcoy,
9500 ft, 61 (KU); 3.5 mi S San Juan Ixcoy, 10,120 ft, 8 (KU);
27 mi N Chiantla, 4 (NM); Todos Santos, 12 (NM), 1 (AM);
5.5 mi N and 1 mi E Chiantla, 9700 ft, 3 (KU); Hda. El ln-
jerto, 1600 m, 38 (UM). Quezaltenango: Finca Helvetia, 5500
ft, 8 (NM); Calel, 22 (NM); 1 mi S Quezaltenango, 2 (NM);
Zunil, 1 1 (NM); Volcan Santa Maria, 1 1 (NM). San Marcos:
Volcan Tajumulco, 10,000 ft, 11 (UM), 10 (FM); 13.5 mi N
and 0.75 mi E San Marcos, 9500 ft, 10 (KU); 4 mi W San
Marcos, 4 (AM). Solola: 3.2 mi E Panajachel, 1 (NM); Volcan
San Lucas, 51 (AM); San Lucas, 75 (AM). Totonicapan: 8 mi
S Momostenango, 9000 ft, 7 (KU); 10 mi E and 4 mi S
Totonicapan, 10,000 ft, 3 (KU); Cumbre Maria Tucum, 2770
m, 4 (UM). MEXICO. Chiapas: Pinabete, 8 (NM); Volcan
Tacana, 8 km N Union Juarez, 2000 m, 32 (KU); Cerro Mozo-
tal, 2850 m, 31 (UM); Triunfo, 1950 m, 21 (UM).

Peromyscus grandis

Giant Mouse

SYNONYMY. Peromyscus grandis Goodwin 1932:4.

HOLOTYPE. An old adult female, skin and skull, in good
condition, AM 79341, collected 16 June 1928, Finca Concep-
cion, ca. 3 miles S San Miguel Tucuru, Alta Verapaz, Guate-
mala, 3750 ft.

DIAGNOSIS. A very large species of the subgenus Per-
omyscus with unexpanded nasals; nonbeaded supraorbital
ridges; moderately complex teeth usually with undivided
anterocone; basically brown dorsal pelage occasionally with
brown color extending onto venter but no ochraceous ventral
color; no pectoral mammae; a fully pouched stomach; a rela-
tively long glans penis with a long protractile tip, large spines,
and undivided dorsal lappets; and a cylindrical baculum with

no great distal enlargement and a large cartilaginous tip.

DISTRIBUTION. Forests of northeastern Baja Verapaz and
adjacent Alta Verapaz, Guatemala (Fig. 9). Limits of range
unknown. Allopatric to all other species here considered.
Southern range probably limited by drier habitat in central
Baja Verapaz and the gorge of the Rio Chixoy o Negro.

VARIATION. Too few specimens, all taken within a small
area, preclude any analysis of geographic variation of this form
at present.

IDENTIFICATION. At present only known to occur with P.
oaxacensis, a much smaller species. Large size (Appendix) sep-
arates this species from the geographically adjacent P. mex-
icanus. Its larger size in most dimensions together with its
brown pelage distinguish it from the smaller, grayer P. guate-
malensis. In the genus, only P. thomasi (subgenus Megadon-
tomys) equals the dimensions of P. grandis and only P. flavidus
and pirrensis (subgenus Isthmomys ) really exceed it.

SPECIMENS EXAMINED (30). GUATEMALA. Alta
Verapaz: Hda. Concepcion, 3 (AM), 18 (UM); 11.5 mi S and
3.5 mi E Coban, 4925 ft, 4 (KU). Baja Verapaz: 2 mi W Pu-
rulha, 4950 ft, 3 (KU); 12.5 mi N Salama, 2 (NM).

Peromyscus gymnotis

Naked-eared Mouse

SYNONYMY. Peromyscus gymnotis Thomas 1894:365.
Peromyscus allophylus Osgood 1904:71. Huehuetan, Chiapas,
Mexico.

HOLOTYPE. A young adult male, in fluid with extracted
and cleaned skull. British Museum 86.5.13.4, collected in
“Guatemala” by Bernoulli.

DIAGNOSIS. A moderately small species of the subgenus
Peromyscus with unexpanded nasals; nonbeaded supraorbital
ridges; moderately complex teeth usually with undivided
anterocone; basically dark brown dorsal pelage with whitish
ventral pelage; no pectoral mammae; a fully pouched stomach;
a relatively long glans penis with a long protractile tip, large
spines, and undivided dorsal lappets; and a cylindrical baculum
with no great distal enlargement and a large cartilaginous tip.

DISTRIBUTION. Foothills and adjacent coastal plain of
southern Chiapas and southwestern Guatemala (Fig. 9). Limits
of range probably known and likely determined by the presence
of P. mexicanus to the north and south of it on the coastal plain
and by the presence of P. guatemalensis in the mountains above
it.

VARIATION. I can detect no geographic variation within
the small known range of this species.

IDENTIFICATION. The known range of P. gymnotis con-
tacts the known range of five other Peromyscus. P. gymnotis
differs from P guatemalensis and mexicanus primarily in its
smaller size in most dimensions (Appendix) and from P.
guatemalensis in having short brown rather than long gray pel-
age with monocolored or ventrally blotched as opposed to usu-
ally bicolored tail. Positive separation of P. mexicanus and P.
gymnotis may frequently require direct comparison with known
specimens. Peromyscus boylii differs in having an hourglass-
shaped interorbital area and usually a bicolored tail. Per-
omyscus oaxacensis normally has very reddish pelage in adults
and heavier teeth with more accessory lophs and cusps. Per-
omyscus lophurus has a very hairy tail usually with a tuft on
the end. In addition, the glans penis of P. gymnotis differs con-
siderably from those of P. boylii, oaxacensis, and lophurus (see
Hooper 1958 and Hooper and Musser 1964 for figures and de-

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 326:1-24

Huckaby: Species of Peromyscus mexicanus Group

15

scription), and the females of these three species possess a pair
of pectoral mammae.

SPECIMENS EXAMINED (208). GUATEMALA.
Escuintla: 11 mi SW Escuintla, 615 ft, 2 (FM); 4 mi W Es-
cuintla, 880 ft, 8 (KU); Finca El Salto, 4 (UM); Finca El
Zapote, 2400 ft, 1 (FM); Finca Santa Cristina, 330 ft, 6 (FM);
Tequisate, 4 (FM); Astillero, 65 ft, 5 (KU); Hda. El Rosario,
950 m, 12 (UM); San Jose, 6 (NM); 45 km S Guatemala, 2950
ft, 1 (KU); 48 km S Guatemala, 2300 ft, 1 (KU); 50 km S
Guatemala, 2000 ft, 6 (KU); 52 km S Guatemala, 1650 ft, 1 1
(KU). Quezaltenango: Finca Helvetia, 3500 and 5500 ft, 7
(NM). Retalhuleu: 20 km NW Retalhuleu, 3 (TCWC). San
Marcos: Hda. California, 12 (AM); Finca El Porvenir, 3700 ft,

1 (UM), 13 (FM). Suchitepequez: Finca El Cipres, 1 (AM); 12'
mi NE Mazatenango, 19 (NM). MEXICO. Chiapas: Huehue-
tan, 4 (NM); Finca Esperanza, 1 (UM); Mt. Ovando, I (UM);
Pijijiapan, 9 (UM); Mapastepec, 20 (UM); 13 km N Huixtla,
14 (UM); Tapachula, 2 (AM); 7 mi ENE Tapachula, 7 (KU);

1 5 mi S Tapachula, 1 (AM); Talisman, 6 (AM); Chicharras, 20
(NM).

Peromyscus mexicanus

Mexican Mouse

SYNONYMY. Hesperomys mexicanus Saussure 1860:103.
Peromyscus mexicanus, Thomas 1894:364. Hesperomys
nudipes J.A. Allen 1891:213. La Carpintera, Cartago, Costa
Rica. Peromyscus nudipes, Thomas 1894:365. Peromyscus
mexicanus totontepecus Merriam 1898:120. Totontepec, Oa-
xaca, Mexico. Peromyscus mexicanus orizabae Merriam
1898:121. Orizaba, Veracruz, Mexico. Peromyscus tehuan-
tepecus Merriam 1898:122. Near Tehuantepec (8 mi up Rio
Tehuantepec, Cerro Giengola), Oaxaca, Mexico. Peromyscus
cacabatus Bangs 1902:29. Boquete, Chiriqui, Panama. Per-
omyscus banderanus angelensis Osgood 1904:69. Puerto Angel,
Oaxaca, Mexico. Peromyscus mexicanus teapensis Osgood
1904:69. Teapa, Tabasco, Mexico. Peromyscus nicaraguae J.A.
Allen 1908:658. Matagalpa, Nicaragua. Peromyscus mex-
icanus philombrius Dickey 1928:3. Los Esesmiles, Chal-
atenango, El Salvador. Peromyscus mexicanus salvadorensis
Dickey 1928:4. Mt. Cacaguatique, San Miguel, El Salvador.
Peromyscus guatemalensis tropicalis Goodwin 1932:3.
Chimoxan, Alta Verapaz, Guatemala. Peromyscus nudipes ori-
entalis Goodwin 1938:3. El Sauce Peralta, Cartago, Costa
Rica. Peromyscus nudipes hesperus Harris 1940:1. Hacienda
Santa Maria, Guanacaste, Costa Rica. Peromyscus banderanus
sloeops Goodwin 1955:2. Rio Mono Blanco, Oaxaca, Mexico.
Peromyscus megalops azulensis Goodwin 1956:6. Cerro Azul,
Oaxaca, Mexico. Peromyscus banderanus coatlanensis Good-
win 1956:7. Agua Zarca, Oaxaca, Mexico.

HOLOTYPE. A mounted skin with separate skull, Geneva
Museum 510/95, Mirador, Veracruz, Mexico (locality re-
stricted to 10 km E Mirador by Dalquest 1950:8).

DIAGNOSIS. A large species of the subgenus Peromyscus
with unexpanded nasals; nonbeaded supraorbital ridges; moder-
ately complex teeth usually with undivided anterocone (divided
anterocone more common from Guatemala south); light to dark
brown dorsal pelage and whitish ventral pelage; no pectoral
mammae; a fully pouched stomach; a relatively long glans pe-
nis with a long protractile tip, large spines, and undivided dor-
sal lappets; and a cylindrical baculum with no great distal
enlargement and a large cartilaginous tip.

DISTRIBUTION. Tropical lowlands of the Gulf side of
Mexico from San Luis Potosi south through Veracruz and

northern Oaxaca into the Isthmus of Tehuantepec; Pacific coast
of Mexico at least from the Guerrero-Oaxaca border south to
the vicinity of Tonala, Chiapas; the northern and eastern low-
lands of Chiapas and adjacent Tabasco probably eastward
through the foothills of the Guatemalan highlands to the
Puerto Barrios area; central valley of Chiapas and adjacent
Guatemala; southeastern volcanoes and highlands of
Guatemala throughout El Salvador, Honduras, and Nicaragua
south to the moderate to high elevations of Costa Rica and ex-
treme western Panama (Fig. 6). Apparently absent from El Pe-
ten, Guatemala, and Belice. Limits of range probably known
except possibly northern extension on Pacific coast in Guerrero
and eastern extension in highlands of western Panama. Taken
sympatrically or closely parapatrically with all other species
herein considered except P. ochraventer, grandis, yucalanicus,
and megalops (taken lower on same slopes as P. megalops).

VARIATION. Musser (1969) provided evidence that the
named forms P. banderanus angelensis Osgood, P. b. coatlanen-
sis Goodwin, P. b. sloeops Goodwin, and P. guatemalensis trop-
icalis Goodwin consist solely of specimens of the wide-ranging
P mexicanus. I have examined the specimens concerned and
agree completely. In addition, I would arrange P. megalops
azulensis Goodwin as a species-level synonym of P. mexicanus
(Saussure). Goodwin ( 1956) described this form from one spec-
imen from the mountains in the Isthmus of Tehuantepec, but
later (1969:191) he referred another specimen from further
west in Oaxaca to it. I have examined both specimens and con-
sider them fairly ordinary examples of P mexicanus. Neither
exhibits the beaded supraorbital ridges characteristic of P.
megalops.

Overall size varies locally in P. mexicanus (Appendix) with
no particular macrogeographic trend. I can detect geographic
variation in no other characters except dorsal coloration. Ac-
cordingly, I will limit my admittedly cursory discussion of geo-
graphic variation in this species to the variation in dorsal
coloration.

Specimens of P. mexicanus from warm areas have a pale
buffy brown dorsum, and those from moist warm areas have a
dark reddish brown dorsum. Specimens from cool areas tend to
have a grayish cast to the pelage, which ranges from light gray
in dryer areas to nearly black in humid regions. The palest
specimens come from the Pacific coastal region of Oaxaca —
from near Tehuantepec west to the Guerrero border. Somewhat
darker specimens come from the dry interior valley of Chiapas
and the southeastern highlands of Guatemala and adjacent El
Salvador. All of these areas have a long dry season and open
scrubby deciduous vegetation with some open savannah. The
mice exhibit a seasonal difference in pelage color in that the
wet season pelage appears considerably darker than that of the
dry season. I could not satisfactorily determine whether this
difference results from molting or from wear and fading of the
dark, wet season pelage as the dry season progresses.

The darkest specimens of P. mexicanus come from the north-
ern slopes of the Sistema Montanoso in Oaxaca, the northern
lowlands of Chiapas and adjacent Tabasco, and Costa Rica.
These areas have a short dry season, high rainfall, and ever-
green forests. Specimens intermediate in color occur in the
other areas, apparently associated with intermediate habitats in
regard to climate and type of vegetation.

Most mammalian taxonomists, in dealing with species such
as P. mexicanus that exhibit considerable variation in size and/
or color over a wide range of habitats, have broken these spe-

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 326:1-24

16

Huckaby: Species of Peromyscus mexicanus Group

cies up into numerous subspecies. I have chosen not to do so
primarily because of the apparent complexity of the variation,
and because I do not think that I had enough large samples to
adequately describe the morphometric variation. Size does not
correlate well with color in this species, which would necessi-
tate the recognition of a fairly large number of subspecies. In
fact, most of the large samples available to me differ from one
another on the basis of some combination of characters. This
microgeographic variation seems best described, for the pres-
ent, as I briefly attempted above and not further complicated
with the addition of many new and poorly defined trinomials.

IDENTIFICATION. The appropriate species accounts con-
tain the characteristics that distinguish P. mexicanus from
other species herein considered. The known ranges of P.
aztecus, boylii, oaxacensis, evides, melanophrys, banderanus,
leucopus, and flavidus overlap that of P. mexicanus. Per-
omyscus aztecus, evides, and oaxacensis closely resemble one
another and may represent geographical variants of the same
species. All three differ from P. mexicanus in having three pairs
of mammae, a short glans penis with a flared and scalloped dis-
tal end that lacks dorsal lappets (Hooper 1958, Plate VII), usu-
ally smaller dimensions, heavier more complicated teeth,
usually more parallel tooth rows and a decidedly reddish cast to
old adult pelage. Peromyscus boylii and P. leucopus have much
smaller dimensions in most measurements than P. mexicanus,
three pairs of mammae, and usually very hourglass-shaped in-
terorbital areas. Peromyscus melanophrys has a very long and
relatively more hairy tail, three pairs of mammae and a rela-
tively shorter rostrum than P. mexicanus. Peromyscus ban-
deranus has a narrower skull with very strongly beaded
supraorbital ridges, a short, relatively simple glans penis
(Hooper 1958, Plate X), and a completely naked as opposed to
slightly hairy heel. Peromyscus flavidus has much larger di-
mensions in most measurements.

SPECIMENS EXAMINED (2381). COSTA RICA.
Alajuela: Palmira de Zarcero, 2 (UM); Volcan Poas, Rio
Poasito, 2000 m, 10 (UM); Lajas Villa Quesada, 22 (AM);
Zapote, 4 (AM); Tapesco, 1 (AM). Cartago: San Ramon de
Tres Rios, 1 (NM); El Sauce Peralta, 3 (AM), 8 (NM); Cer-
vantes, 13 (NM), 2 (AM); La Carpintera, 1 (AM); Volcan
Irazu, 9400 ft, 30 (AM); 2350 m, 7 (UM); 2850 m, 3 (UM);
Rancho de Rio Jimenez, 5 (AM); Islo Nievo Irazu, 1 (AM); El
Muneco, 8 (UM); Las Vueltas, 5 (UM); Santa Teresa Peralta,
8 (AM); Juan Vinas, 4 (AM); 3 mi SE Turrialba, 602 m, 11
(UM); Tapanti, 2100 m, 3 (UM); Moravia, 1116 m, 15 (UM).
Guanacaste: Hda. Santa Maria, 5 (UM); Cerros de San Juan,
2 (UM). Puntarenas: Monteverde, 1400 m, 1 8 (UM). San Jose:
San Joaquin de Dota, 1 (UM); Los Higuerones Escazu, 20
(AM); San Jose, 1 (AM); Cerro de la Muerte, 5 (UM). EL
SALVADOR. Chalatenango: Vi mi N San Ignacio, 3200 ft, 3
(MVZ); Los Esesmiles, 1 (UCLA), 6 (UM), 98 (MVZ); San
Jose del Sacare, 3600 ft, 9 (MVZ). La Union: Pine Peaks, 31
(MVZ). San Miguel: Mt. Cacaguatique, 1 (UCLA), 6 (UM),
106 (MVZ); Volcan de San Miguel, 33 (MVZ). San Vicente:
Hacienda El Carmen, Volcan San Vicente, 3300 ft, 10 (MVZ).
Santa Ana: V2 mi NE Cerro del Aguila, 5900 ft, 7 (MVZ);
Cerro de los Naranjos, Volcan de Santa Ana, 5800-6150 ft, 60
(MVZ). Sonsonate: Hacienda Chilata, 6 (UM), 43 (MVZ).
GUATEMALA. Alta Verapaz: Chimoxan, 23 (AM). Es-
cuintla: Volcan Agua, 1200 m, 1 (UM); 1520 m, 1 (UM); 1990
m, 1 (UM); 2100 m, 12 (UM); 2300 m, 22 (UM); 1800 m, 16
(UM). Guatemala: 4 mi S Guatemala, 4700 ft, 10 (KU); 5 mi

S Guatemala, 4950 ft, 1 8 (KU); 6 mi S Guatemala, 4680 ft, 1 3
(KU); 7 mi S and 6 mi E Guatemala, 5800 ft, 5 (KU); 24 km S
Guatemala, 4100 ft, 18 (KU). Huehuetenango: Hda. Guailia,

14 (UM); Hda. El Reposito, 5 (UM); Nenton, 4 (NM); Jac-
altenango, 33 (NM); Chanquejelve, 7 (AM). Izabal: Escobas,

3 (FM). Jalapa: 6 mi E Mataquescuintla, La Soledad Grande,
8600 ft, 9 (FM). Jutiapa: 1 mi SE Jutiapa, 2950 ft, 11 (FM).
Sacatepequez: Finca San Rafael, 7000 ft, 33 (FM), 1 (UM).
Santa Rosa: Finca El Progresso, 10 (FM). HONDURAS. Dis-
trito Central: Cantoral, 18 (AM), 1 (UM); La Flor Archaga, I
(UM); La Cueva Archaga, 5 (AM); Hatillo, 1 (UM); Rancho
Quemado, 2 (UM). Cortes: La Lima, 9 (AM). Francisco Mor-
azan: El Caliche Orica, 9 (AM); Cerro Santa Maria, 4 (UM);
Cerro Uyuca, 10 (UM). La Paz: Muye, 37 (AM). Lempira:
Cerro Puca, 60 (AM); Monte Linderos, 12 (AM); Cementerio,
5 (AM); Puca, 11 (AM); Las Flores, 2 (AM). Ocotepeque:
Monteverde, 5 (AM). Olancho: 40 km E Catacamas, 1
(TCWC). Santa Barbara: Santa Barbara, 3 (AM). MEXICO.
Chiapas: Villa Corzo, 4 (AM); Tres Picos, 4 (AM); Morelia,
4700 ft, 2 (KU); 4 mi S Altamirano, 10 (KU); 2 mi W Agua
Escondido, 2 (KU); 2 mi SE Las Tasas, 1 (KU); El Paraiso,
4050 ft, 1 I (KU); Finca San Antonio, 1 (KU); Tonina Ruins, 7
(KU); La Victoria, 3 (KU); San Vicente, 3 (NM); Canjob, 14
(NM); San Bartoleme, 7 (NM); Catarina, 1300 m, 12 (UM);
Prusia, 1100 m, 21 (UM); Bochil, 1320 m, 3 (UM); 17 mi W
Bochil, 14 (AM); 1 1 mi E Bochil, 4 (AM); Palenque, 1 1 (KU);
El Salto, 2 (NM); Tumbala, 1 (NM); 5 km S Solusuchiapa, 6
(UM), 2 (KU); 4 mi S Pichucalco, 3 (KU); mts. near Tonala, 8
(NM); 9 mi SE, 10 mi NE Tonala, 21 (LACM); Ocozocoautla,

4 (NM), 3 (UM); Cinco Cerros, 20 (AM); Cintalapa, 10
(AM); Cerro Pecho Blanco, 6 (AM); 3 km E Riso de Oro, 3
(UM); 5 mi N Berriozabal, 26 (UM); Tuxtla, 4 (NM), 13
(KU); 1 mi N Tuxtla Gutierrez, 14 (UM); 5 mi NW Tuxtla
Gutierrez, 34 (KU); 10 mi W Tuxtla Gutierrez, 2 (UM); 11
km W Tuxtla Gutierrez, 1 (UM). Guerrero: 9 mi SE Omete-
pec, 1 (NM). Oaxaca: Comaltepec, 2 (NM); Choapam, 1
(NM); Vista Hermosa, Tarabundi, 20 (AM); Puerto Elijio, 1
(AM); 4 mi S Valle Nacional, 2600 ft, 8 (MSU); Totontepec,
10 (NM); Estancia, 1 (AM); Ixcuintepec, 39 (AM); 12 de
Julio, 8 (AM); Lagunas, 2 (NM), 6 (FM); 2 mi S Tollocita, 26
(KU); Santa Efigenia, 13 (NM); Guichicovi, 8 (NM); Santo
Domingo, 19 (NM); 7 mi E Santa Maria Chimalapa, 5 (AM);

15 mi N Tapanatepec, Cerro Baul, 14 (AM); 20 mi N Tap-
anatepec, 2 (AM); 22 mi N Tapanatepec, 1 (AM); 25 mi N
Tapanatepec, 6 (AM); Rio Mono Blanco, 4 (AM); Tap-
anatepec, 1 (AM); Zanatepec, 8 (AM); mts. N Zanatepec,
5000 ft, 35 (AM); Ubero, 1 (AM); 8-9 mi S Veracruz border, 6
(AM); Mogone, 1 (AM); Laguna Sol y Luna, 3 (AM); Arroyo
Encantado, 1 (AM); Chivela, 1 (AM); Nizanda, 5 (AM);
Cerro Azul, 25 mi NW Zanatepec, 2 (AM); 18 mi N Matias
Romero, 1 (AM); Arroyo Cardon, 3 (AM); 8 mi up Rio
Tehuantepec, Cerro Giengola, 5 (NM); 10 mi W Tehuantepec,
2 (UM); Santiago Lachiguiri, 7000 ft, 6 (AM); Agua Zarca, 5
(AM); Santa Lucia, 44 (AM); Cerro Arenal, 5 (AM); San An-
tonio, 1 (AM); Arroyo San Juan, 3 (AM); Tenango, 7 (AM);
Tres Cruces, 5 (AM); Las Cuevas, 2 (AM); Escuranos, 4
(AM); Mixtequilla, 1 (AM); Cerro San Pedro, 10 (AM); Sal-
azar, 3 (AM); Guigovalaga, 3 (AM); Media Loma, 3 (AM); El
Companario, 1 (AM); Santa Maria Ecatepec, 7 (AM); San
Felipe Lachillo, 1 (AM); Puerto Angel, 19 (NM); Chacalapa,
650 ft, 6 (KU); 4 mi S Candelaria, 1300 ft, 6 (KU); 3 mi S
Candelaria, 1200 ft, 12 (KU); Pluma Hidalgo, 1 (NM); Escon-

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 326:1-24

Huckaby: Species of Peromyscus mexicanus Group

17

dido Bay, 3 (UM); Nopala, 3 (AM); 2 mi E San Gabriel Mix-
tepec, 10 (AM); 5 mi E Rio Grande, 2 (MSU); 10-11 mi N
Puerto Escondido, 2000 ft, 4 (MSU); Pinotepa, 1 (NM); Rio
Verde, 1 (AM); 4 mi SSW Cacahuatepec, 1000 ft, 2 (MSU); 2
mi SE Jamiltepec, 900 ft, 1 (MSU). Puebla: 6 km E, I km N
Villa Juarez, 17 (KU); Metlaltoyuca, 7 (NM); Pahuatlan,
1 100 m, 7 (UM). San Luis Potosi: 3 km N Tamazunchale, 18
(LSU); 2 mi W Tamazunchale, 2 (UM). Tabasco: 12 mi S
Villa Hermosa, 1 (KU), 6 mi S Cardenas, 4 (KU); 5 mi SE
Macuspana, 8 (KU); Montecristo, 2 (NM); vicinity of Teapa,
25 (KU), 34 (LSU); Teapa, 17 (NM). Veracruz: Papantla, 8
(NM); Xico, 7 (FM); Barranca Texcolo at Puente Texcolo, 1 1
(UM); 2 mi N Teocelo, 1000 m, 13 (UM); Teocelo, 2 (NM);
0.5 mi E, 3 mi NW Plan del Rio, 2 (UM); Carrizal, 9 (NM);
Mirador, 4 (NM); Orizaba, 20 (NM); Barranca Metlac, 4
(UM); 9 km WNW Potrero Viejo, 1700 ft, 9 (UM); Cordoba,
9 (AM); Presidio, 23 (UM); Motzorongo, 11 (NM); San An-
dres Tuxtla, 4 (NM); Volcan Tuxtla, 6 (NM); Catemaco, 5
(NM); Lake Catemaco, 31 (AM); 15 mi N San Andres, 19
(AM); 18 mi N San Andres, 17 (AM); Achotal, 18 (FM); 2
km SSW Tenochtitlan, 9 (UM); Suchil, 7 (AM); Pasa Nueva,
9 (AM). NICARAGUA. Matagalpa: Matagalpa, 20 (AM);
Santa Maria de Ostuma, 1400 m, 26 (UM). PANAMA. Chiri-
qui: Boquete, 4 (UM), 26 (FM); Rio Chiriqui Viejo, 1700 m, 7
(UM); Quebrada El Chiquero, 1800 m, 2 (UM); Cerro Punta,
1825 m, 6 (UM); Volcan de Chiriqui, 2000 m, 7 (UM).

SUMMARY

The Peromyscus mexicanus species group of Hooper (1968)
consists of 12 species with largely allopatric or parapatric
ranges. Peromyscus mexicanus has by far the widest distribu-
tion, and its range contacts or closely approaches the ranges of
all the other species. Peromyscus mexicanus occurs in lowland
areas of Mexico and northern Central America and occurs
from moderate to high elevations in southern Central America.
Peromyscus mexicanus ranges into highland areas only in
places where the other species of this group do not occur,
mainly in the southern part of Central America.

Of the other 1 1 species, only P yucatanicus, stirtoni, and
gymnotis occur at low elevations. Peromyscus yucatanicus oc-
curs on the Yucatan Peninsula where its range does not contact
that of any other member of this group. Peromyscus stirtoni
occurs in low valleys on the Pacific slopes of Guatemala, El Sal-
vador, and Honduras where it comes in contact with P. mex-
icanus alone of the species herein considered. Peromyscus
gymnotis occurs on the Pacific coastal plain of eastern Chiapas
and western Guatemala; this species contacts P. mexicanus on
its western and eastern edge and contacts P. guatemalensis on a
broad front in the mountains to the north.

The remaining eight species inhabit upland areas within or
adjacent to the northern part of the range of P mexicanus. Per-
omyscus ochraventer has a small known range immediately ad-
jacent to the northern known limit of the range of P. mexicanus
on the eastern side of Mexico. Peromyscus furvus occurs on the
eastern slopes of the Sierra Madre Oriental from San Luis Po-
tosi south to Oaxaca. Peromyscus mexicanus occurs lower
down the same slopes. Peromyscus melanocarpus occurs on the
northeastern slopes of the mountains of northern Oaxaca where
its range appears as an extension of the range of P. furvus (the
deep canyon of the Rio Santo Domingo/Quiotepec separates
the two), and where it also occurs at higher elevations than the

more lowland P. mexicanus. Peromyscus megalops occupies
disjunct patches on the Pacific slopes of the mountains of south-
ern Oaxaca and Guerrero. In Oaxaca, P. megalops contacts P.
melanurus, which occurs lower down these same Pacific slopes.
Peromyscus melanurus occupies a range on these Pacific slopes
of Oaxaca sandwiched between P. megalops at higher and P.
mexicanus at lower elevations.

Further south, P. zarhynchus occupies part of the higher ele-
vations of the mountains of northern Chiapas. Peromyscus
mexicanus occurs at lower elevations on these same mountains.
P. guatemalensis occurs in the higher elevations of the western
part of the mountains of southern Guatemala and the moun-
tains of southern Chiapas. Peromyscus mexicanus contacts P.
guatemalensis at the eastern, northern, and western edge of this
area but not in the southern part. Peromyscus gymnotis oc-
cupies the coastal plain and slopes below the southern edge of
the range of P. guatemalensis. Finally, P. grandis occupies a
small range in the northeastern part of the main highland mass
of southern Guatemala; P. grandis does not contact any of the
other species herein considered, but P. mexicanus occurs in the
adjacent lowlands to the north.

RESUMEN

Para estimar los limites dentro del grupo de Peromyscus mex-
icanus de Hooper (1968), modificado por Musser (1969, 1971)
compare mas de 4000 especimenes a traves de tecnicas conven-
cionales y multivariadas. Base la comparacion en una serie de
caracteres elegidos del craneo, del sistema reproductive mas-
culino y de la morfologia externa de especimenes preservados.

Hooper (1968) reconocio 14 formas nominales (P. al-
lophylus, furvus. grandis. guatemalensis. gymnotis. latirostris,
megalops, melanocarpus. mexicanus, nudipes, ochraventer, stir-
toni, yucatanicus, and zarhynchus ). Sigo a Hall (1971 ) en iden-
tificar a P. latirostris con furvus y a Musser (1971) en
identificar P. allophylus con gymnotis. Reconozco todo el resto
como especies en si con excepcion de P. nudipes que hago si-
nonimo de P. mexicanus. Tambien reconozco P. melanurus (an-
terioramente considerado como una susespecie de megalops )
como una especie en si.

ACKNOWLEDGMENTS

This report derives mostly from a dissertation submitted to the
Horace H. Rackham School of Graduate Studies, The Univer-
sity of Michigan, 1973, in partial fulfillment for the degree of
Doctor of Philosophy in Zoology. 1 wish to thank C.W. Hib-
bard, E.T. Hooper, A.G. Kluge, D.M. Lay, G.R. Smith, and
R.W. Storer for services rendered in this regard.

E. Barriga-B., M.D. Carleton, and A.L. Gardner gave valu-
able field assistance. J.A. Lackey collected valuable specimens
from Tamaulipas and Campeche. J.A. Arnold collected valu-
able specimens from Oaxaca proving the sympatry of P. mega-
lops and melanurus. M.D. Carleton illustrated the glandes
penis, and A. Solis photographed the skulls.

I appreciate the loan of specimens from the following indi-
viduals and their respective institutions: J.R. Patton, Museum
of Vertebrate Zoology, University of California, Berkeley; R.L.
Orr and L.C. Binford, California Academy of Science; D.R.
Patten and L. Lester, Natural History Museum of Los Angeles
County; T.R. Howell, Dickey Collection, University of Califor-
nia, Los Angeles; J.K. Jones, Jr., and R.S. Hoffmann, Museum

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 326:1-24

18

Huckaby: Species of Peromyscus mexicanus Group

of Natural History, University of Kansas; L. de la Torre, Field
Museum of Natural History; R.A. Baker, The Museum, Michi-
gan State University; G.H. Lowery, Jr., Museum of Zoology,
Louisiana State University; D.J. Schmidly, Texas Cooperative
Wildlife Collection, Texas A & M University; G.G. Musser,
American Museum of Natural History; C. Jones, National
Museum of Natural History; R. Hardy, California State Uni-
versity, Long Beach.

I acknowledge support of the National Science Foundation
through NSF GB-6230 to N.G. Hairston, Museum of Zoology,
The University of Michigan. R.H. Corzo, El Director General
de la Fauna Silvestre, Departemento de Conservacion y Pro-
pagacion de la Fauna Silvestre provided collecting permits for
Mexico.

Finally I thank M.D. Carleton, G.G. Musser, and D.R. Pat-
ten for their helpful reviews of this paper. I accept, of course, all
responsibility for the final result.

LITERATURE CITED

Allen, J. A. 1891. Notes on a collection of mammals from
Costa Rica. Bull. Am. Mus. Nat. Hist. 3:201-18.

.1908. Mammals from Nicaragua. Bull. Am. Mus. Nat.

Hist. 24:647-70.

Allen, J.A., and F.M. Chapman. 1 897a. On mammals from
Yucatan, with description of new species. Bull. Am. Mus.
Nat. Hist. 9:1-12.

.1897b. On a collection of mammals from Jalapa and

Las Vigas, state of Veracruz, Mexico. Bull. Am. Mus. Nat.
Hist. 9:197-208.

Anderson, S.1968. A new craniometer and suggestions for
craniometry. J. Mammal. 49:221-28.

Baker, R.H. 1951. Mammals from Tamaulipas. Univ. Kans.

Pubis. Mus. Nat. Hist. 5:207-18.

Bangs, 0.1902. Chiriqui mammals. Bull. Mus. Comp. Zool.
Harv. 39:15-51.

Carleton, M.D. 1973. A survey of gross stomach morphology
in New World Cricetinae (Rodentia, Muroidea), with
comments on functional interpretations. Misc. Pubis. Univ.
Mich. Mus. Zool. 146:1-43.

.1977. Interrelationships of populations of the Per-
omyscus boylii species group (Rodentia, Muridae) in west-
ern Mexico. Occ. Pap. Mus. Zool. Univ. Mich. 675:1-44.
Carleton, M.D., and D.G. Huckaby. 1975. A new species of
Peromyscus from Guatemala. J. Mammal. 56:444-51.
Dalquest, W.W.1950. Records of mammals from the Mex-
ican state of San Luis Potosi. Occ. Pap. Mus. Zool. La. St.
Univ. 23:1-15.

.1953. Mammals of the Mexican state of San Luis Pot-
osi. La. St. Univ. Stud. Biol. Sci. Series 1:1-229.

Dickey, D.R. 1928. Five new mammals of the genus Per-
omyscus from El Salvador. Proc. Biol. Soc. Wash. 41:1-6.
Farris, J.S.1966. Estimation of conservatism of characters by
constancy within biological populations. Evolution
20:587-91.

Goldman, E.A.1951. Biological investigations in Mexico.

Smithson. Misc. Colins. 115:1-476.

Goodwin, G.G. 1932. Three new Reithrodontomys and two
new Peromyscus from Guatemala. Am. Mus. Novit.
560:1-5.

— .1938. Four new mammals from Costa Rica. Am. Mus.

Novit. 987:1-5.

.1941. A new Peromyscus from western Honduras. Am.

Mus. Novit. 1121:1

.1955. Two new white-footed mice from Oaxaca, Mex-
ico. Am. Mus. Novit. 1732:1-5.

.1956. Seven new mammals from Mexico. Am. Mus.

Novit. 1791:1-10.

.1964. A new species and a new subspecies of Per-
omyscus from Oaxaca, Mexico. Am. Mus. Novit. 2183:1 —
8.

.1969. Mammals from the state of Oaxaca, Mexico, in

the American Museum of Natural History. Bull. Am.
Mus. Nat. Hist. 141:1-269.

Hall, E.R.1971. Variation in the blackish deer mouse, Per-
omyscus furvus. Ann. Inst. Biol. Univ. Me: c., 39. Ser. Zool.,
1:149-54.

Hall, E.R., and T. Alvarez. 1961. A new species of mouse
{Peromyscus ) from northwestern Veracruz, Mexico. Proc.
Biol. Soc. Wash. 74:203-6.

Harris, W.P., Jr. 1940. A new subspecies of Peromyscus from
Costa Rica. Occ. Pap. Mus. Zool. Univ. Mich. 423:1-2.

Hooper, E.T.1957. Dental patterns in mice of the genus Per-
omyscus. Misc. Pubis. Mus. Zool. Univ. Mich. 99:1-59.

.1958. The male phallus in mice of the genus Per-
omyscus. Misc. Pubis. Mus. Zool. Univ. Mich. 105:1-24.

.1968. Classification. In Biology of Peromyscus (Roden-
tia), ed. J.A. King, pp. 27-74. Spec. Pubis., no. 2. Am.
Soc. Mammal ., 21-1 A.

Hooper, E.T., and G.G. Musser. 1964. Notes on classification
of the rodent genus Peromyscus. Occ. Pap. Mus. Zool.
Univ. Mich. 635:1-13.

Huckaby, D.G. 1973. Biosystematics of the Peromyscus mex-
icanus group. Unpublished dissertation, Univ. of
Michigan.

Lawlor, T.E.1965. The Yucatan deer mouse, Peromyscus
yucatanicus. Univ. Kans. Pubis. Mus. Nat. Hist. 1 6:42 1 —
38.

Linzey, A.V., and J.N. Layne.1969. Comparative morphology
of the male reproductive tract in the rodent genus Per-
omyscus (Muridae). Am. Mus. Novit. 2355:1-47.

Merriam, C.H.1898. Descriptions of twenty new species and a
new subgenus of Peromyscus from Mexico and Guate-
mala. Proc. Biol. Soc. Wash. 12:115-25.

Morrison, D.F. 1967. Multivariate statistical methods. New
York: McGraw-Hill.

Musser, G.G. 1964. Notes on geographic distribution, habitat,
and taxonomy of some Mexican mammals. Occ. Pap. Mus.
Zool. Univ. Mich. 636:1-22.

.1969. Notes on Peromyscus (Muridae) of Mexico and

Central America. Am. Mus. Novit. 2357:1-23.

.1971. Peromyscus allophylus Osgood: a synonym of

Peromyscus gymnotis Thomas (Rodentia, Muridae). Am.
Mus. Novit. 2453:1-10.

Osgood, W.H.1904. Thirty new mice of the genus Peromyscus
from Mexico and Guatemala. Proc. Biol. Soc. Wash.
17:55-77.

.1909. Revision of the mice of the American genus Per-
omyscus. N. Am. Fauna 28:1-285.

Saussure, M.H. de. 1860. Note sur quelques mammiferes du
Mexique. Rev. et Mag. de Zool. 1:20-22.

Thomas, 0.1894. Descriptions of some new neotropical
Muridae. Ann. Mag. Nat. Hist., Ser. 6 14:346-66.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 326:1-24

Huckaby: Species of Peromyscus mexicanus Group

19

Appendix. Descriptive statistics (in millimeters) of representative samples of species of Peromyscus.

No. of

Speci- Std. Std.

mens Mean Dev. Error Range

Peromyscus ochraventer.
Rancho del Cielo

Head and body

14

104

6.01

1.61

94-113

Tail

14

118

7.51

2.01

103-130

Hind foot

17

24.4

1.27

0.31

23-28

Skull length

16

30.3

0.63

0.16

28.7-31.2

Rostral length

16

9.2

0.31

0.08

8. 5-9.8

Braincase length

17

13.8

0.23

0.06

13.3-14.1

Interorbital width

17

4.8

0.14

0.03

4. 6-5.0

Braincase width

17

13.2

0.25

0.06

12.6-13.6

Incisive foramen length

17

6.3

0.27

0.06

5. 6-6. 6

Molar row

15

4.4

0.13

0.03

4. 1 -4.6

Interpterygoid fossa length

16

5.0

0.23

0.06

2.4-5. 3

Intermolar width

15

3.0

0.14

0.09

2. 7-3. 2

Interpterygoid fossa width

17

1.9

0.10

0.02

1. 7-2.0

Molar width

14

1.3

0.04

0.01

1.2-1. 4

Peromyscus stirtoni,
all El Salvador specimens

Head and body

8

101

4.70

1.57

93-110

Tail

8

95

6.22

2.20

92-108

Hind foot

8

23.4

0.75

0.26

22-24

Skull length

8

28.9

0.86

0.31

27.4-29.9

Interorbital width

8

4.9

0.09

0.03

4. 8-5.0

Braincase width

8

13.0

0.23

0.08

12.7-13.3

Molar row

8

4.0

0.12

0.04

3.8-4. 1

Intermolar width

8

3.2

0.08

0.03

3. 1-3. 3

Peromyscus yucatanicus, Escarcega

Head and body

20

99

7.39

1.65

80-1 10

Tail

20

100

8.12

1.81

80-110

Hind foot

24

22.0

0.97

0.19

20-24

Skull length

25

28.2

1.15

0.23

25.0-29.7

Rostral length

25

8.3

0.50

0.10

7. 0-8. 9

Braincase length

25

13.2

0.34

0.07

12.4-13.8

Interorbital width

25

4.8

0.17

0.03

4. 5-5. 2

Braincase width

25

12.5

0.27

0.05

12.1-12.9

Incisive foramen length

25

5.6

0.34

0.07

4. 7-6. 3

Molar row

25

3.7

0.13

0.03

3. 4-3. 9

Interpterygoid fossa length

25

4.9

0.30

0.06

4. 2-5. 4

Intermolar width

25

3.1

0.22

0.04

2.7-3. 5

Interpterygoid fossa width

25

1.9

0.15

0.03

1. 7-2.3

Molar width

25

1.2

0.04

0.01

1.0-1. 2

Peromyscus furvus, Xilitla

Head and body

13

128

1 1.71

3.25

114-152

Tail

13

131

10.03

2.78

112-142

Hind foot

9

23.4

1.01

0.34

22-25

Skull length

14

34.5

1.58

0.42

31.9-36.8

Rostral length

14

10.8

0.78

0.21

9.4-11.9

Braincase length

14

15.7

0.38

0.10

15.2-16.2

Interorbital width

14

5.4

0.16

0.04

5. 1-5.7

Braincase width

14

14.4

0.34

0.09

13.6-14.8

Incisive foramen length

14

7.0

0.38

0.10

6.4-7. 8

Molar row

14

5.1

0.14

0.04

4.9-5. 3

Interpterygoid fossa length

14

5.7

0.36

0.10

4. 8-6. 3

Intermolar width

14

3.4

0.18

0.05

3.2-3. 7

Interpterygoid fossa width

14

2.2

0.20

0.05

1 .9-2.6

Molar width

14

1.5

0.07

0.02

1.4- 1.6

Continued

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 326:1-24

20

Huckaby: Species of Peromyscus mexicanus Group

Appendix continued.

No. of
Speci-
mens

Mean

Std.

Dev.

Std.

Error

Range

Peromyscus furvus, Hidalgo

Head and body

28

129

7.25

1.37

115-143

Tail

28

130

7.51

1.42

115-147

Hind foot

28

29.5

0.96

0.18

27-32

Skull length

28

33.8

0.96

0.18

31.6-35.6

Rostral length

28

10.5

0.45

0.09

9.4-1 1.4

Braincase length

28

15.1

0.35

0.07

14.4-15.7

Interorbital width

28

5.2

0.19

0.04

4. 8-5. 5

Braincase width

28

14.1

0.32

0.06

13.6-14.9

Incisive foramen length

28

6.8

0.32

0.06

6. 4-7. 4

Molar row

28

5.1

0.13

0.02

4.8-5. 3

Interpterygoid fossa length

28

5.7

0.28

0.06

5. 3-6. 2

Intermolar width

28

3.3

0.16

0.03

3.0-3. 5

Interpterygoid fossa width

28

2.2

0.13

0.02

2. 0-2. 6

Molar width

28

1.5

0.06

0.01

1.4-1. 7

Peromyscus furvus, Xico

Head and body

15

125

6.1

1.60

113-136

Tail

15

133

9.7

2.50

115-145

Hind foot

15

28.3

0.98

0.25

27-30

Skull length

15

33.9

1.25

0.32

31.5-35.6

Rostral length

15

10.7

0.52

0.13

9.8-11.6

Braincase length

15

15.0

0.50

0.13

14.1-15.9

Interorbital width

15

5.2

0.1 1

0.03

5. 0-5. 4

Braincase width

15

13.9

0.20

0.05

13.5-14.2

Incisive foramen length

15

6.9

0.43

0.11

6. 1-7.4

Molar row

15

5.0

0.18

0.05

4. 8-5. 3

Interpterygoid fossa length

15

5.7

0.35

0.09

5. 2-6.2

Intermolar width

15

3.5

0.21

0.05

3. 1-3.8

Interpterygoid fossa width

15

2.2

0.14

0.04

2. 0-2. 5

Molar width

15

1.5

0.05

0.01

1.4- 1.6

Peromyscus megalops, Rio Molino

Head and body

32

131

7.49

1.32

112-141

Tail

32

137

9.19

1.62

1 16-151

Hind foot

32

29.7

0.79

0.14

28-31

Skull length

32

33.7

1.03

0.18

31.2-35.3

Rostral length

32

10.8

0.56

0.10

9.4-11.5

Braincase length

32

15.1

0.32

0.06

14.4-15.8

Interorbital width

32

5.6

0.18

0.03

5. 2-5.9

Braincase width

32

14.3

0.35

0.06

13.5-14.9

Incisive foramen length

32

7.2

0.27

0.05

6. 6-7.8

Molar row

31

4.9

0.12

0.02

4.6-5. 1

Interpterygoid fossa length

32

5.9

0.31

0.06

5. 1-6.4

Intermolar width

31

3.5

0.19

0.03

3. 1-3.9

Interpterygoid fossa width

32

2.2

0.16

0.03

1. 9-2.5

Molar width 31

Peromyscus melanocarpus, 116 km SW of Tuxtepec

1.5

0.15

0.01

1.4- 1.6

Head and body

1 1

124

9.70

2.92

100-133

Tail

11

122

8.81

2.66

100-130

Hind foot

1 1

29.1

0.83

0.25

28-30

Skull length

1 1

32.9

1.46

0.44

29.8-34.4

Rostral length

11

10.2

0.83

0.25

8.4-1 1.2

Braincase length

11

15.1

0.34

0.10

14.7-15.8

Interorbital width

1 1

5.3

0.13

0.04

5. 1-5.6

Braincase width

11

14.2

0.44

0.13

13.7-15.0

Incisive foramen length

11

7.1

0.42

0.13

6. 0-7. 6

Molar row

10

4.8

0.1 1

0.03

4. 6-5.0

Interpterygoid fossa length

10

5.5

0.47

0.15

4.6-6. 1

Intermolar width

1 1

3.3

0.12

0.04

3. 1-3. 5

Interpterygoid fossa width

11

2.1

0.24

0.07

1. 7-2.5

Molar width

1 1

1.5

0.03

0.01

1.4-1. 5
Continued

Contrib. Sci. Natur. Hist. Mus. Los Angeles County 1980. 326:1-24

Huckaby: Species of Peromyscus mexicanus Group

21

Appendix continued

No. of

Speci-

mens

Mean

Std.

Dev.

Std.

Error

Range

Peromyscus melanurus, Km 1 83— P. E.
Head and body

road

36

123

9.01

1.50

95-138

Tail

36

127

9.18

1.53

111-146

Hind foot

36

27.3

0.88

0.15

26-29

Skull length

33

32.4

1.22

0.21

29.9-34.3

Rostral length

33

9.8

0.54

0.09

8.4-10.6

Braincase length

33

14.9

0.35

0.06

14.3-15.6

Interorbital width

33

5.2

0.14

0.02

4. 9-5. 4

Braincase width

33

14.2

0.29

0.05

13.6-14.7

Incisive foramen length

33

6.8

0.30

0.05

6. 1-7.4

Molar row

29

4.6

0.14

0.03

4. 3-4.9

Interpterygoid fossa length

33

5.5

0.35

0.06

4.7-6. 2

Intermolar width

29

3.3

0 13

0.02

3. 1-3.6

Interpterygoid fossa width

33

2.0

0.13

0.02

1. 8-2.3

Molar width

29

1.4

0.06

0.01

1.3-1. 5

Peromyscus zarhynchus, Cerro Tzontehuitz
Head and body 40

140

5.77

0.91

130-153

Tail

40

146

10.08

1.59

129-165

Hind foot

40

32.8

0.89

0.14

31-35

Skull length

36

36.4

0.77

0.13

34.5-37.0

Rostral length

36

11.6

0.63

0.1 1

10.7-14.4

Braincase length

36

15.9

0.39

0.07

15.0-16.9

Interorbital width

36

5.3

0.15

0.02

4. 9-5. 6

Braincase width

36

14.9

0.32

0.05

14.2-15.6

Incisive foramen length

36

7.9

0.58

0.10

5. 0-8. 6

Molar row

35

5.4

0.16

0.03

5. 0-5. 6

Interpterygoid fossa length

36

6.5

0.34

0.06

5.8-7. 3

Intermolar width

33

3.8

0.17

0.04

3.4-4. 3

Interpterygoid fossa width

36

2.4

0.16

0.03

2. 1-2. 7

Molar width

33

1.6

0.06

0.01

1. 5-1.7

Peromyscus guatemalensis, Injerto

Head and body

38

128

6.74

1.09

112-141

Tail

38

137

13.12

2.93

1 10-156

Hind foot

38

30.3

0.63

0.10

29-32

Length of skull

40

34.0

0.83

0.13

32.4-36.2

Rostral length

40

10.4

0.47

0.07

9.5-1 1.7

Braincase length

40

15.1

0.44

0.07

14.3-16.2

Interorbital width

40

5.2

0.16

0.03

4.9-5. 5

Braincase width

40

14.4

0.25

0.04

13.9-14.9

Incisive foramen length

40

7.3

0.33

0.05

6. 6-8.0

Molar row

38

4.9

0.22

0.04

4. 5-5.3

Interpterygoid fossa length

40

6.1

0.25

0.04

5. 6-6. 6

Intermolar width

38

3.6

0.15

0.02

3. 2-3.9

Interpterygoid fossa width

40

2.3

0.15

0.02

2. 0-2. 6

Molar width

38

1.5

0.08

0.01

1 .4-1.7

Peromyscus guatemalensis, Volcan Lucas
Head and body 40

129

1 1.59

1.83

80-149

Tail

40

124

12.51

1.98

104-180

Hind foot

'40

30.2

1.32

0.21

27-34

Skull length

40

33.3

1.04

0.16

31.2-36.0

Rostral length

40

10.1

0.50

0.08

9.1-11.4

Braincase length

40

14.7

0.33

0.05

14.0-15.5

Interorbital width

40

5.2

0.24

0.04

4. 3-5. 7

Braincase width

39

14.2

0.34

0.05

13.6-15.0

Incisive foramen length

40

6.8

0.36

0.06

6.1-7. 5

Molar row

40

5.0

0.18

0.03

4.6-5. 5

Interpterygoid fossa length

39

5.8

0.33

0.05

5. 1-6.7

Intermolar width

40

3.4

0.25

0.04

2. 7-3. 9

Interpterygoid fossa width

40

2.2

0.14

0.02

1. 8-2.4

Molar width

39

1.5

0.08

0.01

1.3-1. 7

Continued

Conlrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 326:1-24

22

Huckaby: Species of Peromyscus mexicanus Group

Appendix continued.

No. of
Speci-
mens

Mean

Std.

Dev.

Std.

Error

Range

Peromyscus grandis, Concepcion

Head and body

14

143

13.12

3.50

132-165

Tail

14

152

5.48

1.47

142-160

Hind foot

14

33.2

1.37

0.37

30-35

Skull length

14

37.2

1.18

0.31

36.0-39.0

Rostral length

14

12.0

0.57

0.15

11.1-13.0

Braincase length

14

15.9

0.56

0.15

15.3-16.9

Interorbital width

14

5.6

0.26

0.07

5.0-6. 1

Braincase width

14

15.4

0.58

0.16

14.6-16.3

Incisive foramen length

14

7.8

0.31

0.08

7. 2-8. 4

Molar row

1 1

5.4

0.15

0.05

5.2-5. 7

Interpterygoid fossa length

14

6.7

0.24

0.06

6. 3-7. 2

Intermolar width

1 1

3.7

0.21

0.06

3.4-4. 1

Interpterygoid fossa width

13

2.5

0.15

0.04

2. 3-2. 8

Molar width

1 1

1.7

0.04

0.01

1.6-1. 8

Peromyscus gymnotis, El Rosario

Head and body

1 1

109

1 1.36

3.42

93-125

Tail

11

108

13.25

3.99

86-130

Hind foot

11

24.8

0.60

0.18

24-26

Skull length

11

29.4

1.95

0.59

25.9-32.1

Rostral length

1 1

8.8

0.85

0.26

7. 2-9. 9

Braincase length

1 1

13.6

0.58

0.17

12.8-14.7

Interorbital width

1 1

4.9

0.23

0.07

4. 5-5. 4

Braincase width

1 1

13.0

0.55

0.17

12.3-14.1

Incisive foramen length

11

5.9

0.44

0.13

5.0-6. 3

Molar row

10

4.3

0.18

0.06

4.0-4. 5

Interpterygoid fossa length

9

5.2

0.30

0.10

4. 6-5. 6

Intermolar width

1 1

3.2

0.26

0.08

2.8-3. 5

Interpterygoid fossa width

10

2.0

0.27

0.08

1. 7-2.5

Molar width

1 1

1.3

0.04

0.01

1. 3-1.4

Peromyscus mexicanus, Teocelo

Head and body

22

115

10.23

2.18

101-136

Tail

22

117

8.12

1.73

100-132

Hind foot

22

26.0

1.25

0.27

24-29

Skull length

22

32.1

0.65

0.14

31.0-33.1

Rostral length

22

9.8

0.33

0.07

9.1-10.3

Braincase length

22

14.5

0.27

0.06

14.0-14.8

Interorbital width

21

4.9

0.19

0.04

4.6-5. 3

Braincase width

22

13.5

0.24

0.05

12.9-13.9

Incisive foramen length

22

6.4

0.30

0.06

5. 9-7.0

Molar row

20

4.5

0.14

0.03

4. 2-4. 8

Interpterygoid fossa length

22

5.6

0.29

0.06

5.0-6. 3

Intermolar width

22

3.5

0.16

0.03

3.2-3. 8

Interpterygoid fossa width

22

2.0

0.12

0.03

1. 8-2.4

Molar width

22

1.4

0.05

0.01

1. 3-1.5

Peromyscus mexicanus, Ixcuintepec

Head and body

19

115

6.90

1.58

104-128

Tail

19

124

6.29

1.44

113-136

Hind foot

19

27.3

0.99

0.23

25-29

Skull length

19

32.7

0.58

0.13

31.2-33.4

Rostral length

19

10.1

0.37

0.09

9.2-10.7

Braincase length

19

14.7

0.24

0.06

14.2-15.0

Interorbital width

19

5.2

0.15

0.03

4.8-5. 5

Braincase width

19

13.8

0.27

0.06

13.3-14.4

Incisive foramen length

19

6.4

0.25

0.06

5. 9-6. 8

Molar row

19

4.6

0.14

0.03

4.4-4. 8

Interpterygoid fossa length

19

5.9

0.27

0.06

5. 4-6. 4

Intermolar width

18

3.5

0.14

0.03

3. 3-3.7

Interpterygoid fossa width

19

2.1

0.16

0.04

1.9-2. 5

Molar width

19

1.5

0.05

0.01

1.4-1. 5
Continued

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 326:1-24

Huckaby: Species of Peromyscus mexicanus Group

23

Appendix continued.

No. of

Speci-

mens

Mean

Std.

Dev.

Std

Error

Range

Peromyscus mexicanus. Santa Lucia

Head and body

35

112

1 1.00

1.86

87-128

Tail

35

122

12.30

2.08

101-152

Hind foot

35

26.7

0.70

0.12

25-28

Skull length

35

32.3

1.36

0.23

29.5-34.1

Rostral length

35

9.9

0.62

0.10

8.7-10.8

Braincase length

35

14.7

0.36

0.06

13.9-15.5

Interorbital width

35

5.2

0.16

0.03

4.9-5. 5

Braincase width

35

13.7

0.32

0.05

13.0-14.1

Incisive foramen length

35

6.2

0.31

0.05

5.6-6. 7

Molar row

35

4.3

0.09

0.01

4. 1-4.4

Interpterygoid fossa length

35

5.6

0.37

0.06

4. 8-6.3

Intermolar width

35

3.6

0.14

0.02

3. 3-4.1

Interpterygoid fossa width

35

2.0

0.14

0.02

1.6-2. 2

Molar width

35

1.3

0.04

0.01

1 .3-1.4

Peromyscus mexicanus. Teapa

Head and body

30

109

8.24

1.50

94-123

Tail

30

117

9.92

1.81

98-136

Hind foot

30

26.3

1.11

0.20

25-28

Skull length

31

31.8

1.04

0.19

29.5-33.8

Rostral length

31

9.8

0.45

0.08

8.8-10.8

Braincase length

31

14.4

0.35

0.06

13.5-15.3

Interorbital width

31

5.1

0.17

0.03

4.8-5. 5

Braincase width

31

13.4

0.35

0.06

12.8-14.2

Incisive foramen length

31

6.2

0.26

0.05

5. 5-6. 8

Molar row

29

4.5

0.12

0.02

4. 2-4. 6

Interpterygoid fossa length

31

5.7

0.30

0.05

5. 2-6.3

Intermolar width

31

3.3

0.16

0.03

3. 0-3. 6

Interpterygoid fossa width

31

2.1

0.13

0.02

1.9-2. 4

Molar width

31

1.4

0.04

0.01

1 .3-1.5

Peromyscus mexicanus. Volcan Agua

Head and body

43

122

8.65

1.32

95-135

Tail

43

125

9.04

1.38

102-139

Hind foot

43

26.4

1.08

0.16

24-29

Skull length

43

32.5

1.22

0.19

29.7-35.0

Rostral length

43

9.9

0.57

0.09

8.4-10.7

Braincase length

43

14.8

0.42

0.06

14.2-16.0

Interorbital width

44

5.2

0.21

0.03

4. 7-5. 6

Braincase width

44

14.1

0.35

0.05

13.1-14.7

Incisive foramen length

44

6.7

0.35

0.05

6. 0-7. 4

Molar row

43

4.7

0.18

0.03

4.3-5. 1

Interpterygoid fossa length

44

5.8

0.40

0.06

5. 1-6.6

Intermolar width

43

3.6

0.21

0.03

3. 1-3.9

Interpterygoid fossa width

44

2.2

0.17

0.02

1.9-2. 5

Molar width

43

1.5

0.08

0.01

1. 3-1.6

Peromyscus mexicanus. Mt. Cacaguatiq
Head and body

ue

59

1 1 1

6.84

0.89

94-125

Tail

60

1 16

8.77

1.13

95-132

Hind foot

62

25.3

0.87

0.11

23-27

Skull length

61

30.5

1.05

0.13

26.7-32.2

Rostral length

61

9.2

0.44

0.06

79-9.9

Braincase length

61

14.0

0.41

0.05

12.6-14.7

Interorbital width

60

5.2

0.14

0.02

4.6-5. 5

Braincase width

61

13.4

0.35

0.04

12.4-14.2

Incisive foramen length

61

6.2

0.29

0.04

5. 5-6.8

Molar row

61

4.4

0.13

0.02

4. 0-4. 7

Interpterygoid fossa length

61

5.4

0.28

0.04

4. 6-6.0

Intermolar width

61

3.3

0.17

0.02

2. 9-3. 8

Interpterygoid fossa width

61

2.0

0.11

0.01

1. 8-2.3

Molar width

61

1.4

0.05

0.01

1.2-1. 5

Continued

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 326:1-24

24

Huckaby: Species of Peromyscus mexicanus Group

Appendix continued.

No. of

Speci-

mens

Mean

Std.

Dev.

Std.

Error

Range

Peromyscus mexicanus, Santa Maria de Ostuma
Head and body 19

120

4.91

1.13

111-131

Tail

19

125

7.28

1.67

1 14-138

Hind foot

19

26.7

0.73

0.17

25-28

Skull length

23

32.1

0.89

0.19

30.8-33.9

Rostral length

23

9.6

0.42

0.09

9.0-10.6

Braincase length

23

14.6

0.33

0.07

13.9-15.1

Interorbital width

23

5.0

0.15

0.03

4. 8-5. 4

Braincase width

23

13.7

0.26

0.04

13.1-14.3

Incisive foramen length

23

6.7

0.29

0.06

6. 1-7. 2

Molar row

23

4.5

0.12

0.03

4. 3-4. 8

Interpterygoid fossa length

23

5.8

0.22

0.05

5. 5-6.4

Intermolar width

23

3.3

0.20

0.04

2. 9-3. 6

Interpterygoid fossa width

23

2.0

0.14

0.03

1.8-2. 4

Molar width

23

1.4

0.05

0.01

1.3-1. 5

Peromyscus mexicanus, Monteverde

Head and body

17

118

7.94

1.93

105-131

Tail

17

119

4.76

1.15

107-127

Hind foot

17

26.0

0.71

0.17

25-27

Skull length

16

31.8

0.59

0.15

30.7-32.9

Rostral length

16

9.7

0.41

0.10

8.9-10.2

Braincase length

16

14.3

0.15

0.04

13.0-14.6

Interorbital width

16

5.1

0.14

0.03

4. 9-5. 4

Braincase width

16

13.6

0.28

0.07

13.1-14.0

Incisive foramen length

16

6.6

0.26

0.07

6.2-7. 1

Molar row

16

4.6

0.12

0.03

4. 3-4. 7

Interpterygoid fossa length

16

5.6

0.28

0.07

5. 0-5. 9

Intermolar width

16

3.4

0.12

0.03

3. 1-3.6

Interpterygoid fossa width

16

2.0

0.14

0.03

1. 7-2.3

Molar width

16

1.4

0.03

0.01

1.3-1. 5

Peromyscus mexicanus, Volcan Irazu

Head and body

21

125

7.68

1.68

111-137

Tail

21

125

9.71

2.12

103-140

Hind foot

21

28.0

1.14

0.25

26-30

Skull length

22

32.9

0.97

0.21

30.8-34.5

Rostral length

22

10.1

0.43

0.09

9.4-11.0

Braincase length

22

15.1

0.34

0.07

14.3-15.8

Interorbital width

22

5.3

0.22

0.05

4. 8-5. 7

Braincase width

22

14.2

0.43

0.09

13.6-15.4

Incisive foramen length

22

6.9

0.34

0.07

6.2-7. 5

Molar row

22

4.9

0.1 1

0.02

4. 7-5.0

Interpterygoid fossa length

22

5.7

0.34

0.07

5. 0-6. 2

Intermolar width

22

3.6

0.18

0.04

3. 2-3.8

Interpterygoid fossa width

22

2.2

0.14

0.03

2.0-2. 5

Molar width

22

1.5

0.05

0.01

1. 5-1.7

Submitted 20 September 1977. Accepted for publication 29 May 1979.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 326:1-24

The scientific publications of the Natural History Museum of Los Angeles County have been issued
at irregular intervals in three major series; the articles in each series are numbered individually, and
numbers run consecutively, regardless of subject matter:

• Contributions in Science, a miscellaneous series of technical papers describing original

research in the life and earth sciences.

• Science Bulletins, a miscellaneous series of monographs describing original research in the
life, and earth sciences. This series was discontinued in 1978 with the issue of Numbers 29
and 30; monographs are now published by the Museum in the Contributions in Science

series.

• Science Series, long articles on natural history topics, generally written for the layman.
Copies of the publications in these series are available on an exchange basis to institutions and
individual researchers. Copies are also sold through the Museum Bookshop.

1

THE BEE GENUS BICORN ELIA
(HYMENOPTERA: COLLETIDAE)1

By Roy R. Snelling2

Abstract. The diphaglossine genus Bicornelia was described by Friese in 1899 from two males from
Mexico and has remained essentially unknown since then. The genus is redescribed from recently collected
material that includes males of the type species, B. serraia, from Mexico and both sexes of a newly de-
scribed species, B. inusitata, from Panama. Pertinent structures of both species are illustrated.

The genus Bicornelia was proposed by Friese (1899) for a new
species, B. serraia, based on two males from “Tuzantlu Lau-
rel,” Mexico, collected by Bilimek. It was placed near
Caupolicana from which it differs in details of wing venation
and in the modifications of the legs. Bicornelia has remained
unstudied and known only from the types, which have not been
seen by recent workers. Michener (1966) included it in the
tribe Mydrosomini on the basis of the original description and
its presumed relationship to Mydrosoma. Several males of B.
serrata are now available, as well as both sexes of an un-
described Panamanian species. The following generic descrip-
tion is patterned after those of Michener (1966) to facilitate
comparison with other diphaglossine genera.

SYSTEMATICS

Bicornelia friese

DIAGNOSIS. Diphaglossinae, Mydrosomini. MALE: Api-
cal segments of flagellum strongly serrate beneath; mid
trochanter with lamelliform apical spine; hind femur moder-
ately to strongly swollen; hind tibia strongly swollen or moder-
ately swollen and with median tubercle at midiength.
FEMALE: Flocculus of hind femur absent from dorsal face,
hairs sparse on anterior and ventral faces, the longer hairs
plumose apically only; inner hind tibial spur finely, evenly
serrate.

DESCRIPTION. (1) Clypeus evenly convex, weakly ele-
vated above adjacent parts of face, not continuous with su-
praclypeal area in profile. Clypeoantennal distance in female
slightly greater than diameter of antennal socket. (2) Pre-
stigma about twice as long as stigma; marginal cell not pro-
longed basally as a narrow sinus to apex of stigma. (3) Hind
basitarsus of female a little more than twice longer than wide;
second hind tarsal segment of female expanded, but a little
longer than wide, third much narrower than second. (4) Outer
hind tibial spur of male normal, articulated at base like inner
spur. (5) Abdomen without metallic tints, or with extremely

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 327:1-6
ISSN: 0459-81 1 3

weak bluish to greenish tints basally on terga (B. inusitata). (6)
Lateral extremities of terga of male without areas of short,
dense erect hair. (7) Sixth sternum of male with posterior mar-
gin neither thickened nor sulcate. (8) Seventh sternum of male
with complex paired lobes. (9) Eighth sternum of male uni-
formly and moderately pigmented, median process down-
curved, longest hairs of distal half shorter than width of
process. (10) Gonoforceps swollen apically, truncate.

The only other genus in the tribe Mydrosomini is My-
drosoma F. Smith (1879), which I have not seen. The flagellum
of male Mydrosoma is simple, not serrate beneath as in Bicor-
nelia. Neither Smith’s description of Mydrosoma, nor Moure’s
(1945) description of the synonymous Dissoglotta, mention the
presence of a spine on the mid trochanter. However, the trans-
parent, lamelliform spine present in Bicornelia is easily over-
looked, and it is possible that such a spine may also be present
in males of Mydrosoma. The very superficial description of the
female of M. metallicum is not very helpful. The metasomal
dorsum is stated to have a metallic green luster. Apparently,
also, the first recurrent vein of the forewing is interstitial with
Rs. These differences are weak; it may be that, when more ma-
terial is available, Bicornelia will prove to be a synonym of
Mydrosoma.

Subsequent to the original description of Bicornelia and its
type species, Friese (1925) described four additional species in
this genus, all from South America. As no material of any of
these species has been available to me, the following comments
are based wholly upon the original descriptions.

Bicornelia sericata was described from a female collected at
Guayaquil, Ecuador, by von Buchwald, April 1923, at flowers
of Cucurbita. The specimen apparently resembles the female of
B. inusitata (described below), but the thorax is weakly punc-

1 Review Committee for this contribution
Charles L. Hogue
Charles D. Michener
Robbin W. Thorp

2EntomoIogy Section, Natural History Museum of Los Angeles
County, Los Angeles, California 90007

2

Snelling: Bee Genus Bicornelia

Figures 1 through 5. Head and antennae, Bicornelia species. Figure 1, B. serrata male, head, frontal view. Figure 2, B. serrata male, antenna
Figure 3, B. inusitata male, head, frontal view. Figure 4, B. inusitata male, antenna. Figure 5, B. inusitata female, head, frontal view. Scale lines
Figures 2 and 4, 1.00 mm; Figures 1, 3, and 5, 0.50 mm.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 327:1-6

Snelling: Bee Genus Bicornelia

3

tate, the first flagellar segment is equal to the second, and the
first abdominal segment is shinier than the second.

Bicornelia aterrima was described from a female from Tar-
ata, Bolivia, collected by Priewasser. This bee, if correctly
placed in Bicornelia, may be recognized by the coarsely sculp-
tured abdomen, the dark wings and pilosity, and the short,
dense, plumose (?) hairs covering the gastric terga.

Bicornelia andina was described from males from Sierra Par-
ime, Venezuela, and Tarata, Bolivia; the type locality is here
restricted to Sierra Parime. Although Friese stated that this is a
typical Bicornelia, the flagellum is described as simple and the
abdomen almost nude. If this is truly a Bicornelia (not My-
drosoma ?), these features will render it easily recognized. The
legs are apparently not much modified.

Bicornelia longitarsis was described from one male from
Blumenau, Brazil. It is said to be similar to B. serrata but with
the abdomen black, with weak bluish reflections and with broad
bands of appressed hairs on the first four segments.

Nothing is known of the biology of Bicornelia. The yellowish
color and enlarged ocelli of the two species studied suggest that
these species are crepuscular, nocturnal, or matinal. The three
specimens of B. serrata collected near Tequila were taken in the
evening.

Floral preferences are unknown. The type of B. sericata was
collected at flowers of Cucurbita. Pollen grains from males
taken in Mexico are apparently congeneric with those from pol-
len loads removed from females from Panama. These have not
been identified but apparently are of a legume.

The heads and antennae of B. serrata and B. inusitata, new
species, are shown in Figures 1 through 5. Legs and sternum
segments of both species are illustrated in Figures 6 through
11, and genital capsules are compared in Figures 12 and 13.

Bicornelia serrata friese
Figures 1 and 2, 10 through 12

DIAGNOSIS. MALE. Flagellar segments 7-10 serrate be-
neath; hind femur without transverse lamella beneath in
middle; hind tibia swollen but without median tubercle ante-
riorly. FEMALE. Unknown.

DESCRIPTION: Measurements (in millimeters) are as fol-
lows: Head length (HL), 3.18-3.28; head width (HW), 3.85—
4.10; wing length (WL), 10.6-11.2; total length (TL, head 4-
thorax + terga 1 and 2), 12.2-13.0.

(1) Inner eye margin sinuate; lower interocular distance 0.88
to 0.92 times greatest interocular distance. Ocelli large; inter-
ocellar distance almost twice diameter of anterior ocellus;
ocellocular distance a little greater than diameter of anterior
ocellus; ocelloccipital distance about equal to diameter of ante-
rior ocellus. (2) Extreme basal part of labrum with one or two
weak transverse ridges. (3) First flagellar segment less than
half as long as scape, about half as long as following segment;
flagellar segments 1-6 strongly bulging beneath, segments 7-
10 serrate beneath; apical segment a little longer than second
flagellar segment, apex pointed, with low median convexity be-
neath (Fig. 2). (4) Anterior femur, seen from above, about
twice longer than thick. Middle trochanter beneath with a thin,
apical, transparent spine which is about half as long as greatest
thickness of trochanter. Middle femur distinctly flattened be-
neath and about 2.7 times longer than thick. (5) Hind femur, in

dorsal view, about twice as long as it is thick, ventral surface
slightly concave. Hind tibia in lateral view much broadened
over lower half, length about 2.9 times apical width. (6) Hind
basitarsus about three-fourths as long as hind tibia, about four
times longer than it is wide, widest beyond middle, narrowed
somewhat toward apex. (7) Propodeal triangle without trans-
verse ridges. (8) Posterior margins of sterna 2-4 transverse;
sternum 5 evenly and shallowly concave on posterior margin.

(9) Apical margin of sternum 6 weakly convex; disc with a low,
shiny prominence on either side of middle at about mid length.

(10) Hidden sterna and genitalia shown in Figures 10 through
12. (11) Integument of head black; pale yellowish color on
basal two-thirds of mandible, labrum, clypeus, supraclypeal
area, side of face to slightly above level of upper margin of an-
tennal socket, lower genal and hypostomal area. Antenna pale
ferruginous, with blackish blotches on outer side of flagellar
segments 2-7. Thorax black. Legs, tegula, and abdomen red-
dish yellow; abdomen with sublateral brownish marks on basal
face of first tergum and dorsally at about mid length on terga
1-5; tergum 6 with transverse, preapical brownish fascia,
which is strongly curved cephalad in middle to join median lon-
gitudinal bar; tergum 2 with small lateral spot; tergum 7
mostly brownish; with medioapical yellowish spot. Sterna
mostly reddish yellow, but with large, transverse, median
brownish blotches. Wings light yellowish brown, slightly darker
beyond cells; veins and stigma darker yellowish brown. (12)
Pubescence abundant, yellowish, paler on sides and venter.
Hairs of tergum I long, plumose, erect; remainder of terga with
subdecumbent, shorter, simple reddish yellow hairs which be-
come longer caudad. Sterna 2-4 with abundant suberect to
erect, long, weakly plumose, yellowish hairs, discs with scat-
tered appressed, simple, short hairs; sternum 5 with conspic-
uous apicolateral tufts of erect, reddish, simple hairs; sternum 6
with patches of sparse, simple, subappressed to suberect yellow-
ish hairs and a weak row across apical margin; sterna 2-5 with
marginal fringe of very short, widely spaced, whitish hairs.

TYPE MATERIAL. “2d'cf im Mus. Wein von Mexico
(Bilimek 1871, Tuzantlu Laurel).” This may be the village of
Tuzantla, Michoacan, situated about 48 km SSW of Zitacuaro.
The types have not been studied.

SPECIMENS STUDIED. MEXICO, Jalisco: 3o”cf, 7 km
NW of Tequila, 1275 m elev., 10 September 1974 (E.M.
Fisher; LACM).

DISCUSSION. The three males from near Tequila are the
basis for the above description. These males correspond closely
with the original description of B serrata. The bees were taken
at dusk flying about plants. The yellowish color and large ocelli
suggest the possibility that the species is matinal, crepuscular,
or both.

Bicornelia inusitata , new species
Figures 3 through 9, 11 and 13

DIAGNOSIS. MALE. Flagellar segments 5-10 serrate be-
neath; hind femur with transverse lamella beneath in middle;
hind tibia with conspicuous tubercle on anterior margin a little
beyond middle. FEMALE. Scutum and scutellum with scat-
tered fuscous hairs; hairs of scopa mostly dark brownish; meta-
soma mostly brownish; anterior width of second and third
submarginal cells subequal.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 327:1-6

4 Snelling: Bee Genus Bicornelia

Figures 6 through 10. Legs and sternites of males of Bicornelia species. Figure 6, B. inusitata, apex of mesotrochanter. Figure 7, B.
inusitata, metafemur, metatibia, and metabasitarsus. Figure 8, B. inusitata. sternum 8, left half. Figure 9, B. inusitata. sternum 8, lateral
view. Figure 10, B. serrata. sternum 8, right half. Scale line. Figures 8 through 10, 0.50 mm.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 327:1-6

Snelling: Bee Genus Bicornelia

5

Figures 11 through 13. Sternum 7 and genital capsules of males of Bicornelia species. Figure 1 1, Sternum 7 of B. serrata (left) and B
inusitata (right). Figure 12, B. serrata genital capsule, ventral (left) and dorsal (right) views. Figure 13, B. inusitata genital capusle, ventral
(left) and dorsal (right) views. Scale lines: Figure 11, 0.50 mm; Figures 12 and 13, 1.00 mm.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 327:1-6

6

Snelling: Bee Genus Bicornelia

DESCRIPTION.

MALE. Measurements (in millimeters) are as follows
(values in parentheses refer to holotype): HL, 2.87; HW, 3.49-
3.59 (3.49); WL, 9. 5-9. 7 (9.5); TL (HL + thorax + terga 1
and 2), 9.9-10.1 (9.9).

(1) Inner eye margin sinuate; lower interocular distance
about 0.90 times greatest interocular distance. Ocelli large; in-
terocellar distance 1.9 to 2.0 times diameter of anterior ocellus;
ocellocular distance about 1.5 times diameter of anterior
ocellus; ocelloccipital distance a little greater than diameter of
anterior ocellus. (2) Extreme basal part of labrum with two dis-
tinct, short transverse ridges in middle. (3) First flagellar seg-
ment less than one-fifth as long as scape, about half as long as
following segment; segments 1-4 strongly bulging beneath, seg-
ments 5-10 serrate beneath; apical segment about twice length
of second segment, strongly bulging beneath, profile sinuate
(Fig. 4). (4) Anterior femur, seen from above, about 2.8 times
longer than thick. Middle trochanter (Fig. 6) flattened beneath,
apically with a thin, broad, transparent projection that is about
as long as maximum thickness of trochanter. (5) Hind femur, in
dorsal view, about twice longer than thick, flattened ventrally,
with a low, weakly convex, transverse lamella in middle and
preapical, flattened, spiniform process. Hind tibia moderately
expanded toward apex in lateral view, length about 3 times api-
cal width; anteriorly with a large, blunt tubercle a little below
middle (Fig. 7). (6) Hind basitarsus about 5 times longer than
wide, widest in basal third. (7, 8 and 9) About as in B serrata.

(10) Hidden sterna and genitalia shown in Figs. 8, 9, 11, 13.

(11) Head as in B serrata but flagellar segments 2-10 dark
brownish above. Thorax black, with extremely faint greenish
reflection on side of propodeum. Coxae, trochanters, and hind
femur light brownish, remainder of legs reddish yellow. Meta-
somal terga mostly brownish, with faint bluish green reflec-
tions, segments 1-5 with broad, transparent, yellowish apical
margins, tergum 7 with medioapical reddish spot. Sterna
mostly brownish, with translucent yellowish apical margins.
Wings light brownish yellow, veins and stigma yellowish brown.

(12) Pubescence about as in B. serrata.

FEMALE. Measurements (in millimeters) are as follows
(values in parentheses refer to allotype): HL, 2.77-2.92 (2.92);
HW, 3.61-3.64 (3.64); WL, 8.7; TL, 9. 8-9.9 (9.9).

(13) Inner eye margin weakly sinuate, lower interocular
distance about 0.90 times greatest interocular distance. Inter-
ocellar distance 1.8 to 2.1 times diameter of anterior ocellus;
ocellocular distance 1.8 to 1.9 times diameter of anterior
ocellus; ocelloccipital distance slightly greater than diameter of
anterior ocellus. (14) First flagellar segment about one-fifth as
long as scape. Second segment broader than long, remaining
segments a little longer than broad. (15) Labrum with two fine
transverse ridges at extreme base. (16) Base of clypeus more
finely and closely punctate than remainder of clypeus; su-
praclypeal area with scattered fine punctures; integument
slightly shiny and sharply tessellate. (17) Anterior coxa with a
short, blunt, hairy apical process. (18) Propodeal triangle with
very weak traces of transverse wrinkles along midline. (19)
Head and thorax blackish brown, side of propodeum with ex-
tremely faint greenish reflections, middle of mandible reddish;
underside of flagellum yellowish. Legs brown, tibial spurs yel-
lowish, tarsal claws red. Metasoma brown, terga 1-4 with
translucent yellowish apical bands, wider in middle; sterna

paler brown than terga, with narrow translucent apical bands
on segments 2-5. Wings clear, brownish, darker beyond cell
area, veins and stigma brown. (20) Hairs on head mostly
whitish, some yellowish hairs apically on clypeus, hairs of oc-
ciput dusky. Hairs of thoracic dorsum dark ochreous, becoming
paler, whitish, beneath; mesoscutum and scutellum with numer-
ous fuscous hairs. Hairs white on femora and anterior margin
of hind tibia and hind basitarsus; remainder of hair of legs light
to dark brown. Hairs of basal face of tergum 1 whitish,
plumose; on summit of tergum 1 and on terga 2-4, yellowish,
simple, appressed to subdecumbent, with scattered erect
fuscous hairs, terga 5-6 with hairs suberect, plumose, fuscous;
terga 1-4 with narrow apical fringe of appressed, short,
plumose, whitish hairs, tergum 5 with dense fringe of long,
coarse, apically branched, yellowish red hairs. Hairs of sterna
short at base of segment, becoming longer toward margin;
mostly whitish to yellowish, but with numerous fuscous hairs on
segments 4-6, hairs simple mesially, plumose laterad; sterna 2-
5 with weak marginal fringe of very short, whitish, simple
hairs.

TYPE MATERIAL. Holotype male. Madden Forest Pre-
serve, Canal Zone, PANAMA, 12 December 1957 (W.J.
Hanson). Allotype female, 1 male and 1 female paratypes,
same locality, 1 January 1958 (W.J. Hanson). Holotype and al-
lotype in Snow Entomological Museum, University of Kansas;
paratypes in Natural History Museum of Los Angeles County
(LACM).

ETYMOLOGY. From Latin, inusitata. strange or extraordi-
nary, due to the modified antenna and hind leg.

ACKNOWLEDGMENTS

I am indebted to the following for their assistance: E.M. Fisher,
for donation of Bicornelia serrata to the LACM; C.D.
Michener, for the loan of, and permission to describe, B. in-
usitata. Pollen samples were examined by A.R. Moldenke, and
his assistance is gratefully acknowledged.

LITERATURE CITED

Friese, H. 1899. Monographic der Bienengattungen Mega-
cilissa, Caupolicana und Oxaea. Ann. K. K. Hofmus.
Naturh. Wein 14:239-246.

1925. Neue neotropischen Bienenarten, zugleich II.

Nachtrag zur Bienenfauna von Costa Rica (Hym.). Stett.
Entomol. Zeit. 86:1-41.

Michener, C.D. 1966. The classification of the Diphaglossinae
and North American species of the genus Caupolicana
(Hymenoptera, Colletidae). Univ. Kans. Sci. Bull. 46:717-
751.

Moure, J.S. 1945. Contribuicao para o conhecimento dos Di-
phaglossinae, particularmente Ptiloglossa (Hym. — Ap-
oidea). Mus. Paranaense Arq. 4:137-178.

Smith. F. 1879. Description of new species of Hymenoptera in
the collection of the British Museum, London.

Accepted for publication 23 April 1980.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County 1980. 327:1-6

SRI LANKA AND INDIA
(HYMENOPTERA: COLLETIDAE)

Roy R. Smelling

Natural History Museum of Los Angeles County * 9(90 Exposition Boulevard f, Los Angeles, CaESoraia 904)07

at irregular intervals in three major series; the articles in each series are numbered individually, and
numbers run consecutively, regardless of subject matter:

- Contributions in Science, a miscellaneous series of technical papers describing original
research in the life and earth sciences.

• Science Bulletins, a miscellaneous series of monographs describing original research in the
life and earth sciences. This series was discontinued in 1978 with the issue of Numbers 29
and 30; monographs are now published by the Museum in the Contributions in Science

series.

• Science Series, long articles on natural history topics, generally written for the layman.
Copies of the publications in these series are available on an exchange basis to institutions and
individual researchers. Copies are also sold through the Museum Bookshop.

NEW BEES OF THE GENUS f IV LA US FROM
SRI LANKA AND INDIA
(HYMENOPTERA: COLLETIDAE)1

Roy R. Snelling2

ABSTRACT. The following new species of Hylaeus Fabricius 1793 are
described and illustrated: H. krombeini from Sri Lanka; H oresbius,
H. porcatus, H thyreus, H. peltates, and H. parmatus from males
only, all from India; H. eurygnathus, H. sedens, and H. pannuceus,
from females only, from India. A key is given for the separation of these
species.

INTRODUCTION

Hylaeine bees appear to be rare, or at least are rarely encoun-
tered, in southern India and Sri Lanka. Most of the species de-
scribed below are represented by single specimens that I have
had for many years. More recently, Karl V. Krombein3 sent a
small series from Sri Lanka. That species is described below so
that the name might be available for Dr. Krombein’s use. The
other species are described at this time in the hope that some
interest in the species of southern India might be generated.

Two species have been described, under the old generic name
Prosopis Fabricius 1 804; from Sri Lanka: P. monilicornis Mor-
awitz and P. mixtus F. Smith. As Meade- Waldo (1923) has in-
dicated, P. mixtus is an allodapine anthophorid. Morawitz’s
description of P. monilicornis is so vague that nothing about the
bee seems certain other than that it may be Hylaeus.

A few forms have been previously described from India and
Pakistan; these seem to be related to the Palearctic fauna and
do not appear to have much in common with those treated here.
This opinion may, of course, be modified if more material be-
comes available. The types of most of these species are in the
British Museum (Natural History) and should be subjected to
modern study and redescribed. These previously described spe-
cies are:

H. advocatus (Nurse 1903). 9 . Kashmir.

H. basimacula (Cameron 1904). 9. Darjiling.

H. bellicosus (Cameron 1896). 9 . Barrackpore.

H. buddhae Meade- Waldo 1923. 9 . Barrackpore.

— Prosopis absoluta Cameron 1896, preoccupied.

= Prosopis butea Warncke 1970. Unnecessarily proposed
to replace P. absoluta Cameron. NEW SYNONYMY.

H.feae (Vachal 1895). 9 . Deesa.

= Prosopis striatifrons Cameron 1896. 9 . Barrackpore.

H. gujaraticus (Nurse 1903). 9 . Deesa.

H. kashmirensis (Nurse 1903). 9 . Kashmir.

H. montanus (Nurse 1903). 9 . Mt. Abu.

H. repent ens (Nurse 1903). 9 . Deesa.

Contributions in Science, Number 328, pp. 1-18
Natural History Museum of Los Angeles County, 1980

H. secretus (Nurse 1903). 6 . Kashmir.

H. strenuus (Cameron 1896). 9. Barrackpore.

H. vetustus (Nurse 1903). 9. Kashmir.

Warncke (1970) dealt with the western Palearctic Hylaeus,
under the name Prosopis. In that paper, he proposed new names
for some species, the original names of which were secondary or
tertiary homonyms. Since Meade- Waldo (1923) had already
proposed new names for some of the involved species, the
Warncke names become objective synonyms. Those names not
considered elsewhere are disposed of below.

H. binominatus Meade- Waldo 1923

= Prosopis laticeps Perkins 1899, not Morawitz 1876
= Prosopis avara Warncke 1970. NEW SYNONYMY.

H. promontorii Meade-Waldo 1923

= Prosopis longula Friese 1913, not Perez 1903
= Prosopis corpana Warncke 1970. NEW SYNONYMY.

H. nivicola Meade-Waldo 1923

= Prosopis nivalis Perkins 1899, not Morawitz 1867
= Prosopis farinosa Warncke 1970. NEW SYNONYMY.

H.simplior Meade-Waldo 1923

= Prosopis simplex Bingham 1912, not Perkins 1899
= Prosopis postica Warncke 1970. NEW SYNONYMY.

H. tristissimus Meade-Waldo 1923

= Prosopis tristis Schrottky 1906, not Frey-Gressner
1900

= Prosopis tritica Warncke 1970. NEW SYNONYMY.

H. insulae Meade-Waldo 1923

= Prosopis vicina Perkins 1899, not Sichel 1867
= Prosopis trigona Warncke 1970. NEW SYNONYMY.

TERMINOLOGY

Most of the terms utilized in the descriptions below are those
established in my earlier papers on this genus, such as Snelling
(1970). However, those earlier descriptions are clumsy and, too
often, characters are described in vague or subjective terms.
Much of the terminology can be standardized, and I propose to

1. Review committee for this contribution: Jack C. Hall, Charles L.
Hogue, and F.D Parker.

2. Entomology Section, Natural History Museum of Los Angeles
County, Los Angeles, California 90007.

3. National Museum of Natural History, Washington, D.C.

ISSN 0459-0113

do so here. Descriptions can also be simplified by omitting
many nonessential statements.

Houston (1975) proposed a few modifications of terminology
within the hylaeines and introduced a few new terms, some of
which require comment. Houston proposed to call the dorsal
and posterior lobes of the pronotum the “collar” and “posterior
lobes,” respectively. Since Michener (1965) has already uti-
lized the term “pronotal collar,” following a well-established
tradition, the usage is continued here. “Pronotal tubercule,”
however, seems an unnecessary substitution for the older term,
“pronotal lobe.”

For the elevated and marginate area of the face between and
above the antennal sockets, Houston has proposed “elevations
of the interantennal area.” I find the term cumbersome and will
use, instead, “frontal shield.” The frontal shield is elevated
above the frons and is often sharply marginate. The length of
the frontal shield is measured, somewhat arbitrarily, from the
level of the lower margin of the antennal socket to the upper
termination of its lateral margin on the frons. Houston has
noted that the width of the frontal shield (FSW) at its apex on
the frons. when compared to the width of the antennal socket
(ASD), is a useful specific character.

Except for noting the obvious differences related to puncta-
tion and other macrosculpture, 1 have not devoted much atten-
tion to superficial difference of texture. Surface texture, in my
experience, varies greatly within a species, and the differences
between closely allied species may be too subtle for good char-
acterization. Instead, I find that simple statements on the re-
flectiveness of the surface work well enough. General terms
such as dull, slightly shiny, moderately shiny, shiny, etc., seem
to be adequate. The elaborate terminology proposed by
Houston (1975) may work well enough for the Australian fauna
but seems difficult to apply to the bees described below.

I continue to use “puncture” in preference to Houston’s
“pit”; the former is long established. Houston illustrated the
relative sizes of punctures, but I find the system difficult to ap-
ply, even though it is a great improvement over the old, wholly
subjective terminology. I propose to use a different system.
Puncture diameters were measured with an ocular micrometer
at 120X and the following terminology developed:
minute 0.010-0.019 mm

fine 0.020-0.035 mm

moderate 0.036-0.055 mm

coarse 0.056-0.070 mm

very coarse over 0.070 mm

Since punctures are rarely of one size on a given segment or
stipulated area, they may be described as “fine to moderate”
(puncture diameter varying between 0.020 and 0.055 mm),
though usually a more limited size range, such as “moderate,”
prevails.

Areas that are rugose, reticulorugose, or reticulate are de-
scribed on the same basis as the punctation. Thus, “coarse re-
ticulae” have a transverse measurement of between 0.056 and
0.070 mm. More or less quadrate reticulae, such as found on
the basal area of the propodeum, over about 0.090 mm are
areolae.

The relative density of punctation has typically been de-
scribed in subjective terms and has often been a source for con-

fusion. For Hylaeus I propose the following standardization:

Contiguous — punctures so close that they are often de-
formed; interspaces are compressed and sharp-edged.

Subcontiguous — punctures separated by more or less flat in-
terspaces up to about 0.30 times a puncture diameter; some
punctures may be deformed.

Dense — punctures separated by more or less flat interspaces
between 0.30 and 0.70 times a puncture diameter; punctures
usually round but may be elongate.

Close — punctures separated by more or less flat interspaces
0.70 to 1.50 times a puncture diameter.

Sparse — punctures separated by more or less flat interspaces
1 .50 to 3.00 times a puncture diameter.

Scattered — puncture interspaces are very irregular and
range from about 3.00 to 6.00 or more times a puncture
diameter.

Variation in puncture density is expressed by combining
terms, as for example, sparse to scattered. Clypeal punctation
is described from the middle one-third of the segment.

The “face” includes all areas below the vertex except the
clypeus, supraclypeal area, and frontal shield. The “vertex” in-
cludes the area between the tops of the eyes, on a tangent to the
lower margin of the anterior ocellus, and back to the preoccipi-
tal ridge.

Mesoscutal punctation is described from the area between
the parapsides at the level of the tegula. Scutellar and mesono-
tal sculpture is described from the mesal one-third of each seg-
ment. The middle of the mesopleural disc is the standard for
that segment as is true for the side of the propodeum.

The first tergum has an oblique anterior or basal face and a
more or less horizontal posterior face; the latter is referred to as
the disc of that segment.

The second tergum is crossed, usually at about its anterior
one-third, by the curved and depressed gradulus. Anterior to
the gradulus is the pregradulus, usually duller and more finely
punctate than the postgradulus or disc. Punctation of the disc is
described from the mesal one-third, anterior to the apical de-
pression of the segment.

The distribution of the hairs, whether simple or plumose, is
monotonously uniform in most Hylaeus. and pilosity is here ac-
corded minimal descriptive attention. Only those differences
that are noteworthy are mentioned here, such as the presence
or absence of apicolateral patches on the first tergum or the
presence of long, erect hairs across the basal face of that
segment.

Some terms used in the descriptions have been abbreviated
to the following:

ASD — Antennal socket diameter. The maximum diameter,
between outer margins, perpendicular to the longitudinal axis
of the head.

BCW — Basal clypeal width. The distance between the sub-
antennal sutures along the basal margin of the clypeus.

COD — Clypeocular distance. Distance from laterobasal an-
gle to nearest point on eye margin.

CW — Clypeal width. The maximum width of the clypeus,
near its lowermost point.

HL — Head length. Maximum length between highest point
of the vertex and lowermost extremity of the clypeus.

2 Contributions in Science, Number 328

Sneiling: Hylaeus of Sri Lanka and India

HW — Head width. Maximum width of the head, across the
eyes.

LFW — Lower facial width. The minimum distance between
the eyes at their lower end. This term is utilized in its relation-
ship with UFW (q.v.) to express degree of convergence of the
inner eye margins: weakly convergent — UFW 1.01-1.29 times
LFW; moderately convergent — UFW 1.30-1.49 times LFW;
strongly convergent — UFW 1.50-1.70 times LFW; very
strongly convergent — UFW more than 1.70 times LFW.

OD — Ocellar diameter. Transverse diameter of anterior
ocellus.

SL — Scape length. The usual method, length of scape shaft,
exclusive of basal condyle.

TL — Total length. This is the least satisfactory of the mea-
surements used; it is certainly the least exact. The method used
here differs from the conventional but seems less subject to the
vagaries resulting from wide variations in death posture of the
specimen. TL is the sum of the following: HL + thoracic
length (in dorsal view, from anterior margin of pronotal collar
to posterior extremity of propodeum) 4- length of first tergum
(dorsal view, along midline with the summit of the anterior or
basal face just occluding the basal attachment) 4- length of
second tergum (along midline, from gradulus to apical margin).

UFW — Upper facial width. The minimum distance between
the upper ends of the eyes, at about the level of the anterior
ocellus, not at a point of greatest width as Houston (1975) has
it; consistent with my use of the term in earlier papers (e.g.,
1970).

WL — Wing length. The length of the anterior wing from
margin of tegula to apical extremity.

SYSTEMATICS

Hylaeus (Paraprosopis) krombeini new species

Figures 1-5

DIAGNOSIS

Lateral and oblique propodeal carinae present; disc of first
tergum uniformly and densely punctate; lower end of eye sub-
contiguous with margin of clypeus. Male first flagellar segment
about half as long as second; second and third together about as
long as scape; scape about twice longer than thick.

DESCRIPTION

Male

Measurements (mm): HL 1.07-1.20; HW 1.20-1.35; WL 2.7—
2.9; TL 3. 4-4.2.

Head. Broad, HW 1.12-1.13 x HL; scape short, SL 0.19-0.24
x HL; SL 1.88-2.29 x SW. Eyes very strongly convergent be-
low, UFW 2.30-2.67 x LFW. Clypeus narrow, CW 0.92-0.96 x
CL; BCW 0.43-0.50 x CW, 1.11-1.50 x ASD, 0.83-1.00 x
IAD, 1.11-1.50 x COD.

Clypeus slightly shiny between fine, subcontiguous to dense
punctures; supraclypeal area slightly shinier, with fine, close to
sparse punctures. Margins of frontal shield abruptly reflexed.

apex narrow, about 0.50 times ASD, disc slightly shiny between
fine, contiguous to subcontiguous punctures. Face with fine,
contiguous punctures, which become slightly coarser and sub-
contiguous on maculate areas at side. Interocellar area dull,
minutely and contiguously punctate. Gena moderately shiny
between fine, subcontiguous punctures, which become slightly
larger and elongate below. Preoccipital carina not attaining hy-
postomal carina.

Thorax. Mesoscutum about 1.3 times wider than long.
Scutellum weakly convex, 0.3-0. 4 times length of mesoscutum.
Metanotum weakly convex. Mesoscutum slightly shiny be-
tween fine, subcontiguous punctures; scutellum a little shinier
between fine, subcontiguous to dense punctures; metanotum
dull, finely rugosopunctate; mesopleuron moderately shiny be-
tween fine, dense to close punctures; metapleuron finely, and
contiguously punctate. Side of propodeum moderately shiny,
minutely and contiguously punctate; stigmatal area moderately
shiny, coarsely reticulorugose; basal area coarsely areolate; disc
slightly shiny, finely rugosopunctate.

Abdomen. Disc of first tergum smooth and shiny between min-
ute, subcontiguous to dense punctures; disc of second tergum
moderately shiny between ultraminute, close to sparse
punctures.

Pilosity. Sides of propodeum not densely pubescent; first and
second terga without apicolateral patches of appressed
pubescence.

Color. Blackish. The following pale yellowish: mandible, ex-
cept reddish apex; small spot on labrum; clypeus; side of face,
ending narrowly slightly above level of upper margin of anten-
nal socket; minute supraclypeal spot (if present); underside of
scape; pronotal collar, broadly interrupted in middle; most of
pronotal lobe; tegular spot; complete stripe on protibia; basal
and apical spots on mesotibia; basal stripe on metatibia; all
basitarsi. Remaining tarsal segments yellowish ferruginous.
Flagellum brownish, paler beneath. Wings clear, slightly brown,
veins and stigma medium brown.

Female

Measurements (mm): HL 1.20-1.27; HW 1.33-1.37; WL 3.1-
3.2; TL 4. 1 -4.4.

Head. Broad, HW 1 .08-1 . 11 x HL; scape short, SL 0.24-0.25
x HL; SL 3.0 x SW. Eyes very strongly convergent below, UFW
1 .73-1 .80 x LFW. Clypeus, narrow, CW 0.88-0.97 x CL; BCW
0.60-0.64 x CW, 1.80-2.25 x ASD, 1.00-1.29 x IAD, 1.80-
2.00 x COD.

Clypeus slightly shiny between fine, dense punctures; su-
praclypeal area slightly shiny between fine, dense to close punc-
tures; margins of frontal shield weakly bowed, slightly reflexed,
FSW about equal to ASD, disc slightly shiny, finely and con-
tiguously punctate; face dull, minutely and contiguously punc-
tate, becoming slightly shiny and finely, contiguously punctate
in area of sinus; maculate area slightly shiny between fine, sub-
contiguous punctures; vertex slightly shiny, finely and con-
tiguously punctate; gena slightly shiny, punctures fine, slightly
elongate, contiguous to subcontiguous, becoming coarser below.

Facial fovea very short, ending on eye margin below summit
of eye.

Thorax. Mesoscutum 1.2-1. 3 times wider than long.

Contributions in Science, Number 328

Smelling: Hylaeus of Sri Lanka and India 3

Figures 1 through 5. Hylaeus krombeini. Figure 1, male head, frontal view; Figure 2, female head, frontal
view; Figure 3, male, sternite 7; Figure 4, male, sternite 8; Figure 5, male, genitalic capsule, ventral view.

4 Contributions in Science, Number 328

Snelling; Hylaeus of Sri Lanka and India

Scutellum slightly flattened about 0.3 times length of meso-
scutum. Metanotum narrow and convex.

Mesoscutum slightly shiny between fine, subcontiguous
punctures; scutellum a little shinier, punctures slightly larger
and less close; metanotum dull, moderately rugosopunctate;
mesopleuron slightly shiny between fine, subcontiguous punc-
tures; metapleuron dull, finely rugosopunctate. Side of pro-
podeum dull, minutely rugosopunctate; stigmatal area coarsely
reticulorugose; basal area coarsely to very coarsely areolate;
disc dull, minutely rugosopunctate.

Abdomen. Disc of first tergum polished and shiny between
minute to fine, subcontiguous to dense punctures; disc of sec-
ond tergum moderately shiny between minute, close punctures.

Pilosity. First and second terga without apicolateral patches of
dense appressed pubescence.

Color. Blackish. The following pale yellowish: longitudinal
median bar on clypeus, not attaining base or apex; lateral face
mark, terminating above at about middle of antennal socket;
pronotal collar, broadly interrupted in middle; most of pronotal
lobe; tegular spot; long stripe on protibia; base of mesotibia;
basal half, or less, of metatibia; metabasitarsus. Remaining tar-
sal segments light to dark brownish ferruginous. Flagellum
brownish, paler beneath. Wings clear, light brownish, veins and
stigma light to dark brown.

TYPE MATERIAL

Holotype male, allotype female, five male and four female par-
atypes: Hunuwiligama, Anuradhapura Dist., SRI LANKA,
22-26 May 1976 (K.V. Krombein, P.B. Karunaratne, S. Ka-
runaratne, D.W. Balasooriya). Holotype, allotype, and most
paratypes in U.S. National Museum of Natural History; one
paratype of each sex in Natural History Museum of Los An-
geles County.

ETYMOLOGY

This species is dedicated to Karl V. Krombein, who kindly
made the specimens available for study.

DISCUSSION

The small size and dense punctation of the first tergum are very
suggestive of species of the Holarctic subgenus Paraprosopis,
and male genitalic structure confirms this placement. This spe-
cies is most like H. oresbius, described below. But, in H. krom-
beini, the punctures of the first tergum of the male are both
coarser and denser than those of the second tergum. The op-
posite is true of the male of H. oresbius, of which the female is
unknown.

Hylaeus oresbius new species

Figures 6-9

DIAGNOSIS

Male. Disc of first tergum with punctures close to sparse, ul-
traminute to minute; lateral and oblique propodeal carinae
present; mandible and tegula immaculate; first flagellar seg-

ment half as long as second, second and third together about as
long as scape.

Female. Unknown.

DESCRIPTION

Measurements (mm). HL 1 .20; HW 1 .37; WL 3.5; TL 4.2.

Head. Broad, HW 1.14 x HL; scape short SL 0.25 x HL; SL
2.00 x SW. Eyes very strongly convergent below, UFW 2.33 x
LFW. Clypeus narrow, CW 0.89 x CL; BCW 0.56 x CW, 1 .75 x
ASD, 1.00 x I AD, 1.40 x COD.

Clypeus moderately shiny between moderate to coarse, sub-
contiguous punctures; supraclypeal area moderately shiny and
finely lineolate with subcontiguous, moderate punctures above
and at side. Margins of frontal shield slightly reflexed, apex
narrow, about 0.5 times ASD, disc finely and contiguously
punctate. Face finely, contiguously punctate, punctures shiny
within, becoming moderate on maculate area; vertex finely,
contiguously punctate, grading to moderate punctures on pre-
occiput and upper gena, interspaces shiny on lower gena. Preoc-
cipital carina not attaining occipital carina.

Thorax. Mesoscutum about 1.3 times wider than long. Scutel-
lum slightly flattened, about 0.4 times length of mesoscutum.
Metanotum weakly convex.

Mesoscutum and scutellum dull between fine to moderate,
contiguous to subcontiguous punctures, which are shiny within;
metanotum dull, coarsely rugosopunctate; mesopleuron slightly
shiny between fine to moderate, subcontiguous punctures,
metapleuron dull, finely rugosopunctate. Side of propodeum
slightly shiny, minutely to finely rugosopunctate; stigmatal area
very coarsely reticulorugose; basal area very coarsely areolate
and with prominent longitudinal median ridge; disc moderately
to coarsely rugulose.

Abdomen. Disc of first tergum with ultraminute to minute
punctures of variable spacing, from subcontiguous to sparse, in-
terspaces smooth and shiny; disc of second tergum smooth and
shiny between minute, dense, sharply defined punctures.

Pilosity. Side of propodeum without dense appressed hairs;
first and second terga with apicolateral patches of appressed
pubescence.

Color. Blackish. The following bright yellow: minute spot at
base of mandible; mediobasal spot on labrum; clypeus, except
apical margin; lateral face mark to about level of middle of
antennal socket; underside of scape; pronotal collar, broadly in-
terrupted in middle; most of pronotal lobe; complete external
stripe on protibia; basal spot on mesotibia; basal third of meta-
tibia; meso- and metabasitarsi. Remaining tarsal segments
brownish ferruginous. Flagellum brownish, paler beneath.
Wings clear, brownish, veins and stigma dark brown.

TYPE MATERIAL

Holotype male: Cinchona, 3500 ft, Anamalai Hills, S. INDIA,

Contributions in Science, Number 328

Smelling: Hylaeus of Sri Lanka and India 5

Figures 6-9. Hylaeus oresbius, male. Figure 6, head, frontal view; Figure 7, sternite 7; Figure 8, sternite 8; Figure 9, genitalic
capsule, ventral view, gonobase omitted.

6 Contributions in Science, Number 328

Snelling: Hylaeus of Sri Lanka and India

May 1956 (P.S. Nathan), in Natural History Museum of Los
Angeles County.

ETYMOLOGY

From the Greek oresbios (mountain dweller).

DISCUSSION

This bee is superficially very similar to H. krombeini, but the
genitalic structures are quite different, and this species cannot
be placed in subgenus Paraprosopis. Externally, H. oresbius has
distinct apicolateral patches of appressed pubescence on the
first and second terga, lacking in H. krombeini. The first two
terga of both species are sharply punctate, but the punctures
are coarser on the second tergum than on the first in H. ores-
bius, opposite to the condition in H krombeini.

The female of H. oresbius is unknown but should have abun-
dantly punctate first and second terga, each of these segments
with an apicolateral patch of appressed pubescence.

Hylaeus porcatus new species

Figures 10-13

DIAGNOSIS

Male. First flagellar segment shorter than pedicel; basal area
of propodeum transversely strigose; scape less than twice longer
than wide.

Female. Unknown.

DESCRIPTION

Measurements (mm). HL 1 .40; HW 1.50; WL 3.9; TL 4.7.

Head. Broad, HW 1.07 x HL; scape very short, SL 0.19 x HL;
SL 1.59 x SW. Eyes strongly convergent below, UFW I 61 x
LFW. Clypeus narrow, CW 0.93 x CL; BCW 0.50 x CW, 1 .60 x
ASD, 1.33 x I AD, 1.45 x COD. Clypeus distinctly separated
from eye.

Clypeus and supraclypeal area moderately shiny, sharply and
longitudinally lineolate and with sparse to scattered, minute
and line punctures. Frontal shield very short, narrow and high,
terminating abruptly immediately above level of upper margin
of antennal socket, apex less than 0.25 times ASD; disc rough-
ened and with a few dense, fine punctures. Face shiny, finely
rugosopunctate in middle, punctures a little coarser, subcon-
tiguous laterad; maculate area lineolate and with scattered
minute punctures; gena shiny between fine, dense punctures.

Thorax. Mesoscutum about 1.3 times wider than long.
Scutellum weakly convex, about 0.4 times length of meso-
scutum. Metanotum moderately convex.

Mesoscutum slightly shiny between moderate, subcontiguous
to dense punctures; scutellum slightly shiny between contiguous
to dense, coarse punctures; metanotum moderately rugosopunc-
tate; mesopleuron moderately shiny between moderate to coarse,
contiguous to subcontiguous punctures, which are shiny within;
metapleuron moderately shiny, finely rugosopunctate. Lateral
propodeal carina present to stigmatal area; side and disc moder-
ately shiny between fine, contiguous to subcontiguous punctures;
stigmatal area coarsely reticulorugose; entire basal area crossed
by about six high, sharp, transverse ridges.

Abdomen. First tergum shiny, nearly impunctate, the few punc-
tures very widely scattered, ultraminute; disc of second tergum
shiny, with only very scattered ultraminute piligerous punctures.

Pilosity. First and second terga without apicolateral patches of
appressed pubescence.

Color. Blackish. The following yellow: large trefoil mark on
clypeus; lateral face mark, ending narrowly on eye margin
slightly above level of upper margin of antennal socket; large
mark on underside of scape; pronotal collar, broadly interrupted
in middle; most of pronotal lobe; tegular spot; outer stripe on
protibia; basal spot on mesotibia; basal one-third of metatibia;
metabasitarsus. Remaining tarsal segments reddish. Flagellum
brownish, lighter beneath. Wings clear, slightly brownish, veins
and stigma brown.

TYPE MATERIAL

Holotype male: Cinchona, 3500 ft, Anamalai Hills, S. INDIA,
May 1957 (P.S. Nathan), in Natural History Museum of Los
Angeles County.

ETYMOLOGY

From the Latin porca (ridge between furrows) because of the
transversely ridged basal area of the propodeum.

DISCUSSION

The sharply, transversely ridged basal area and virtually impunc-
tate first two abdominal segments of this species are distinctive.
Although H. pannuceus may be the female, such seems unlikely,
as noted in the discussion of that species.

Of previously described species, only H. bellicosus is men-
tioned as having a transversely rugose propodeal base. In that
species, the abdomen is, apparently, more clearly punctate than
in H porcatus, and the scape is wholly blackish.

Hylaeus thyreus new species

Figures 14-17

DIAGNOSIS

Male. Mandible broad, without obvious external ridges and
grooves; frontal shield longer than greatest width; distance be-
tween clypeus and lower end of eye about equal to OD.

Female. Unknown.

DESCRIPTION

Measurements (mm). HL 1 .50; HW 1 .60; WL 4.6; TL 5.7.

Head. Broad, HW 1.07 x HL. Scape short, SL 0.24 x HL; SL
2.75 x SW. Eyes moderately convergent below, UFW 1.38 x
LFW. Clypeus narrow, CW 0.83 x CL; BCW 0.67 x CW, 1.82 x
ASD, 1.18 x I AD, 1.25 x COD. Clypeus separated from lower
end of eye by about OD.

Mandible about twice longer than broad, lower margin
curved, cutting margin oblique and broadly rounded into upper
margin, preapical notch weak; outer face without obvious ridges
or grooves, with fine longitudinal striae.

Clypeus moderately shiny, longitudinally striate, with close,
moderate punctures. Frontal shield long, about 1.6 times longer
than wide, the lateral ridges reaching clypeal base; punctures

Contributions in Science, Number 328

Snelling: Hylaeus of Sri Lanka and India 7

Figures 10 13. Hylaeus porcatus, male Figure 10, head, frontal view; Figure 11, sternite 7; Figure 12, sternite 8; Figure 13,
genitalic capsule, ventral view, gonobase omitted.

8 Contributions in Science, Number 328

Snelling: Hylaeus of Sri Lanka and India

Figures 14-17. Hylaeus thyreus, male. Figure 14, head, frontal view; Figure 15, sternite 7; Figure 16, sternite 8; Figure 17, genitalic capsule,
ventral view.

Contributions in Science, Number 328

Snellimg: Hylaeus of Sri Lanka and India 9

Figures 18-21. Hvlaeus parmatus, male. Figure 18, head, frontal view; Figure 19, sternite 7; Figure 20, sternite 8; Figure 21,
genitalic capsule, ventral view.

10 Contributions in Science, Number 328

Snelling; Hylaeus of Sri Lanka and India

moderate, in subcontiguous rows; apical width about 1.3 times
ASD. Face and vertex shiny between moderate, subcontiguous
punctures (in rows in middle); maculate area slightly shiny be-
tween fine, dense punctures; gena shiny between moderate, sub-
contiguous punctures.

First and second flagellar segments each about equal to ped-
icel in length; third a little longer; combined, second and third
shorter than scape. Interocellar and ocellocular distances
subequal.

Thorax. Mesoscutum about 1.3 times wider than long;
scutellum flattened about 0.4 times length of mesoscutum;
metanotum flattened.

Mesoscutum shiny between fine, subcontiguous to dense
punctures; scutellum shiny between fine to moderate, dense
punctures; metanotum shiny between fine, subcontiguous to
dense punctures; mesopleuron shiny between moderate, dense
punctures; metapleuron with fine punctures in subcontiguous
rows. Side of propodeum, stigmatal area, and disc shiny be-
tween dense, moderate punctures; basal area very irregularly
reticulorugose at base, but mostly free of sculpture, shiny.

Abdomen. Disc of first tergum shiny between close to sparse,
fine punctures; disc of second tergum shiny between sparse,
minute to fine punctures; disc of third tergum with conspicuous
minute, sparse, piligerous punctures.

Pilosity. First tergum with apicolateral patch of appressed
pubescence.

Color. Blackish. The following bright yellow: large preapical
clypeal spot; lateral face mark, ending on inner eye margin at
about level of upper end of frontal shield; pronotal collar, nar-
rowly interrupted in middle; most of pronotal lobe; tegular spot;
outer stripe on basal three-fifths of protibia; basal spot on meso-
tibia; basal one-third of metatibia. Mandible, labrum, and un-
derside of scape reddish. Flagellum dark brownish, a little
lighter beneath. Wings clear, slightly brownish, veins and
stigma dark brown.

TYPE MATERIAL

Holotype male: Singara, 3400 ft, Nilgiri Hills, S. INDIA, May
1954 (P.S. Nathan), in Natural History Museum of Los An-
geles County.

ETYMOLOGY

From the Greek thyreos (an elongate shield), in reference to the
long frontal shield.

DISCUSSION

The similarities in mandibular structure and punctation suggest
rather strongly that H. eurygnathus is the female of this spe-
cies. It is also possible, however, that H. sedens may belong
here. Since there clearly are several species in this complex, no
attempt has been made to associate sexes with the limited ma-
terial available.

The elongate frontal shield of H. thyreus will separate it
from similar forms such as H. peltates and H. parmatus, both

described below. The mandibular structure of H. thyreus is also
distinctive.

Hylaeus parmatus new species

Figures 18-21

DIAGNOSIS

Male. Mandible broad, without obvious external ridges or
grooves, preapical notch deep; frontal shield about as wide as
long; macula whitish, scape immaculate.

Female. Unknown.

DESCRIPTION

Measurements (mm). HL 1 .43; H W 1 .60; WL 4.4; TL 5.3.

Head. Broad, HW 1 .12 x HL. Scape short, SL 0.26 x HL; SL
2.00 x SW. Eyes very strongly convergent below, UFW 1.93 x
LFW. Clypeus narrow, CW 0.74 x CL; BCW 0.46 x CW, 1.09 x
ASD, 0.86 x IAD, 0.75 x COD. Clypeus separated from lower
end of eye by about 0.3 x OD.

Mandible a little more than twice longer than broad, lower
margin weakly curved, large preapical tooth on cutting margin;
outer face without obvious ridges or grooves, smooth and shiny
over much of disc.

Clypeus and supraclypeal area moderately shiny between
moderate, subcontiguous to dense punctures. Frontal shield
about as wide as long, margin strongly bowed and distinctly
reflexed, apex about equal to ASD. Face and vertex moderately
shiny between subcontiguous, moderate, irregular punctures;
maculate area moderately shiny between dense to close, moder-
ate, round punctures; gena moderately shiny between subcon-
tiguous, fine to moderate punctures.

First flagellar segment about equal to pedicel in length, sec-
ond about 1.4 times first, third about 1.8 times first. Interocel-
lar distance slightly greater than ocellocular distance.

Thorax. Mesoscutum about 1.3 times wider than long;
scutellum flattened about 0.4 times length of mesoscutum;
metanotum flattened.

Mesoscutum slightly shiny between fine, subcontiguous
punctures; scutellum moderately shiny, punctures fine and sub-
contiguous in middle, moderate and close on either side; meta-
notum slightly shiny between minute to fine, dense punctures;
mesopleuron moderately shiny between fine, close punctures;
metapleuron shiny, with a few obscure, moderate punctures in
center, otherwise smooth. Side and disc of propodeum moder-
ately shiny between subcontiguous, fine punctures grading into
moderately reticulopunctate stigmatal area; basal area, in mid-
dle, about as long as scutellum, shiny, with a few irregular,
transverse reticulae near anterior margin.

Abdomen. Disc of first tergum smooth and shiny between fine,
close punctures; disc of second tergum slightly shiny between
sparse, minute punctures.

Pilosity. First tergum with long, erect, sparse hairs across
basal face; with apicolateral patch of appressed pubescence.

Color. Blackish. The following whitish: most of outer face of
mandible; large preapical clypeal mark; lateral face mark, ter-
minating on eye margin at level of sinus; pronotal collar,
broadly interrupted in middle; approximately one-half of prono-

Contributions in Science, Number 328

Smelling: Hylaeus of Sri Lamka and India 1 1

Figures 22-25. Hylaeus peltates, male. Figure 22, head, frontal view; Figure 23, sternite 7; Figure 24, sternite 8; Figure 25,
genitalic capsule, ventral view.

12 Contributions in Science, Number 328

Snelling; Hylaeus of Sri Lanka and India

tal lobe; tegular spot; outer stripe on protibia; basal spot on
mesotibia; basal half of metatibia; metabasitarsus. Remainder
of tarsi reddish brown. Flagellum dark brownish above, reddish
brown beneath. Wings clear, brownish, veins and stigma dark
brown.

TYPE MATERIAL

Holotype male: Singara, 3400 ft, Nilgiri Hills, S. INDIA, May
1954 (P.S. Nathan), in Natural History Museum of Los An-
geles County.

ETYMOLOGY

The specific epithet is a Latin word for one armed with a small
shield.

DISCUSSION

Some of the differences between this species and the similar H.
peltates have been noted in the section on that species. The
mostly smooth metapleuron is unique among the species I have
seen from southern India. If consistent, it would prove a good
diagnostic feature. The minute punctures on the disc of the sec-
ond tergum are distinctive within the group of species with
broad mandibles.

Hylueus peltates new species

Figures 22-25

DIAGNOSIS

Male. Mandible broad, without obvious external ridges or
groove, preapical notch weak; frontal shield about as wide as
long; distance between clypeus and lower end of eye about 0.5
times OD.

Female. Unknown.

DESCRIPTION

Measurements (mm). HL 1 .45; HW 1 .57; WL 4.5; TL 5.3.

Head. Broad, HW 1.08 x HL. Scape short, SL 0.24 x HL; SL
2.10 x SW. Eyes very strongly convergent below, UFW 1.93 x
LFW. Clypeus narrow, CW 0.78 x CL; BCW 0.57 x CW, 1.60 x
ASD, 0.94 x IAD, 1.33 x COD. Clypeus separated from lower
end of eye by about 0.5 x OD.

Mandible about twice longer than broad, lower margin
curved, cutting margin oblique and with weak preapical notch;
outer face without distinct ridges or grooves, smooth and shiny.

Clypeus and supraclypeal area moderately shiny, finely line-
olate between dense, fine, linearly arranged punctures, which
become contiguous laterad. Frontal shield about as broad as
long, margins distinctly bowed, not reflexed, apex about 1.25
times ASD; disc moderately shiny between fine, subcontiguous
to dense punctures. Face, vertex, and gena moderately shiny
between fine to moderate, subcontiguous punctures; punctures
of maculate area in rows, somewhat elongate, with extensive
impunctate areas.

First flagellar segment about equal to pedicel in length, sec-

ond about 2.3 times first, third about 3.0 times first. Interocel-
lar distance slightly greater than ocellocular distance.

Thorax. Mesoscutum about 1.4 times wider than long;
scutellum flattened, about 0.4 times length of mesoscutum;
metanotum flattened.

Mesoscutum slightly shiny between moderate, subcontiguous
punctures; scutellum similar but punctures become dense on
either side of middle; metanotum slightly shiny between fine,
dense punctures; mesopleuron moderately shiny between dense,
moderate punctures; metapleuron moderately shiny between
fine, subcontiguous punctures. Side and disc of propodeum
moderately shiny between fine to moderate, subcontiguous to
dense punctures grading into coarsely reticulopunctate stigma-
tal area; basal area moderately shiny, with a few coarse re-
ticulae basomedially and few weak, oblique rugulae laterad
that converge toward posterior middle.

Abdomen. Disc of first tergum smooth and shiny between fine
to moderate, close to sparse punctures; disc of second tergum
moderately shiny between sparse, fine punctures.

Pilosity. First tergum with long, erect, sparse hairs across
basal face and with apicolateral patch of appressed pale hairs.

Color. Blackish. The following bright yellow: outer face of
mandible (which has reddish apex); large, irregular preapical
clypeal spot; lateral face mark, ending acutely on eye margin
slightly above ocular sinus; large apical spot on scape; pronotal
collar, broadly interrupted in middle; most of pronotal lobe;
outer stripe on protibia; small basal spot on mesotibia; basal
half of metatibia; basal one-third of metabasitarsus. Remainder
of tarsi reddish brown. Flagellum brown, slightly paler beneath.
Wings clear, brownish, veins and stigma brown.

TYPE MATERIAL

Holotype male: Naduvatam, 6000 ft, Nilgiri Hills, S. INDIA,
May 1958 (PS. Nathan), in Natural History Museum of Los
Angeles County.

ETYMOLOGY

From the Greek peltates, a warrior with a shield, in reference to
the large, broad frontal shield of the species.

DISCUSSION

This bee is related to H. thyreus and H. parmatus. It differs
from H thyreus in the shorter frontal shield, the shorter dis-
tance between the clypeus and lower end of the eye, and the
longer second flagellar segment. It is much more similar to H
parmatus, in which the mandible, however, has a sharp preapi-
cal tooth, the macula are distinctly whitish, the scape immac-
ute, etc.

Hylaeus sedens new species

Figures 26-27

DIAGNOSIS

Female. Mandible broad, outer face without distinct ridges.

Contributions in Science, Number 328

Snelling: Hylaeus of Sri Lanka and India 13

29

Figures 26 30. Hylaeus spp., females. Figures 26 and 27, H. sedens, head in frontal view and
head in frontal view and mandible. Figure 30, H. pannuceus, head in frontal view.

14 Contributions in Science, Number 328

mandible. Figures 28 and 29, H. eurygnathus.

Snelling: Hylaeus of Sri Lanka and India

inner tooth strong; first tergum with apicolateral pubescent
patch.

Male. Unknown.

DESCRIPTION

Measurements (mm). HL 1.65; HW 1.75; WL 4. 6; TL 6. 1 .

Head. Broad, HW 1.06 x HL; scape short, SL 0.26 x HL; SL
3.25 x SW. Eyes moderately convergent below, UFW 1.42 x
LFW. Clypeus narrow, CW 0.77 x CL; BCW 0.61 x CW, 1.83 x
ASD, 1.10 x I AD, 1.22 x COD; clypeus distinctly separated
from eye margin.

Mandible broad, less than twice as long as greatest width,
lower margin strongly curved; outer face without obvious ridges
or grooves, obliquely and finely striate over apical half; cutting
margin oblique and subangulate with upper margin; preapical
notch deep, defining preapical tooth.

Clypeus coarsely lineolate, slightly shiny, with moderate
punctures in rows, dense basally, coarser and subcontiguous
preapically; supraclypeal area very short, lineolate, and with
subcontiguous, moderate punctures. Margins of frontal shield
bowed, flat, apex 2.00 times ASD; disc with distinct rows of
moderate, subcontiguous punctures, interspaces shiny. Face
shiny between fine to moderate, subcontiguous punctures, finer
in middle; gena shiny between moderate, irregularly shaped,
subcontiguous punctures.

Interocellar distance equal to ocellocular distance. Facial
fovea ending less than halfway between eye and lateral ocellus.

Thorax. Mesoscutum about 1.4 times wider than long.
Scutellum flattened, about 0.4 times length of mesoscutum.
Metanotum weakly convex.

Mesoscutum moderately shiny between fine, subcontiguous
punctures; scutellum with fine, subcontiguous punctures in
middle and moderate, dense punctures on either side, inter-
spaces moderately shiny; metanotum dull, with fine, subcon-
tiguous to dense punctures; mesopleuron shiny between
moderate, subcontiguous punctures; metapleuron similar, but
punctures mostly contiguous. Propodeum without transverse,
oblique, and lateral carinae; side shiny between moderate, sub-
contiguous punctures, which grade to contiguous punctures on
stigmatal area and disc; basal area slightly depressed in middle
and with a few irregular longitudinal and oblique rugae, sur-
face shiny.

Abdomen. Disc of first tergum shiny between irregularly
spaced (subcontiguous to close) punctures of several sizes (ul-
traminute to moderate); disc of second tergum moderately
shiny between sparse, minute punctures.

Pilosity. First tergum with apicolateral patch of appressed,
plumose pubescence; first tergum with patch of sparse, erect
hairs in middle of anterior face.

Color. Blackish. The following whitish; small preapical spot on
clypeus; lateral face mark, extending narrowly along inner orbit
to lower end of fovea; interrupted stripe on pronotal collar; nar-
row posterior margin of pronotal lobe; tegular spot; short basal
stripe on protibia; basal spot on mesotibia; large basal spot on
metatibia. Mandible piceous, paler near apex. Flagellum dark

brownish above, medium brownish beneath. Wings clear,
faintly brownish, veins and stigma dark brown.

TYPE MATERIAL

Holotype female; Karikal, Pondicherry State, INDIA, January
1956 (P.S. Nathan). Paratypes: 2 2 9, Singara, 3400 ft. elev.,
Nilgiri Hills, S. INDIA, May 1956 (PS. Nathan). All types in
collection of Natural History Museum of Los Angeles County.

ETYMOLOGY

From the Latin prefix se (apart) and dens (tooth), in reference
to the distinct preapical mandibular tooth.

DISCUSSION

Although obviously related to H. eurygnathus, described below,
this species is distinct. The principal differences include man-
dibular structure, pale maculae, and greater interocellar
distance.

One additional specimen that seems to be conspecific with
the types has been available: Peak View Motel, Kandy, elev.
1800 ft., Kandy Dist., SRI LANKA, 15-24 January 1970
(Davis and Rowe; U.S. National Museum of Natural History).
Aside from the fact that the mandibles are blackish brown,
rather than ferruginous, this seems to be a normal specimen of
H. sedens.

Hylaeus eurygnathus new species

Figures 28-29

DIAGNOSIS

Female. Mandible broad, outer face without distinct ridges,
inner tooth weak; first tergum with apicolateral pubescent
patch.

Male. Unknown.

DESCRIPTION

Measurements (mm). HL 1.13—1.75; HW 1 .27-1 .95; WL 3.6—
5.5; TL 4.4-6. 8.

Head. Broad, HW 1.1 1 — 1.12 x HL; scape short, SL 0.29-0.41
x HL; SL 2.80-3.00 x SW. Eyes moderately convergent below,
UFW 1 .42 x LFW. Clypeus narrow, CW 0.76-0.80 x CL; BCW
0.74-0.76 x CW, 2.33-2.55 x ASD, 1.17 x IAD, 1.56 x COD.
Clypeus distinctly separated from eye.

Mandible broad, about twice longer than greatest width, lower
margin strongly curved; outer face without obvious ridges and
obliquely and finely striate over apical half; cutting margin
oblique and broadly rounded into upper margin; notch between
apical and preapical teeth weak.

Clypeus moderately shiny, lineopunctate, width of punctures
less than 0.020 mm; punctures of supraclypeal area fine, in sub-
contiguous rows. Margins of frontal shield bowed, flat, apex
about 2.0-2. 1 times ASD, disc finely punctate, punctures in sub-
contiguous rows. Face and vertex moderately shiny between fine
to moderate, subcontiguous punctures of irregular shape that
become round and dense on maculate areas; gena shiny between
fine, irregularly shaped, subcontiguous punctures.

Contributions in Science, Number 328

Smelling: Hylaeus of Sri Lanka and India 15

Interocellar distance less than ocellocular distance. Facial
fovea ending less than halfway between eye and lateral ocellus.

Thorax. Mesoscutum about 1.3-1. 4 times wider than long.
Scutellum flattened, about 0.4 times length of mesoscutum.
Metanotum weakly convex.

Mesoscutum slightly shiny between moderate, subcontiguous
punctures; scutellum moderately shiny, punctures fine and sub-
contiguous in middle, moderate and dense on either side, meta-
notum slightly shiny between fine, dense punctures; mesopleuron
moderately shiny between fine to moderate, subcontiguous to
dense punctures; metapleuron moderately shiny, lineolate, and
finely, subcontiguously punctate. Propodeum without lateral,
oblique, and transverse carinae; side, stigmatal area, and disc
shiny between moderate, subcontiguous punctures; basal area
shiny, with a few longitudinal rugulae, which are evanescent
posteriorly.

Abdomen. Disc of first tergum shiny between ultraminute and
minute punctures, which vary widely in spacing, from close to
scattered; disc of second tergum moderately shiny between min-
ute, sparse punctures.

Pilosity. First tergum with apicolateral patch of dense, ap-
pressed pubescence and with erect hairs across anterior face.

Color. Blackish. The following bright yellow: large preapical
spot on clypeus; lateral face mark, upward along inner orbit to
lower end of facial fovea; pronotal collar (interrupted in middle
in holotype); most of pronotal lobe; tegular spot; basal three-
fourths of protibia externally; basal spot on mesotibia; basal half
of outer face of metatibia. Mandible mostly ferruginous.
Flagellum brownish, paler beneath. Wings clear, brownish, veins
and stigma dark brown.

TYPE MATERIAL

Flolotype female: Naduvatam, 6000 ft, Nilgiri Hills, S. INDIA,
May 1 958 (P.S. Nathan). Paratype female: Ammatti, 3 1 00 ft, S.
Coorg, S. INDIA, March 1952 (P.S. Nathan). Both specimens in
Natural History Museum of Los Angeles County.

Head. Broad, HW 1.10 x HL; scape short, SL 0.28 x HL; SL
3.25 x SW. Eyes moderately convergent below, UFW 1.43 x
LFW. Clypeus narrow, CW 0.89 x CL; BCW 0.68 x CW, 2.09 x
ASD, 1.21 x IAD, 1.44 x COD. Clypeus distinctly separated
from eye.

Clypeus dull to slightly shiny between fine to moderate, sub-
contiguous to dense punctures; supraclypeal area similar.
Frontal shield short and narrow, about 1.5 times longer than
wide; margins very weakly convex, subcarinate, not reflexed;
disc dull between fine, contiguous punctures; apex very narrow,
about 0.3 times ASD. Face slightly to moderately shiny be-
tween moderate, contiguous to subcontiguous punctures; inter-
ocellar area tumescent, weakly depressed along midline, area
on either side dull, densely tesselate, and nearly impunctate;
gena moderately shiny between fine to moderate, somewhat
elongate, contiguous to subcontiguous punctures.

Interocellar distance greater than ocellocular distance. Facia!
fovea ending about one-third of distance between eye and lat-
eral ocellus.

Thorax. Mesoscutum about 1.3 times wider than long. Scutel-
lum flattened, about 0.4 times length of mesoscutum. Meta-
notum weakly convex.

Mesoscutum slightly shiny between fine to moderate, dense
punctures; scutellum a little shinier, punctures fine to moderate,
dense to close, but subcontiguous in middle; metanotum dull
between subcontiguous to close, minute punctures; meso-
pleuron moderately shiny between moderate, subcontiguous to
dense punctures, which are shiny within; metapleuron slightly
shiny between fine, contiguous punctures. Transverse and
oblique propodeal carinae absent, lateral carina present but
weak in upper two-thirds; side and disc slightly shiny, moder-
ately to coarsely rugosopunctate, grading into coarsely rugose
stigmatal area; basal area shiny, with a transverse row of
coarse, quadrate areolae, behind which is a series of widely
spaced, very irregular, transversely oriented, coarse to very
coarse areolae.

ETYMOLOGY

From the Greek eurys (broad) and gnathos (jaw), because of the
peculiar mandibles.

DISCUSSION

The only known species with similar mandibular structure is H.
sedens, described above. In that species, the mandibular struc-
ture is different, and the macula are paler whitish.

Hylaeus pannuceus new species

Figure 30

DIAGNOSIS

Female. Interocellar area with a pair of low, densely tesselate
tumescences; frontal shield short, narrow; basal area of pro-
podeum transversely rugose.

Male. Unknown.

Abdomen. Disc of first tergum shiny between minute to fine
punctures, very irregularly spaced (subcontiguous to scattered);
disc of second tergum moderately shiny between minute sparse
to scattered punctures.

Pilosity. First tergum with long apicolateral patch of dense ap-
pressed white pubescence.

Color. Blackish. The following very pale yellowish: large pre-
apical clypeal spot; large transverse supraclypeal mark; lateral
face mark, ending on eye margin at lower end of fovea; prono-
tal collar, narrowly interrupted in middle; almost entire prono-
tal lobe; large tegular spot; stripe on outer three-fourths of
protibia; stripe on outer half of mesotibia; basal half of meta-
tibia. Tarsal segments brownish. Preapical one-third of clypeus
and most of mandible dark reddish. Flagellum reddish brown
above, lighter beneath. Wings clear, slightly brownish, veins
and stigma medium to dark brown.

DESCRIPTION TYPE MATERIAL

Measurements (mm). HL 1 .55; HW 1.70; WL 7.3; TL 5.5. Holotype female: Jabalpur, 1600 ft, central INDIA, October

16 Contributions in Science, Number 328

Snelling: Hylaeus of Sri Lanka and India

1957 (PS. Nathan), in Natural History Museum of Los An-
geles County.

ETYMOLOGY

The specific name is a Latin word for wrinkled, in allusion to
the transversely wrinkled propodeal base.

DISCUSSION

The propodeal structure and tumescent interocellar area seem
to be unique within the Indian fauna. It is possible that H. por-
catus, described above, is the male, but this seems unlikely. In
that species, the metanotum is rugulose, the first tergum lacks
an apicolateral hair patch, and the propodeal side is not rugoso-
punctate. The opposite sexes of Hylaeus usually more or less
agree in these characters.

The species treated and described above may be separated
by the following key.

KEY TO HYLAEUS
OF SOUTH INDIA AND
SRI LANKA

la. Antenna 13 segmented (male) 2

b. Antenna 12 segmented (female) 7

2a. Lateral margin of clypeus contiguous with inner eye mar-
gin (Figs. 1, 6); oblique propodeal carina present 3

b. Lateral margin of clypeus separated from inner eye mar-
gin by at least 0.3 times OD (Figs. 10, 14); oblique pro-
podeal carina absent 4

3a. Mandible and labrum mostly yellow; tegular spot present;
disc of first tergum subcontiguously to densely punctate;

mesoscutal punctures fine krombeini

b. Mandible and labrum with minute yellow spot; tegular
spot absent; disc of first tergum closely to sparsely punc-
tate; mesoscutal punctures moderate oresbius

4a. Basal area of propodeum with a few weak longitudinal

rugulae 5

b. Basal area of propodeum with sharp, transverse ridges ....
porcatus

5a. Distance between lateral margin of clypeus and inner eye
margin no more than 0.5 times OD; supraclypeal shield no
longer than greatest width; length of second flagellar seg-
ment at least 1.5 times first 6

b. Distance between lateral margin of clypeus and inner eye
margin subequal to OD; supraclypeal shield longer than
greatest width; second flagellar segment slightly longer

than first thyreus

6a. Preapical tooth large, distinct, macula whitish, scape im-
maculate; metapleuron mostly smooth and shiny

parmatus

b. Preapical tooth very low, obscure; macula yellowish, scape
maculate; metapleuron finely, subcontiguously punctate. ..
pel fates

7a. Lateral margin of clypeus distinctly separated from inner
eye margin; oblique propodeal carina absent 8

b. Lateral margin of clypeus contiguous with inner eye mar-
gin; oblique propodeal carina present krombeini

8a. Interocellar area uniformly and subcontiguously punctate,
not swollen; outer face of mandible without ridges or

grooves, minutely striate in part 9

b. Interocellar area swollen and bearing a dull, sparsely punc-
tate area on each side; outer face of mandible with distinct
ridges and grooves, without minute striations . . .pannuceus

9a. Preapical notch of mandible weak; markings yellow

eurygnathus

b. Preapical notch of mandible strong, a distinct preapical
tooth present; markings whitish sedens

GENERAL DISCUSSION

Although the material treated here obviously involves several
subgenera, any attempt to define subgenera within this fauna
would be premature. When more material becomes available
and sexes are definitely associated, such a study could be
initiated.

The Holarctic subgenus Paraprosopis is clearly represented
by H. krombeini. Superficially similar to H. krombeini is H.
oresbius and, in nearly every feature defining subgenera in
Hylaeus, this species, too, would be placed in Paraprosopis.
The eighth sternum, however, is definitely not like that of other
Paraprosopis since it is not apically bifurcate (Fig. 8). For the
present, H. oresbius may be tentatively assigned to Paraproso-
pis. but may ultimately prove to be an aberrant species of the
subgenus Prosopis.

None of the remaining species appear to belong to existing
subgenera as they are currently defined. Among the females,
H. eurygnathus and H. sedens clearly belong together; H pan-
nuceus is not related and would form another group.

The broad-mandible males (H. thyreus, H. parmatus, and H
peltates) are very similar in external features and at first would
seem to form a natural group in which the females would be
represented by such species as H. eurygnathus and H. sedens.
Unfortunately, H. thyreus does not fit this scheme, for the sev-
enth sternum (Fig. 15) is very different from that of the other
two species known only from males.

Possibly two subgenera are involved. One would include H.
thyreus, H. eurygnathus, and H. sedens. The second would ac-
commodate H. peltates and H. parmatus: females for this
group are unknown if I have correctly associated the two broad-
mandible female species with H. thyreus. Considerably more
material will be necessary before this problem can be clarified;
there are likely more species with broad mandibles, and possi-
bly there are annectant forms between what now appear, based
on males, to be separate groups.

A final group would be represented by H. porcatus. It is pos-
sible that this species belongs to the Holarctic subgenus Pro-
sopis and that the peculiarly narrow, elongate eighth sternum
(Fig. 12) is merely an exaggeration of that of Prosopis. How-
ever, the distally narrowed gonocoxites of H. porcatus (Fig. 13)
are also unlike those of Prosopis.

Although the material available for study does include spe-
cies with unusual characteristics (broad mandibles, elongate

Contributions in Science, Number 328

Snelling: Hylaeus of Sri Lanka and India 17

gonocoxites), the affinities of these species seem to be with the
Palearctic hylaeine fauna rather than with groups originating in
southeast Asia or the Australian-Papuan faunas.

LITERATURE CITED

Houston, T.F. 1975. A revision of the Australian hylaeine bees
(Hymenoptera: Colletidae). 1 Austr. Jour. Zool., Suppl.
ser. 36:1-135.

Meade-Waldo, G. 1923. Fam. Apidae, Subfam. Prosopidinae.
In P. Wytsman, Genera Insectorum, Brussels, fasc. 181:1 —
45.

Michener, C.D. 1965. A classification of the bees of the Aus-
tralian and South Pacific Regions. Amer. Mus. Nat. Hist.
Bull. 130:1-362.

Snelling, R.R. 1970. Studies on North American bees of the
genus Hylaeus. 5. The subgenera Hylaeus, s. str. and Para-
prosopis. Los Angeles County Mus., Contrib. Sci. 180:1-
59.

Warncke, K. 1970. Beitrag zur systematik und verbreitung der
bienengattung Prosopis F. in der Westpalaarktis
(Hymenoptera, Apoidea, Colletidae). Bull. Rech. Agron.
Gembloux 5:745-768.

Accepted for publication 19 June 1980.

18 Contributions in Science, Number 328

Snelling: Hylaeus of Sri Lanka and India

Q

11

L52X

NH

Number 325)
29 August 1980

III

i

L : : . . . .<

I

EOCENE NEOSELACHIANS FROM THE LA MESETA FORMATION,

SEYMOUR ISLAND, ANTARCTIC PENINSULA

El

Bruce J. Welton

and William J. Zinsmeister

B

■ ■

SERIAL PUBLICATIONS
OF THE NATURAL HISTORY MUSEUM OF LOS ANGELES COUNTY

at irregular intervals in three major series; the articles in each series are numbered individually, and
numbers run consecutively, regardless of subject matter:

• Contributions in Science, a miscellaneous series of technical papers describing original
research in the life and earth sciences.

• Science Bulletins, a miscellaneous series of monographs describing original research in the
life and earth sciences. This series was discontinued in 1978 with the issue of Numbers 29
and 30; monographs are now published by the Museum in the Contributions in Science

series.

• Science Series, long articles on natural history topics, generally written for the layman.
Copies of the publications in these series are available on an exchange basis to institutions and
individual researchers. Copies are also sold through the Museum Bookshop.

■

ill

ll!i

Kill

■ i

asm

if

\\ :

Milii

,:il

nil M

‘iraiill

Hill

iliillli

“li

*

'fev Castle Press-, Pasadena, Catift

' ■

■

1 1

EOCENE NEOSELACHIANS FROM THE LA MESETA FORMATION,
SEYMOUR ISLAND, ANTARCTIC PENINSULA1

Bruce J. Welton2 and William J. Zimsmeister3

ABSTRACT: Neoselachian teeth from the La Meseta Formation,
Seymour Island, Antarctic Peninsula, include those of the following
taxa: Carcharodon auriculatus (Blainville 1816), Eugomphodus ma-
crota (Agassiz 1843), Lamnidae indeterminate, Squalidae indetermi-
nate, Squatina sp., and Myliobatoidea indeterminate.

The lamniforms Eugomphodus macrota and Carcharodon au-
riculatus are cosmopolitan species and indicate a middle to late Eocene
and possibly early Oligocene age for the La Meseta Formation.

INTRODUCTION

During the austral summer of 1974-1975, a joint field party
from the Institute of Polar Studies at The Ohio State Univer-
sity and the Argentine Antarctic Institute visited Seymour Is-
land, located on the northeast side of the Antarctic Peninsula.
A geological survey of the Tertiary sequence at the northern
end of Seymour Island was undertaken by William Zinsmeister
and, during the course of the season, a small collection of shark
and ray teeth was made from Unit II of the La Meseta Forma-
tion. The fossils are deposited in the Museum of Paleontology
of the University of California, Berkeley (UCMP) and in the
Institute of Polar Studies at The Ohio State University, Colum-
bus, Ohio. The description and taxonomic evaluation of these
fossils constitute the basis for this paper.

REVIEW OF

PREVIOUSLY DESCRIBED SHARKS
FROM SEYMOUR ISLAND

Cretaceous and Eocene fishes were first collected from
Seymour Island by the Swedish South Polar Expedition be-
tween the years 1901 and 1903. Woodward (1906) described
several large selachian vertebrae from Cretaceous material
contained in the above collection and referred them to the
genus Ptychodus Agassiz (1835). He based this identification
upon shared characters of internal and external calcification
observed in the Seymour Island specimens and in vertebrae
found in association with fossil teeth of Ptychodus decurrens
Agassiz 1835 from the English Chalk (Woodward 1902:228-29,
pi 52, fig. 6). The presence of large, biconcave vertebrae with a
very short anteroposterior length, thin concentric calcified
lamellae with perforated walls within each intermedialia,
weakly developed radii, and basilia for neural and hemal arch
cartilages are characters that Woodward (1906:1-2) used to di-
agnose vertebrae of Ptychodus. As Woodward (1902:229)
noted, this calcification pattern is also very similar to that
found in vertebrae of Corax (= Squalicorax). In addition, con-
centric calcification patterns are also known in modern and fos-

Contributions in Science, Number 329, pp. 1-10
Natural History Museum of Los Angeles County, 1980

sil species of the basking shark Cetorhinus (Hasse, 1882, pi. 32,
figs. 1-8; Ridewood 1921:360, fig. 13a; White 1935:79), the an-
gel shark Squatina (Ridewood 1921:376-78, fig. 25b), and the
whale shark Rhincodon typus (White 1930:136, pi. 29, fig. m).

Vertebrae found in association with teeth of Ptychodus chap-
pelli Reinhart (1951) from the Turonian of San Luis, Tolima,
Colombia (UCMP 39525), possess numerous thin concentric
lamellae; however, the vertebrae are proportionately antero-
posteriorly longer and dorsoventrally shorter than those at-
tributed to Ptychodus by Woodward (1902:228-29) from either
the English Chalk or the Cretaceous rocks of Seymour Island.
At present, articulated skeletons of Ptychodus are not known,
and the finding of teeth and vertebrae in association should not
be regarded as demonstrable proof that all are from the same
genus.

The presence of relatively undistorted concentric calcifica-
tions in the vertebrae from Seymour Island (Woodward 1906,
pi. 1 , figs. 1 -3) suggests that their short anteroposterior length is
not due to compression from sediment load. In most respects,
the Seymour Island vertebrae resemble those found in an artic-
ulated skeleton of Squalicorax cf. S. kaupi from the Niobrara
Formation of Kansas (Section of Vertebrate Paleontology, Natu-
ral History Museum of Los Angeles, County, LACM 120090).
In light of the above data, continued allocation of the Seymour
Island vertebrae to the genus Ptychodus must be considered
tenuous.

Elliot et al. (1975) were the first to mention the occurrence
of fossil shark teeth in the early Tertiary rocks of Seymour Is-
land. Subsequently, del Valle et al. (1976:8) identified Scapano-
rhynchus raphiodon (Agassiz) 1844, S', subulatus (Agassiz)
1844, Isurus mantelli Agassiz 1843, Isurus sp., and Carcharias
sp. based upon a small number of fragmentary shark teeth from
the La Meseta Formation. Unfortunately, their taxonomy was
based largely upon Ameghino’s incorrect identifications of
teeth from the Patagonian marine formation of Argentina
(Ameghino 1906). Cione et al. (1977:9-14) published the first
comprehensive description of sharks from the La Meseta For-
mation and recognized three genera and two species: Eu-

1. Review Committee for this contribution: Lawrence G. Barnes,
Leonard J.V. Compagno, and David J. Ward.

2. Section of Vertebrate Paleontology, Natural History Museum of
Los Angeles County, Los Angeles, California 90007

3. Institute of Polar Studies, The Ohio State University, Columbus,
Ohio 43210, and Research Associate, Natural History Museum of Los
Angeles County

ISSN 0459-01 1 3

gomphodus macrota (Agassiz 1843), Isurus novusl (Winkler
1873), and Procarcharodon Casier 1960. As a result of the pre-
sent study, five taxa of neoselachians are recognized from Unit
II of the La Meseta Formation. In addition to Carcharodon au-
riculatus and Eugomphodus macrota , the angel shark Squat-
ina, an indeterminate squaloid close to Centrophorus Muller
and Henle 1837, and a myliobatoid ray are recorded for the
first time.

LOCALITY

UCMP V77014, Seymour Island, Antarctic Peninsula,
64°15'S, 56°45'W. Fossil shark and ray teeth were collected
from a coarse, pebbly, fossiliferous shell bank in Unit II of the
La Meseta Formation, approximately 1 50 m above the base of
the Seymour Island Series on the north side of the island, di-
rectly across from Corkburn Island (6.5 km east, 3 degrees
from Bodman Point and 4.1 km south, 61 degrees from Cape
Wiman). The shell bank, being slightly more resistant to weath-
ering, forms a distinct break in slope that can be traced for
several kilometers along strike (Fig. 1).

STRATIGRAPHY

The sequence of tertiary rocks on Seymour Island was first de-
scribed by Anderson (1906). He proposed the name “Seymour
Island Series” for the loosely consolidated sands and sandy silt-
stones of Tertiary age at the north end of the island (Fig. 1).
Elliot and Trautman (in press) have proposed that the term
“Seymour Island Series” be dropped in favor of “Seymour Is-
land Group,” which includes the Cross Valley and La Meseta
Formations.

Since the present paper is primarily concerned with the fossil
shark assemblage in the La Meseta Formation, only a brief de-
scription of the Cross Valley Formation is given. For a more
detailed description of the stratigraphy of the Cross Valley For-
mation, see Elliot and Trautman (in press). The Cross Valley
Formation crops out in a series of fault splinters in Cross Valley
(Fig. 1). It consists of approximately 105 cm of immature sand-
stones and pebbly sandstones unconformably overlying the Up-
per Cretaceous “Snow Hill Island Series.” The sands of the
Cross Valley Formation contain a high percentage of volcanic
glass and pumice with locally abundant concentrations of fossil
wood and plant debris near the base, and the Cross Valley For-
mation is interpreted as representing a predominantly non-
marine deltaic deposit. The presence of marine molluscs near
the top of the formation indicates that its uppermost part is
marine.

The La Meseta Formation consists of at least 450 m of
loosely consolidated sandstones, sandy siltstones, and interbed-
ded conglomerates. Discontinuous concretionary horizons are
present throughout the section. Elliot and Trautman (in press)
have informally divided the formation into three lithologic
units. Unit I, near the base of the formation, consists of at least
160 m of unconsolidated fine sands and silty sands. Unit II con-
sists of approximately 200 m of fine laminated sands with
prominent discontinuous fossiliferous conglomerates that may
extend for a kilometer along strike. These highly fossiliferous

conglomerates are interpreted as representing accumulations in
a nearshore, high-energy marine environment. All the shark
material described in this paper was surface-collected from the
lowermost shell banks at the base of Unit II. Local cross-bed-
ding, oscillation ripple marks, and small to large cut and fill
channels are present in the finer grained facies of Unit II.

Unit III, the upper 80 to 125 m of the La Meseta Formation,
is characterized by fine sands with intervals of fine clays and
sandy gravel horizons. All fossil penguin material collected dur-
ing the 1974-1975 field season came from Unit III (Elliot et al.
1975). The invertebrate fauna of this unit differs significantly
from that of Units I and II; the differences are of such a magni-
tude (Zinsmeister and Camacho, in press) that a hiatus of un-
known duration may have occurred between the deposition of
material in Units II and III.

AGE AND CORRELATION

There has been considerable controversy over the age of the
Tertiary rocks on Seymour Island. Until recently, Antarctic
workers have accepted Wilckens (1911) assignment of the se-
quence to the Miocene. Ihering (1927) suggested that the mol-
luscan faunas from Seymour Island were Eocene, and this
interpretation has been confirmed for at least part of the se-
quence in a recent series of papers (Simpson 1971; Zinsmeister
1977, 1978; Hall 1977). It appears from the available data that
the Tertiary rocks on the island range in age from Paleocene
(Cross Valley Formation) to late Eocene to possibly lowermost
Oligocene (La Meseta Formation), with the Early Eocene
missing.

The presence of two species of nautiloids (Aturia sp. and Eu-
trephoceras argentinae del Valle and Fourcade 1976) supports
Simpson’s 1971 assignment of the La Meseta Formation to the
Eucene. Aturia sp. from Unit II is very similar to A. bruggeni
Ihering from the Eocene of Tierra del Fuego. The occurrence of
the genus Eutrephoceras in the La Meseta Formation is an
important indication of the age of the fauna. Except for a single
species of Eutrephoceras in the Oligocene from the west coast
of North America, the genus is confined to the Paleocene and
Eocene of the western hemisphere (Zinsmeister and Camacho,
in press). In a discussion of some remarkably well-preserved
specimens of Eutrephoceras from the La Meseta Formation,
Zinsmeister (1978) showed that E. argentinae is very similar to
the Eocene E. allani Fleming from New Zealand.

The data summarized in this study and the joint occurrence
of two cosmopolitan lamnoid species, Carcharodon auriculatus
and Eugomphodus macrota , in Unit II of the La Meseta For-
mation strongly suggest a middle to late Eocene age (Lutetian
to Bartonian Correlative) for Unit II deposits. A middle to late
Eocene age is also indicated for Unit III by the occurrence of a
mandible of an archaeocete whale at this horizon (Elliot et al.
1975).

The shark and ray assemblage of Unit II of the La Meseta
Formation falls within the stratigraphic interval included in the
provisional Antarctodarwinella nordenskjoldi molluscan zone
of Zinsmeister and Camacho (in press). Examination of pal-
ynomorphs from Seymour Island suggests that the La Meseta
Formation is correlative with the Rio Turbio Formation of

2 Contributions in Science, Number 329

Welton and Zinsmeister: Seymour Island Neoselachians

southwest Patagonia and the Lena Dura Formation of Tierra
del Fuego (Hall 1977).

SYSTEMATIC PALEONTOLOGY

Family Squalidae Leach 1818
Genus and species indeterminate

Figure 2

REFERRED SPECIMEN: UCMP 121795, one worn lower left
anterolateral tooth lacking most of the root and labial crown
flange.

DESCRIPTION: Tooth large, mesodistal crown length 6.02
mm; cusp broad-based, triangular, and distally inclined at an
angle of 53 degrees; mesial cutting edge weakly sinuous, almost
straight and smooth; distal cutting edge weakly convex and
basally serrated; distal blade well developed, apically flat with
low distobasal inclination, and mesially serrated; cusplets ab-
sent; labial flange present, but size, shape, and basal develop-
ment indeterminate because of poor preservation and breakage;
labial crown root forming moderately strong ledge and bor-
dered basally by five foramina on the distal root lobe; lingual

crown face very badly eroded, obscuring all details of the cen-
tral lingual foramen, lingual protuberance, and crown-root rela-
tionships; distal crown depression well developed as in
Centrophorus and many other squaloids; root almost com-
pletely absent and lacking all taxonomically significant
characters.

DISCUSSION: The intermediate apicobasal crown height,
moderate distal cusp inclination, presence of a well-developed
distolingual depression for the articulation of the adjacent (next
distal) tooth, and absence of cusplets are characters that, when
taken in combination, distinguish this specimen from upper or
lower teeth belonging to species of Centroscyllium , Etmop-
terus , Dalatias, Scymnodon , Heteroscymnoides , Scym-
nodalatias, some Centroscymnus, Oxynotus , Squaliolus , Eu-
protomicrus , Euprotomicroides , Aculeola, Somniosus , and
Isistius. The absence of a distinct short and detached flange,
the height of the distal root lobe (as preserved), and the pres-
ence of a strong distolingual depression separate this specimen
from Squalus, Cenlrosqualus , and Cirrhigaleus. In most char-
acters, UCMP 121795 most closely resembles the lower antero-
lateral teeth of Centrophorus or Deania.

— i —

0°

Atlantic

Ocean

Seymour-"-;

Island

South | .» ^//

o '?!$ / Warl Ha II

America 'l Antarctica

%

Africa

Australia

La Meseta Formation
(A— A1)

Cross Valley
Formation
^ (B-B1)

100-

Tr~T^ Upper Tertiary to Quaternary glacial
deposits on top of the meseta
| La Meseta Formation

t-V-l Cross Valley Formation

Y/X Cretaceous strata

^ Tertiary Cretaceous undifferentiated

0 Km 5

1 I I J I I

Shell bank

S Pebble horizon

l/'/'v ; •] Conglomerate

| Resistant sandstone

| Unconsolidated sand

i— j Unconsolidated sand, silty
^ — clay, clayey sand

Figure 1. Index map and geologic map showing the distribution of the Cretaceous, Tertiary, and Quaternary rocks on Seymour Island, Antarctic
Peninsula. The collecting locality, UCMP V77014, for the fossil sharks described in this paper is shown on both the map and the stratigraphic section
A-A1.

Contributions in Science, Number 329

Welton and Zinsmeister: Seymour Island Neoselachians 3

Figure 2. Squalidae, genus and species indeterminate, UCMP 121795;
incomplete lower left anterolateral tooth: a, lingual view; b, labial view.
Scale line - 1 mm.

Root morphology is extremely important for determination
of specific, generic, and subfamilial levels of identification
among the Squalidae. The incomplete root in the Seymour Is-
land specimen precludes a more specific identification.

Family Squatinidae Bonaparte 1838
Genus Squatina Risso 1810
Squatina sp.

Figure 3

REFERRED SPECIMENS: UCMP 121796, one incomplete
tooth with a worn cusp apex and lacking tip of mesial root lobe;
UCMP 121797, one incomplete tooth with a badly eroded root
and lacking cusp apex.

DESCRIPTION: Teeth small, mesodistal root length 5.16 +
mm (UCMP 121796) and 4.94+ mm (UCMP 121797); api-
cobasal tooth height 3.58+ mm (UCMP (121796) and 2.59 +
mm (UCMP 121797); cusp short, broad based, weakly inclined
distally, and weakly recurved labially; lingual and labial faces
strongly convex, smooth; transverse ridges and grooves absent;
labial flange weak, basally rectangular (as preserved, UCMP

121796) or weakly rounded (UCMP 121797), with a strong,
short median apicobasal ridge on the labial flange face that
does not extend onto cusp; mesial and distal cutting edges of
cusp continuous from apex to mesial and distal edges of blades;
blades low and narrow; root typically Squatina ; central lingual
protuberance prominent with a strongly developed labiolingual
ridge extending to crown foot and laterally below each blade;
lateral root ridges bordered by numerous irregular foramina
(UCMP 121796); large foramina on apicolingual surface of
root prominence; basal face of root moderately concave with a
very large central basal foramen; labial root face apicobasally
low with deep grooves paralleling crown foot both mesially and
distally.

DISCUSSION: UCMP 121796 is distinguished from Orec-
tolobus Bonaparte 1837 and similar to Squatina in having a
short tapering cusp, in lacking an elevated mesial or distal
blade, in having mesodistally narrow rather than expanded and
tabular root lobes; and in possessing a deep pit rather than a
shallow, generally triangular, lingually pointed transverse
groove on the midline of the labiobasal root surface.

Features of the crown shape — its short broad base, apically
acute cusp, short and rectangular labial flange, and moderately
concave basal root face — are characters that appear to separate
these specimens from the Paleocene-Eocene Squatina prima
(Winkler 1873), as well as from all other nominal fossil species
of Squatina.

Difficulties in recognizing dental differences between species
of Squatina are best overcome by analysis of large numbers of
teeth. It is our contention that the poorly preserved Seymour
Island specimens do not provide a sufficient morphological
basis for such an analysis and that it would not be prudent at
this time to make a specific determination.

Family Odontaspididae Muller and Henle 1841
Genus Eugomphodus Gill 1861
Eugomphodus macrota (Agassiz, 1843)

Figures 4g through p

SYNONYMS: Otodus macrotus Agassiz 1843:273. Lamna ele-
gans Agassiz 1843:289. Odontaspis (Synodontaspis) macrota
Casier 1946:66. Casier 1958:18. Striatolamia macrota Glick-
man 1964:102. Odontaspis ( Synodontaspis ) macrota Casier
1966:69-71. Odontaspis macrota Applegate 1968:32-36. Odon-
taspis ( Synodontaspis ) macrota Edwards and Stinton
1971:449-53. Odontaspis macrota Cione et al. 1977:10. For a
more extensive synonomy encompassing the years 1889-1946,
see Casier 1946:66.

REFERRED SPECIMENS: UCMP 1 16454-58, incomplete an-
terior teeth; UCMP 116459 and 116460, incomplete lateral
teeth; and 225 broken tooth crowns in the collection of The
Ohio State University, Institute of Polar Studies.

DESCRIPTION: The sample includes numerous broken ante-
rior and lateral teeth (tooth group terminology follows Apple-
gate 1965 and Compagno 1970). Symphyseals, intermediates,
and posteriors are not present in the sample. Crowns of the an-
terior teeth are small (15 mm, UCMP 116456) to very large
(54+ mm, UCMP 1 16454); A2, A3, and ?L' mesiodistally very
broad and labiolingually compressed. Labial faces weakly con-

4 Contributions in Science, Number 329

Welton and Zinsmeister: Seymour Island Neoselachians

Figure 3. Squalina sp., UCMP 121796; anterolateral tooth; a, basal view; b, labial view; c, distal view; d, apical view; e, lingual
view. Scale line = 1 mm.

vex or almost flat; crown foot nearly smooth (UCMP 1 16454
and 116458) or with well-developed, short, shallow to deep
transverse grooves extending from the crown foot in an apical
direction for a distance of approximately 5 to 13 percent of the
crown height (UCMP 116455, 1 16459, and 116460); lingual
face moderately to strongly convex just apical to crown foot;
lingual transverse ridges parallel to subparallel, strong (UCMP
1 16455, 1 16457) to weakly developed (UCMP 116454,
1 16458-60) and extending from crown foot almost to cusp apex
(UCMP 1 16458) or well developed near crown foot and becom-
ing very faint toward distal half of the crown; mesial and distal
cutting edges extend from crown apex to crown foot in laterals
(UCMP 1 16459 and 1 16460) but are not continuous to crown
foot in any of the anterior tooth positions; cusplets of anteriors
very small, conical — there is only one on the mesial and distal
edges of the crown foot; lingual neck (term after Glickman
1964a and b) narrow but well developed on all anterior teeth;
roots robust, anteriors with weak to strong lingual pro-
tuberance; root morphology variable according to tooth
position.

DISCUSSION: Although none of the teeth are complete, most
of the important morphological characters are well enough pre-
served to allow for definitive taxonomic assignment. In com-
bination, the broad flat lingual crown face with weakly devel-
oped transverse ridges, one pair of small conical lateral
cusplets, large and robust lingual protuberance, nearly flat or
weakly concave labial crown face, and extremely large anterior
and lateral teeth are characters consistent with Eugomphodus

macrota (Agassiz 1843) and separate this taxon from all other
species of Eugomphodus.

Family Lamnidae Muller and Henle 1838
Genus Carcharodon Muller and Henle 1841
Carcharodon auriculatus (Blainville 1816)
Figures 4a through c

SYNONYMS: Squalus auriculatus Blainville 1816:384. Car-
charodon auriculatus Agassiz 1843:254. Procarcharodon au-
riculatus Casier 1960:13. Otodus auriculatus Glickman
1 964a : 1 14-16. Carcharodon auriculatus Keyes 1972:237-40.
Procarcharodon sp. Cione et al. 1977:13.

REFERRED SPECIMEN: UCMP 116453, an incomplete ante-
rior tooth lacking one cusplet, portions of both root lobes, and
part of the ingual crown foot and neck.

DESCRIPTION: Tooth large, greatest apical basal length 70 +
mm (as preserved, apex of cusp worn); crown high and narrow,
labial face smooth, moderately convex on mesial and distal
edges, nearly flat or slightly concave along the central axis just
above crown foot; labial face of crown strongly convex; trans-
verse grooves and ridges absent; mesial and distal cutting edges
coarsely serrated; serrations badly worn but appear to have
been pointed; cutting ridges extend basally to crown foot and
are continuous with one large coarsely serrated, bladelike, bi-
convex cusplet (one cusplet missing); root bilobate and narrow
but too poorly preserved to warrant further description.

DISCUSSION: In combination, a high narrow crown, coarse

Contributions in Science, Number 329

Welton and Zinsmeister: Seymour Island Neoselachians 5

pointed serration, and large lateral cusplets are characteristic
of Paleogene and early Neogene species of Carcharodon ( = C.
auriculatus and C. angustidens Agassiz). Carcharodon au-
riculatus has higher and narrower crowned anterior teeth than
C. angustidens, and UCMP 1 16453 is tentatively referred to
the former species.

Since the early 1800’s, approximately 54 nominal fossil spe-
cies of Carcharodon have been described, and many have sub-
sequently been recognized as junior synonyms of C. megalodon
Agassiz. Many of the remaining nominal species are of ques-
tionable validity, and their phylogenetic relationships to one an-
other and to the modern species of the genus Carcharodon, as

well as the interrelationships of the latter genus to all other
members of the Lamniformes, are poorly understood.

Casier (1960) deduced that the fossil species of Carcharodon
could be divided into two genera ( Procarcharodon and Car-
charodon) based entirely on the morphology of the teeth. As a
result, he removed all Paleogene and Miocene and some
Pliocene species from the genus Carcharodon, leaving it com-
posed of “phyletically interrelated” Pliocene to Recent species.
Based on this interpretation, Casier derived modern Car-
charodon from early Isurus hastalis ( Oxyrhina hastalis) and
then recognized a separate lineage for the remaining species of
Procarcharodon. Casier’s study necessarily emphasized the

Figure 4. Carcharodon auriculatus (Blainville 1816), UCMP 1 16453; a, lingual view; b, labial view; c, distal view; Myliobatoidea family indetermi-
nate, UCMP 1 16461; d, lingual view; e, basal view; f, apical view; Eugomphodus macrota (Agassiz 1843); g, UCMP 1 16459, lingual view; h, UCMP
1 16456, lingual view; i, UCMP 1 16460, labial view; UCMP 1 16454; j, lingual view; k, distal view; 1, labial view; UCMP 1 16455, m, lingual view; n,
labial view; o, mesial view; p, UCMP 1 16458, lingual view. Scale line = 10 mm.

6 Contributions in Science, Number 329

Welton and Zinsmeister; Seymour Island Neoselachians

dentition, because teeth form the major part of the fossil rec-
ord, but no attention was paid to details of ontogeny and com-
parison of homologous tooth positions, or most importantly,
dental formulae. We do not find Casier’s interpretations con-
vincing, and we agree with Keyes (1972) that the similarities
between /. hastalis and Carcharodon (in the restricted sense of
Casier 1960) might be indication of convergence rather than
phylogenetic relationship. The phylogeny of Carcharodon is
best interpreted through detailed analysis of tooth morphology
and dental formulae using specimens collected only under con-
ditions of demonstrable superpositional control. A detailed
study of ontogenetic heterodonty in the living C. carcharias and
Isurus spp. would be extremely useful in resolving these prob-
lems. We have chosen to retain the older nomenclature of Car-
charodon and reject, at this time, placement of C. auriculatus
in the genus Procarcharodon Casier.

Superfamily Myliobatoidea Compagno 1973
Family Indeterminate
Figures 4d through f

REFERRED SPECIMEN: UCMP 1 16461, one incomplete me-
dial tooth lacking distal end of crown and one side of root.

DESCRIPTION: Isolated, worn medial plate of typical my-
liobatoid morphology with a smooth (as preserved) occlusial
crown surface and polyaulacorhizous root (as defined by Casier
1947c). Tooth not chevron shaped in occlusial or basal view as
in the upper dentition of Aetobatis Blainville (see Garman
1913, pi. 49, figs. 1-3) but mesiodistally straight as in My-
liobatis Cuvier 1817, Aetomylaeus Garman 1908,
Pteromylaeus Garman 1913, and Rhinoptera Cuvier 1829. Oc-
clusial surface of tooth weakly concave in labial or lingual
view — apicobasal height only slightly greater in middle of tooth
than at either mesial or distal end.

DISCUSSION: Approximately 118 nominal fossil species of
myliobatoid rays have been described from Cenozoic marine
deposits throughout the world. Of the five extant genera of my-
liobatoid rays, Aetomylaeus has never been reported from the
fossil record. The apparent absence of a paleontological record
for this genus is a result of our past and present inability to
recognize generic differences in myliobatoid dentitions. Most
isolated or associated ray plates have been assigned to a species
in the genus Myliobatis, and what appears to be high specific
diversity among Cenozoic Myliobatis (approximately 91 nomi-
nal species) is in reality an artifact of this procedure. No doubt,
the 91 species are indicative of extreme oversplitting brought
about by (1) misinterpretation of individual variation for spe-
cific differences, (2) failure to decipher patterns of heterodonty,
and (3) failure to understand the comparative dental morphol-
ogy of extant myliobatoids.

A preliminary analysis of over 300 individual jaws of the ex-
tant California bat stingray Myliobatis californicus Gill from
two populations along the central and northern California
coasts revealed that the arrangement of tooth row-groups and
the number and morphology of teeth in each row are highly
variable (results of research by Leonard J.V. Compagno,
Tiburon Center for Environmental Studies, and Welton). In
fact, the range of variation within the tooth row-group pattern

found in Myliobatis californicus transcends virtually all sup-
posed “generically distinct” dental patterns of other my-
liobatoid rays (including the lower but not the upper dentition
of Aetobatis). In addition to displaying a surprisingly high level
of polymorphism, the M. californicus specimens have dental
abnormalities that range from total disruption of the typical
row-group pattern (some jaws bore a superficial resemblance to
those of the dasyatid “ Hypolophus Muller and Henle”) to the
presence of sinuous medials bearing the primary generic char-
acters of the Maestrichtian genus Igdabatis Cappetta 1972
(this same developmental abnormality also occurs in Rhinop-
tera steindachneri Evermann and Jenkins (1892), and a weakly
developed version of it was illustrated by Gudger (1933:75, fig.
10) for a Pliocene specimen of Myliobatis crassus Gervais
1859). Analysis of the large sample also revealed the presence
of dental sexual dimorphism: Males had labiolingually shorter
teeth than females of equivalent size. This finding has particu-
larly strong implications, as length-width ratios of fossil teeth
have been used as a method for distinguishing one taxon from
another. In M. californicus , at least, the length-width ratios of
teeth in the same row in the same jaw may vary widely, suggest-
ing, among other things, seasonal variation in the rate of tooth
production (Leonard J.V. Compagno, personal communi-
cation).

It appears at this time that specific assignment of UCMP
116461 by comparison with fossil dentitions of equivalent or
different age would have practically no validity and that its
placement in the superfamily Myliobatoidea without generic
allocation provides the best arrangement.

DISCUSSION AND CONCLUSIONS

The sharks from the La Meseta Formation that were de-
scribed but not illustrated by del Valle et al. (1976) were all
incorrectly identified according to the redescription and reiden-
tification of the same specimens by Cione et al. (1977). Accord-
ing to the latter authors, the shark assemblage of the La
Meseta Formation includes Eugomphodus macrota, Eu-
gomphodus sp., Isurus novusl, Isuridae indet., and Pro-
carcharodon sp.

We recognize the occurrence of Eugomphodus macrota and
Carcharodon ( = Procarcharodon of Cione et al. 1977) in the La
Meseta Formation, and a well preserved tooth of Carcharodon
is more specifically referred to Carcharodon auriculatus. A
tooth shown by Cione et al. (1977, fig. 3g) and identified as
Isurus novusl is too poorly preserved in our opinion to warrant
specific or generic identification and should be included under
Lamnidae indeterminate. Teeth referred to Eugomphodus sp.
(Cione et al. 1977, fig. 3a and b) are probably Eugomphodus
macrota. One indeterminate squaloid tooth, two poorly pre-
served teeth of Squatina sp., and disassociated pavement teeth
of a myliobatoid or rhinopterid stingray are new additions to
the neoselachian assemblage of the La Meseta Formation. The
total fossil shark assemblage from the La Meseta Formation,
therefore, now includes Eugomphodus macrota , Carcharodon
auriculatus, Lamnidae indeterminate, Squalidae indetermi-
nate, Squatina sp., and Myliobatoidea genus and species
indeterminate.

Contributions in Science, Number 329

Welton and Zinsmeister: Seymour Island Neoselachians 7

The geologic range of Carcharodon auriculatus differs de-
pending on whether the species is broadly or narrowly inter-
preted. According to Casier (1960:13), C. auriculatus ranged
approximately from middle to upper Eocene, and the species
ultimately evolved into C. angustidens in the Oligocene. Car-
charodon auriculatus was a cosmopolitan taxon and has been
reported from the Eocene and Oligocene of New Zealand
(Keyes 1972:238), from the middle and upper Eocene of North
America, Europe, and Africa (Avnimelech 1959:36-37, Casier
1960:13), and the Oligocene of Europe (Leriche 1910:291) and
Patagonia (Ameghino 1906:181). According to Glickman
(1964b), “ Otodus auriculatus ” occurs in deposits of Paleocene
to middle Oligocene age in the Volga area and the Ukraine,
Caucases, and Kazakhstan. Reports of Carcharodon au-
riculatus from the Oligocene to lower Miocene of Australia
(Chapman and Pritchard 1904; Chapman and Cudmore 1924;
Pledge 1967) may actually represent records for C.
anguistidens.

According to Casier (1946:67), Eugomphodus macrota
(Agassiz) has been reported from deposits of Ypresian to Barto-
nian age in Belgium, Northern France, the Paris Basin, and
England and from deposits of Lutetian age in Africa. Teeth of
Eugomphodus macrota are abundant in deposits of Lutetian
and Bartonian age in California and Oregon (Welton, un-
published data) and have also been reported from the Eocene
of the east coast of North America (Leriche 1942), Chile
(Schneider 1936), and the Volga Area, Ukraine, Central Asia,
and Kazakhstan (Glickman 1964a and b).

The three nonlamnoid taxa in the La Meseta Formation are
too incompletely known at this time to allow comparisons with
faunas elsewhere. Generally, myliobatoid rays and Squatina are
well represented in most Cenozoic neoselachian assemblages.
The tooth referable to ?Squalinae is of interest as it appears to
represent a non -Squalus squaloid — a taxon that has not been
recorded previously from Cenozoic rocks of South America,
New Zealand, or Australia.

ACKNOWLEDGMENTS

We thank the University of California’s Museum of Paleon-
tology for support of this study and the University’s
Department of Paleontology for use of their photographic
darkroom.

This study was aided by the staff of the Section of Vertebrate
Paleontology, Natural History Museum of Los Angeles County.
We are particularly grateful to Mrs. Dianne McDonald and
Mrs. Betty Marsh for sorting and picking fossils from bulk ma-
trix samples of the La Meseta Formation. The line drawings
and photographs were prepared by Welton. We also express our
gratitude to Argentina’s Direccion Nacional del Antartico and
Instituto Antartico Argentine for their field support of
Zinsmeister during the 1974-1975 field season. This work was
also supported by National Science Foundation Grant
OPP74-21 509 to The Ohio State University and the Institute of
Polar Studies.

We also thank Mr. David Ward, Dr. Leonard J.V. Compagno,

and Dr. Lawrence G. Barnes, for critically reading the

manuscript.

LITERATURE CITED

Agassiz, L. 1833-1845. Recherches sur les poissons fossile.
3:390.

Ameghino, F. 1906. Les formations sedimentaires du Cretace
superieur et du Tertiaire de Patagonia. An. Mus. Nac.
Buenos Aires, Ser. 3 8:1-508.

Anderson, J.G. 1906. On the geology of Graham Land. Upp-
sala Univ. Geol. Inst. Bull. 7:19-71

Applegate, S.P. 1965. Tooth terminology and variation in sharks
with special reference to the sand shark, Carcharias taurus
Rafinesque. Los Angeles County Museum, Contrib. Sci.
86:1-18.

. 1968. A large fossil sand shark of the genus Odon-

taspis from Oregon. State of Oregon Department of Geol-
ogy and Mineral Industries. The Ore Bin 30(2):32-36.

Avnimelech, M. 1959. Sur la presence de Carcharodon an-
gustidens Agassiz dens l’Eocene moyen mediterraneen.
Compt. Rend. Soc. geol. Fr. 2:36-37.

Blainville, H.M. de. 1816. Prodrome d’une nouvelle distribu-
tion systematique du regen animal. Bull. Sci. Soc. Phil.
Paris, pp. 1 1 3-24.

Cappetta, H. 1972. Les Poissons cretaces et tertiaires du Bassin
des Iullemmeden (Republique du Niger). Palaeover-
tebrata, Montpellier 5:179-251.

Casier, E. 1946. La faune ichthyologique de l’Ypresien de la
Belgique. Mem. Mus. roy. Hist. nat. Belg. 104:1-267.

1947a. Constitution et evolution de la racine den-

taire des Euselachii, 1: Note preliminaire. Bull. Mus. roy.
Hist. nat. Belg. 23(1 3): 1 - 1 5.

. 1947b. Constitution et evolution de la racine den-

taire des Euselachii, 2: Etude comparative des types. Bull.
Mus. roy. Hist. nat. Belg. 23( 1 4): 1 -32.

. 1947c. Constitution et evolution de la racine den-

taire des Euselachii, 3: Evolution des principaux carac-
teres morphologiques et conclusions. Bull. Mus. roy. Hist,
nat. Belg. 23(1 5): 1-45.

1958. Contribution a l’etude des Poissons fossiles

des Antilles. Mem. suisses de Paleontol. 74:1-95.

. 1960. Note sur la collection des Poissons Paleocene

et Eocenes de l’Enclave de Cabinda (Congo) recueillis par
J. Bequaert au cours de sa mission en 1913. Ann. Mus. roy.
Congo. Belg., Ser. 3 1(2): 1-48.

. 1966. Faune ichthyologique du London Clay.

Mem., Brit. Mus. Nat. Hist. London 1-496.

8 Contributions in Science, Number 329

Welton and Zinsmeister: Seymour Island Neoselachians

Chapman, F., and F.A. Cudmore. 1924. New or little-known
fossils in the National Museum, Part 27: Some Cainozoic
fish remains, with a revision of the group. Proc. Roy. Soc.
Viet. 36(2): 1 07-1 76.

Chapman, F., and G.B. Pritchard. 1904. Fossil fish remains
from the Tertiaries of Australia, part 1. Proc. Roy. Soc.
Viet. 17(1 ):267-97.

Cione, A.L., R.A. del Valle, C.A. del Valle, C.A. Rinaldi, and
E.P. Toni. 1977. Nota preliminar sobre los pinguinas y
tiburones del Terciario inferior de la isla Vicecomodoro
Marambio, Antartida. Inst. Antartico Argentino 213:3-21.

Compagno, L.J.V. 1970. Systematics of the genus Hemitriakis
(selachii: Carcharhinidae), and related genera. Proc. Calif.
Acad. Sci. (Ser. 4) 38:63-98.

1973. Interrelationships of living elsasmobranchs.

In Interrelationships of fishes, PH. Greenwood, R.S.
Miles, C. Patterson, eds. Suppl. No. 1, Zool. Jour. Linn.
Soc. 53:1 5-61 .

del Valle, R.A., H.H. Fourcade, and C.A. Rinaldi, 1976. Nota
sobre el hallazgo de una nueva especies de nautiloideo del
genero Eutrephoceras Owen, in la isla Vicecomodoro Mar-
ambio, Antartida. Inst. Antartico Argent. Contrib.
208:3-10.

Edwards, N., and F.C. Stinton. 1971. A fish fauna from the
lower Tertiary marine bed, Clapham Common, West Sus-
sex. Proc. Geol. Assoc. (London) 84(2):449-53.

Elliot, D.H., and T.A. Trautman. Lower Tertiary strata on
Seymour Island, Antarctic Peninsula. Proceedings of the
Third Symposium on Antarctic Geology and Geophysics,
22-27 August 1977, Madison, Wisconsin, in press.

Elliot, D.H., C. Rinaldi, W.J. Zinsmeister, T.A. Trautman,
W.A. Bryant, and R. del Valle. 1975. Geological investiga-
tions in Seymour Island, Antarctic Peninsula. Antarctic
Jour. U.S. 10:182-86.

Garman, S. 1908. New Plagiostomia and Chismopnea. Bull.
Mus. Comp. Zool. Harv. Coll. 51:251-56.

1913. The Plagiostoma. Mem. Mus. Comp. Zool.,

Harvard 36:1-515.

Glickman, L.S. 1964a. Akuly paleogena i ikh strat-
igrafischeskoe znachenie. Moskow-Laningrad: “Nauka”
Press. (Russian).

1964b. Subclass Elasmobranchii (Sharks). In Fun-
damentals of Paleontology, Yu. A. Orlov and D.V.
Obruchev, eds., vol. 11, pp. 292-352.

Gudger, E.W. 1933. Abnormal dentitions in rays, Batoidei.
Jour. Elisha Mitchell Sci. Soc. 49(l):57-96.

Hall, S.A. 1977. Cretaceous and Tertiary dinoflagellates from
Seymour Island, Antarctica. Nature 267(5608):239-41 .

Hasse, C. 1882. Das Natiirliche System der Elasmobranchier
auf Grundlage des Baues und der Entwicklung iher Wir-
belsaule, ’Jena, Besonderer Theil, ’Lief. 3:183-285.

Ihering, H. von. 1927. Die Kreide-Eocan Alblagerungen de
Antarktis. N. Jahrb. Min. Geol. Palaontol.:240-301 .

Keyes, I.W. 1972. New records of the elasmobranch C. mega-
lodon (Agassiz) and a review of the genus Carcharodon in
the New Zealand fossil record. N. Z. Jour. Geol. Geophys.
1 5(2):229-42.

Leriche, M. 1910. Les Poissons oligocenes de la Belgique.
Memoires Mus. roy. Hist. nat. Belg. 5:229-363.

. 1942. Contribution a l’etude des faunes ich-

thyologiques marines des terrains tertiaires de la plaine
cotiere atlantique et du centre des Etats Unis. Le syn-
chronisme des formations tertiaires des deux cotes de
l’Atlantique. Mem. Soc. geol. France 45:1-1 10.

Muller, J., and F.G.J. Henle. 1837. Gattungen der Haifische
und Rochen nach ihrer Arbeit, “Ueber die
Naturgeschichte der Knorpelfische”. Ber. Konig. preuss.
Akad. Wiss., Berlin.

Pledge, N.S. 1967. Fossil elasmobranch teeth of South Aus-
tralia and their stratigraphic distribution. Trans. Roy. Soc.
S. Aust. 91:135-60.

Reinhart, R.H. 1951. A new shark of the family Ptychodon-
tidae from South America. Univ. Calif. Pub. Geol. Sci.
28( 8): 1 95-202.

Ridewood, W.G. 1921. On the calcification of the vertebral
centra in sharks and rays. Phil. Trans. Roy. Soc. Lond. Ser.
B. 210:31 1-407.

Schneider, C.O. 1936. El Odontaspis elegans (Agassiz) en el
Terciaro Eoceno de Chile. Com. Mus. Conception
1-4:69-70.

Simpson, G.G. 1971. Review of fossil penguins from Seymour
Island. Trans. Roy. Soc. London, Ser. B. 178:357-87

White, E.G. 1930. The whale shark, Rhineodon typus. Descrip-
tion of the skeletal parts and classification based on the
Marathon specimen captured in 1923. Bull. Amer. Mus.
Nat. Hist. 61(4): 1 29-60.

. 1935. Interrelationships of the elasmobranchs with

a key to the Order Galea. Bull. Amer. Mus. Nat. Hist.
74(2):25-l 38.

Wilckens, O. 1911. Die mollusken der Antarkischen Tertiarfor-
mation. Wiss. Ergebn. Schwed. Sudpolar exped.
1901-1903 3( 1 3): 1 -62.

Winkler, T.C. 1873. Memoire sur des dents de poissons du ter-
rain bruxellien. Arch. Mus. Teyler, Haarlem 3(4): 1-10.

Woodward, A.S. 1902. The fossil fishes of the English Chalk,
Part 1. The Paleontolographical Society, London: 1-264.

1906. On fossil fish-remains from Snow Hill and

Seymour Islands. Schwed. Sudpolar Exped. 1901-1903
3(4): 1 -4.

Zinsmeister, W.J. 1977. Note on a new occurrence of the
Southern Hemisphere aporrhaid gastropod Struthioptera
Fenlay and Marivick on Seymour Island, Antarctica. Jour.
Paleontol. 51(2):399-404.

1978. Eocene nautiloid fauna from the La Meseta

Formation of Seymour Island, Antarctic Peninsula. Ant-
arctic Jour. U.S., 13(4).

Contributions in Science, Number 329

Welton and Zinsmeister: Seymour Island Neoselachians 9

Zinsmeister, W.J., and H.H. Camacho. Preliminary report on
the upper Eocene (to possibly lower Oligocene) molluscan
fauna of the La Meseta Formation of Seymour Island. Pro-
ceedings of the Third Symposium on Antarctic Geology
and Geophysics, August 22-27, 1977, Madison, Wisconsin,
in press.

Accepted for publication 19 June 1980.

10 Contributions in Science, Number 329

Welton and Zinsmeister: Seymour Island Neoselachians

If i to n * M

.::::. :>

• i;i;;i!

11

L52X

NH

Si

!!jfjfiiiili!i!ij!l

■

"fi-:

mmMMmrn

: '•' li

Hia

Hiiii'irH
KlSi'iti

ij’iii jiKfctfa : Hi

Number 330
IS September 1980

iMi

-ii’ ^ H11* • ’ *• : ?1 5t:i : ? v ! : :!>;f : f

- , • -jj m sjl $1 $5M ’•'■ Wpiil ! &: 128 ;

‘ '.; , !

jjl:lijww' *" ■'* - ! '

PM

f

m-4 1

MIM

r 1

i1 | '

1 :m

a

M

MlpISi® ii

pIMMM

hW

VI OF LGS ANGELES COUNTY

Cvr

ii

9

II

I'Cij ;

Itil

■

fE

mmm

Bi,i T
BfeVfe}!

SHliiSH

Biiigliii'i

jffKto : ■■ aw j

fjfrHir tH? fHl/tiBifiji! iMi! $![ 1 1

liliii

AVIAK PALEONTOLOGY
m.DEGARDE HOWARD

©ill

Mm...,

fc;l

m

11
ill 1, 1

Campbell,: Jr.

;ii»p

a

Ml

lIliiliiliHiti

few

||§|j|jfi “hDi! i4' M 111) 1 Vi

wmi'WItS loijii + 1 1 ;'!' 1 1 1 ■ . ' j

HwiFi® *1 I '

psij;: : . ? : i 'W^jWJWW' -wj--

;.I:J :

■=' ■■ !-:‘ : •:: ; ' it!":1;'-.

lliiliii

r !!,l : ? !'ur^Sls i

Ii

fehii

111 life1,

IS

m

' :|ii ifc iiif::'' '

|| WMMmiPisBai;

'ibfe few

wlilE rfw;1 nf,r

HH9:

bttTWn* *4 '! '(h ' \ ,!« ...I1 (1 > t«l»! \

H BlEKSlpfflif# n>? IS »i| P^!i!i| e #»

: f- ! Ah

'

■ ^ojE«? Ji 1 ,iu iji JI 1 8 ft,] 1 1 ' jdflJ1

Mnnil

|! ; I .

H

I

m \m ii i!!!i!iiji}{j;j VL
IHKi, mm iwraw$vts
iiv. i ,^1™

i;|3i mi mramnua

W

hilSiilS1

«i

JM OF LOS ANGELES COUNTY . 900 EXPOSITION BOULEVARD • COS ANGELKS, CALIFORNIA S0007

■

SERIAL PUBLICATIONS
OF THE NATURAL HISTORY MUSEUM
OF LOS ANGELES COUNTY

The scientific publications of the Natural History Museum of Los Angeles County have been
issued at irregular intervals in three major series; the articles in each series are numbered
individually, and numbers run consecutively, regardless of subject matter:

• Contributions in Science, a series of miscellaneous technical papers describing
original research in the life and earth sciences.

® Science Bulletins, a series of miscellaneous monographs describing original research
in the life and earth sciences. This series was discontinued in 1978 with the issue
of Numbers 29 and .50; monographs are now published by the Museum in the
Contributions in Science series.

• Science Series, long articles on natural history topics, generally written for the
layman.

Copies of the publications in these series are available on an exchange basis to institutions
and individual researchers. Copies are also sold through the Museum Bookshop.

Number 330
15 September 1980

CONTRIBUTIONS IN SCIENCE

NATURAL HISTORY MUSEUM OF LOS ANGELES COUNTY

PAPERS IN AVIAN PALEONTOLOGY
HONORING HILDEGARDE HOWARD

edited by Kenneth E. Campbell, Jr.

Published by the NATURAL HISTORY MUSEUM OF LOS ANGELES COUNTY • 900 EXPOSITION BOULEVARD • LOS ANGELES, CALIFORNIA 90007

ABSTRACT: Nineteen papers in avian paleontology — including theoretical as-

pects, faunal studies, reviews of specific groups, the description of several new forms,
and archaeological studies — are presented here to honor Hildegarde Howard. Preceding
these are appreciations by Theodore Downs, Jean Delacour, and Herbert Friedmann;
a review of the contributions of Hildegarde Howard by Kenneth E. Campbell, Jr.; the
bibliography of Hildegarde Howard; an index to avian taxa described by Hildegarde
Howard; and the illustrations of avian osteology from “The Avifauna of Emeryville
Shellmound” by Hildegarde Howard. George G. Simpson reviews the development of
the field of paleornithologv, discusses the evolution of birds, and reviews the evolution
of penguins. Joel Cracraft reviews the principles of cladistic analysis and discusses their
application to studies in avian paleontology. Cecile Mourer-Chauvire reviews the ar-
achaeotrogonids of the Eocene and Oligocene Phosphorites du Quercy, France, erecting
a separate family for the group and describing the fourth known species of Archaeo-
trogon. Ella Hoch reviews the middle Eocene oilshale deposits of Messel, West Ger-
many, and describes a new genus and species of shorebird with columboid features
from that site. Storrs L. Olson rediagnoses the family Plotopteridae Howard, describes
a new genus and species of plotopterid from late Oligocene deposits of Washington,
and discusses plotopterid adaptations. Kenneth E. Campbell, Jr., and Eduardo P.
Tonni review the family Teratornithidae L. Miller, describe a new genus and species
of teratorn from the Huavquerian (late Miocene) of Argentina, and briefly review ter-
atorn cranial adaptations. E.N. Kurochkin describes new species of Palaeoaramides,
Rallus, and Crex from middle Pliocene deposits of Western Mongolia, and comments
on what these fossil rails indicate as to the paleoclimate and paleoecology of that region.
Larry D. Martin and Robert M. Mengel describe a new species of Anser from the late
Pliocene (Blancan) Broadwater Local Fauna of western Nebraska, and reconstruct the
body proportions of the new goose by comparing the size of its limb bones to those of
Recent geese. Pierce Brodkorb describes a new species of Ardea from the Plio/Pleisto-
cene deposits of Shungura, Ethiopia, and changes the systematic position of four fossil
species previously assigned to the Ardeidae. Pat Vickers Rich reviews the family Dro-
mornithidae, extinct large ratites known from Miocene to late Pleistocene deposits of
Australia, and comments on relationships between the various groups of ratites. Ed-
uardo P. Tonni reviews the present state of knowledge of Cenozoic birds of Argentina
and presents preliminary data on new finds. Alan Feduccia describes a new species of
Burhinus from Pleistocene (Sangamon) deposits of Kansas and discusses the paleoeco-
logical and paleoclimatic implications of the presence of this tropical genus in North
America during the Pleistocene. Kenneth E. Campbell, Jr., reviews the Rancholabrean
avifauna of the Itchtucknee River, Florida, and notes the presence of the tropical genus
Milvago in Florida. David W. Steadman reviews the osteology and paleontology of all
known species of living and fossil turkeys, concluding that the Meleagridinae is com-
prised of three genera, Rhegminornis Wetmore, Proagriocharis Martin and Tate, and
Meleagris Linnaeus; that all diagnostic specimens of Blancan and younger ages, in-
cluding both living species, are referable to Meleagris \ and that M. gallopavo has been
present in southeastern United States since at least the Blancan. Amadeo M. Rea
analyzes turkey remains from 17 southwestern late Quaternary sites; concludes that all
pre-agricultural turkeys are referable to Meleagris crassipes ; and proposes that M.
gallopavo was imported into the southwest by paleoindians and became a feral popu-
lation, M . g. merriami, with the breakdown of southwestern cultures. Charmion R.
McKusick analyzes remains of three different forms of turkeys from southwestern ar-
chaeological sites, and discusses how the three forms may have developed. Paul Par-
malee analyzes the avian remains from Archaic and Fremont sites of Utah, and dis-
cusses how birds may have been utilized by the prehistoric inhabitants of Utah. Pat
Vickers Rich and A.R. McEvev describe a fossil Plain Wanderer (Pedionomidae), and
use the specimen to date the Morwell Fire-hole deposits of southeastern Victoria, Aus-
tralia. Lyndon L. Hargrave and Steven D. Emslie discuss the first Holocene records
of the Passenger Pigeon, Ectopistes migratorius, from New Mexico.

ISSN 0459-8113

PREFACE

In July 1980, Hildegarde Howard entered her fifty-second year with the Natural His-
tory Museum of Los Angeles County. She began her career by spending many of her
student years working with the Museum’s collections, and went on to become one of
the Museum’s best known and most respected scientists. Her interests in the Museum
and her chosen field, avian paleontology, have never diminished, and her continuing
research is an inspiration to us all. To honor her past achievements and to show our
appreciation for her continuing contributions to avian paleontology, we present this
festschrift.

The editor is especially indebted to Joanne Baker who dealt so effectively with the
preparation of the manuscripts and the voluminous correspondence related to the
festschrift. He also expresses his appreciation to the contributors who have worked so
hard to make this festschrift possible, and thanks Lidia Lustig and Antonia Tejada-
Flores for translation of foreign language manuscripts. The frontispiece is by Lawrence
Reynolds, Museum Photographer, who also provided copies of the earlier photographs
of Dr. Howard. The editor gratefully acknowledges the assistance of the following
people who formed the review committee for this publication:

William A. Akersten
Pierce Brodkorb
Charles T. Collins
Joel Cracraft
Theodore Downs
Alan Feduccia
Larry D. Martin
Robert M. Mengel
G. Victor Morejohn
Storrs L. Olson
John H. Ostrom

Paul W. Parmalee
Amadeo M. Rea
Pat Vickers Rich
George G. Simpson
David W. Steadman
Fred S. Truxal
Eduardo P Tonni
Stuart L. Warter
Elizabeth Wing
Glen Woolfenden

Grateful acknowledgment for funding of this publication is given to:

Mr. Ed N. Harrison

The Giles W. and Elise G. Mead Foundation

Natural History Museum of Los Angeles County Foundation

in

CONTENTS

Appreciations, by Theodore Downs, Jean Delacour, and Herbert Friedmann ... vii

The Contributions of Hildegarde Howard, by Kenneth E. Campbell, Jr xi

Bibliography of Hildegarde Howard xvii

Index to Avian Taxa Described by Hildegarde Howard xxv

Illustrations of Avian Osteology Taken from “The Avifauna of Emeryville Shell-

mound” by Hildegarde Howard xxvii

Fossil Birds and Evolution, by George Gaylord Simpson 3

Phylogenetic Theory and Methodology in Avian Paleontology: A Critical Ap-
praisal, by Joel Cracraft 9

The Archaeotrogonidae of the Eocene and Oligocene Phosphorites du Quercy

(France), by Cecile Mourer-Chauvire 17

A New Middle Eocene Shorebird (Aves: Charadriiformes, Charadrii) with

Columboid Features, by Ella Hoch 33

A New Genus of Penguin-like Pelecaniform Bird from the Oligocene of Wash-
ington (Pelecaniformes: Plotopteridae), by Storrs L. Olson 51

A New Genus of Teratorn from the Huayquerian of Argentina (Aves: Tera-

tornithidae), by Kenneth E. Campbell, Jr., and Eduardo P. Tonni 59

Middle Pliocene Rails from Western Mongolia, by E. N. Kurochkin 69

A New Goose from the Late Pliocene of Nebraska with Notes on Variability
and Proportions in Some Recent Geese, by Larry D. Martin and Robert

M. Mengel 75

A New Fossil Heron (Aves: Ardeidae) from the Omo Basin of Ethiopia, with
Remarks on the Position of Some Other Species Assigned to the Ardeidae,

by Pierce Brodkorb 87

The Australian Dromornithidae: A Group of Extinct Large Ratites, by Pat

Vickers Rich 93

The Present State of Knowledge of the Cenozoic Birds of Argentina, by

Eduardo P. Tonni 105

A Thick-knee (Aves: Burhinidae) from the Pleistocene of North America, and

Its Bearing on Ice Age Climates, by Alan Feduccia 115

A Review of the Rancholabrean Avifauna of the Itchtucknee River, Florida,

by Kenneth E. Campbell, Jr 119

A Review of the Osteology and Paleontology of Turkeys (Aves: Meleagridinae),

by David W. Steadman 131

Late Pleistocene and Holocene Turkeys in the Southwest, by Amadeo M. Rea . . 209

Three Groups of Turkeys from Southwestern Archaeological Sites, by Charmion

R. McKusick 225

Utilization of Birds by the Archaic and Fremont Cultural Groups of Utah,

by Paul W. Parmalee 237

A Fossil Plain Wanderer (Aves: Pedionomidae) from Fire-hole Deposits, Morwell,

Southeastern Victoria, Australia, by Pat Vickers Rich and A. R. McEvey . . 251

Passenger Pigeon Bones from Archaeological Sites in New Mexico, by Lyndon L.

Hargrave and Steven D. Emslie 25 7

v

VI

APPRECIATIONS

HILDEGARDE HOWARD

Hildegarde Howard first became involved in the activities
of the Natural History Museum of Los Angeles County in
1921, just 11 years after the ground-breaking ceremony for the
original Museum structure. Her very distinguished career in
avian paleontology began in 1924 when she initiated her stud-
ies of the fossil birds from Rancho La Brea under the careful
tutelage of Loye Miller. Her enthusiasm for and participation
in the field of avian paleobiology is undiminished to this day.

Dr. Howard was born 3 April 1901 in Washington, D.C.,
and in 1906 moved to Los Angeles with her parents. Her father
was a writer, often composing and editing scripts for the movie
studios in Hollywood; her mother was a musician and com-
poser. Dr. Howard published the first of her 140 papers on
avian paleontology, general science, curation, and other mat-
ters in an international high school natural history bulletin in
1923. In 1924, she began working at the Museum with the
title of “Day Laborer.” She met her husband, Henry Anson
Wylde (who became Chief of Exhibits at the Museum), during
that year, when they were both assigned to sorting La Brea
fossils in the basement of the original Museum building.

From 1924 to 1928, Dr. Howard received her BA., M.A.,
and Ph.D. degrees at the University of California, Berkeley.
Dr. Howard worked as an assistant in Zoology at the Univer-
sity of California, Los Angeles, and as a research associate at
the Los Angeles County Museum during the same period. In
Berkeley and in Los Angeles, she was greatly influenced by
association and study with Loye Miller, Pirie Davidson (later
Mrs. Maverick), Chester Stock, William Diller Matthew, Jo-
seph Grinnell, and William H. Burt (to name a few). Loye
Miller was especially inspirational to her in her scientific and
philosophical outlook. Dr. Howard assumed her first official
permanent position with the Museum, entitled Junior Clerk,
in February 1929, although she actually began working full-
time in the Museum in 1928. Despite the titles, she was in
reality a curator, perhaps the first true specialist in avian pa-
leontology.

Dr. Howard has been the most productive curator-scientist
associated with this museum. It is fortunate for avian paleon-
tology that she has resided for so long in the Los Angeles area,
thus being readily available to identify and study the fossil
birds constantly being uncovered in the environment of erosion
and man’s development in the southern California marine
coastal sediments.

Throughout her career, Dr. Howard (along with the late
Chester Stock) championed the scientific, educational, and
historical aspects of the Rancho La Brea site in Hancock Park.
Unlikely as it may seem, defenses had to be constantly imple-
mented to preserve the site in the proper manner. As a result
of Dr. Howard’s efforts, the birds from Rancho La Brea were
by far the best curated fossil vertebrates in the Museum’s col-
lection.

During her tenure as Chief Curator of Science in the 1950’s,
Dr. Howard was largely responsible for the important increase
in the professional staff of the Museum. I have always been
impressed by her ability to effectively set aside time for con-
centration on research in spite of the days in her career that
demanded administrative function or service to the public.

After retirement in 1961, Dr. Howard became a Guggen-
heim Fellow, completing the research for her paper entitled
“Fossil Birds from the Anza-Borrego Desert,” an important
southwest avifauna of the early to mid-Pleistocene, and car-
rying on research on other fossil birds of the western United
States. Her research on Rancho La Brea birds continues: for
example, in 1974 she described new elements of the relatively
rare La Brea Condor, Breagyps clarki. We count on her work-
ing one day per week at the Museum as Chief Curator Emer-
itus, and she has a complete study at her home in Laguna
Hills, California.

Dr. Howard has long been a member of the American As-
sociation for the Advancement of Science (fellow); Society of
Vertebrate Paleontology (Honorary Life Member); California
and Southern California Academies of Sciences (fellow); Amer-
ican Ornithologists’ Union (fellow); Cooper Ornithological So-
ciety (Honorary Life Member); Geological Society of America
(fellow); Phi Sigma; Phi Beta Kappa; and Sigma Xi. For her
outstanding contribution to avian paleontology, she was
awarded the distinguished Brewster Memorial Award in 1953
by the American Ornithologists’ Union. She is an honorary
member of the Soroptimists Club of Miracle Mile near Rancho
La Brea, a member of the Church of the Brethren, and she is
active in a diversity of group programs in the community in
which she lives. She is also proud to be a Research Associate
of the Santa Barbara Museum of Natural History.

In 1973, the California Academy of Sciences honored Dr.
Howard as a distinguished California Scientist and featured
a special public exhibit of her works. This not only substan-
tiated our pride in Dr. Howard but fulfilled a rarely recognized
need for joining the layman and scientist in an effective yet
simple way. The Hildegarde Howard Cenozoic Hall in the
Natural History Museum of Los Angeles County was opened
in 1977 and honors her as this museum’s most eminent pale-
ontologist. The imaginative new exhibits at the George C.
Page Museum of La Brea Discoveries, which appropriately
highlight the diversity of the La Brea avifauna, also boldly
accentuate the results of Dr. Howard’s scientific work.

Dr. Howard is relatively little known as a private person,
except to a few close colleagues and friends. She is not a dev-
otee of meetings or conferences and seems to treasure privacy
and select friends or associates (part of this relates to a long-
standing hearing problem). However, when asked, she always
presents a clearly stated view of her opinion or observation
and rarely spends time in argumentative ramblings. She has
a good feeling for the problems of a museum and the people
in the trenches — for many years she was there. There are some
people whose presence seems to add respectability to any sit-
uation or organization; this is certainly true of Hildegarde
Howard.

My original introduction to fossils was a master’s thesis
study of a late Pleistocene avifauna from Kansas in 1947-1948.
I benefited from the counsel of Hildegarde Howard in that
study, as I have many times since. Perhaps this early study
unconsciously influenced her decision to hire me for my first
job as Curator of Vertebrate Paleontology at the Museum in
1952, even though I had already wandered “astray” into the

field of paleomammalogy. I consider it a privilege to be a part
of this recognition of the outstanding contributions made by
Hildegarde Howard to the science of avian paleontology and
to the growth and stature of all the sciences at the Natural
History Museum of Los Angeles County during the past SO
years. Reviewing the papers in this volume has rekindled my
appreciation of our “unfeathered friends” and the investigators
of their fragile remains. And it reminds me of how Hildegarde
has often declared, when in less serious mood, that she pre-

ferred to see the birds without the feathers when identifying
a specimen.

It was in 1976 that papers were published to honor another
eminent scholar, the late Dr. Alexander Wetmore. Again, in
1980, we have published records herein that further support
Dr. Storrs Olson’s comment in the Wetmore volume that
“avian paleontology is truly experiencing a renaissance.” We
are very grateful for Dr. Howard’s contribution to the foun-
dation of the renaissance and for her continued participation.

THEODORE DOWNS j

Chief Curator Emeritus
Earth Science Division
Natural History Museum of

Los Angeles County

A TRIBUTE TO
DR. HILDEGARDE HOWARD

Mv memory of Dr. Hildegarde Howard is a long one. I first
met her in 1936, when she was a young assistant curator at
the Natural History Museum of Los Angeles County. I was
spending the winters in Pasadena, in the company of the late
Masauji Hachisukse, a Japanese ornithologist educated in
France and in England, and, at that time, we met many of
our colleagues in southern California. Most of the younger
generation had been trained by Love Miller, a prominent pro-
fessor at the University of California, Los Angeles, who was
their patriarchal mentor. They called him “Padre,” and Hilde-
garde was one of his preferred students. Her interests in or-
nithology are specialized: she is a paleontologist, one of the
world authorities on fossil birds. I can only say that my ap-
preciation of her knowledge is such that I asked her, some 25
years ago, to contribute a special chapter on all known fossil
anatids for my four volume work, “The Waterfowl of the
World.”

I want, however, to state particularly here my appreciation
of Dr. Howard’s achievements as Chief Curator of Science
(Natural History) at the Natural History Museum of Los An-

geles County, a position she held during the 9 years of my
directorship of that institution (1952-1960). Science in those
days was one of the four divisions of the Museum, the others
being History, Art, and Education. I had to attend to all of
them. Although I am a biologist, I could only devote a part
of my time to the Science Division. I therefore relied upon Dr.
Howard for its management. I sincerely believe that no one
could have done it better — her experience, her authority, and
her understanding of people and problems were perfect. Dur-
ing all those years, we worked together in complete harmony,
and I trust that our combined efforts resulted in a definite
improvement of the collections and of their presentations to
the public, as well as in a better standing of our Museum in
the scientific world.

Dr. Howard retired 1 year after I, having reached the man-
datory retirement age, had myself left the Museum. I cannot
help feeling grateful that she was still there at the time of my
departure. I would have missed her tremendously.

I am happy to pay here a tribute to a prominent scientist,
to an outstanding administrator, and also to a very dear friend.

JEAN DELACOUR
Director Emeritus
Natural History Museum of

Los Angeles County

viii

HILDEGARDE HOWARD AND THE
MUSEUM: FIFTY YEARS

The prestige and scientific importance of a great natural
history museum are made by, and depend on, two main assets:
the scope and quality of its research collections, and the ex-
pertise and devotion of its professional staff. Without these,
no matter how extensive and excellent its exhibits and related
programs may be, the museum would be purely a local edu-
cational institution and would never command a position of
eminence in the learned world as a center of scholarship and
as a treasure house of irreplaceable, original materials awaiting
elucidation.

Seldom has the development of any major museum been so
closely related to one individual member of its curatorial staff
as has that of the Natural History Museum of Los Angeles
County to the presence and the work of Dr. Hildegarde How-
ard. For half a century, from August 1929 to the present, she
has been constantly concerned with the study of its ever grow-
ing collections of fossil birds. Indeed, for some of the early
years of her association with the institution, she was practically
the entire scientific staff of the museum, and later, as its pro-
grams and collections expanded and specialists in fields other
than her own were added to the staff, she became the ac-
knowledged head of the museum’s scientific faculty. Although
Dr. Howard officially retired in 1961, she has remained an
active and productive contributor to the museum’s research
work, coming in to examine specimens at least once a week,
and often taking them home for further study during the in-
tervening days. In a very real sense the museum has been
identified with her career, her scientific life, and although she

has modestly kept from public acclaim, her colleagues on the
museum staff and the knowledgeable members of its Board of
Governors and of that of the Museum Alliance are well aware
of how much the museum owes to her. The importance of her
research on fossil birds is not only abundantly recognized by
the enthusiastic response of the worldwide contributors to the
present “Festschrift,” but was signally acknowledged by the
American Ornithologists’ Union many years ago, in 1953, by
their bestowal on her of their prized Brewster Medal. Also,
shortly after her retirement from her position in the museum,
the John Simon Guggenheim Memorial Foundation awarded
her a fellowship with a travel and study stipend to enable her
to continue and to extend her researches in paleornithology.
As another testimonial to her work and influence, the museum
officially named in her honor “The Hildegarde Howard Ce-
nozoic Hall,” an exhibition gallery devoted to a display of
Cenozoic vertebrate fossils, the fauna of one of the geological
periods to which she has devoted much study over many years.

It says much for Dr. Howard’s ability to handle the many,
and sometimes irksome, problems of people and events that
inevitably arise in any sizeable institution that after her long
association with the museum she is able to look back on 50
years remarkably free of personal animosities or institutional
dilemmas. Her path was not always easy, but she knew how
to make it not only smooth but steadily progressive. Her many
friends and colleagues thank her for all she has done and ex-
tend their best wishes for a further continuation of this fine
relationship.

HERBERT FRIEDMANN
Director Emeritus
Natural History Museum of

Los Angeles County

IX

THE CONTRIBUTIONS OF HILDEGARDE HOWARD

By Kenneth E. Campbell, Jr.

Within the broad held of vertebrate paleontology, paleor-
nithology was long considered to be a relatively minor sub-
discipline. Few workers contributed to the held in a regular
manner, and almost all of the early contributors worked with
avian fossils as an aside to their main interests. Studies were
conducted only when particularly interesting or complete spec-
imens were found, or a notable collection of avian fossils was
made from a single site. This rather haphazard growth in our
knowledge of fossil birds continued well into the hrst half of
this century. Slowly, however, there began a trend among a
few workers to devote more and more of their efforts to the
fledgling held of avian paleontology.

The earliest of the American workers who went on to be-
come leaders in avian paleontology were the late Drs. Alex-
ander Wetmore and Love H. Miller. Their studies of fossil
birds will continue as models of scientihc integrity for gener-
ations to come. But these remarkable men were equally, if not
more, prolihc in their research and writing on modern birds.
In 1976, a volume such as this present was dedicated to Dr.
Wetmore in honor of his contributions to avian paleontology.
That volume can be considered as marking the coming of age
of paleornithologv. But it is to Dr. Love Miller that we owe
a debt of gratitude for the hrst true specialist in avian paleon-
tology: Dr. Hildegarde Howard.

Dr. Love Miller began his work in fossil birds when the
large collections of fossil vertebrates from the asphalt deposits
at Rancho La Brea became available. The great quantity of
bird fossils in these collections undoubtedly played an enor-
mously influential role in the development of Dr. Miller’s
methods and approaches to the study of avian fossils, just as
they were to play such an important role in Dr. Howard’s
career. By the time Dr. Miller became acquainted with a
young student by the name of Hildegarde Howard, he had
already published 16 papers on fossil birds from the Pacihc
states. Although most of these papers were concerned with the
fossil birds from Rancho La Brea, he had barely scratched the
surface of this large collection.

When Hildegarde Howard began attending the Southern
Branch of the University of California (now known as Uni-
versity of California at Los Angeles, or UCLA) in 1920, she
was not the least bit inclined toward a career in biology. Her
first biology instructor, Miss Pirie Davidson, made the subject
so interesting, however, that Dr. Howard not only became
deeply interested in the subject but also began to work as a
laboratory assistant in the class. At that time, Dr. Love Miller
was the chairman of the Biology Department. Through the
efforts of Miss Davidson, Dr. Howard obtained a part-time
job working for Dr. Chester Stock, a well-known mammalian
paleontologist. Beginning in 1921, Dr. Howard worked for
Dr. Stock sorting bones from Rancho La Brea in the basement
of the Los Angeles Museum of History, Science and Art (now
known as the Natural History Museum of Los Angeles Coun-
ty), even though he was at the time teaching at the University
of California, Berkeley. In 1922, Dr. Howard went to Berkeley

to finish her degree (UCLA was a two-year school at the time).
At Berkeley, she took classes from Dr. Stock, while continuing
to work for him.

When Dr. Howard completed her B.A. degree in 1924, Dr.
Love Miller offered her a position working part time at UCLA
and part time at the Museum. During the school year 1924—
25, her work for Dr. Miller at the Museum consisted primarily
of research on the California Turkey from Rancho La Brea,
Meleagris ( =Parapavo ) californicus. She obtained credit to-
ward her Master’s degree at Berkeley for this work, and it
became the subject of her first major publication. It was this
year that set Hildegarde Howard firmly toward her career in
avian paleontology and a long period of collaboration with
Loye Miller.

Dr. Howard returned to Berkeley in the fall of 1925 to con-
tinue her graduate work; she obtained her M.S. degree in 1926,
her Ph.D. degree in 1928. Her dissertation, entitled “The Avi-
fauna of Emeryville Shellmound,” was not only a landmark
achievement for her, but when published it became one of her
most popular works. The work was a model of careful com-
parative research, and it has become a classic. One of the
reasons for its impact was a series of drawings illustrating the
bones of a bird skeleton, with clearly labeled osteological fea-
tures (see p. xxvii, this vol.). For the first time avian paleon-
tologists had a standard terminology, a clear point of reference
for the works of different authors. This paper remains as the
principle reference of its kind.

After returning to Berkeley in the fall of 1925, Dr. Howard
continued to work at the Museum part time during breaks in
her academic schedule. Upon receiving her Ph.D. degree she
returned to Los Angeles where she began working fulltime at
the Museum in 1928. She obtained a permanent position with
the Museum in 1929: her title, Junior Clerk; her initial assign-
ment, the curation of the fossils from Rancho La Brea and
research on the birds of this collection. Hildegarde Howard’s
achievements in avian paleontology, including those that pre-
date her formal association with the Museum, have made her
one of the most recognized and respected scientists on the staff
of this museum. Her works have contributed significantly to
the status of this museum as a major research center, and for
this we extend our deepest appreciation.

Over the span of her long career, Dr. Howard has published
on a wide variety of problems in avian paleontology, but her
papers can be grouped into general topics. While it is not
possible to survey her many diverse achievements in detail,
we can bring focus upon her major contributions. Throughout
the remainder of this text, her papers are referred to by a
number enclosed in parentheses; the numbers correspond to
those found in her bibliography (see p. xvii, this vol.).

It was the tremendous collection of bird fossils from the as-
phalt deposits of Rancho La Brea, a collection numbering over
100,000 specimens, that formed Dr. Howard’s training ground.
Indeed, the names Hildegarde Howard and Rancho La Brea
are readily recognized and connected by paleontologists of all

XI

Hildegarde Howard at work on the birds of Rancho La Brea, 1939.

specialities the world over. It was this large collection that
taught Dr. Howard the caution, restraint, and thoroughness
in methodology that came to characterize her works. For, as
many paleontologists have learned, it is far easier to describe
a species when only one or two specimens are available than
it is when hundreds of specimens are available. Few, however,
have had the opportunity to learn this so early in their career
as did Dr. Howard. She learned this lesson even before she
began her graduate studies when she studied the fossil turkey
from Rancho La Brea (2). Working with over 800 specimens
representing all the major bones of the body, she discovered
the critical importance of considering variability within a
species before drawing any hard and fast conclusions. Consid-
ering the osteological variability found in turkeys (see the pa-
pers by Steadman, Rea, and McKusick herein), one can only
speculate that perhaps Love Miller had just this lesson in mind
when he assigned her this group for her first research project.

Many of Dr. Howard’s later works on the fossil birds from
Rancho La Brea also involved studies of large numbers of
specimens. For example, her studies of the eagles and eagle-
like vultures of Rancho La Brea (12) involved the analysis of
over 14,000 fossil specimens, and the study of the Rancho La
Brea caracara (24) involved over 900 bones. These studies
clearly reinforced the lessons of osteological variability that
she had learned earlier.

Although much of Hildegarde Howard’s career, and Love
Miller’s as well, was devoted to the study of the fossil birds
from Rancho La Brea, these collections are far from complete-
ly studied While over 133 species have been reported from

the site, major groups, such as the anatids, small raptors, and
shorebirds, have yet to be analyzed in detail. This is not a
result of a lack of continuing interest on Dr. Howard’s part,
but rather, of a lack of what she considered to be an adequate
series of comparative material of extant species. The lack of
sufficient series of skeletons of modern birds has always been
the bane of paleornithologists, with the result that taxa are
often described without due consideration for intraspecific
variability. Although still true today, this problem was partic-
ularly acute during the early days of paleornithology. And
when you have far more fossil specimens of a species than
modern specimens, as did Dr. Howard, you learn to proceed
with caution.

Dr. Howard’s technical works on the fossil birds of Rancho
La Brea fall into two categories: descriptive (2, 3, 5, 12, 14,
18, 22, 23, 24, 26, 38, 109, 137) and synthetic (7, 19, 96, 105).
Her descriptive papers range from short notes (e.g., 3, 22) to
monumental works encompassing more fossil specimens than
any other paleornithologist has ever been privileged to study
in a lifetime (e.g., 2, 12, 24). Characteristic of these papers,
and all of her later descriptive papers, is the care with which
she documents the assignment of a fossil to a species. Diag-
nostic osteological characters were always presented to justify
the assignment of specimens to a species, along with an ex-
planation as to why she considered the characters to be im-
portant. And, importantly, the explanations often carried ref-
erences to the functional aspects of the features she discussed.
If Dr. Howard felt any hesitancy in making her identifications,
the reasons for this were clearly stated, thereby facilitating the
labors of later workers.

The initial synthetic papers concerning Rancho La Brea pre-
sented analyses of the paleoavifauna as if it were representa-
tive of a single deposit, even though the collections from Ran-
cho La Brea actually came from many different excavations,
termed “pits.” Many years after the early excavations ceased,
it was discovered that the pits were not of the same age (96),
nor were the compositions of the various pits necessarily sim-
ilar (105). This discovery led to a program, now nearing com-
pletion, of radiocarbon dating of specimens from various levels
of numerous pits. The completed series of dates will allow us
to look for trends in avian evolution over the past 40,000 years.
The fact that such trends exist and can be documented was
first observed by Howard in her studies of the fossil birds from
Rancho La Brea. This work led to the development of her
concept of chronoclines, or temporal subspecies (see below).
Howard’s comparison of avian assemblages from the various
pits of Rancho La Brea (105) has provided significant infor-
mation pertaining to the paleoecology of the Los Angeles area
and, by inference, much of southern California during the late
Pleistocene. These studies have also provided information con-
cerning the timing of late Pleistocene extinctions.

The second major focus of Dr. Howard’s career has been
the Tertiary marine birds of southern California. The explo-
sive development of southern California as a major urban cen-
ter, which began in earnest in the 1920’s, has proven to be
quite beneficial to paleornithology, although perhaps not to
neornithologv. The numerous new road cuts and excavations
for industrial and housing developments have provided the
paleontologists associated with the Natural History Museum
of Los Angeles County and other institutions with the oppor-
tunity to collect fossils from deposits that would have other-

Hildegarde Howard in 1951.

wise remained inaccessible. Regrettably, most of the sites are
covered shortly after exposure, thus limiting the size of the
collections from any one site, but the ever-increasing rate of
population growth in southern California is assurance that new
sites will continue to be found. And, most fortunately, the
increasing importance being attached to paleontological re-
mains by local governmental institutions will ensure that more
fossils will be collected in the future.

Most of the Tertiary deposits in southern California from
which fossil birds have been collected are Miocene and Plio-
cene strata of marine origin. Consequently, all but a few of
the Tertiary avian fossils found are seabirds. Most of the
groups represented in collections from these deposits are com-
mon to the Pacific coast of North America today: e.g., loons,
grebes, albatrosses, shearwaters, boobies, and auks. Fasci-
nating extinct groups also formed an important part of the
Tertiary avifauna of coastal California. These groups include
the flightless auks of the subfamily Mancallinae, the “toothed”
odontopterygiform birds of the genus Osteodontornis, and the
flightless pelecaniform plotopterids. It is no exaggeration to
state that Hildegarde Howard has led the way for our under-
standing of all of these groups, although she would be quick
to point out that Love Miller was also responsible for much of
our knowledge of these seabirds.

It has indeed been fortunate for avian paleontology that Dr.
Howard has been actively involved with the paleoavifaunas
of the Pacific coast for such a long time. The continuity thus
obtained has undoubtedly been of great assistance in devel-
oping her (and our) understanding of many of the fossil groups.
For example, Dr. Howard’s first published mention of the
flightless diving auks of the genus Mancalla was in 1939 (2 7),

Love Miller presents Hildegarde Howard with the first fossil he col-
lected at Rancho La Brea. The specimen, a vertebra of a sabertooth
cat, was rescued from the spoilbank of J.C. Merriam’s first excavation
at Rancho La Brea in 1906, and presented to Dr. Howard in 1957.

her most recent work on members of the group was published
in 1976 (138), and she is presently involved in new studies of
these auks.

Through Dr. Howard’s careful work, it is now possible to
visualize the evolution of the flightless diving auks of the
subfamily Mancallinae. The most primitive form known, de-
scribed by Howard in 1966 (119), is the late Miocene Prae-
mancalla lagunensis. This species shows distinct specializa-
tion toward flightlessness, but not to the degree found in the
later species of Mancalla. A second species of Praemancalla,
P wetmorei, was described by Howard in 1976 (138). This
late Miocene species was intermediate in characters between
P. lagunensis and the early species of Mancalla.

Of the four known species of Pliocene Mancalla, Howard
described two, M. milleri and M. cedrosensis (132), and Love
Miller described one, M. diegense (Miller 1937). The first
known species of Mancalla, M. calif orniensis Lucas 1901, was
the first fossil bird to be described from California. In 1968,
Dr. Howard described a third genus of mancalline auk, Al-
codes (123). The continual collection of new material of Man-
calla has resulted in hundreds of specimens of this genus. As
the collections grew, two reviews of the genus were published.
The first was by Miller and Howard in 1949 (49), the second
by Howard in 1970 (129). And, as mentioned above, Dr. How-
ard is presently hard at work on additional aspects of these
auks. We look forward with great anticipation to her update
on these interesting species of flightless birds.

But the mancalline auks were not the only species of flight-
less Tertiary marine birds revealed to us by Dr. Howard. Per-
haps one of the most remarkable of Dr. Howard’s achieve-
ments was her correct diagnosis of the group of flightless diving
birds belonging to the family Plotopteridae from only the hu-
meral end of a coracoid (126). Her diagnoses of plotopterid
relationships and adaptations have been fully substantiated by
recent discoveries of associated partial skeletons and single
elements of at least three genera of plotopterids (see Olson,
this vol.).

Many of the avian fossils from the Tertiary of the Pacific
coast occur as skeletal impressions, or molds, on slabs of shale

Hildegarde Howard at the time of her retirement in 1961

or diatomite. These specimens are difficult to work with be-
cause, although much of the skeleton may be represented by
the impressions, it is not possible to obtain the fine details of
structure necessary for description or comparison between
specimens. Partial skeletons also occur, and although these
provide more detailed structural information, the bones are
only exposed on one side and they are quite often crushed. It
was from just such a specimen as the latter type, and a most
remarkable specimen at that, that Howard described the first
odontopterygiform, or “toothed,” bird from North America.

Osteodontornis orri was described by Howard (86) on the
basis of an associated skeleton preserved on opposing surfaces
of a shale slab. Although feather impressions were visible on
the slab, osteological details were not well preserved. The
unique “toothed” skull did, however, provide much infor-
mation about the species and its relationships with other odon-
topterygiform species. Based on the lengths of the wing bones
and feather impressions, Howard calculated the wingspread
of 0. oiri to be upwards of 5—5 Vz meters, making it one of
the largest flying birds known.

It was only a few years later that a second specimen of
Osteodontornis was found. This specimen, which consisted of
several skull fragments, was described by Howard and White
(101) and referred to O. orri. In 1969, a third specimen of
“toothed” bird was described by Howard, in conjunction with
Stuart Warter (127). This specimen, from New Zealand, was
also part of a skull, although of a different genus (Pseudodon-
tornis)\ it provided additional information concerning the re-
lationships of the odontopterygiform birds. We are indebted
to Dr. Howard for her part in developing our understanding
of these unique birds.

A number of Dr. Howard’s papers on Tertiary marine birds
(e.g., 8, 20, 48, 121, 123, 132, and 139) were faunal studies
from particularly important fossil sites. Faunal studies are
often difficult and time-consuming because of the variety of
taxa involved and the small number of specimens of each tax-
on usually available from each site. Nevertheless, whenever
a sufficient number of specimens accumulated from a site, or
a site was only exposed for a short while before being lost, Dr.
Howard felt it important that the available specimens be put

on record. As a result of her persistent work with these small
collections, we now have a basic, albeit limited, knowledge of
the Tertiary avifaunas of the west coast, including information
on many extinct species. For example, Dr. Howard described
five extinct species of shearwaters in the papers listed above,
along with new species of loons, albatrosses, sulids, and auks.

A distinctive feature of these works, as in all of Dr. How-
ard’s studies, is the caution she used in describing new forms.
If a specimen did not possess good, solid diagnostic characters,
it was not given a name, even if she was convinced it repre-
sented a new form. Rather, such specimens were simply de-
scribed, thus being put on record in the event similar, more
diagnostic material should appear in the future. This approach
has kept other workers abreast of new finds without cluttering
the literature with names based on undiagnostic material.

Non-marine Tertiary avian fossils are normally relatively
rare. An early exception to this rarity was a large collection of
avian fossils from the Eocene of Patagonia, from which Dr.
Howard described (80) a water bird of phoenicopterid char-
acter, now known as Presbyornis antiquus (Howard). Al-
though recent work on Presbyornis by Alan Feduccia has re-
sulted in the synonymy of the genus Howard erected for the
species she described, her interpretations as to the nature of
the species and its systematic relationships have been borne
out.

Howard’s contributions to our knowledge of other Tertiary
birds include the description of the first Eocene birds known
from California (116); the description of a Miocene hawk (35)
and a Miocene thrush (87) from California; the description of
a Miocene raptor and quail from South Dakota (120); and
reports on Pliocene birds from Mexico (114, 118).

The major Pleistocene and prehistoric avifaunas that Dr.
Howard described (excluding Rancho La Brea) came from sev-
eral sources, including marine deposits, Pleistocene lake de-
posits, and cave deposits.

In western North America, many large lakes were formed
and maintained by the climatic conditions that prevailed dur-
ing periods of Pleistocene glacial activity. The lacustrine de-
posits that accumulated in these lakes have produced large
collections of avian fossils. The most notable of these collec-
tions described by Dr. Howard are those from Fossil Lake,
Oregon (39), and Manix Lake, California (83).

The paleoavifauna from Fossil Lake was particularly chal-
lenging because not only was it very large (over 2500 specimens
identified to the family level, over 1800 of these to species), it
was scattered through seven separate collections that had been
made over a period of 60 years. Portions of this paleoavifauna
had been described by E.D. Cope and R.W. Shufeldt in the
late 1800’s and early 1900’s. The latter was not a particularly
careful worker, and many of his species assignments were in
error. Also, types were missing from the collections, or they
had not been clearly identified in the original description of the
species. Through her careful, meticulous work, Dr. Howard
brought order to what had been a rather chaotic situation, and
she placed the Fossil Lake avifauna in perspective with the
other Pleistocene avifaunas known at the time.

It was in this paper that Dr. Howard first used the trinomial
to designate chronoclinal variation. She knew that some late
Pleistocene forms varied in predictable but relatively minor
ways from their living counterparts and that overlap often
existed between the fossil and living forms. In such cases,

xiv

separation of the forms at the species level was considered
unwise even though a difference clearly existed. Dr. Howard
chose to make note of these differences by designating the fossil
forms in question that had been previously named as new
species as temporal subspecies. This represented a turnabout
from an earlier opinion expressed by Dr. Howard that such
subspecific relationships were “wholly untenable” (24:239).
Although she had described a fossil subspecies in an earlier
paper (28), she considered that case to be one of geographic,
not temporal, variation (the second subspecies she named (122)
was also an example of geographic variation). She discussed
the chronocline concept in a later note (43), and it has played
an important role in her analyses of late Pleistocene taxa (41).

In the discussion of the Fossil Lake avifauna, Dr. Howard
described a number of specimens of a “pigmy goose” of the
genus Anabernicula, but left the question as to their species
assignment until she could give more time to the problem. In
her later thorough review of the genus (112), she described the
Oregon form as a new species and clarified the relationships
of the genus within the Anseriformes.

Although the paleoavifauna of Manix Lake was small, it
was important because it contained two species of flamingos.
These were the first records of that group for California, al-
though one of the species had been previously recorded from
Fossil Lake, Oregon.

Another large collection described by Dr. Howard from
Pleistocene deposits came from the Anza-Borrego Desert of
southern California (107). She described several new species
in this paper, and she also applied a technique for dating the
paleoavifauna that she had used earlier in dating the various
“pits” from Rancho La Brea (105). Based on relative numbers
of extinct species, the technique cannot provide an exact age,
but it can assist in correlating one paleoavifauna with another.
For collections where absolute dating techniques are unavail-
able, this technique can prove to be useful in establishing rel-
ative ages.

The Pleistocene marine deposits, including those that occur
on the Channel Islands off the coast of southern California,
have proven to be important sources of avian fossils. Sum-
maries of the avian fossils from these marine deposits were
presented by Dr. Howard in 1949 (46) and 1958 (89). From
these deposits have come large collections of the flightless div-
ing geese of the genus Chendytes. In 1955 (82), Dr. Howard
described the second known species of the genus Chendytes,
C. milleri, from early Pleistocene deposits of San Nicolas Is-
land. Interestingly, C. milleri was structurally more primitive,
i.e. , with less advanced wing reduction, than the late Pleis-
tocene species, C. lawi L. Miller. Both before (42, 46) and
after (111) her description of C. milleri, Howard described
various limb elements of Chendytes. Recently, very large col-

lections of Chendytes have become available, and they will
undoubtedly tell us much more about the evolution of these
flightless geese.

Cave deposits in New Mexico and Nevada provided large
collections of avian fossils that were described by Dr. Howard
(9, 13, 17, 66, 107, 131). The great majority of the specimens
from these caves were assigned to living species, but there was
also a consistent representation of extinct species. Thus, the
cave deposits were considered to be of late Pleistocene, or
possibly early Holocene, age. Dr. Howard pursued the de-
scription of these collections because she believed the fossil
birds they contained would provide important distribution rec-
ords for the species represented, living as well as extinct. The
collections also contained an occasional new species, the most
surprising of which was Teratornis incredibilis Howard (66).
This species, to which Dr. Howard later referred two addi-
tional specimens (107, 135), was about 40% larger than Ter-
atornis merriami Miller from Rancho La Brea, and about
twice as large as the condor, Gymnogyps calif ornianus . As she
clearly stated, however, the relationships of this species cannot
be accurately determined because no truly diagnostic element
has been found.

On three occasions Dr. Howard published general reviews
(21, 53, 140) of advances in avian paleontology, bringing non-
paleornithologists up to date on progress in the field. She also
undertook an even more extensive and difficult review of all
fossil species of the order Anseriformes (110); several years
later she updated this work (136). This review was not just a
listing of all known fossil anseriform species. Rather, in the
Howard tradition, all known osteological details of each fossil
species were presented along with measurements and com-
ments on possible relationships.

Hildegarde Howard also contributed numerous articles of
a non-technical nature to the Museum’s publications. These
served to inform the public as to the Museum’s activities and
some of the intriguing fossils she and others were working on.
In 1945, she published a general review of fossil birds with an
emphasis on the birds of Rancho La Brea (37). This work was
updated and expanded in 1955 (81) and 1962 (100). It remains
a very popular publication with visitors to the Museum.
Through these efforts, and her generous willingness to spend
time with students and interested members of the general pub-
lic, she engendered public support for the Museum.

Such a brief overview as this of a career as long and pro-
ductive as that of Dr. Howard’s can only hint at its depth and
breadth. To survey Hildegarde Howard’s contributions to pa-
leornithology is to see a perfection of technique, the evolution
of ideas, and devotion to a science. As the preeminent student
of paleornithology, she has served her chosen field well It is
heartening and reassuring to know that her work continues.

xv

BIBLIOGRAPHY OF HILDEGARDE HOWARD

1923

1. The fossil beds of La Brea. International Assoc. High School Nat. Hist. Clubs,
Bull. 2. (November)

1927

2. A review of the fossil bird, Parapavo californicus (Miller) from the Pleistocene
asphalt beds of Rancho La Brea. Univ. California Publ., Bull. Dept. Geol. Sci
17(1): 1—62 , 13 pis. (September 30)

1928

3. The beak of Parapavo californicus (Miller). Bull. So. California Acad. Sci. 27:90-
91. (December 31)

1929

4. The avifauna of Emeryville shellmound. Univ. California Publ Zool. 32(2):301-
394, 4 pis., 55 text figs. (July 19)

5. Additional bird records from the Pleistocene of Rancho La Brea. Condor 31:251—
252. (November 15)

1930

6. An abnormal wing development in a Pintail Duck Condor 32:58-61, 2 text figs.
(January 20)

7. A census of the Pleistocene birds of Rancho La Brea from the collections of the
Los Angeles Museum. Condor 32:81-88, 3 text figs. (March)

1931

8. Pliocene bird remains from Santa Barbara, California. Condor 33:30-31. (Janu-
ary 15)

9. A new species of Road-runner from Quaternary cave deposits in New Mexico.
Condor 33:206-209, 3 text figs. (September 15)

10. Cryploglaux funerea in New Mexico. Condor 33:216. (September 15)

1932

11. A new species of Cormorant from Pliocene deposits near Santa Barbara, Califor-
nia. Condor 34:118-120, 1 text fig. (May 16)

12. Eagles and eagle-like vultures of the Pleistocene of Rancho La Brea, California.
Carnegie Inst, of Wash., Publ. 429:1-82, 29 pis., 3 text figs. (October)

1933

13. Bird remains from cave deposits in New Mexico. Condor 35:15-18. (With Alden
H. Miller) (January 16)

14. A new species of owl from the Pleistocene of Rancho La Brea, California. Condor
35:66-69, 1 text fig. (March 17)

15. Bird remains from an Indian shellmound near Point Mugu, California. Condor
35:235. (With Leigh Marian Dodson) (November 15)

1934

16. Characters differentiating certain species of Stercorarius. Condor 36:158-160.
(With George Willett) (July 16)

1935

17. A new species of eagle from a Quaternary cave deposit in eastern Nevada. Condor
37:206-209, 1 text fig. (July 15)

18. The Rancho La Brea Wood Ibis. Condor 37:251-253, 1 text fig. (September 16)

XVII

1936

19. Further studies upon the birds of the Pleistocene of Rancho La Brea. Condor
38:32-36. (January IS)

20. A new fossil bird locality near Plava del Rev, California, with description of a
new species of sulid. Condor 38:211-214, 1 text fig. (September IS)

21. A new record for Parapavo californicus. Condor 38:249-250. (November 16)

1937

22. A Pleistocene record of the Passenger Pigeon in California. Condor 39:12-14, 1
text fig. (January 15)

1938

23. The status of the extinct condor-like birds of the Rancho La Brea Pleistocene.
Publ. Univ. California at Los Angeles, Biol. Sci. 1:169-176, 1 pi., 2 text figs.
(With Love Miller) (February 18)

24. The Rancho La Brea Caracara, a new species. Carnegie Inst. Wash. Publ.
487:217-240, 3 pis., 1 text fig. (July 7)

1939

25. A prehistoric record of Holboell Grebe in Nevada. Condor 41:32. (January 17)

26. The avifauna associated with human remains at Rancho La Brea, California.
Carnegie Inst. Wash. Publ. 514:39-48, 4 text figs. (With Alden H. Miller)
(May 18)

27. Aves. Fortschritte de Palaontologie 2(1937/1 938):309— 3 2 2 .

1940

28. A new race of Caracara from the Pleistocene of Mexico. Condor 42:41-44. (Jan-
uary 19)

1941

29. The horse, a biography. Los Angeles County Mus. Quarterly 1:10-13, 2 text figs.
(April)

1942

30. A review of the American fossil storks. Carnegie Inst. Wash. Publ. 530:187-203,
1 pi., 2 text figs. (January 19)

31. Rancho La Brea birds. Los Angeles County Mus. Quarterly 2:16-20, 2 text figs.
(January)

1943

32. Review of “Fossil Birds of California” by Love Miller and Ida de May. Condor
45:79-80. (March 19)

1944

33. The ascent of Equus. A story of the origin and development of the horse. Los
Angeles County Mus., Sci. Ser. 8, Paleon. Publ. 5:1-38, 30 text figs. (With Chester
Stock) (March 31)

34. The Rancho La Brea bird mounts. The Museum News 21(19):8. (April 1)

35. A Miocene hawk from California. Condor 46:236-237, 1 text fig. (September 27)

36. Miscellaneous avian fossil records from California. Bull. So. California Acad. Sci.
43:74-77, 1 pi. (September 30)

1945

37. Fossil birds. With especial reference to the birds of Rancho La Brea. Los Angeles
County Mus., Sci. Ser. 10, Paleon. Publ. 6: 1 — 40, 18 text figs. (May)

38. Observations on young tarsometatarsi of the fossil turkey, Parapavo californicus
(Miller). Auk 62:596-603, 1 pi., 1 text fig. (October)

xviii

1946

39. A review of the Pleistocene birds of Fossil Lake, Oregon. Carnegie Inst. Wash.
Publ. 551:141-195, 2 pis. (January 25)

40. George Willett: May 28, 1879-August 2, 1945. Condor 48:49-71, 11 text hgs. and
frontispiece portrait. (April 2)

1947

41. A preliminary survey of trends in avian evolution from Pleistocene to Recent time.
Condor 49:10-13. (February 6)

42. Wing elements assigned to Chendytes. Condor 49:76-77, 1 text fig. (March 31)

43. An ancestral Golden Eagle raises a question in taxonomy. Auk 64:287-291. (April)

44. California’s flightless birds. Los Angeles County Mus. Quarterly 6(2): 7— 1 1 , 3 text
figs. (Summer)

1948

45. Later Cenozoic avian fossils from near Newport Bay, Orange County, California.
Abstract. Bull. Geol. Soc. Amer. 59:1372-1373.

1949

46. Avian fossils from the marine Pleistocene of Southern California. Condor 51:20-
28. (January)

47. The spring “Robin.” Los Angeles County Mus. Quarterly 7(3): 1 7 . (March)

48. New avian records for the Pliocene of California. Carnegie Inst. Wash. Publ.
584(6): 177—199, 3 pis. (June 22)

49. The flightless Pliocene bird, Mancalla. Carnegie Inst. Wash. Publ. 5 84( 7):20 1 —
228, 6 pis., 1 text fig. (With Love Miller) (June)

50. Adaptations. Los Angeles County Mus. Quarterly 7(4): 1 1-15, 5 text figs. (October)

51. The California Turkey. Los Angeles County Mus. Quarterly 7(4): 15-16. (October)

1950

52. Personalities in Paleontology (E.E. Hadley). Soc. Vert. Paleon. Bull. 27:26-27,
1 text fig. (October)

53. Fossil evidence of avian evolution. Ibis 92:1-21, 5 text figs. (January)

54. Foreword. Pp. v-vi in Lifelong Boyhood by Love Miller. Univ. California Press,
Berkeley and Los Angeles.

55. Gayle Benjamin Pickwell. Auk 67:280-281. (April) See also No. 75, 1954.

56. Charles Dean Bunker. Auk 67:425-426. (July) See also No. 75, 1954.

57. Thomas Tonkin McCabe. Auk 6 7 :42 7 — 42 8 . (July) See also No. 75, 1954.

58. The Hall of Evolving Life. Los Angeles County Mus. Quarterly 8(l):6-7, 1 text
fig. (Spring)

59. Teratornis, the wonder bird of the Ice Age. Los Angeles County Mus. Leaflet
Ser., Sci. 3:1-4. (December)

1951

60. Chester Stock (1892-1950). Soc. Vert. Paleon. Bull. 31:33-34, 1 text fig. (Feb-
ruary)

61. Pleistocene duck bones from Ohio. Condor 53:205. (July)

62. Chester Stock. January 28, 1892-December 7, 1950. Los Angeles County Mus.
Quarterly 8(3-4): 15-18, 1 text fig. (May)

63. E.E. Hadley (1864-195 1). Soc. Vert. Paleon. Bull. 33:32. (October)

64. Louis Bennett Bishop. 1865-1950. Auk 68:440-446, 1 text fig. (October)

65. Elza Ellsworth Hadley, 1864-1951. Bull So. California Acad. Sci. 50(3): 172 — 173,
1 text fig. (December).

1952

66. The prehistoric avifauna of Smith Creek Cave, Nevada, with a description of a
new gigantic raptor. Bull. So. California Acad. Sci. 5 1(2):50— 54, 1 pi. (May)

67. Dorothea Minola Alice Bate. Auk 69:491. (October) See also No. 75, 1954.

68. Mary Louise Fossler. Auk 69:493. (October) See also No. 75, 1954.

XIX

1953

69. Paintings depict “Life through the Ages.” Los Angeles County Mus. Quarterly
10(l):6-8, 6 text figs. (March)

70. Forty years at Rancho La Brea. Los Angeles County Mus. Quarterly 10(2):6-12,
17 text figs. (June)

71. Southern California Academy of Sciences. Science 1 18(3072):3. (November 13)

72. An early bird. Los Angeles County Mus. Quarterly 10(4): 1 2— 1 3 , 3 text figs. (De-
cember)

1954

73. Albert Ernest Colburn. Auk 71:111. (January) See also No. 75, 1954.

74. John M. Davis. Auk 71:236. (April) See also No. 75, 1954.

75. Biographies of members of the American Ornithologists’ Union. Reprinted from
“The Auk,” 1884-1954. Pp. 35, 100, 135, 161, 212, 380, 455. Lord Baltimore
Press, Washington, D C. (With T.S. Palmer and others)

76. James Charlwood Marsh, 1867-1954. Bull. So. California Acad. Sci. 53(3): ISO-
182, 1 text fig. (December 31)

77. The Hall of Evolving Life in the Los Angeles County Museum. Museum 7(4): 209—
217, 11 text figs. (Including French translation) (December)

1955

78. Rancho La Brea Tar Pits. Los Angeles County Employee 28( 1 ): 10—1 1 , 50, cover
photo and 3 text figs. (January)

79. History of scientific work at Rancho La Brea. Issued with Miracle Mile Milestones
No. 4. (January 6)

80. A new wading bird from the Eocene of Patagonia. Amer. Mus. Nov. 1710:1-25,
8 text figs. (March 11)

81. Fossil Birds. With especial reference to the birds of Rancho La Brea. (Revised
edition) Los Angeles County Mus., Sci. Ser. 17, Paleon. Publ. 10:1-40, 22 text
figs. (April 27)

82. New records and a new species of Chendytes, an extinct genus of diving geese.
Condor 57:135-143, 3 text figs. (May 25)

83. Fossil birds from Manix Lake, California. U.S. Geol Survey Prof. Paper No.
2 64J : 199—205 , 1 pi., 1 text fig. (June 8)

84. Job analysis of curatorial positions. Clearing House for Western Museums, News-
letter 189:877-880. (November)

85. Summary of vertebrate remains. Pp. 14-18 in Preliminary report of the Schuiling
Cave by Gerald A. Smith. Quarterly, San Bernardino Co. Mus. Assoc. 3(2). (With
Theodore Downs) (Winter)

1957

86. A gigantic “toothed” marine bird from the Miocene of California. Santa Barbara
Mus. Nat. Hist., Bull. Dept. Geol. 1:1-23, 8 text figs. (February 1)

87. A new species of passerine bird from the Miocene of California. Los Angeles Co.
Mus., Contrib. Sci. 9:1-16, 2 text figs. (June 28)

1958

88. A hundred million years of California’s prehistory in a famous collection. Los
Angeles County Mus. Quarterly 1 4( 1 ):2— 5 , 5 text figs. (February)

89. Further records from the Pleistocene of Newport Bay mesa, California. Condor
60(2): 136. (March)

90. Standards for curatorial positions. Committee Report (Chaired by Hildegarde
Howard). Clearing House for Western Museums Newsletters 217-218:2032-2051.
(First issued in mimeograph form to members of Western Mus. Conference, May
1957.)

91. Miocene sulids of southern California. Los Angeles Co. Mus., Contrib. Sci. 25:1 —
16, 3 text figs. (August 15)

92. Condensed version of chapter on “Origin and Evolution of Birds” from “Fossil
Birds,” Los Angeles County Mus., Sci Ser. 17, Paleon. Publ. 10:1-40. Published

xx

in program of 1958 Annual Pheasant Show, Arcadia, California, November 22-
23, 1958.

93. An ancient cormorant from Nevada. Condor 60:411-413. (November 26)

1959

94. Quaternary animals from Schuiling Cave in the Mojave Desert, California. Los
Angeles Co. Mus., Contrib. Sci. 29:1-21, 8 text figs. (With Theodore Downs,
Thomas Clements, and Gerald A. Smith) (April 14)

95. What is Hancock Park? Los Angeles County Mus. Quarterly 15(4): 10—12 . (Au-
tumn)

1960

96. Significance of Carbon-14 dates for Rancho La Brea. Science 13 1(3402): 7 12—7 14.
(March 11)

97. The Division of Science. Los Angeles County Mus. 50th Anniversary Quarterly
16(2): 19-23. (Spring)

98. What about standards for small museums? Clearing House for Western Mus.
1(1): 7 — 10. (July)

1961

99. Howard Rice Hill, 1891-1961. Bull. So. California Acad. Sci. 60:193-195. (July)

1962

100. Fossil Birds. With especial reference to the birds of Rancho La Brea. (New edition
with addendum) Los Angeles County Mus., Sci. Ser. 17, Paleon. Publ. 10:1-44,
23 text figs.

101. A second record of Osteodontornis, Miocene “toothed” bird. Los Angeles Co.
Mus., Contrib. Sci. 52:1-12, 5 figs. (With John A. White) (February 26)

102. Bird remains from a prehistoric cave deposit in Grant County, New Mexico.
Condor 64:241-242. (May)

103. A comparison of avian assemblages from individual pits at Rancho La Brea,
California. Abstracts of Papers, XIII Internat. Ornithol. Congress. June 17-24,
1962. Ithaca, New York.

104. A new Miocene locality record for Puffin us diatomicus and Sw/a willetti. Condor
64:512-513. (November)

105. A comparison of prehistoric avian assemblages from individual pits at Rancho La
Brea, California. Los Angeles Co. Mus., Contrib. Sci. 58:1-24, 5 text figs. (De-
cember 2 1)

106. Contributions from the Los Angeles Museum-Channel Island Biological Survey.
36. A fossil bird, Caracara, from Santa Rosa Island. Bull. So. California Acad.
Sci. 6 1 (4): 2 2 7 —228. (December 31)

1963

107. Fossil birds from the Anza-Borrego Desert. Los Angeles Co. Mus., Contrib. Sci.
73:1-33, 3 pis., 1 text fig. (December 30)

108. Ascent of Equus. Second edition. Los Angeles County Mus., Sci. Ser. 22, Paleon.
Publ. No. 12:1-38, 15 text figs. (With Chester Stock) (July)

1964

109. A fossil owl from Santa Rosa Island, California, with comments on the eared
owls of Rancho La Brea. Bull. So. California Acad. Sci. 63( 1 ):2 7 —3 1 , 1 text fig.
(April 21)

110. Fossil Anseriformes. Pp. 233-326 (Chapter 10) in Waterfowl of the World by Jean
Delacour. Vol. 4. Country Life Ltd., London. Second edition, 1973.

111. Further discoveries concerning the flightless “diving goose” Chendytes lawi. Con-
dor 66:372-376, 1 text fig. (September)

112. A new species of “Pigmy Goose,” Anabernicula, from the Oregon Pleistocene,
with a discussion of the genus. Amer. Mus. Nov. 2200:1-14, 2 text figs. (Decem-
ber 15)

113. Hilda Wood Grinnell. Auk 81:586. (October)

XXI

1965

114. A new species of cormorant from the Pliocene of Mexico. Bull. So. California
Acad. Sci. 64( 1):50— 5 5 , 1 text fig. (April 26)

115. Laurence Markham Huey. Auk 82:323. (April)

116. First record of avian fossils from the Eocene of California. J. Paleon. 39(3):350-
354, 1 pi. (“May,” distributed June (late, undated))

117. Egmont Zachary Rett. Auk 82:686. (October)

1966

118. Pliocene birds from Chihuahua, Mexico. Los Angeles Co. Mus., Contrib. Sci.
94:1-12, 1 text fig. (April 4)

1 19. A possible ancestor of the Lucas Auk (Family Mancallidae) from the Tertiary of
Orange County, California. Los Angeles Co. Mus., Contrib. Sci. 101:1-8, 1 text
fig. (May 5)

120. Two fossil birds from the Lower Miocene of South Dakota. Los Angeles Co.
Mus., Contrib. Sci. 107:1-8, 1 text fig. (July 22)

121 Additional avian records from the Miocene of Sharktooth Hill, California. Los
Angeles Co. Mus., Contrib. Sci. 114:1—11, 1 text fig. (December 28)

1968

122. Limb measurements of the extinct vulture, Coragyps occidentalism with a descrip-
tion of a new subspecies. Papers Archaeol. Soc. New Mexico 1:115-128. (May 9)

123. Tertiary birds from Laguna Hills, Orange County, California. Los Angeles Co.
Mus., Contrib. Sci. 142:1-21, 2 text figs. (June 14)

124. Fossil Birds. Pp. 42-45 in Prehistory of Santa Rosa Island by Phil C. Orr. Santa
Barbara Mus. Nat. Hist. Publ. (September)

125. A preliminary report of Pleistocene birds of Central Mexico. Abstracts, Annual
Meeting Geol. Soc. Amer., Mexico City, 1968:142.

1969

126. A new avian fossil from Kern Co., California. Condor 71:68-69, 1 text fig. (Feb-
ruary 14)

127. A new species of bonv-toothed bird (Family Pseudodontornithidae) from the Ter-
tiary of New Zealand. Rec. Canterbury Mus. 8(4):345-35 7, 4 pis. (With Stuart
L. Warter) (May 31)

128. Avian fossils from three Pleistocene sites in central Mexico. Los Angeles Co. Mus.,
Contrib. Sci. 172:1-11, 1 text fig. (June 30)

1970

129. A review of the extinct avian genus, Mancalla. Los Angeles Co. Mus., Contrib.
Sci. 203:1-12, 1 text fig. (November 24)

1971

130. In Memoriam: Love Holmes Miller. Auk 88:276-285, photo. (April)

131. Quaternary avian remains from Dark Canyon Cave, New Mexico. Condor
73:237-240. (May 21)

132. Pliocene avian remains from Baja California. Los Angeles Co. Mus., Contrib.
Sci. 217:1-17, 2 text figs. (November 12)

1972

133. Type specimens of avian fossils in the collections of the Natural History Museum
of Los Angeles Co. Los Angeles Co. Mus., Contrib. Sci. 228:1-27. (June 7)

134. The bibliography of Love Holmes Miller. Condor 74:268-271. (September)

135. The Incredible Teratorn again. Condor 74:341-344, 1 text fig. (September 18)

1973

136. Fossil Anseriformes. Pp. 233-326 (Chapter 10), and New General Corrections
and Additions, pp. 371-378 (Chapter 12) in Waterfowl of the World by Jean
Delacour. Second Edition. Vol. 4. Hamlyn Publ. Group, Ltd. (Country Life Ltd.),
London.

xxu

1974

137. Postcranial elements of the extinct condor, Breagyps clarki (Miller). Los Angeles
Co. Mus., Contrib. Sci. 256:1-24, 9 text figs. (May 22)

1976

138. A new species of flightless auk from the Miocene of California (Alcidae: Mancal-
linae). Smithsonian Contrib. to Paleobiology 27:141-146, 1 text fig. (May 21)

1978

139. Late Miocene marine birds from Orange Co., California. Los Angeles Co. Mus.,
Contrib. Sci. 290:1-28, 4 text figs. (March 21)

1979

140. Aves. Pp. 60-70 in The Encyclopedia of Paleontology (R.W. Fairbridge and D.
Jablonski, Eds.). Dowden, Hutchinson, and Ross, Inc., Stroudsburg, Penn.

xxiii

INDEX TO FOSSIL AVIAN TAX A DESCRIBED
BY HILDEGARDE HOWARD

Listed below in alphabetical order are the fossil avian taxa described by Hildegarde
Howard. Species are listed in the genera to which they were originally referred. Fol-
lowing each name is the publication number (from Dr Howard’s bibliography) and
page in which the name was proposed.

Families

Palaeoscinidae 87:15 Telmabatidae 80:23

Plotopteridae 126:69

Genera

Alcodes 123:18
Arikarornis 120:2
Brantadorna 107:8
Breagyps 23:171
Miohierax 35:236
Osteodontornis 86:3

Palaeoscinis 87:6
Paleosula 91:12
Plotopterum 126:68
Praemancalla 119:4
Telmabates 80:3
Wasonaka 118:5

Species

aldeni, Miortyx 120:5
antiquus, Telmabates 80:3
anza, Agriocharis 107:19
barnesi, Puffinus 139:7
bessomi, Oxyura 107:13
brea, Strix 14:66
brodkorbi, Gavia 139:4
calhouni, Puffinus 123:6
calif orniensis, Protostrix 116:350
cedrosensis, Mancalla 132:11
conklingi, Geococcyx 9:208
downsi, Brantadorna 107:8
felthami, Puffinus 48:194
fossilis, Bucepliala 107:11
goletensis, Phalacrocorax 114:51
hammeri, Fulmarus 123:9
hesterna, Fulica 107:22
incredibilis, Teratornis 66:51
joaquinensis, Plotopterum 126:68
kanakoffi, Puffinus 48:187
kennelli, Phalacrocorax 48:188
lagunensis, Praemancalla 119:4
macdonaldi, Arikarornis 120:2
magnus, Morus 139:17
milleri, Chendytes 82:137
milleri, Diomedea 121:2

milleri , Mancalla 129:7
milleri, Urubitinga 12:25
minor, Cerorhinca 132:9
minutus, Phoenicopterus 83:202
oregonensis, Anabernicula 112:5
oregonus, Falco 39:178
oiri, Osteodontornis 86:3
pliocenus, Brachyramphus 48:191
pohli, Sula 91:4
prelutosus, Polyborus 24:226
priscus, Asio 109:28
recentior, Miosula 48:190
reyana, Moris 20:213
rogersi, Phalacrocorax 11:118
rossmoori, Aethia 123:16
sliufeldti, Stercorarius 39:184
stirtoni, Pseudodontornis 127:348
stocki, Miohierax 35:236
tedfordi, Puffinus 132:2
turdirostris, Palaeoscinis 87:6
ulnulus, Alcodes 123:18
vallecitoensis, Neophrontops 107:17
wetmorei, Mycteria 18:253
wetmorei, Praemancalla 138:142
willetti, Spizaetus 17:207
vepormerae, Wasonaka 118:5

Subspecies

grinnelli, Polyborus prelutosus 28:41

mexicanus, Coragyps occidentalis 122:124

ILLUSTRATIONS OF AVIAN OSTEOLOGY TAKEN FROM
“THE AVIFAUNA OF EMERYVILLE SHELLMOUND”

Of the many significant and invaluable contributions Hildegarde Howard has made
to the field of avian paleontology, her paper entitled “The Avifauna of the Emeryville
Shellmound” was one of the most important. This paper was published in 1929, and
is especially cherished by those fortunate enough to obtain a copy. This work was not
only a particularly valuable early contribution to avian paleontology, it contained a
series of illustrations of the major bones of the avian skeleton with the major diagnostic
features of each bone indicated and named. Over the past 50 years, these illustrations
have proven very valuable, especially to new students of avian osteology. To this day
they have not been surpassed for their usefulness as the terminology used in current
studies of avian osteology remains based on that introduced by Dr. Howard. And
anyone who has tried to orient a bone to determine view designations without the help
of a mounted skeleton has often had cause to give thanks for the illustrations. That
such a work remains so important after a period of 50 years testifies to its thoroughness
and accuracy, two characters that have typified Dr. Howard’s works through the years.

Even after 50 years Dr. Howard continues to receive many requests for copies of the
Emeryville Shellmound paper; it is perhaps her most sought-after paper. For this reason
we reproduce here the illustrations of avian osteological features from that paper. When
speaking of the illustrations, Dr. Howard always credits William H. Burt for working
with her in devising the nomenclatural system used in the illustrations, and Frieda
Abernathy for executing the drawings. Quoted below are the explanatory notes for the
illustrations, taken from page 325 of “The Avifauna of Emeryville Shellmound,” by
Hildegarde Howard, 1929, Univ. of California Publ. Zool. 32(2):301— 394:

Description of Species

The terms employed in describing the diagnostic characters of the various rep-
resented species will be found in the accompanying series of labeled figures, drawn
by Mrs. Frieda Abernathy.

The system of nomenclature here set forth was devised by the writer in collab-
oration with Mr. William H. Burt, of the University of California. Papers by the
following authorities were consulted: Furbringer (1888), Heilmann (1926), Lam-
brecht (1914), Lowe (1928), Miller (1925a, 1925b, 1927a), Milne-Edwards ( 1867—
68), Owen (1866), Shufeldt (1890, 1909), Stresemann (1927), and Wetmore (1922,
1923). Dr. Miller and Dr. Wetmore were also consulted personally.

The Golden Eagle (Aquila chrysaetos) and the Snow Goose ( Chen hyperboreus)
have been used for illustration. Such parts as cannot well be shown on Aquila are
labeled on Chen, and vice versa. Of the Golden Eagle, Museum of Vertebrate
Zoology specimen no. 28884 has been used except for figures 5 and 8, where MVZ
no. 40866 was substituted; of the Snow Goose, MVZ no. 45555 has been drawn,
except in figure 12 where MVZ no. 22446 has been used.

Grateful acknowledgment for permission to reprint the illustrations is given the Uni-
versity of California Press. Larry Reynolds provided unblemished photographs of the
illustrations for reproduction here from a very well-worn copy of the original publi-
cation.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 330:xxvii-xxxviii

premaxi I lary

frontal
parietal
supraoccipital

external naris

nasal process
of premaxillary

maxillary
maxillo palatine

palatine
vomer-
lachry mal

ba si pterygoid process
pterygoid
Sphenoidal rostrum
postorbital process

^ff|| openings for

Eustachian tubes— ^
quadrate
mm basitemporal plate

basioccipital
occipital condyle-
foramen magnum

supraoccipital-^ lexoccipitol
orbitosphenoid-i
postorbital process-,
interorbital septurrc
ethmoid^
achrym at

opisthotic

parietal-

premaxillary-

palatine 5phenoidal rostrum-1

opisthotic
squamosal
quadrate

alisphenoid ^quadratojuga

Skull of Chen hyperboreus. Fig. 1, dorsal view; fig. 2, ventral view;
fig. 3, lateral view. X 1.

xxviii

neural spine
neural cana
transverse process

NjDOStzyg apophysis
centrum

pterygoid

articulation

mandibular

articulation'

haem apophyses

pleurapophysis
diapophysis
parapophysis
prezygopophysis
articular surface of centrum

. orbital process

otic process

socket for

quadratojuga.1

neural canal

squamosal\ A

articulation driK'/lrf'

articular surface

crest

haemal orifice

haemal canal

diapophysis

neural arch

costal facets
-neural canal

neurapophysis^

centrum
hypapophysis

neural spine

anapophy sis

postzygapophysis

prezyg apophysis
odontoid process
vertebrarterial canal
10 hypapophysis

Figs. 4-11, Aquila clirysaetos. Fig. 4, 6th cervical vertebra, dorsal view;
fig. 5, caudal vertebra, anterior view; fig. 6, 6th cervical vertebra, ventral view;
fig. 7, left quadrate, external view; fig. 8, pvgostvle; fig. 9, atlas, posterior view;
fig. 10, 4th thoracic vertebra, anterior view; fig. 11, axis, left side; fig. 12,
Chen hyperboreus, mandible. X 1.

xxix

carinal apex

ventral manubrial spine

ventral lip] , ... .

, . ./[orcoracoidal sulcus

dorsal lip]

dorsal manubrial spine
pneumatic foramen
sterno-coracoidal process

costal process
costal margin

intercostal space

manubrium

dorsal manubrial spine-^
sterno-coracoidal process -

corocoidal sulcusj ^or^al J'P
[ventral lip

ventral labial prominence
sterno-coracoidal impression

intercostal space

costal margi
costal process

ventral

m anubrial spme

anterior

carinal

margin

sternal plate

intermuscular line

posterior lateral
process

carinal apex
cari na
Sternal notch

-

^p^-^-post pectoral line

-xiphial area

Sternum of Chen lijipcrboreus. Fig. 13, dorsal view; fig. 14, lateral view. X 1.

XXX

■Scapular tuberosity-

clavicle

Furcular process
coroco-humeral
surface

left clavicle

symphysis

furcular facet-

brachial
tuberosity

pneumatic

foramina

triossea!

canal

coracoid a I
articulation

Aquila chrysaetos. Fig. 15, furcula, dorsal view; fig. 16, coracoid, internal
view; fig. 17, coracoid, dorsal view; fig. 18, scapula, ventral view; fig. 19, scapula,
dorsal view. X 1.

XXXI

external

tuberosity

pectoral

attachment

median crest-

attachmentof,
latissimusdorsj.
posterioris

deltoid crest

line of-
latissimus dorsi
anterioris

capital groove
internal tuberosity
pneumatic fossa
pneumatic foramen
attachment of infraspinatus
attachment of supraspinatus

external
tuberosity

bicipital

furrow

\ ,i 'deltoid crest

VI

shaft-

shaft

ectepicondylar ft' ...
prominence

ecte pi condyle

,^^h,^=fexternal condyle
trochlea^ internal condyle

attachment of pronator brevis

attachment of
anterior articular ligament

entepicondy lar prominence

internal condyle
entepicondyle

impression of
brachialis antlcus
brachi al

depression

intercon dyl ar
furrow

ectepicondylar
prominence

external
condyle
ectepicondyle

Humerus of Chen hyperboreits. Fig. 20, anconal view; fig. 21, palmar view. X 1.

XXX11

internal cotyla
i ntercotylar area-
prominence for
anterior articuloc.
ligament

bicipital attachment

impression of
brachiolis anticus

external condyle^

internal
condyle

corps' >^tlculalion

27

ligamental
prominence

Aquila chrysaetos. Fig. 22, radius, anconal view; fig. 23, ulna, palmar view;
fig. 24, radius, palmar view; fig. 25, ulna, anconal view; fig. 26, cuneiform; fig. 27,
scapholunar. Figs. 22-25 X %; figs. 26-27 X 1.

xxxiii

carpaltrochlea,
internal

I iga mental Fossa]

ligamental
attachment-"6f
pisiform process

. , TTT -metacarpal II

metacarpal III-

intermetacarpal

space

distal

metacarpal

symphysis

metacarpal facet-

anterior carpal fossa
extensor
attachment

process of
metacarpal I

pollical facet

metacarpal facet

29

process of
metacarpal I

metacarpal 1

s\/ tuberosity of
metacarpal n

* facet for digit II

v facet for digit III

metacarpal II

[digital facet

A tuberosity of
1 J metacarpal II

facet for digitl!

digital facet

.carpal trochlea

external
llgamenta!
attachment

flexor

attachment

intermetacarpal

"tuberosity

tendinal groove-f
< digital facet

intermetacarpal

space'

metacarpal III

distal metacarpal
symphysis

facet for digit III

Aquila chrysaetos. Pig. 28, carpometacarpus, internal view. figs. 29-32,
phalanges of manus: fig. 29, digit 2, phalanx 1; fig. 30, digit 3; fig. 31, pollex;
fig. 32, digit 2, phalanx 2; fig. 33, carpometacarpus, external view. X 1.

XXXIV

median

dorsal

ridge

prezygapophysis

•anterior portion of synsacrum

costal facets

synsacral
thoracic
vertebrae

lumbar

vertebrae

jftj^parapophysis

^ 'isacrai

■vertebrae

renal depression

ischial angle

Pelvis of Chen hyperboreus. Fig. 34, lateral view; fig. 35, ventral view. X 1.

XXXV

trochanter-

iliac facet

obturator

ridge

nutrient

foramen

shaft

fibular

condyle

fibular

groove

neck

attachment of
round ligament

head

posterior

intermuscular

ine

popliteal area

internal condyle
intercondylar fossa

external condyle

tubercle

uncinate

process

articular surface
for sternal rib

-costal

facet

first

phalanx

second

phalanx

third
phalanx

digital condyle

attachment of
round ligament-

head

43

neck

sternal
— f acet

42

Figs.
Fig. 36,
42, digit
view. X

ungual

phalanx

nternal condyle

ligamental
attachment

pit for tibialis antic

-trochanter

■trochanteric

ridge

anterior

intermuscular

ine

■rotular groove

flexor
attachment

external
condyle

36-42 and fig. 45, Chen hyperboreus; figs. 43-44, Aquila chrysaetos.
femur, posterior view; figs. 37-38, rib and sternal rib no. 4; figs. 39-
3^of pes; fig. 43, patella; fig. 44, metatarsal I; fig. 45, femur, anterior

XXXVI

inner cnemial crest
head of fibula

external
articular
surface

fibula
fibular crest

foramen for
medullary,
artery

internal /

V articular surfacex^’

interarticular

area

ligamental

attachment

flexor

attachment

head of fibula

external
articular surface

outer

cnemial crest

rotular crest
inner cnemial crest-

outer

cnemial crest

spine of fibula

fibula

fibular crest

intermuscular
ine

posterior
intercondylar
sulcus

external condyle

external ligamental
prominence

groove for
peroneus profundus

external condyle

internal

igamental

prominence

internal

condyle

48

supratendinal
bridge

internal
ligamental
prominence

internal condyle

spine of fibula

i /tendinal groove

Id i/

-groove for _
peroneus*
profundus

anterior inter-
,'fTcondylar fossa

^external condyle

Tibiotarsus and fibula of Chen liyperboreus. Fig. 4(5, posterior view; fig. 47,
proximal end, proximal view; fig. 48, distal end, external view; fig. 49, anterior
view. X 1.

xxxvn

internal cotyla

proximal ligamental
attachment

tubercle For
tibialis anticus

anterior

metatarsal groove

inner

extensor groove

metatarsal Facet-

intertrochlear
notches:
[external
[interna

external cotyla

attachment oF external ligament

proximal Foramen

internal

cotyla

intercotylar

area

calcaneal ridges of hypotarsus

outer
extensor
roove

distal

Foramen

trochleae For digits 2 3 4

intercotylar area

intercotylar

depression

hypotarsus

intercotylar

prominence

innerl proximal
outer] Foramen

calcaneal ridges

^ distal foramen

trochlea
For digit 4

yving of trochlea

For digit 2

J trochlea
For digit 2

Tarsometatarsus. Figs. 50, 51 and 54, Aquila cliri/saetos ; figs. 52-53, Chen
liypcrboreus. Fig. 50, anterior view; figs. 51 and 52, proximal end, proximal
view; fig. 53, proximal end, posterior view; fig. 54, posterior view. X 1.

trochlea
For digit 3

posterior
metatarsal groove

metatarsal Facet

intertrochlear
notches:
external]
internal]

xxxviii

PAPERS IN AVIAN PALEONTOLOGY
HONORING HILDEGARDE HOWARD

FOSSIL BIRDS AND EVOLUTION

By George Gaylord Simpson1

PALEORNITHOLOGY

One of the first textbooks of vertebrate paleontology, that
published in 1898 by A. (later Sir Arthur) Smith Woodward
devoted 14 pages, 3 percent of its text pages, to birds. It dis-
cussed particulars of only Archaeopteryx, Hesperornis, Ichthy-
ornis, Aepyornis (not figured) and three moas. When I studied
vertebrate paleontology at Yale in the mid-1920’s the class
received even shorter coverage of birds. As much time was
devoted to “Tetr apteryx,” a “bird” that never existed, as to
the two real fossil birds that were discussed. It was generally
felt that fossil birds were too rare to have any great evolution-
ary interest beyond that engendered by Archaeopteryx, of
which more later. That depreciative view is still sometimes
encountered, but now rarely and without justification.

A decided change in this subject, and in attitudes toward
it, began in the late 1920’s and has been accelerating ever
since. It is true that the late Alexander Wetmore published a
short paper on a fossil bird as early as 1917 (Wetmore 1917)
and long continued such studies, but he was primarily a neon-
tologist and his career was centered on Recent birds. Hilde-
garde Howard published a long paper on a fossil bird in 1927,
the start of a great career. She was certainly one of the first,
perhaps the very first, to adopt paleornithology as a full-time
specialty and to occupy a salaried position explicitly devoted
to that speciality.

That many fossil birds were in fact known by 1930 is evident
from Lambrecht’s massive Handbuch der Palaeornithologie
(1933). Even so, the first sentence of that work begins (in Ger-
man), “As is known, the number of remains of fossil birds is
comparatively very limited. . . .” The fossil record of birds is
indeed still markedly incomplete, as is that of even such richly
documented groups as, for instance, echinoderms or mam-
mals. Nevertheless it is now far from negligible, as witness
Brodkorb’s Catalogue of Fossil Birds (1963, 1964, 1967, 1971a,
1978) and Fisher’s chapter on Aves in the symposium volume
on The Fossil Record (1969).

At present the fossil record of birds not only throws consid-
erable light on the history of birds, a subject of great interest
in itself, but also provides evidence bearing more broadly on
the principles of evolution. In what follows I shall exemplify
both of those aspects of the subject.

THE EARLY BIRD

A tantalizing and perhaps incorrect reference to Jurassic
birds was published by Schlotheim as early as 1820. A partial

1 The University of Arizona and The Simroe Foundation, S1S1 East

Holmes St., Tucson, Arizona 85711.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 330:3-8.

but considerable skeleton of the Jurassic Archaeopteryx was
found in 1855, but was not recognized as avian until 1970 (see
Ostrom 1972). The first specimen of a Jurassic bird to be rec-
ognized as such was a splendidly preserved, nearly complete
skeleton with impressions of feathers that was found in 1861
and acquired by the British Museum (Natural History). It was
named and briefly described by von Meyer (1862) and more
fully described by Owen (1863). Numerous other studies of
that and a second specimen similar in origin have appeared
since 1863. The definitive study of the British Museum spec-
imen, made after further preparation, was by de Beer (1954a).
It is interesting that this was Sir Gavin’s only excursion into
paleornithology. One might say that he studied this specimen
only because it was there: he was at the time director of the
British Museum (Natural History).

It was at once recognized, and is obvious at first sight, that
Archaeopteryx has resemblances both to birds and to reptiles.
It was early agreed that Archaeopteryx had evolved from some
reptilian stock, but beyond that point opinions long differed.
An occasional minority view was that Archaeopteryx was a
pseudo-bird, independently derived from reptiles with no close
relationship to true birds. However, there now seems to be no
dissent from the majority view that it was in or near the an-
cestry of some, and probably of all, later birds and should
itself be classed in the Aves. As to the reptilian ancestry, it
was suggested as early as 1863 (Weinland) and still maintained
as late as 1950 (Petronievics) that Archaeopteryx was derived
from some lacertilian stock. Owen (1874) hinted, although not
clearly in evolutionary terms, at a pterosaur ancestry. Neither
of those views is tenable in the light of later studies. Abel
(1919) suggested derivation from a pseudosuchian, but possibly
from some dinosaur itself evolved from a pseudosuchian (or
other early thecodont). Heilmann (1926) more positively en-
dorsed derivation from a pseudosuchian. T.H Huxley (1868),
somewhat vaguely, and Marsh (1877) and others following
him, more positively, supported descent from some early di-
nosaurs.

There has long been a strong consensus, now virtually unan-
imous, that birds, including Archaeopteryx, evolved either
from a dinosaurian (theropod) stock, or from a common an-
cestry with such a stock but within prior thecodonts. Ostrom
(e.g., 1975), the most recent to study this question in depth,
is insistent on a dinosaurian origin. He considers the skeleton
of Archaeopteryx more dinosaur-like than bird-like, but con-
tinues to classify the genus as an ancestral, or near-ancestral,
bird.

Whether birds arose from dinosaurs or from the immediate
common ancestry of birds and dinosaurs is a phylogenetic de-
tail of no great importance from a broader view of evolutionary

4

Simpson: Fossil Birds and Evolution

Table 1. Some data on first appearances of families of birds in the
fossil record, based mainly on Fisher (1967).

Geologic
Period
or Epoch

Number of
First Known
Appearances

Percent
Extinct
before the
Holocene

Percent
Surviving
into the
Holocene

Jurassic

1

100

0

Cretaceous

12

75

25

Paleocene

3

66.7

33.3

Eocene

41

31.7

68.3

Oligocene

18

38.9

61.1

Miocene

24

25

75

Pliocene

9

0

100

Pleistocene

38

2.6

97.4

Holocene

54

0

100

theory. In either case it is clear that Archaeopteryx stands in
an intermediate position between the classes Reptilia and
Aves. During the transition from one class to another, evolu-
tion may have been, and quite likely was, accelerated, but
there was a transition, not a saltation as has from time to time
been claimed for the origin of taxa at upper hierarchic levels.
There are no known instances of such origins that cannot have
been transitional, many known cases, of which this is only
one, in which the origin was almost certainly transitional, and
no known cases in which the evidence makes saltation more
probable. The old saw that the first bird was born from a
reptile’s egg is not true.

That is the most important theoretical bearing of the early
bird, but it has another also of some importance. When there
is a transition from one high taxonomic category to another
there are two extreme theoretical possibilities, although some-
thing between the two extremes is also quite possible. At one
extreme, all characteristics of the ancestral form may evolve
uniformly into the different characteristics of the descendant,
so that an animal like Archaeopteryx would be in all respects
intermediate between one high taxon, in this case the Class
Reptilia, and another, here the Class Aves. As a matter of fact
Archaeopteryx is not intermediate in that sense. Many of its
characters had changed hardly at all from the reptilian grade,
although I think that Ostrom, as previously cited, has some-
what overstated that case. On the other hand, some characters
of that genus were already completely avian, notably the fur-
cula, the presence of feathers, and their arrangement on the
wing.

De Beer (1954a) did not discuss just this point in his mono-
graph on the London specimen, but he did in an address to
the British Association for the Advancement of Science (de
Beer 1954b). He proposed the term “mosaic evolution” for the
apparently disharmonious sort of transition exemplified in Ar-
chaeopteryx. He also gave other examples, and many more
have been pointed out since then. In fact it had long been
recognized, although not always so clearly, that different char-
acteristics of organisms often, indeed usually, evolve at quite
different rates even within a single lineage. (Although I was
not the first to notice this, I did clearly state it in 1944, 10
years before the restatement by de Beer.) De Beer’s term is apt
and is a handy designation for this phenomenon. De Beer did
not himself claim that his observation of the phenomenon was
original, although some subsequent users of the term have

mistakenly ascribed the principle, and not only the term, to
him.

Two other points involving Archaeopteryx are to be men-
tioned here only briefly. It is fairly obvious that/ 4 rchaeopteryx
could not have been capable of long, sustained flight in the
manner of most modern birds. There was, however, a clear
consensus that it was capable of brief gliding or leaping flight
and that its strongly feathered forelimbs were a stage in the
evolution of sustained flight. Recently, however, Ostrom
(1976) has maintained that the origin of those feathered fore-
limbs had nothing to do with flight but were adaptations of a
running animal for garnering insects. If that were true, those
forelimbs would be only adventitiously preadapted for flight.
I do not pretend to authority on this point, but I do find Os-
trom’s hypothesis incredibly bizarre. (See Feduccia 1979 — note
added after completion of this manuscript.)

The other point is that it has several times been suggested
that various birds without aerial flight (although many of them
with wings) were primarily flightless either because they
evolved from reptiles independently of true Aves or because
the ancestral Aves were flightless (for example, Lowe 1944,
and earlier papers there cited). With special reference to pen-
guins, but incidentally to other supposedly flightless birds, I
(Simpson 1946) strongly opposed that view, and I do not know
of any more recent adherence to it.

BITS OF AN OUTLINE OF HISTORY

There have been several fairly recent reviews of the whole
history of birds, most notably that by Brodkorb (1971b). I am
not capable of writing a review in equal or greater depth and
have no intention of trying. There are, however, some points
bearing on evolutionary principles and on the interpretation
of the fossil record that suggest brief comment here.

Some data on the first appearances of families of birds are
given in Table 1 . I have based these on Fisher (1967), primarily
because Brodkorb’s catalogue was not complete when this pa-
per was written. Even now the earlier parts (at least Brodkorb
1963, 1964, and 1967) are out of date. The data from Fisher,
more complete than Brodkorb’s when this paper was written,
seem to be sufficient for the general points here made.

It is not surprising that the percentage of pre-Holocene ex-
tinctions decreases, and that of survival into the Holocene
increases almost regularly from Jurassic to Holocene. (A few
families known only from the Holocene but now extinct are
here counted as Holocene survivals.) The only somewhat ev-
ident irregularity is in the Oligocene, and this is probably a
sampling error. For one thing, the Oligocene was shorter than
either the Eocene or the Miocene, and so would have fewer
first appearances even if the rate per annum were constant.

The very high numbers of first appearances in the Pleisto-
cene and Holocene are a measure of the incompleteness of the
record. It is highly improbable that these families actually
originated in either of those epochs. Thus with no probable
and few possible exceptions, their pre-Pleistocene members
simply have not yet been found, to put the matter optimisti-
cally. To put it pessimistically, in many instances pre-Pleis-
tocene representatives may not exist as accessible fossils. (Even
for vertebrates it is certain that not all species or genera, prob-
able that not all families, and possible that not all orders were

Simpson: Fossil Birds and Evolution

5

fossilized and are now present in rocks accessible for explo-
ration.)

It is a reasonable conclusion from these figures and from the
more detailed data on which they are based that most and
perhaps all of the families of birds that have ever existed, and
hence of course those now surviving, had arisen by the end of
the Miocene. That agrees with the well-informed opinion of
Brodkorb (197 1 b:43), who wrote that, “By the end of the Mio-
cene all of the nonpasserine families were probably estab-
lished, as well as most, if not all, of the passerines.” He then
estimated that there were about 155 families extant in the
Miocene, the number being reduced moderately to 148 in the
Holocene.

For comparison, I have given in Table 2 similar data for
Mammalia, a class with a better but still quite incomplete
fossil record. The figures are tentative only, because there is
no recent and reliable listing of all known mammalian families
and their distribution in the Cenozoic, although Lillegraven
(1972) has published graphs based on a fairly recent tabulation.
(There is one by Lillegraven, Lindsay, and Simpson, as yet
unpublished, for the Mesozoic.) My arrangement is conser-
vative, with fewer families than are now sometimes recognized
in the Tertiary, but I believe that the pattern is significant.
Even so my arrangement for mammals has many more families
(259) than Fisher’s for birds (200). The patterns are similar in
some respects but strikingly different in others. A considerable
number of bird families first known in the Cretaceous, Paleo-
cene, and Eocene — 32 families or 57 percent of those first ap-
pearing during those times — survived into the Holocene. For
mammals the corresponding figures are 116 families and 26.7
percent. Both proportionately and absolutely, many more
mammalian than bird families first appear in the record at
those times, but fewer of them survived into the Holocene.
For both classes most of the Holocene families had appeared
by the end of the Miocene, but some of the mammalian fam-
ilies probably did become differentiated in the Pliocene where-
as it is not clear that any bird families did. In both cases it is
unlikely that any family emerged after the Pliocene. The much
lower numbers and percentages of first appearances of mam-
malian than of bird families in the Pleistocene and Recent is
evidence that the fossil record for mammals, although still
incomplete, is better than that for birds.

As Brodkorb (1971b) has pointed out, more living families of
birds appear in the record for the Eocene than at any other
time. (It is understood that comparison with the higher num-
bers for the Pleistocene and Holocene is not valid.) For mam-
mals there is a marked difference: the greatest number of living
families appear in the record for the Miocene. There are in
fact many more Miocene first appearances than Pleistocene or
Holocene. As relatively few Eocene mammalian families are
still living, it is clear that there has been a much more marked
faunal turnover since the Eocene for mammals than for birds.

The bird record is strongly biased both taxonomically and
geographically. The most striking taxonomic bias is that rel-
atively far fewer passeriform families than nonpasseriform
families are known before the Pleistocene. On Fisher’s data
only 22.8 percent of recognized passeriform families are known
before the Pleistocene but for nonpasseriforms the figure is 67
percent. That may be a sampling bias, caused in part by non-
passeriforms (such as many shore birds) being more likely to
be preserved in sediments, by a higher proportion of nonpas-

Table 2. Some data on first appearances of families of mammals in
the fossil record.

Geological
Period
or Epoch

Number of
First Known
Appearances

Percent
Extinct
before the
Holocene

Percent
Surviving
into the
Holocene

Rhaeto-Lias

4

100

0

Jurassic

11

100

0

Cretaceous

19

94.7

5.3

Paleocene

33

100

0

Eocene

63

82.5

17.5

Oligocene

44

56.8

43.2

Miocene

38

28.9

71.1

Pliocene

21

9.5

90.5

Pleistocene

7

0

100

Holocene

26

0

100

seriforms in regions that have been sampled, or by smaller
average size of passeriforms making them harder to find and
identify. However it is also evident that the differentiation of
passeriform families probably occurred, on an average overall,
at later dates than for nonpasseriforms.

The geographic bias largely, although not entirely, follows
the intensity of paleontological field work. Fossil birds are
fairly well known in North America and Europe but less so
in South America, Asia, Africa, and Australia. Yet even in
Australia there is a fair sampling from the Miocene onward,
as was recently tabulated by Rich (1975). The evidence sug-
gests that by mid-Miocene, at latest, the Australian fauna was
fairly modernized and largely endemic. Virtually all the known
fossils are nonpasseriform. From Antarctica some fossil pen-
guins are known, but no deposits likely to contain nonmarine
birds have yet been found.

EVIDENCE FOR SUCCESSIVE
RADIATIONS

Descriptions by Marsh (1872, 1880) of Hesperornis and
Ichthyornis, supposedly toothed birds, created a sensation and
these have been the most discussed fossil birds except Ar-
chaeopteryx. It was already known that Archaeopteryx had
teeth, but Marsh’s “Odontornithes” were much later, and some
authorities did not consider Archaeopteryx wholly (or at all) a
bird. More recently Gregory (1952) suggested that, although
Hesperornis had teeth, Ichthyornis probably did not. Bock
(1969) still later questioned whether Hesperornis had teeth.
Brodkorb (1971b) attacked “the fable of the toothed birds.” The
fable was simply the claim that all Mesozoic birds had teeth.
In fact both Hesperornis and Ichthyornis did have teeth (Gin-
gerich 1972, 1973; Martin and Stewart 1977). Although pos-
sibly tooth-bearing parts are not known in the likewise Cre-
taceous genera Baptornis (referred by Brodkorb 1963 to the
Podicipediformes), Enaliornis (referred bv Brodkorb to the
Gaviiformes), or Neogaeornis (referred by Brodkorb to the
Podicipediformes), Martin and Tate (1976) have established
that these genera, too, probably belong in the Hesperornithi-
formes.

Added indication of the archaic nature of the genera listed
in the preceding paragraph is given by evidence that the skull
of Hesperornis was in fact palaeognathous (Gingerich 1973,
1976) although faulty reconstruction had led to belief that it

6

Simpson: Fossil Birds and Evolution

was neognathous. Although the skull structure of Archaeop-
teryx is not known in clear detail, Gingerich has also mar-
shalled evidence that a palaeognathous skull was probably
ancestral for birds in general, and hence probably was present
in Archaeopteryx. (I am, however, informed that Martin and
Whetstone, in a study not published when this paper went to
press, deny that Hesperornis was palaeognathous, which
would also cast doubt on the possible palaeognathy of Ar-
chaeopteryx. )

Thus in the Cretaceous there was a group of archaic birds
apparently sharing ancestral characters, although divergent to
the ordinal level in derived characters. Among the Hesperor-
nithiformes and the Ichthyornithiformes long known, some,
at least, and possibly all were palaeognathous, and some and
possibly all retained teeth. To them may now be added Go-
biopteryx from the late Cretaceous of Mongolia (Elzanowski
1977). It, too, was palaeognathous, but it was toothless. El-
zanowski proposed for it a new order, Gobiopterygiformes,
but it might well be put in the still living order Casuariiformes,
or the Struthioniformes if, as has been defended by Bock (1963)
among others, all the ratites were put in one order. (The def-
inition of such an order becomes difficult if some palaeognath-
ous birds are excluded from it.) Brodkorb (1978:224) has ex-
pressed his belief that Gobiopteryx is not a bird, but a small
dinosaur.

The most economical hypothesis is that the living palaeog-
nathous birds, the ratites (whether classified as one order or
as up to six) and the tinamous, are survivors of an archaic
radiation. Most of the known Mesozoic members of that ra-
diation were aquatic or at least littoral and most were found
in marine rocks. Of earlier known members of the radiation,
Archaeopteryx and Ichthyornis were most likely to have been
land birds, but they have been found only in definitely marine
beds. The known later, Eocene to Holocene, palaeognathous
birds are land birds; all but the tinamous are flightless, and
the tinamous are poor fliers.

Thus we can return, with Gingerich (1976), to the essence
of views already expressed by T.H. Huxley (1868) and by-
Marsh (1880) long ago. The palaeognathous birds are the relics
(“waifs and strays” of Huxley) from an archaic (mainly Cre-
taceous) radiation of the Aves.

Although Brodkorb’s view that almost all the known Cre-
taceous birds were referable to, or near the ancestry' of, Ce-
nozoic neognathous birds is an overstatement, it seems estab-
lished that, near the end of the Cretaceous, some were
(Brodkorb 1976). Because of the bias of sampled environ-
ments, the known Cretaceous members of probably neognath-
ous groups are almost all aquatic, marine, or shore birds. They
strongly suggest that a major radiation of neognathous non-
passeriforms was under way before the end of the Cretaceous,
reaching its height in the early Cenozoic. Starting within that
radiation, one basic line, that of the passeriforms, underwent
its own radiation from mid-Cenozoic to Holocene and became
the dominant group in later Cenozoic and Recent avifaunas.

A WORD ABOUT PENGUINS

The oldest known penguins are late Eocene in age (not early
Eocene, as indicated by Fisher 1967; Fisher also errs in listing
Palaeeudyptes marplesi as a neospecies). At that time they
already had all the derived characteristics of the family Sphe-

niscidae as a whole. Some, at least, of the known late Eocene
through Miocene species had a few characters that seem to
have been more primitive than recent penguins, but at the
generic level they also had derived characters that make them
all quite distinct from any recent genus. Some of them, even
in the late Eocene, had quite specialized generic characters.
It is unlikely that any of the known forms of those ages were
closely related to recent penguins at the generic level, and
those that are adequately known were probably not ancestral
to known post-Miocene penguins. Only in the late Pliocene of
New Zealand do two species occur in the known record that
are close to, and have been referred to, living genera: Pygos-
celis and Aptenodytes. (On those two see Simpson 1972, and
on fossil penguins in general Simpson 1975, and earlier pub-
lications there cited; for a less technical discussion see also
Simpson 1976.) It is curious that those two genera now live
much farther south than where their known fossil species were
found, although by the late Pliocene New Zealand must have
been in nearly the same latitudes as now. No pre-Pleistocene
fossils are known for the genera now breeding in New Zealand:
Megadyptes, Eudyptes, and Eudyptula.

The family Spheniscidae and order Sphenisciformes must
have evolved before the late Eocene when they first appear in
the record, and some, if not all, Holocene genera must have
had distinguishable ancestors before the late Pliocene. As pen-
guins are marine and littoral, they would seem well-suited for
preservation as fossils. Nevertheless two special circumstances
make the almost complete lack of ancestral or transitional se-
quences explicable. First, penguins are predominantly insular.
One genus each now occurs on the coasts of three continents:
Africa and South America (Spheniscus), and Australia (Eu-
dyptula). Only two genera (Aptenodytes and Pygoscelis) occur
in continental Antarctica, where, furthermore, no appropriate
fossil-bearing post-Eocene rocks are known. All six living gen-
era are now much more common on islands than on continents,
and twelve of the (nominally) sixteen to eighteen living species
are almost or quite confined to islands when ashore anywhere.
The prolific polvtypy of the group now, and even more its
speciation in the past, are evidently the result of the isolation
of island populations, with some subsequent dispersal. The
islands on which ancestral speciation leading to later genera
occurred probably no longer exist for the older part of the
record, at least, and for the later part those that exist are not
known to have fossiliferous rocks of appropriate ages. A sec-
ond point is that all known fossil penguins are well within the
geographic ranges of Recent penguins, and the whole order
has probably always been almost entirely restricted to areas
now in the Southern Hemisphere. But the known fossil record
of birds in general in that hemisphere is exceptionally poor. It
is surprising that so many, rather than so few, fossil penguins
are known.

Until recently penguins were usually considered particularly
primitive birds. That view is evident even in the the fairly
recent compendious summary by Fisher previously cited. Pen-
guins are there listed in the heart of orders belonging to the
earliest radiation, between the Ichthyornithiformes and the
Struthioniformes. That and similar arrangements may be a
not wholly conscious hangover from the speculation that pen-
guins are primitively (ancestrally) flightless. In fact they are
carinate and neognathous and they fly with great power, but
in water rather than in air. They quite surely had ancestors

Simpson: Fossil Birds and Evolution

that did fly in air. The picture of avian evolution here adopted
is a succession of three radiations of differing character and
scope: ancient and largely or wholly palaeognathous, neog-
nathous nonpasseriform, and neognathous passeriform. It is
clear where penguins belong in that scheme: in the neogna-
thous nonpasseriform radiation. Within that group their de-
rived characters are unique in detail and association. They
make the penguins among the most specialized birds. Only in
some of the Alcidae (including Mancallinae), another branch
of the neognathous nonpasseriform radiation, did some similar
derived characters evolve (but see Olson and Hasegawa 1979;
Olson this vol — Ed.). That development was clearly indepen-
dent and convergent on the part of sea birds that were geo-
graphic, Northern Hemisphere, vicars of the Southern Hemi-
sphere penguins.

NOTE

The manuscript was written early in 1978. Although a few
changes have been made since that time, it has not been pos-
sible to update fully.

LITERATURE CITED

Abel, O. 1919. Die Stamme der Wirbeltiere. Verein. Wis-
sensch. Verleger, Berlin and Leipzig. 914 pp.

Beer, G. de. 1954a. Archaeopteryx lithographic a. A study
based upon the British Museum specimen. British Mu-
seum (Nat. Hist.), London. 68 pp.

. 1954b. Archaeopteryx and evolution. The Advance-
ment of Science, 11:160-170.

Bock, W.J. 1963. The cranial evidence for ratite affinities.
Proc. XIII Internat. Ornith Cong. 1:39-54.

. 1969. Origin and adaptive radiation of birds. Ann.

New York Acad. Sci. 169:147-155.

Brodkorb, P. 1963. Catalogue of fossil birds. Part 1 (Ar-
chaeoptervgiformes through Ardeiformes). Bull. Florida
State Mus., Biol. Sci. 7:179-293.

. 1964. Catalogue of fossil birds. Part 2 (Anseriformes

through Galliformes). Bull. Florida State Mus., Biol. Sci.
8:195-335.

. 1967. Catalogue of fossil birds Part 3 (Ralliformes,

Ichthvornithiformes, Charadriiformes). Bull. Florida
State Mus., Biol. Sci. 11:99-220.

. 1971a. Catalogue of fossil birds. Part 4 (Columbi-

formes through Piciformes). Bull. Florida State Mus.,
Biol. Sci. 15:163—266.

. 1971b. Origin and evolution of birds. Pp. 19-55 in

Avian Biology (Farner, D.S., and J.R. King, Eds.). Ac-
ademic Press, New York and London.

. 1976. Discovery of a Cretaceous bird apparently an-
cestral to the orders Coraciiformes and Piciformes (Aves
Carinatae). Smithsonian Contrib. Paleobiol. No. 2 7:67—
73.

. 1978. Catalogue of fossil birds. Part 5 (Passeri-
formes). Bull. Florida State Mus., Biol. Sci. 2 3(3): 139—
228. Dec. 15, 1978. (I had not received this publication
when the present manuscript was completed.)
Elzanowski, A. 1977. Skulls of Gobiopteryx (Aves) from the
upper Cretaceous of Mongolia. Palaeontologia Polonica
No. 37:153-165.

Feduccia, A. 1979. Feathers of Archaeopteryx : asymmetric

vanes indicate aerodynamic function. Science 203:1021 —
1022. (Received after this manuscript was completed.)

Fisher, J. 1967. Aves. Pp. 733-762 in The Fossil Record
(Harland, W.B. et ah, Eds.) Geol. Soc. London.

Gingerich, P.D. 1972. A new partial mandible of Ich-
thyornis. Condor 74:471-473.

. 1973. Skull of Hesperornis and early evolution of

birds. Nature 243:70-73.

. 1976. Evolutionary significance of the Mesozoic

toothed birds. Smithsonian Contrib. Paleobiol. No. 27:23-
33.

Gregory, J.T. 1952. The jaws of the Cretaceous toothed
birds, Ichthyornis and Hesperornis. Condor 54:73-88.

Heilmann, G. 1926. The origin of birds. Witherbv, London.

208 pp.

Howard, H. 1927. A review of the fossil bird Parapavo cal-
ifornicus (Miller) from the Pleistocene asphalt beds of
Rancho La Brea. Univ. California Publ., Bull. Dept.
Geol. 17:1-30.

Huxley, T.H. 1868. Remarks upon Archaeopteryx litho-
graphica. Ann. Mag. Nat. Hist. (4)1:220-224.

Lambrecht, K. 1933. Handbuch der Palaeornithologie.
Borntraeger, Berlin. 1024 pp.

Lillegraven, J.A. 1972. Ordinal and familial diversity of
Cenozoic mammals. Taxon 1:261-274.

Lowe, P R. 1944. Some additional remarks on the phylogeny
of the Struthiones. Ibis 86:37-43.

Marsh, O.C. 1872. Discovery of a remarkable fossil bird.
Amer. Jour. Sci. 3:56-57.

. 1877. Introduction and succession of vertebrate life

in America. Proc. Amer. Soc. Advan. Sci. 1877:21 1-258.

. 1880. Odontornithes: a monograph on the extinct

toothed birds of North x^merica. Report of the geological
exploration of the 40th parallel by Clarence King: i-xv,
1-201. Government Printing Office, Washington. (Also
published as Yale University Peabody Museum Memoirs,
No. 1.)

Martin, L.D., and J.D. Stewart. 1977. Teeth in Ich-
thyornis (Class: Aves). Science 195:133 1-1332. (Received
after this manuscript was completed.)

Martin, L.D., and J. Tate, Jr. 1976. The skeleton of Bap-
tornis advenus (Aves: Hesperornithiformes). Smithsonian
Contrib. Paleobiol. No. 27:35-66.

Meyer, H. von. 1862. Archaeopteryx lithographica aus dem
lithographischen Schiefer von Solenhofen. Palaeonto-
graphica 10:53-56.

Olson, S.L. 1980. A New Genus of Penguin-like Pelecani-
form Bird from the Oligocene of Washington (Pelecani-
formes: Plotopteridae). (this vol.)

Olson, S.L., and Y. Hasegawa. 1979. Fossil Counterparts
of Giant Penguins from the North Pacific. Science
206(44 1 9):688 — 689.

Ostrom, J.H. 1972. Description of the Archaeopteryx spec-
imen in the Tevler Museum, Haarlem. Proc. Konikl.
Nederl. Akad. Wetensch. Amsterdam B75:289-305.

. 1975. The origin of birds. Ann. Rev. Earth and

Planetary Sci. 3:55-77.

. 1976. Some hypothetical stages in the evolution of

avian flight. Smithsonian Contrib. Paleobiol. No. 27:1-20.

Owen, R. 1863. On the Archaeopteryx of von Meyer, with
a description of the fossil remains of a long-tailed species

Simpson: Fossil Birds and Evolution

from the lithographic stone of Solenhofen. Phil. Trans.
Roy. Soc. London 1 5 3 : 3 3 — 4 7 .

. 1874. Monograph of the fossil reptiles of the Liassic

formations II. Pterosauria. (Pterodactylus). Paleont. Soc.
Monogr. 2 7 :i— vii, 1-14.

Petronievics, B. 1950. Les deux oiseaux fossiles les plus
anciens (Archaeopteryx et Archaeornis). An. Geol. Pen.
Balkan 18:89-127.

Rich, P.V. 1975. Antarctic dispersal routes, wandering con-
tinents, and the origin of Australia’s non-passeriform avi-
fauna. Mem. Nat. Mus. Victoria 36:63-126.

Schlotheim, E.F. von. 1820. Die Petrefaktenkunde auf
ihrem jetzigen Standpunkte durch die Beschreibung sei-
ner Sammlung versteinerter und fossiler Ueberreste des
Thier- und Pflanzenreichs der Vorwelt erlautert. Gotha.
437 pp.

Simpson, G.G. 1944. Tempo and mode in evolution. Colum-
bia Univ. Press, New York. 237 pp.

. 1946. Fossil penguins. Bull. Amer. Mus. Nat. Hist.

87:1-99.

. 1972. Pliocene penguins from North Canterbury,

New Zealand. Rec. Canterbury Mus. 9:159-182.

. 1975. Fossil penguins. Pp. 19-41 in The Biology of

Penguins (B. Stonehouse, Ed.). Macmillan, London and
Basingstoke.

. 1976. Penguins, past and present, here and there.

Yale Univ. Press, New Haven and London. 150 pp.

Weinland, D.F. 1863. Der Greif von Solenhofen (Archaeop-
teryx lithograpliica H. von Meyer). Zool. Gart. Frankfurt
4:118-122.

Wetmore, A. 1917. The relationships of the fossil bird Pa-
leochenoides miocenus. Jour. Geol. 25:555—557.

Woodward, A.S. 1898. Outline of vertebrate palaeontology
for students of evolution. Cambridge Univ. Press, Cam-
bridge. 470 pp.

PHYLOGENETIC THEORY AND METHODOLOGY IN AVIAN
PALEONTOLOGY: A CRITICAL APPRAISAL

By Joel Cracraft1

ABSTRACT: The thesis of this paper is that the application of the theory and method of cladistic

analysis (phylogenetic systematics) will greatly improve systematic practices within avian paleontology.
Specifically, cladistic analysis will (1) facilitate the formulation of more precise phylogenetic hypotheses
of Recent taxa, and these in turn will clarify the array of hypotheses that must be considered when
analyzing the systematic position of fossil taxa; (2) draw attention to the concept that phylogenetic rela-
tionships are postulated on the basis of shared derived characters and to the realization that the mor-
phology of fossil taxa will have to be studied in these terms; and (3) de-emphasize the importance of
considering intermediate taxa as possible ancestors and focus attention instead on assessing their cladistic
relationships. The major methodological problem in avian paleontology is the belief that relationships can
be determined by some measure of overall similarity. Cladistic theory and methodology provides a solution
to this problem: similarity must be partitioned into primitive and derived conditions at each hierarchical
level. Consequently, there is a nested pattern of derived similarities for any set of taxa, and the primary
methodological goal of systematics is the search for this pattern.

There has been scant discussion about the theory and meth-
od of phylogenetic analysis in avian paleontology. Avian pa-
leontologists seem to operate comfortably within the concep-
tual framework established by post-Darwinian vertebrate and
invertebrate paleontology. In general this can be characterized
by the assumption that the phylogenetic process is slow and
gradual, with species being arbitrary segments of an evolu-
tionary continuum. This transformational or gradualistic phi-
losophy engenders the view that phylogenetic analysis is pri-
marily an empirical endeavor, with fossils our only recourse
to reconstructing the history of life:

The morphology, physiology, zoogeography, and be-
havior of living birds tempt us to deduce phylogenetic
relationships, but without paleontological support such
conclusions must remain hypothetical. Only the fossil re-
cord will teach us, eventually, what has in fact happened.
(Brodkorb 1971:20)

As mentioned, the gradualistic philosophy constrains our
approach to phylogenetic methodology, and manifestations of
this philosophy are common in the literature of avian paleon-
tology and phvlogeny. Because fossil data are often considered
superior to neontological data and because it is assumed that
fossils are the best evidence for discerning the geometry of
phylogeny, paleontologists frequently do not attempt to ana-
lyze the relationships of fossil taxa within some prior phylo-
genetic hypothesis of Recent taxa. The time dimension itself
is emphasized, and fossil taxa, simply because of their age,
serve as the basis for speculations about morphological trans-

1 Department of Anatomy, University of Illinois at the Medical Cen-
ter, P.O. Box 6998, Chicago, Illinois 60680.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 330:9-16.

formation sequences or changes in geographical distribution;
such analyses are seldom, if ever, carried out within the con-
text of testing alternative phylogenetic hypotheses. Finally, the
gradualistic philosophy emphasizes a search for ancestors.
Unfortunately, it is doubtful whether a single case within the
literature of avian paleontology approaches the problem of the
identification of ancestral taxa within a testable framework.
On the contrary, ancestors are specified either because they
occur earlier in time and seem to possess some primitive fea-
tures or because they seem to be morphologically intermediate
between two or more Recent taxa. In this paper I shall outline
some theoretical aspects of phylogenetic analysis that are cur-
rently being discussed in the systematic literature, discuss their
implications for paleontological analysis within ornithology,
and apply them to a critique of some current paleontological
practices in order to suggest that theorectical ideas can have
a significant impact on real-world data analysis. One of my
conclusions is that, if practicing paleontologists paid more at-
tention to theory, their methodology would be improved sub-
stantially.

HYPOTHESES ABOUT PHYLOGENETIC
PATTERN

The Nature and Expectation of Pattern

Species taxa can be hypothesized to be discrete evolutionary
units in space and time if it is assumed that the period of
differentiation is itself short relative to the period of species
existence. The evidence for this seems relatively strong (El-
dredge and Gould 1972; Gould and Eldredge 1977; Stanley
1978). Methodologically, the majority of fossil vertebrate

10

Cracraft: Phylogenetic Theory

Species 1 Species 2 Species 3

a' a' ab'

Figure 1. The expectation that evolutionary novelties (derived char-
acters) exhibit a nested pattern can be assumed from evolutionary
theory. Species 1 and 2 are hierarchically nested within a larger group
(species 1 + species 2 + species 3) on the basis of sharing a derived
character, a'. No other aspect of morphological comparison that is
used to form nested sets of taxa seems consistent with evolutionary
theory.

species taxa, and certainly nearly all those of birds, can be
viewed as discrete. If so, then the adoption of discrete species
in phylogenetic analysis, particularly within paleornithology,
would seem to be logically and empirically well-founded.

A proper analysis of phylogenetic pattern is a prerequisite
for all subsequent discussion about the nature of evolutionary
trees. By phylogenetic pattern, I mean the nested pattern of
evolutionary novelties (derived characters) exhibited by the
taxa in question. Indeed, the existence of such a nested pattern
might be taken as a fundamental deduction of the theory of
evolution (Fig. 1).

The pattern of nested evolutionary novelties for any group
of taxa must be inferred as it is not subject to direct empirical
investigation. Hypotheses about this pattern are termed clado-
grams; this usage of cladogram need not refer specifically to a
statement about evolutionary history, as will be discussed be-
low.

Constructing Cladistic Hypotheses

The methods used to construct cladograms have been dis-
cussed in considerable detail by various workers (Hennig 1966;
Schaeffer et al. 1972; Cracraft 1972, 1974a; Wiley 1975; El-
dredge 1979; Eldredge and Tattersall 1975; Eldredge and Cra-
craft 1980; Bonde 1977; Gaffney 1979), so only the salient
features will be mentioned here.

Clearly, the central methodological problem of cladistic
analysis is the identification of evolutionary novelties. Some
paleornithologists have questioned our ability to recognize de-
rived conditions by comparative analysis:

I doubt that a methodology exists for actually deter-
mining primitive-derived sequences in more than a hand-
ful of cases in the entire class Aves. In comparisons across
broad groups of birds it may be impossible to determine
unequivocally which character states are primitive and
which are derived. . . . (Feduccia 1976:598)

But such an extreme position is clearly unjustified, for many
of the defining characters of countless avian taxa, at all taxo-
nomic levels, are almost certainly derived, despite the fact that

previous workers have not presented extensive corroborative
evidence. Within a cladistic view of phytogeny reconstruction,
the issue is not whether we can “establish unequivocally” the
polarity (i.e., whether primitive or derived) of observed simi-
larities (Feduccia 1976:598; 1977:20), for clearly scientific anal-
ysis cannot establish such issues with certainty. This is not to
deny that evidence for character polarity may be difficult, or
perhaps impossible, to gather in individual cases. Neverthe-
less, the methods of cladistic analysis seek to establish hy-
potheses about character polarity and then use these hypoth-
eses to evaluate alternative phylogenetic hypotheses; these
latter hypotheses, in turn, tell us something about our esti-
mations of character phvlogenv.

Perhaps the most critical cognitive issue in the theory of
phylogeny reconstruction is the realization that monophvletic
groups can be defined only by shared derived characters (svn-
apomorphies). As was illustrated earlier (Fig. 1), this conclu-
sion is a simple expectation of evolutionary theory. If so, then
difficulties in determining polarity would seem to be beside the
point, for no other type of similarity can define monophvletic
taxa and be, at the same time, theoretically compatible with
what we know of the evolutionary process. In cases of diffi-
culty, therefore, it would appear we simply have to work
harder.

Three types of data traditionally have been recommended
as being useful in determining polarity: ontogenetic, paleon-
tological, and the comparative distribution of homologous
characters. Because this discussion is primarily concerned with
the analysis of fossil material, ontogenetic data will not be
considered further (see Nelson 1978).

Because of the importance often attached to paleontology as
the final arbiter of phylogenetic questions (e.g., the quote of
Brodkorb above), data from fossils traditionally have been
considered important in postulating polarity sequences. Those
characters occurring earlier in the stratigraphic record are
thought to be primitive relative to those occurring later. That
primitive characters must occur earlier in time cannot be de-
nied. The relevant question is whether the observed distribu-
tion of characters in the fossil record accurately parallels char-
acter phylogeny. The answer is that we cannot have a priori
knowledge about the degree to which this parallel exists. Rel-
atively greater confidence in the parallel traditionally has ex-
isted when the fossil record is dense or when the alternative
characters are distributed stratigraphicallv in widely separated
time intervals. But most vertebrate fossil records, and certain-
ly that of birds, do not fall into the categories set by these
extremes. It has been repeatedly stated in the literature that
there is no theoretical reason why we should not expect derived
characters to occur sometimes in earlier strata than primitive
conditions: indeed, if species extinctions or survival, proba-
bility of fossilization, and probability of recovery by paleon-
tologists are all statistically independent of whether the species
possessed a primitive or derived condition for a given feature,
then this expectation must be admitted. Thus, there is no rea-
son why the reverse order of discovery cannot be of fairly
common occurrence.

One problem with paleontological inquiry in this regard is
that the fossil record is too often assumed to give us an em-
pirical picture of history. The pattern of the fossil record — the
distribution of taxa and characters in space and time — must
be evaluated critically. Paleontological data can be used to

Cracraft: Phylogenetic Theory

11

hypothesize polarity sequences, but then these hypotheses need
to be evaluated by comparative “out-group” procedures. What
must be avoided is axiomatic acceptance of fossil data as a
true picture of character phylogeny.

Without question, comparative analysis of taxa offers the
best source of data for inferring character phylogeny. As men-
tioned earlier, the justification for comparative analysis follows
from the expectation that evolutionary novelties (derived char-
acters) are nested. Thus, for a given character transformation,
a condition postulated to be primitive within a group, say AB,
may characterize (define) a taxonomic group (ABC) broader
than, and including, group AB, which itself must be defined
by a derived condition of a second character. What this means
is that all postulated homologies are derived (svnapomorphous)
at one level and primitive (symplesiomorphous) at all lower
levels. Out-group comparison has been discussed extensively
in the literature (see references cited above). Only one com-
ment is necessary here: out-group comparison does not neces-
sitate definitive knowledge or acceptance of a higher-level phy-
logeny, because such phylogenetic hypotheses are themselves
open to critical testing (Wiley 1975; Gaffney 1979).

Once primitive-derived sequences are postulated, one or
more phylogenetic hypotheses are usually suggested. It is rare
for a single phylogenetic hypothesis to be compatible with all
the polarity sequences. The problem then becomes one of eval-
uating alternative cladistic hypotheses.

Evaluation of Cladistic Hypotheses

If a postulated synapomorphv is consistent or congruent
with a proposed cladistic hypothesis, say A + B, then that
synapomorphv conflicts or is incongruent with alternative hy-
potheses such as A + C, B + C, A + D, B + D. . . . The
goal of cladistic analysis is to find that hypothesis with the
fewest conflicts, or expressed in more affirmative terms, to find
that hypothesis which best accounts for the pattern of nested
synapomorphv. It is necessary to minimize conflicts in svn-
apomorphy because for each conflict an explanation must be
found, and there seem to be only two: (1) the similarity is
homologous, but not a synapomorph; therefore, it is a shared
primitive similarity (a symplesiomorph), in which case it is not
relevant in evaluating the alternative cladistic hypotheses at
this hierarchical level, or (2) the similarity is not homologous
in the first place and must be explained as a convergence. To
invoke convergence as an explanation of a conflict in a given
cladistic hypothesis is ad hoc for that hypothesis because we
must therefore accept that the taxa sharing the similarity are
not monophvletic, i.e. , we must assume some other cladistic
hypothesis to be true. Thus, the choice of the cladistic hy-
pothesis that minimizes conflicts in postulated convergences is
simply a method of minimizing ad hoc assumptions.

This discussion emphasizes the reciprocal nature of testing
cladistic hypotheses and evaluating hypotheses of character
phylogeny. Although character phytogenies are postulated on
the basis of comparative data, their ultimate evaluation rests
on the extent to which they are nested by a cladistic hypothesis.
If, within a specific cladistic hypothesis, an observed character
does not define a set of taxa at some hierarchical level, then
that character cannot be interpreted as derived. On the other
hand, within the framework of an alternative hypothesis, that
character may be interpreted as derived.

The major problems within ornithological systematics and
paleontology with regard to phylogenetic reasoning are (1) hy-
potheses of relationships are seldom precisely stated, and the
taxa being analyzed are not always strictly monophyletic, or
assumed to be (see Example 1 below), and (2) these hypotheses
are frequently not evaluated by derived characters (see Ex-
amples 1 and 2 below). More often than not, decisions about
relationships are based on overall resemblance, or the char-
acters used to unite groups are primitive, in which case the
argument for relationships is severely weakened. Finally, it is
often not appreciated that the phylogenetic position of a fossil
taxon is impossible to assess without some understanding of
the relationships of the Recent taxa. This, it can be suggested,
is one of the primary reasons we have had difficulty in eval-
uating the relationships of many fossil taxa. And this is also
reflected in the attitude of considering fossils as inherently
primitive or ancestral in morphology (if Recent taxa are mo-
saics of primitive and derived characters, why not also fossil
taxa?). That a knowledge of relationships of Recent taxa is of
critical importance is elementary: a fossil is first identified as
a bird, then perhaps as a nonpasserine, then as a piciform,
then as an “advanced” piciform, and finally as a picid. The
extent to which we do not understand the relationships of
Recent taxa increases the difficulty of testing alternative hy-
potheses of relationships involving fossil taxa (see Examples
1, 3, and 4 below).

The Nature and Importance of
Monophyletic Groups

The delineation of strictly monophyletic groups (Hennig
1966) represents a central goal of systematics in reconstructing
the history of life. The importance of monophyletic groups
cannot be overestimated because they alone have reality in
that such groups are part of the “genealogical nexus” (M. Ghi-
selin’s term). In recent years it has become fashionable in some
circles to speak of “minimal” monophylv or of paraphvletic
groups, but such groups are classificatory constructs (artifacts
of the mind if you will), have no basis in genealogy, and for
this reason are to be avoided.

Strictly monophyletic groups are of special concern to pa-
leontologists beyond their contribution to an obvious under-
standing of cladistic interrelationships. Their recognition is
central to the question of constructing and evaluating hypoth-
eses of ancestry and descent, a subject high in the mind of
most paleontologists.

HYPOTHESES ABOUT EVOLUTIONARY
TREES

What are Evolutionary Trees?

I will begin by distinguishing the concept of evolutionary
trees from that of cladograms. As was noted above, clado-
grams are hypotheses about the pattern of nested synapomor-
phy. Cladograms need not necessarily be interpreted as a direct
expression of phylogenetic history, although certainly most
systematists have a predilection to treat them as such. How-
ever, cladograms can be viewed strictly in terms of the analysis
of pattern, and it therefore becomes necessary to examine the

12

Cracraft: Phylogenetic Theory

Figure 2. Schematic diagram showing the conceptual difference be-
tween a cladogram and an evolutionary tree. In Part a is shown a
cladogram of three taxa depicting the nested patterns of synapomorphv
(dark rectangles). In the cladogram only synapomorphic pattern is
implied. In Parts b through g are shown six evolutionary trees, each
one of which reflects the pattern of the cladogram. In the tree hy-
potheses decisions are made whether to postulate speciation events
(symbolized by branch points) or directly ancestral taxa (elimination
of branch points).

possible evolutionary implications of that pattern. These im-
plications are expressed in terms of evolutionary trees.

Any single cladogram can have a variable number of evo-
lutionary interpretations (Fig. 2). The basic question is to de-
cide whether branch points are to be recognized and a specia-
tion event thus hypothesized, or whether branch points are to
be eliminated and an ancestral species specified. Thus, the
cladogram of Figure 2a has six possible evolutionary interpre-
tations (Fig. 2 b— g). In Figure 2b both branch points are re-
tained and interpreted in terms of speciation events, whereas
in Figures 2c-g one or both branch points are eliminated and
direct ancestry and descent is specified. Note that all six evo-
lutionary trees are fully consistent with the synapomorphv pat-
tern of the cladogram: in all cases species A and B possess
derived characters not shared with species C.

Constructing and Testing
Evolutionary Trees

In attempting to construct evolutionary trees it is essential
that a corroborated species-level cladistic hypothesis is first
proposed. In constructing evolutionary trees all ancestral and
descendant taxa must be of species rank; terminal taxa may
be of any rank. This follows from elementary evolutionary
theory: species, not supraspecific taxa, are considered the
evolving units of the evolutionary process. The “evolution” of
supraspecific taxa is merely the statistical summation of the
evolutionary histories of the included species. Thus, genera,
families, and so on cannot be designated as ancestral to any
other taxon. It has been customary within paleontology to
identify supraspecific ancestral taxa, but such a practice almost
certainly means (1) the cladistic relationships of the included

species are not properly understood, (2) the hypothesis of re-
lationships is imprecisely stated, and/or (3) the supraspecific
ancestral taxon is not strictly monophyletic. All of these are to
be avoided if the goal is to reconstruct evolutionary history.

Given a cladistic hypothesis (cladogram) for a group of taxa,
what might be some of the considerations in evaluating the
possible evolutionary trees (see Engelmann and Wiley 1977,
and Platnick 1977 for extended discussions)? Ancestors usually
have been recognized on two criteria: primitive or intermediate
morphology and/or earlier stratigraphic occurrence. Such fac-
tors might serve as a basis for postulating ancestor-descendant
relationships such as are expressed in Figures 2c-g. How are
such hypotheses to be tested?

Consider, for example, the simple hypothesis for Figure 2f
in which species B is postulated to be the ancestor of species
A. The hypothesis implies that species B is primitive in all
features relative to the condition in A. If species B possessed
a unique derived character (termed an autapomorphy), we
must postulate two evolutionary events to account for its dis-
tribution: the evolution of the derived feature in the lineage
leading to B and its subsequent loss leading to A. This hy-
pothesis is less parsimonious than one postulating that the
autapomorphy evolved after a speciation event producing both
A and B (Fig. 2b). On this basis, then, the presence of autapo-
morphies can be used to reject an ancestor-descendant hy-
pothesis in favor of a hypothesis involving a speciation event.

If we cannot find any autapomorphies in B, does this mean
the hypothesis of Figure 2f is to be accepted? Not necessarily,
because whereas the hypothesis shown in Figure 2f would ap-
pear to be acceptable, so would the hypothesis in Figure 2b.
In fact, it does not appear possible to accept an ancestor-de-
scendant hypothesis without at the same time accepting the
speciation hypothesis. Indeed, seemingly the only way to reject
the latter is to reject the cladogram on which it is based (Plat-
nick 1977). What this means, therefore, is that there are no
theoretical grounds for preferring only an ancestor-descendant
hypothesis.

Stratigraphic data do not help our evaluation of evolution-
ary trees as much as it might first seem. Although paleontol-
ogists often rely heavily on stratigraphic data to specify an-
cestral taxa, clearly the problem of ancestry and descent is first
and foremost a morphological problem. If, as in Figure 2 f, we
assume taxon B to be primitive morphologically and to occur
earlier in the fossil record, then the hypothesis would appear
to be highly acceptable. But, how do we reject the hypothesis
shown in Figure 2b? Indeed, it would seem we cannot (see
Example 5 below). Furthermore, if B occurred later in time
than A, could we therefore reject the hypothesis shown in
Figure 2f? Yes, but only if (1) we had certain knowledge of
the stratigraphic ranges of A and B, and (2) we had certain
knowledge that A and B both occurred only within the same
stratigraphic sequence and were not geographically distributed
elsewhere. But all this seems highly conjectural and would
almost certainly call for ad hoc assumptions.

Perhaps an important consideration of this discussion should
be that paleornithologv does not deal with a dense fossil record
extending over large periods of time. The theoretical and prac-
tical questions posed by such a situation simply do not exist
(even in paleomammalogy such occurrences are very rare). If
so, then concern with identifying ancestors is perhaps a moot
point.

Cracraft: Phylogenetic Theory

13

THE FUTURE OF AVIAN PALEONTOLOGY

The future contributions of avian paleontology in decipher-
ing the evolutionary history of birds seem inescapably linked
to progress in avian systematics in general. Until we have
highly corroborated phylogenetic hypotheses of Recent taxa,
our attempts to understand the phylogenetic significances of
fossil taxa will be only partially successful (see Examples 1 and
4 below). This is a minority viewpoint within vertebrate pa-
leontology in general, and paleornithology in particular. Tra-
ditional opinion holds that only the discovery of more fossil
material will ultimately reveal the course of avian phvlogeny
(Brodkorb 1971:20). I consider this concept to be mistaken for
the theoretical reasons presented above.

The purpose of this paper, up to this point, has been to
stress the importance of systematic theory in the methodology
of avian paleontology. In the final section these theoretical
ideas will be given expression in specific examples in order to
demonstrate that traditional paleontological analysis has some-
times led to questionable conclusions. The purpose of this sec-
tion is not to refute the specific conclusions of the examples,
but to point out that different theoretical approaches call for
alternative hypotheses that generally have not been consid-
ered. Thus, the examples were chosen not for their taxonomic
interest but solely to illustrate the theoretical points raised in
this paper.

SOME EXAMPLES OF
PALEORNITHOLOGICAL METHODOLOGY

EXAMPLE 1. The phylogenetic analysis of fossil taxa: the
relationships of Alexornis (Brodkorb 1976).

Brodkorb (1976) recently described a new species, Alexornis
antecedens, from the Upper Cretaceous of Baja California.
Based on a comparison of six elements, he concludes (1976:70)
that:

The resemblances of Alexornis are closest to certain
members of the Piciformes and Coraciiformes. Within
those two orders the piciform family Bucconidae and the
coraciiform family Momotidae have the most similarity
to the fossil. The fossil shares certain characters with both
Bucconidae and Momotidae, some with Bucconidae
alone, and some with Momotidae alone; but more of its
characters are unique [italics added].

After a tabulation of similarities among Alexornis, Momot-
idae, and the Bucconidae, the hypotheses that Alexornis is
related to the Coraciiformes, on the one hand, or to the Pici-
formes, on the other, are rejected. On the basis of a “mixture”
of “coraciiform” and “piciform” similarities, Brodkorb con-
cludes (1976:73): “Both morphology and the temporal sequence
thus suggest Alexornis as the presumptive ancestor of the or-
ders Coraciiformes and Piciformes.”

There are two separate questions that need to be discussed
when analyzing the phylogenetic position of a fossil taxon such
as Alexornis, neither of which were considered in this study.
First, what are the precise cladistic relationships of the fossil
taxon? Second, once the cladistic relationships have been de-
termined, what can we say about the various hypotheses re-
garding ancestry and descent that might be formulated?

Brodkorb’s analysis of phylogenetic relationships was based

entirely upon an assessment of overall similarity, and no at-
tempt was made to distinguish between primitive and derived
similarities. A second major problem is the attempt to postu-
late relationships of a fossil taxon in the absence of a corrob-
orated cladistic hypothesis for the Recent taxa. Whereas the
Bucconidae are more or less primitive morphologically within
the Piciformes (S. Simpson and J. Cracraft, in prep.), the
Momotidae are relatively advanced within the Coraciiformes
(Cracraft, in prep.; P.J.K. Burton, in prep.; David Maurer,
in prep.); thus the use of these two families to characterize the
two orders is questionable. Furthermore, the precise interre-
lationships of coraciiform groups, the piciforms, and the pas-
seriforms are as yet unsettled. Thus, any fossil such as Alex-
ornis must be evaluated in light of these observations. It would
seem that the phylogenetic position of this fossil is still an open
question.

Finally, what about the hypothesis that Alexornis is an-
cestral to both Coraciiformes and Piciformes? Brodkorb him-
self presents sufficient evidence to reject this hypothesis; his
support for it, on the other hand, is derived from two tradi-
tional paleontological arguments: apparent “intermediate”
morphology and earlier stratigraphic occurrence. But Brod-
korb notes that there are a minimum of 20 features “unique”
to Alexornis. Assuming that these are autapomorphies of Al-
exornis, then if A. antecedens is an ancestor, we must postulate
at least 20 character reversals. This hypothesis is clearly less
parsimonious than assuming Alexornis is the sister-taxon of
some group and that these features evolved only in the Alex-
ornis lineage.

EXAMPLE 2. Character-analysis and the determination of
relationships: the case of Protornis (Olson 1976).

Olson (1976) recently restudied the lower Oligocene fossil
bird, Protornis glariensis, a species known from a slab con-
taining limbbones and various other elements found in Swit-
zerland about 140 years ago. He makes a strong case that
Protornis is related to the Todidae and Momotidae within the
Coraciiformes, and then states (p. 115):

The proportions of the bill and of the hindlimb and toes
preclude its assignment to the Todidae. In all of its im-
portant features it agrees with the Momotidae. It differs
from the modern forms of the family mainly in the shorter
mandibular symphysis and the higher, more expanded
sternocoracoidal process of the coracoid. Protornis gla-
riensis should, therefore, be assigned to the family Mo-
motidae.

Olson concludes (p. 188) from this that “. . . the existence
of Protornis in the lower Oligocene of Switzerland now pro-
vides evidence that the family Momotidae, presently confined
to the New World, actually had its origins in the Old World.”

Although he may be entirely correct in his phylogenetic as-
sessment and zoogeographic conclusions, Olson’s own data
and analysis permit an alternative hypothesis. As mentioned,
Olson presents evidence that Protornis, the Todidae, and the
Momotidae shared a common ancestor, and the latter two fam-
ilies have been considered sister-groups within the Coraci-
iformes by previous workers. An alternative hypothesis to be
considered is a sister-group relationship between Protornis on
the one hand and Todidae + Momotidae on the other. This
hypothesis certainly would make more sense zoogeographicallv

14

Cracraft: Phylogenetic Theory

by restricting the todid-momotid lineage to the New World.
On the basis of the present evidence this hypothesis cannot be
rejected for it has not been shown that Protornis actually
shares one or more derived characters with the Momotidae.
It would seem that the absence of a primitive-derived char-
acter analysis prevents a more specific statement about the
relationships of Protornis.

EXAMPLE 3. The logic of evaluating phylogenetic hypothe-
ses: Gingerich (1976) on palaeognath phvlogenv.

A number of workers have argued for the monophvly of the
ratite birds and tinamous (palaeognaths) by suggesting that
some shared characters, including the palaeognathous palate,
rhamphothecal structure, and enlarged ilioischiatic fenestra,
are derived or unique to these birds (Bock 1963; Parkes and
Clark 1966; Cracraft 1974b). Gingerich (1976:31-32) presents
two opposing arguments: (1) the three similarities are in fact
not derived to the ratites but are primitive, and that therefore
(2) “it is possible, even probable, that the groups of living
ratites and the tinamous are paraphyletic” (1976:32).

The discussion here is not concerned with the evidence Gin-
gerich raises against the hypotheses about character polarity
proposed by previous workers. After all, it is important to
examine and criticize such hypotheses. But it is one thing to
argue that the hypothesized derived characters of a group are
primitive and quite another to conclude that the group is not
therefore monophyletic. The second argument does not follow
necessarily from the first and to link them confuses two sep-
arate aspects of phylogenetic analysis: on the one hand, the
acceptance of a hypothesis of svnapomorphy and its use in
defining monophyletic groups, and on the other, the preference
of one phylogenetic hypothesis over another.

If the three shared similarities discussed by Gingerich are
primitive, then he is correct in stating that evidence for rat-
ite-tinamou monophvly remains to be discovered. Neverthe-
less, this does not mean we must reject that hypothesis of
monophylv, because preference for, or rejection of, any par-
ticular phylogenetic hypothesis is dependent upon its status
relative to alternative hypotheses. Thus, without undermining
the importance of an evaluation of character polarity, the ul-
timate criticism of a phylogenetic hypothesis is the presentation
of evidence that one or more of the taxa are more closely
related to other taxa of birds: in effect to argue preference for
an alternative hypothesis based on shared derived characters.
As far as the tinamous or ratite taxa are concerned, no alter-
native hypotheses were proposed by Gingerich nor are any
given strong support in the literature.

This bring us back to a consideration of the three observed
similarities said by Gingerich to be primitive. If they are prim-
itive within tinamous and ratites, this implies they are svn-
apomorphous (derived) at some higher taxonomic level, per-
haps to birds as a whole or to birds + some reptilian taxon.
Indeed, Gingerich attempts to suggest this, but an identifica-
tion of the taxonomic level is not made explicit (in fact, no
argument is presented against the peculiar rhamphotheca
being derived). Nor was it suggested that the large number of
postulated derived characters interrelating the ratites them-
selves (Cracraft 1974b) are also primitive or convergent, which
logically would be the case if the ratites are not related to one
another.

Thus, Gingerich’s criticism of palaeognath monophylv,
while well-intentioned with respect to examining the validity

of some of the postulated polarity sequences, is theoretically
in error with regard to using those sequences to evaluate al-
ternative phylogenetic hypotheses.

EXAMPLE 4. The analysis of “intermediate” fossils: the case
of Presbyornis (Feduccia 1976, 1977, 1978).

The search for “missing links” has been a relentless preoc-
cupation of paleontology for over a century. Darwin (1859:280)
set the tone for paleontological methodology: “I have found it
difficult, when looking at any two species, to avoid picturing
to myself, forms directly intermediate between them” (italics
in original]. So it has been that paleontologists have sought to
find fossil intermediates between Recent taxa, for it is com-
monly thought that only by such discoveries can the distant
connections of phylogenetic history be discerned. However,
Darwin saw a problem to the search for intermediates between
Recent taxa, and he followed the above statement with the
observation: “But this is a wholly false view; we should always
look for forms intermediate between each species and a com-
mon but unknown progenitor. ...”

Despite Darwin’s exhortation, the temptation to seek inter-
mediate fossils has remained high, and has found some expres-
sion in several recent studies of fossil birds. Perhaps the most
publicized such fossil is Presbyornis, a genus containing sev-
eral species known from the Eocene of western North America
and possibly the Eocene of South America. The importance
of this fossil is stated explicitly by Feduccia (1978:300):

Although almost all the vertebrate groups are replete
with so-called ‘missing links’ that have added greatly in
elucidating their phytogenies, Presbyornis represents the
first known avian fossil to form a link between a number
of major living orders of birds. Presbyornis is an evolu-
tionary mosaic, combining a strange montage of morpho-
logical characteristics of shorebirds, modern ducks and
allies, and modern flamingos.

The arguments and empirical support for regarding Pres-
byornis to be intermediate between shorebirds, ducks, and
flamingos are extremely complex. In essence, Presbyornis is
said to have a cranium very similar to that of a duck and the
postcranial skeleton of a recurvirostrid shorebird and flamin-
go. Postcranial material of Presbyornis is abundant and as-
sociated in the same matrix, but little, if any, of it is directly
articulated. Consequently, an incontrovertible argument for
the conspecificity of the cranial and postcranial material has
not been presented at this time.

It is not the purpose of this example to discuss the morpho-
logical evidence relating to Feduccia’s claim of intermediacy.
Rather, I wish to cite some difficulties with his argumentation
that illustrate potential problems in the phylogenetic analysis
of fossils, in particular the interpretation that certain fossils
may be intermediate in character.

The major problem presented by the example of Presbyornis
is our lack of understanding of the phylogenetic relationships
among Recent taxa. Nevertheless, our present conceptions
about these interrelationships define an array of hypotheses
not considered by Feduccia. Three major criticisms can be
made.

First, Feduccia implies (1976:600) that recurvirostrids (avo-
cets, stilts) are the sister-group of other “shorebirds,” but this
is almost certainly not true; moreover, non-shorebird charad-
riiforms are excluded from his argument. Recurvirostrids have

Cracraft: Phylogenetic Theory

15

often been thought to be among the more advanced charad-
riiforms, and certainly their skeletal anatomy has not been
considered primitive within the order. If so, this creates a
devastating difficulty for Feduccia’s hypothesis, because the
argument for intermediacy depends upon showing that the
similarities of Presbyornis and charadriiforms are primitive,
not advanced. Thus, the acceptance of at least the outlines of
a phylogeny of the Charadriiformes becomes essential, and
what we think is understood about that is not favorable to
Feduccia’s hypothesis. Virtually all previous workers, for ex-
ample, Lowe (1923) and Stresemann (1927-34), have believed
the shorebirds (i. e. , recurvirostrids, scolopacids, charadriids)
to be among the more advanced charadriiforms.

Second, Feduccia’s argument in support of a link between
Presbyornis and flamingos borders on the circular in that some
of his conclusions seem imbedded or implied in his premises.
The conclusion of interest here is that flamingos are not at all
closely related to ciconiiform birds, as has been thought by
many previous workers. He argues that Presbyornis is similar
to flamingos and recurvirostrids but not to ciconiiforms. His
unstated premise is that because there is no perceived similar-
ity between Presbyornis and ciconiiforms, the latter cannot
have a relationship to flamingos. Presbyornis essentially be-
comes the arbiter of relationships among these Recent taxa
before the relationships of the fossil are fully assessed. Any
argument for intermediacy must treat the problem of a ciconi-
iform-flamingo relationship more rigorously, principally by
showing that ciconiiforms are more closely related to some
other taxon.

Third, the hypothesis of a link between Presbyornis and
anatids is weakened, it would appear, by Feduccia’s exclusion
from his argument of the anseriform family Anhimidae and of
the order Galliformes. Few, if any, modern workers have se-
riously doubted a sister-group relationship between anatids
and anhimids. The latter family is generally considered to be
primitive in most of its features relative to the anatids, partic-
ularly in cranial characteristics. The purported skull of Pres-
byornis is compared to an advanced condition within the an-
seriforms and not one that is primitive, a line of argumentation
counter to the concept of intermediacy. Furthermore, many
previous systematists such as Beddard (1898), Simonetta
(1963), and Prager and Wilson (1976) have called attention to
a possible close relationship of the anseriforms and galliforms.
Thus, any argument for a link between Presbyornis and an-
atids must also be presented in the context of an analysis of
these previous hypotheses.

The point of this example is to indicate the complexity of
arguments that are involved with any hypothesis linking Re-
cent taxa with fossils. Most of this complexity, if not all of it,
relates to evaluating the interrelationships of Recent taxa be-
fore assessing the phylogenetic affinities of fossils. If fossil taxa
are to have significance as intermediate links, then they must
share some of the primitive features of the taxa being linked.
This does not seem to be the case with Presbyornis, and given
present evidence it is difficult to assign the material of Pres-
byornis to the charadriiforms, Anatidae, or the Phoenicopter-
idae. Certainly, the assertion that Presbyornis is an important
link between modern groups of birds is in need of reevaluation.

EXAMPLE 5. The identification of ancestors: the case of Lim-
nofregata azygosternon (Olson 1977).

In an excellent descriptive systematic paper Olson (1977)

described a nearly complete skeleton of a frigatebird-like
species from the early Eocene of Wyoming. The species is
similar to frigatebirds in many respects but also shares some
features with Phaethon and Sula. Following a detailed com-
parative analysis Olson concluded (1977:31):

There is nothing that I can detect in the skeleton of
Limnofregata that precludes its being directly ancestral to
Fregata. The fact that by the early Eocene it was already
markedly specialized along much the same lines as the
modern genus renders this possibility plausible.

Olson’s argument for a direct ancestry by L. azygosternon
is one of the better examples of such a claim to be found in
the paleornithological literature. He specifically points out that
most of the features shared between L. azygosternon and other
pelecaniforms are probably primitive, whereas those shared
with Fregata seem derived solely to that lineage. Whether L
azygosternon has any unique features all its own is uncertain,
but Olson seems not to have found any for none are specifically
mentioned.

The hypothesis that L. azygosternon is the direct ancestor
of the modern genus Fregata would not seem to be capable of
rejection based on any evidence presented by Olson. On the
other hand, for the theoretical reasons presented above, nei-
ther can the hypothesis thatL. azygosternon and Fregata share
a sister-group relationship be rejected. It is just as “plausible”
as the hypothesis of direct ancestry, and reasons are not pre-
sented by Olson for preferring the latter. In this case Olson’s
decision to invoke direct ancestry is not significantly mislead-
ing because the relationship between the fossil and Recent taxa
seems to be properly analyzed.

CONCLUSIONS

Very little attention has been paid within avian paleontology
to problems of systematic theory and methodology. The prac-
tices of avian paleontologists are frequently empirical in ap-
proach and are governed by assumptions that fossils are in-
trinsically important in matters of phylogenetic inference,
character phylogeny, or the analysis of intermediate taxa.

However, an alternative perspective is possible, one that
does not eliminate the importance of fossil taxa but integrates
them into a testable method of phylogenetic inference focused
primarily on Recent taxa. This alternative is cladistic analysis
(phylogenetic systematics). Monophyletic groups, whether fos-
sil, Recent, or a mixture of both, are defined in terms of shared
derived characters (synapomorphv). From an analysis of nest-
ed synapomorphv patterns, cladistic hypotheses (cladograms)
are formulated. These hypotheses in turn can be used to eval-
uate hypotheses about evolutionary history (trees).

It can be shown that the application of cladistic theory to
paleontological practice increases the precision of phylogenetic
research. This precision is manifested particularly in discus-
sions about the phylogenetic relationships of fossil and Recent
taxa, the analysis of morphologically intermediate fossil taxa,
and the postulation of ancestral species level taxa.

ACKNOWLEDGMENTS

I thank Bruce Manion, Robert Raikow, and Sharon Simp-
son for their comments on the manuscript. I am also grateful
to the National Science Foundation, through grant DEB 76-

16

Cracraft: Phylogenetic Theory

09661, for their support of the research leading to this paper.

I thank E. Penny Pounder for executing the figures.

LITERATURE CITED

Beddard, F.E. 1898. The structure and classification of
birds. Longmans, Green, London. 548 pp.

Bock, W.J. 1963. The cranial evidence for ratite affinities.
Proc. 13th Intern. Ornitholog. Congress 1:39-54.

Bonde, N. 1977. Cladistic classification as applied to ver-
tebrates. Pp. 741-804 in Major patterns in vertebrate
evolution (M. Hecht, P.C. Goody, and B.M. Hecht,
Eds.). Plenum Press, New York.

Brodkorb, P. 1971. Origin and evolution of birds. Pp. 19-
55 in Avian Biology, vol. 1 (D.S. Farner and J.R. King,
Eds.). Academic Press, New York.

. 1976. Discovery of a Cretaceous bird, apparently

ancestral to the orders Coraciiformes and Piciformes
(Aves: Carinatae). Smithsonian Contrib. Paleobiology
No. 27:67-73.

Cracraft, J. 1972. The relationships of the higher taxa of
birds: problems in phylogenetic reasoning. Condor
74:379-392.

. 1974a. Phylogenetic models and classification. Syst.

Zool. 23:71-90.

1974b. Phytogeny and evolution of the ratite birds.

Ibis 116:494-521.

Darwin, C. 1859. On the origin of species. Harvard Univ.
Press (1964), Cambridge.

Eldredge, N. 1979. Cladism as common sense. Pp. 1 65 —
198 in Phylogenetic analysis and paleontology (J. Cracraft
and N. Eldredge, Eds.). Columbia University Press, New
York.

Eldredge, N., and J. Cracraft. 1980. Phylogenetic pat-
terns and the evolutionary process: method and theory in
comparative biology. Columbia University Press, New
York.

Eldredge, N., and S.J. Gould. 1972. Punctuated equilib-
ria: an alternative to phyletic gradualism. Pp. 82-115 in
Models in paleobiology (T.J.M. Schopf, Ed.). Freeman,
Cooper & Co., San Francisco.

Eldredge, N., and I. Tattersall. 1975. Evolutionary
models, phylogenetic reconstruction, and another look at
hominid phylogeny. Contrib. Primatology 5:218-242.

Engelmann, G.F., and E.O. Wiley. 1977. The place of
ancestor-descendant relationships in phylogeny recon-
struction. Syst. Zool. 26:1-11.

Feduccia, A. 1976. Osteological evidence for shorebird af-
finities of the flamingos. Auk 93:587-601.

. 1977. Hypothetical stages in the evolution of modern

ducks and flamingos. J. Theor. Biol. 67:715-721.

. 1978. Presbyornis and the evolution of ducks and

flamingos. Amer. Sci. 66:298-304.

Gaffney, E.S. 1979. An introduction to the logic of phylo-
genetic reconstruction. Pp. 79-111 in Phylogenetic anal-
ysis and paleontonology (J. Cracraft and N. Eldredge,
Eds.). Columbia Univ. Press, New York.

Gingerich, P.D. 1976. Evolutionary significance of the Me-
sozoic toothed birds. Smithsonian Contrib. Paleobiology
No. 27:23-33.

Gould, S.J., and N. Eldredge. 1977. Punctuated equilib-
ria: the tempo and mode of evolution reconsidered. Pa-
leobiology 3:115-151.

Hennig, W. 1966. Phylogenetic systematics. Univ. Illinois
Press, Urbana.

Lowe, P R. 1923. Notes on the systematic position of Or-
tyxelos together with some remarks on the relationships
of the Turnicomorphs and the position of the seed-snipe
(Thinocoridae) and Sandgrouse. Ibis 65:276-299.

Nelson, G.J. 1978. Ontogeny, phylogeny, paleontology, and
the biogenetic law. Syst. Zool. 27:324-345.

Olson, S.L. 1976. Oligocene fossils bearing on the origins of
the Todidae and the Momotidae (Aves: Coraciiformes).
Smithsonian Contrib. Paleobiology No. 27:111-119.

. 1977. A lower Eocene frigatebird from the Green

River Formation of Wyoming (Pelecaniformes: Fregati-
dae). Smithsonian Contrib. Paleobiology No. 35:1-33.

Parkes, K.C., and G.A. Clark, Jr. 1966. An additional
character linking ratites and tinamous, and an interpre-
tation of their monophvly. Condor 68:459-471.

Platnick, N.I 1977. Cladograms, phylogenetic trees, and
hypothesis testing. Syst. Zool. 26:438-442.

Prager, E.M., and A.C. Wilson. 1976. Congruency of
phvlogenies derived from different proteins. A molecular
analysis of the phylogenetic position of cracid birds. J.
Mol. Evol. 9:45-57.

Schaeffer, B., M.K. Hecht, and N. Eldredge. 1972.
Phylogeny and paleontology. Evol. Biol. 6:31-46.

Simonetta, A M. 1963. Cinesi e morfologia del cranio ncgli
uccelli non passeriformi Studio su varie tendenze evolu-
tive. Parte 1. Arch Zool. Ital. 48:53-135.

Stanley, S.M. 1978. Chronospecies’ longevities, the origin
of genera and the punctuational model of evolution. Pa-
leobiology 4(l):26-40.

Stresemann, E. 1927-1934. Aves in Handbuch der Zool-
ogie, Vol. 7, No. 2 (W. Kiikenthal and Th. Krumbach,
Eds.). W. de Gruvter, Berlin.

Wiley, E.O. 1975. Karl R. Popper, systematics, and clas-
sification: a reply to Walter Bock and other evolutionary
taxonomists. Syst. Zool. 24:233-242.

THE ARCHAEOTROGONIDAE OF THE EOCENE AND OLIGOCENE
PHOSPHORITES DU QUERCY (FRANCE)1

By Cecile Mourer-Chauvire2

ABSTRACT: The genus Archaeotrogon Milne-Edwards was described in the last century from the

fossiliferous deposits of “Phosphorites du Quercy.” New excavations carried out at these sites have resulted
in additional avian specimens that we have been able to assign to the three previously described species
of Archaeotrogon. The temporal distribution of these species is discussed, and a new species is described.

The species of Archaeotrogon do not have the heterodactyl structure of the foot characteristic of modern
trogons, although this structure had already been acquired in some contemporaneous forms. It appears
that archaeotrogons constituted a distinct family, the Archaeotrogonidae, that evolved parallel with the
family Trogonidae, or true trogons.

RESUME: Le genre Archaeotrogon Milne-Edwards a ete decrit au siecle dernier dans les gisements

des Phosphorites du Quercy. De nouvelles fouilles effectuees dans ces gisements ont permis de retrouver
les trois especes precedemment signalees et de leur attribuer un certain nombre d’elements du squelette.
Leur position chronologique a pu etre precisee et une nouvelle espece a ete decrite.

Les Archaeotrogon ne presentent pas la structure du pied heterodactvle caracteristique des trogons
actuels bien que cette structure soit deja acquise chez des formes fossiles du meme age. On peut done
penser que les Archaeotrogon constituent une famille differente ayant evolue parallelement a celle des
Trogonidae ou vrais trogons.

The “Phosphorites du Quercy” are deposits that filled sink-
holes in the karst topography of the departments of Tarn-et-
Garonne, Lot, and Aveyron, to the southwest of the central
French massif. These deposits were very actively exploited for
the extraction of calcium phosphate between approximately
1870 and 1880. During the course of mining, many specimens
of fossil vertebrates, as well as molluscs and insects, were
discovered in these localized deposits. The first discoveries of
bird bones were announced by Lydekker (1891), followed by
Milne-Edwards (1892). The birds of the Phosphorites du Quer-
cy were thereafter the subject of an important work by Gail-
lard (1908). But the bones of the early collections did not bear
precise data as to which sinkhole they were collected from,
and the phosphorite deposits at Quercy include faunas that
extend from the Upper Bartonian (Robiac’s mammal zone) all
the way to the Upper Stampian (Boningen’s mammal zone).

New work was undertaken at Quercy by the group RCP
311 (Recherche cooperative sur programme 311) of the CNRS
(Centre national de la Recherche scientifique), composed of
researchers from the universities of Montpellier, Paris VI, and
Lyon I. In the course of this recent work, the beds were ex-
cavated separately and each was well dated by means of its
mammalian fauna (Crochet et al. 1972; de Bonis et al. 1973;

1 Translated by Antonia Tejada-Flores.

2 Centre de Paleontologie stratigraphique et Paleoecologie de
l’Universite Claude Bernard, associe au CNRS (LA 11), 27-43 Bou-
levard du 11 Novembre, 69622 Villeurbanne Cedex, France.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 330:17-31.

Hartenberger 1973; Cavaille et al. 1974; Hartenberger et al.
1974; Sige 1974, 1976; Vianey-Liaud 1976; Sudre 1977; Cro-
chet 1978).

I had undertaken the revision of the avifaunas of Quercy,
and for that reason I was able to collect together the older
documents kept in the collections of the National Museum of
Natural History in Paris, the Natural History Museum and
the Department of Earth Sciences of Lyon, and the University
of Utrecht, as well as the newer documents amassed by the
researchers of the University of Montpellier and the University
of Paris VI.

The living trogons belong to a single family, the Trogonidae,
a group of eight genera. Five of these live in Central America,
South America, and the Antilles; two in tropical Africa, and
one in southeast Asia (Peters 1945); see Figure 1. Trogons
appear to have been a constant element of the paleoavifauna
of Europe, ever since they were first described by Milne-Ed-
wards (1867-1871) from the lower Miocene (Aquitanian) de-
posits of the department of Allier under the name of Trogon
gallicns. (The generic name of this fossil form was later
changed to Paratrogon by Lambrecht (1933).) Milne-Edwards
(1892) subsequently discovered the presence of trogons in the
Phosphorites du Quercy and created for these forms the genus
Archaeotrogon. The new excavations at Quercy have shown
that the trogons are often the most abundant elements in the
avifauna, particularly in those beds that date from the upper
Oligocene, such as Pech Desse, and above all, Pech du
Fraysse. Futhermore, Olson (1976) has shown that one of the

18

Mourer-Chauvire: The Archaeotrogonidae

birds found in the “Glarner Fischschiefer” in Switzerland (see
Fig. 1), and known as Protornis glaronensis von Meyer 1884,
possessed the heterodactyl foot structure characteristic of the
living trogons. Therefore, that specimen should be considered
to be a member of the Trogonidae, even though the holotype
of the species should be placed among the Momotidae; Peyer
(1957) believed this species to be a member of the Alcedinidae.
The age of the “Glarner Fischschiefer” is believed to be lower
Oligocene (Sannoisian) because of the fish fossils found there.
In addition to extinct forms of the European Tertiary, two
living species of trogonids have been found as fossils in Pleis-
tocene deposits: Trogon surrucura Vieillot in Brazil and Tem-
notrogon roseigaster (Vieillot) in the Dominican Republic
(Brodkorb 1971).

SYSTEMATICS

Order Alcediniformes Feduccia 1977
Superfamily Trogonoidea Feduccia 1977

Archaeotrogonidae new family

TYPE GENUS: Archaeotrogon Milne-Edwards 1892

DIAGNOSIS: Trogons, that differ from all species of the
family Trogonidae by lacking the heterodactyl foot character-
istic of that family.

TEMPORAL AND GEOGRAPHIC DISTRIBUTION:
Upper Eocene to lower Oligocene. Phosphorites du Quercy,
France.

REMARKS: The family Archaeotrogonidae contains only
the type genus. Although the archaeotrogons were character-
ized by the primitive structure of their tarsometatarsus, i.e.,
the lack of heterodactylv, in deposits of the same age as those
at Quercy there existed a trogon whose foot already had a
heterodactyl structure. One may therefore consider the forms
of Quercy as a line parallel to that of the true trogons, which
belong to the family Trogonidae. The Trogonidae contains the
fossil bird from the Glarner Fischschiefer (Olson 1976), the
extinct genus Paratrogon, as well as the living genera ( Phar -
omachrus, Euplilotis, Priotelus, Temnotrogon, Trogon, Apa-
loderma, Heterotrogon, and Harpactes).

Genus Archaeotrogon Milne-Edwards 1892

TYPE SPECIES: Archaeotrogon venustus Milne-Edwards
1892

DESCRIPTION: Archaeotrogon has been described pri-
marily on the basis of its humerus, and humeri are very
abundant in the older collections. In addition to the humeri,
Lydekker (1891) referred some coracoids (not figured) and
Milne-Edwards (1892) some carpometacarpi (not figured) to
the genus. Gaillard (1908) described and figured a tarsometa-
tarsus that he attributed to the species A. cayluxensis. In cer-
tain sites at Quercy, specimens of Archaeotrogon are very nu-
merous and represent more than half of the bird bones found
in these beds. It is logical to assume that, if the most common
humerus belongs to the genus Archaeotrogon, then the most
common ulna, the most common carpometacarpus, the most
common coracoid, etc., should likewise belong to that genus.

I have therefore attributed to that genus a certain number of
skeletal elements collected from all the sites where the humerus
of Archaeotrogon was found, but it is also true that these
elements show analogies with the corresponding bones of liv-

ing trogons. I must point out that I have never found articu-
lated bones in the Phosphorites du Quercy. This is probably
due in part to the way the fossiliferous cavities were filled, and
is partly a result of the methods of excavation, which included
washing and screening techniques. There cannot, therefore,
be an absolute certainty that the bones attributed to Archaeo-
trogon truly belong to that genus, but there is a strong prob-
ability that they do.

Comparison with Living Trogonidae. At the Natural
History Museum of Leiden I was able to study skeletons be-
longing to the genera Harpactes, Trogon, Pharomachrus, and
Prioteles. The genus Archaeotrogon, when compared with the
Trogonidae, shows the following similarities and differences
(characters of the Trogonidae in parentheses).

Humerus. Similarities: (1) same general form; (2) head en-
larged and flattened; (3) internal trochanter very prominent;
(4) tricipital fossa large; (5) pectoral crest lengthened; (6) distal
extremity transversely widened; (7) tricipital grooves forming
a large depression.

Differences: (1) proximal end very wide transversely (prox-
imal end less wide transversely); (2) head rather flattened (head
more swollen); (3) internal trochanter more strongly bent back-
wards and downwards; (4) no pneumatic orifice in the sub-
trochanteric fossa (pneumatic orifices in the sub-trochanteric
fossa); (5) tricipital fossa larger; (6) ligamental groove very long
(ligamental groove rather short); (7) section of shaft flattened
(corresponding section of shaft more circular); (8) impression
of M. brachialis anticus long and shallow (impression of M.
brachialis anticus more circular and sharply marked); (9) radial
condyle rather long and narrow in the center (radial condyle
much more rounded); (10) epitrochlea and epitrochlear promi-
nence little developed (epitrochlea and epitrochlear promi-
nence more developed and prominent on the internal side); (11)
tricipital grooves very large and deep (tricipital grooves wide
but not very deep); (12) external tricipital groove sharply
marked and bordered by two raised crests (external tricipital
groove less marked); (13) epicondyle well developed (epicondyle
not very developed).

Ulna. Similarities: (1) general shape very similar; (2) same
positioning of internal and external glenoid facets; (3) promi-
nence for anterior articular ligament well marked; (4) shaft
circular; (5) same general shape of distal end.

Differences: (1) proportionately shorter and more curved
(longer and less curved); (2) glenoid surfaces oriented obliquely
to long axis of bone (glenoid surfaces oriented almost parallel
to long axis of bone).

Radius. Differences: (1) general form rectilinear (general
form curved at distal end); (2) distal end spatulate, practically
symmetrical to long axis of bone (distal end assymmetric).

Carpometacarpus. Similarity: Short and wide in both
groups.

Differences: (1) radial apophysis forms a spur comparable
to that seen in Hoplopterus spinosus, the Spurwing Plover; this
radial apophysis was noted by Milne-Edwards (1892) (no spur
in the genera Harpactes, Trogon, Pharomachrus, and Pri-
oteles); (2) metacarpal III lies almost parallel to metacarpal II
(metacarpal II at a very oblique angle to metacarpal III; the
gap between the metacarpals is very wide at the distal end,
and the distal end is very wide); (3) internal digital facet lies
in the same plane as the external digital facet (internal digital
facet lies at a different level than external digital facet).

Mourer-Chauvire: The Archaeotrogonidae

19

Figure 1. Geographic distribution of living trogons (hatching) and fossil trogons: (*) Archaeotrogon, Eocene and Oligocene, Phosphorites du
Quercy, France; (*) Paratrogon, Miocene of 1’Allier, France; A Trogonidae, Oligocene, Glarner Fischschiefer, Switzerland.

Coracoid. Similarities: (1) same general shape of the head,
glenoid facet, and scapular facet; (2) no sub-clavicular fora-
men; (3) distal end large and paddle-shaped; (4) sternal facet
almost perpendicular to the long axis of the bone.

Differences: (1) proportionately slightly shorter and more
massive (proportionately longer and more slender); (2) sub-cla-
vicular apophysis usually broken at the extremity, but rather
wide at its origin (sub-clavicular apophysis narrow); (3) sternal
facet short and strongly curved (sternal facet longer and not
very curved); (4) hvosternal apophysis weakly developed on
the external side and barely present on the internal side (hyos-
ternal apophysis very well developed both externally and in-
ternally); (5) strongly marked groove for the ligament of the
sterno-coracoidal muscle on the upper surface of the bone (very
shallow groove for the sterno-coracoidal muscle).

Femur. Similarity: General shape very similar.

Differences: (1) proximal end rather flattened (proximal end
more swollen); (2) no pneumatic orifice under the trochanter
(pneumatic orifice present in the genus Trogon, but not in Har-
pactes)\ (3) fossa present below the articulation on the posterior
side (no fossa present below the articulation on the posterior
side); (4) shaft slender (shaft heavy in the genus Trogon , but

slender in Harpactes)\ (5) distal end flattened (distal end more
swollen).

Tibiotarsus. Similarities: (1) relatively short in both groups;
(2) proximal articulation perpendicular to the long axis of the
bone; (3) tibial crests poorly developed; (4) supratendinal
bridge lies on the internal side of the bone; (5) shallow groove
for the extensor muscle of the digits.

Differences: (1) shaft relatively slender and slightly widened
toward the distal end (shaft heavier and widens toward the
distal end); (2) external rugosity of oblique ligament well de-
veloped (external rugosity of oblique ligament poorly devel-
oped).

Tarsometatarsus. Similarity: Same general proportions as
compared to the femur and tibiotarsus.

Differences: (1) internal trochlea turned slightly backward;
digit I points backwards, digits II, III, and IV forwards (in-
ternal trochlea turned completely toward the rear; digits I and
II point backwards, digits III and IV forwards); (2) hypotarsus
with a channel pointing externally between two subequal cal-
caneal ridges (hypotarsus with a very strong median ridge
(ridge 1) and two canals situated externally to that median
ridge (see Fig. 2); (3) two very evident superior foramina, the

20

Mourer-Chauvire: The Archaeotrogonidae

Figure 2. Diagram showing the position of the calcaneal ridges of
the hvpotarsus in the genus Archaeotrogon (A) and in the living genus
Trogon (B). Right tarsometatarsus, proximal views.

internal foramen larger than the external foramen (two hardly
noticeable superior foramina of equal size); (4) inferior foramen
wide and in a deep groove, foramen lies clearly proximal to the
trochleae (inferior foramen very small and in a shallow groove;
foramen very close to trochleae); (5) shaft flattened anteropos-
teriorlv (shaft rather flattened mediolaterally); (6) internal co-
tvla prominent and sharply defined (internal cotvla weak); (7)
metatarsal facet well marked (metatarsal facet poorly marked).

It is the tarsometatarsus that shows the greatest contrast
between the Archaeotrogonidae and the Trogonidae. Archaeo-
trogon does not have the heterodactyl foot characteristic of the
living trogons and unique among all the birds.

Comparison with the Genus Paratrogon Lambrecht
1933. This genus contains the single species Paratrogon galli-
cus Milne-Edwards 1871, described from the Aquitanian de-
posits of Allier, and is known only from two humeri (Milne-
Edwards 1867-1871:395-396, pi. 177, figs. 18-22).

The genus Archaeotrogon is noticeably different from Par-
atrogon, and contrary to the opinion of Lambrecht (1933), I
believe that Paratrogon is closer to the living trogons than to
Archaeotrogon and should therefore be placed in the family
Trogonidae. When comparing the humerus of Archaeotrogon
to that of Paratrogon one finds almost the same differences as
noted between the humeri of Archaeotrogon and the living
Trogonidae. The characters of the humerus are shown in Ta-
ble 1.

Comparison with the Coraciiformes and the Capri-
mulgiformes. Feduccia (1977) has shown that the shape of
the stapes, the middle ear ossicle in birds, can be used to show
phylogenetic relationships. The trogonid stapes has a bulbous
and hollow basal part with a large orifice on the posterior side,
and a stapedial process arising from the edge of the basal part.
This morphology is very different from the primitive mor-
phology of the stapes, which is that of a flat discoidal plate
with the stapedial process arising from its center. The trogon
morphology of the stapes is found in four avian families pre-
viously assigned to the order Coraciiformes: Meropidae (bee-
eaters), Alcedinidae (kingfishers), Momotidae (motmots), and
Todidae (todys). According to Feduccia (1977:21), this simi-
larity “argues strongly for monophyly of the trogons and bee-
eaters/kingfisher/motmot/tody assemblage.” The earlier clas-
sification has therefore been modified, and the four families
mentioned above have been removed from the Coraciiformes
and joined with the Trogonidae in the new order Alcedini-
formes. It would be interesting to know if the osteology of the
primitive trogonids can support this relationship.

I was unable to compare Archaeotrogon with the Momotidae
or the Todidae, which are restricted to the tropical zones of
Central America and the Antilles, but I did make the com-
parison with the Meropidae and the Alcedinidae. There are
similarities in the bones of the hindlimb of the latter two fam-
ilies and those of Archaeotrogon, but there are very great dif-
ferences in the shape of the humerus.

In Merops, the proximal end of the humerus is not trans-
versely widened, there is no tricipital fossa, the internal tro-
chanter is low, the pectoral crest is short, the distal end is not
very wide transversely and sits obliquely to the long axis of
the shaft, the epitrochlea is very prominent toward the base,
and the tricipital grooves do not occupy a deep and wide
depression.

In the genera Alcedo and Dacelo, the head of the humerus
is globular, the internal trochanter weakly developed, the sub-
trochanteric fossa very small, the pectoral crest very short,
and the distal end is very different from that of Archaeotrogon .

On the other hand, there is a certain similarity between the
humeri of Archaeotrogon and the living Caprimulgiformes
( Caprimulgus and Chordeiles). This resemblance is particularly
strong in the new species of Archaeotrogon, which has a crest
obliquely crossing the tricipital fossa (see Figs. 4t-w, 10) as in
the genus Caprimulgus. There are likewise other characters in
common in both the humerus and other bones of the skeleton.
The ancestral forms of the Caprimulgiformes are unknown,
since the Aegialornithidae of the Eocene and Oligocene that
have previously been placed in this order (Brodkorb 1971;
Collins 1976) should actually belong to the Apodiformes (Har-
rison 1975; Mourer-Chauvire 1978). One may therefore spec-
ulate as to the possibility of Archaeotrogon being related to the
Caprimulgiformes.

Archaeotrogon venustus Milne-Edwards 1892

Figure 3

1891 Genus b Lydekker, p. 78, fig. 3

1892 Archaeotrogon venustus Milne-Edwards, p. 5-7

1908 Archaeotrogon venustus Milne-Edwards, Gaillard, p.

66-67, fig. 14, pi. 3, figs. 20-23

Mourer-Chauvire: The Archaeotrogonidae

21

Table 1. Morphological characters of the humerus of Archaeotrogon, Paratrogon, and the living Trogonidae.

Characters of
the Humerus

Archaeotrogon

Paratrogon

Living Trogons

Proximal end

Very large and very
recurved medially

Not large, slightly
recurved medially

Not large, slightly
recurved medially

Tricipital fossa

Large and usually
shallow

Narrower

Narrower

Bicipital surface

Very wide transversely

Less wide transversely

Less wide transversely

Ligamental groove

Very long

Not very long

Not very long

Head of humerus

Rather flattened

More swollen

More swollen

Sub-trochanteric fossa

Without pneumatic
orifices

Apparently without
pneumatic orifices

Pneumatic orifices
present

Impression of M.
brachialis anticus

Long and shallow

More circular and deep

More circular and deep

Radial condyle

Lengthened and narrowed
in the center

More globular

More globular

Epitrochlea

Poorly developed

Strongly developed

Strongly developed

Olecranal fossa

Very deep

Apparently rather deep

Rather shallow

Tricipital groove

Well marked externally
and bordered by two
prominent ridges

Weakly marked externally

Weakly marked externally
and bordered by a ridge

1933 Archaeotrogon venustus Milne-Edwards, Lambrecht, p.

625

1971 Archaeotrogon venustus Milne-Edwards, Brodkorb, p.

247

1971 Archaeotrogon venustus Milne-Edwards, Crochet, p.

316

MATERIAL: Early collections without provenance: com-
plete left humeri, QU 15797, QU 15799, QU 15805; incomplete
left humeri, QU 15802, QU 15785; complete right humeri, QU
15781, QU 15782; incomplete right humerus, QU 15804; in-
complete left carpometacarpi, QU 15917, QU 15939; complete
right carpometacarpus, QU 15882; incomplete right carpo-
metacarpi, QU 15915, QU 15918, QU 15940 (Museum of
Paris). Incomplete right humerus, PQ 987; incomplete left hu-
merus, PQ 991 (Museum of Lyon). Two left humeri and one
right humerus almost complete (Department of Earth Sciences,
Lyon).

Deposits of Pech du Fraysse: left humeri more or less com-
plete, PFR 577, 578, 11034, 11147, 11186; proximal left hu-
meri, PFR 11018, 11031, 11142, 11164, 11112, 11196; distal
left humeri, PFR 580, 5105, 5109, 11040, 11188, 11187, 11055,
11056, 11123, 11155, 11062, 11160, 11201, 11116, 11117,

11229, 11230, 11231; shafts of left humeri, PFR 11042, 11191,
11093; right humeri more or less complete, PFR 5106, 5108,
11022, 11195; proximal right humeri, PFR 581, 9545, 1 1070,
11071, 11080, 11081, 11102, 11121, 11180, 11232, 11233,

11234; distal right humeri, PFR 579, 582, 583, 5107, 7218,
9802, 11029, 11033, 11045, 11046, 11051, 11061, 11066,
11080, 11108, 11149, 11150, 11157, 11189, 11190, 11191,

11192, 11194, 11197; shaft of right humerus, PFR 11063; left
coracoids more or less complete, PFR 5 1 12, 5113, 8583, 1 1058,
11083, 11084, 11085, 11100, 11109, 11124, 11161, 11235;

proximal left coracoids, PFR 585, 7050, 11129, 11236; distal
left coracoids, PFR 11237, 1 1238; right coracoids more or less

complete, PFR 5111, 11076, 11088, 11095, 11133, 11162,
1 1166, 1 1168, 11172, 1 1205; proximal right coracoids, PFR
584, 8359, 11239; distal right coracoids, PFR 7465, 1 1126,
11174, 11251, 11240, 11241; left ulnae more or less complete,
PFR 3998, 11047; proximal left ulnae, PFR 11098, 11111,
11125, 11204; distal left ulnae, PFR 5118, 9409, 11075, 11130,
11200, 11213; right ulnae more or less complete, PFR 11043,
11136, 1 1 15 1; proximal right ulnae, PFR 8358, 11110, 11114,
11169, 11198, 11212; distal right ulnae, PFR 593, 594, 3999,
8360, 11068, 11170, 11208, 11242; left carpometacarpi more
or less complete, PFR 576, 5110, 11073, 11074, 11167, 11216;
proximal left carpometacarpi, PFR 7560, 1 1243, 1 1244, 1 1245,
11246, 11247; distal left carpometacarpi, PFR 586, 587, 7222,
11248, 1 1249; right carpometacarpi more or less complete,
PFR 574, 575, 9546, 11086, 11089, 11090, 11115; proximal
right carpometacarpi, PFR 1 1099, 11 127, 1 1 128, 11202,
11250, 11251, 1 1 2 5 2 ; distal right carpometacarpi, PFR 1 1105,
1 1253; proximal scapulae, PFR 11103, 1 1254, 1 1255, 11256,
1 1257; distal radii, PFR 11258, 11259, 11260, 1 1261, 1 1262,
11263; left femora more or less complete, PFR 11082, 11105;
proximal left femur, PFR 1 1107; distal left femur, PFR 11264;
right femora more or less complete, PFR 11060, 11113; prox-
imal right femur, PFR 11265; distal right femora, PFR 11211,
11214, 11266; proximal left tibiotarsus, PFR 11131; distal left
tibiotarsus, PFR 11267; almost complete right tibiotarsus,
PFR 11203; distal right tibiotarsi, PFR 11096, 11132, 11268;
almost complete left tarsometatarsi, PFR 11091, 11 175, 11269;
distal left tarsometatarsi, PFR 11270, 11271, 1 1272, 11273;
distal right tarsometatarsi, PFR 11274, 11275, 11276, 11277
(Museum of Paris).

Deposits of Escamps A: proximal scapula.

Deposits of Itardies: distal left humerus ITD 548; distal right
humeri, ITD 569, 617; proximal right coracoids, ITD 542,
573, 704, 709; distal right coracoid, ITD 691; proximal left
ulnae, ITD 678, 684; proximal right ulna, ITD 538; distal right

22

Mourer-Chauvire: The Archaeotrogonidae

Mourer-Chauvire: The Archaeotrogonidae

23

ulna, ITD 673; proximal left carpometacarpi, ITD 696, 710;
distal left carpometacarpus, ITD 541; proximal right carpo-
metacarpus, ITD 575.

Deposits of Mas de Got B: complete right carpometacarpus,
MGB 1545; distal right femur, MGB 1558.

Deposits of Mounavne: proximal right carpometacarpus,
MOU 1.

Deposits of Pech Desse: complete left humeri, PDS 1226,
1236; distal left humeri, PDS 1218, 1227, 1234; almost com-
plete right humerus, PDS 1212; proximal right humerus, PDS
1257; distal right humeri, PDS 1223, 1232, 1274; almost com-
plete left coracoids, PDS 1230, 1237; proximal left coracoid,
PDS 1243; complete right coracoid, PDS 1244; proximal right
coracoid, PDS 1275; distal right coracoid, PDS 1242; proximal
left ulna, PDS 1235; distal left ulnae, PDS 1249, 1252, 1269,
1270; distal right ulnae, PDS 1241, 1260; distal right carpo-
metacarpus, PDS 1264; proximal scapulae, PDS 1271, 1278,
1281, 1289; almost complete left femur, PDS 1277; distal right
femur, PDS 1280; distal left tarsometatarsus, PDS 1273.

Deposits of Perriere: distal left ulnae, PRR 2599, 2609; prox-
imal left carpometacarpus, PRR 2608; proximal left femur,
PRR 2607.

Deposits of La Plante 2: distal left humerus, PLA 1047; shaft
of right humerus, PLA 1062; proximal right coracoid, PLA
1071; distal right coracoid, PLA 1066; proximal right ulna,
PLA 1065; distal right ulna, PLA 1063; complete right car-
pometacarpus, PLA 1064; proximal left carpometacarpus,
PLA 1073; proximal right carpometacarpus, PLA 1070; prox-
imal scapula, PLA 1069; wing phalanx?, PLA 1067.

Deposits of Roqueprune 2: complete left coracoid, ROQ310;
distal right coracoid ROQ 317; proximal left ulna, ROQ 315;
distal right ulna, ROQ 312; distal left femur, ROQ 313; prox-
imal right femur, ROQ 311; proximal scapula, ROQ 318 (Uni-
versity of Montpellier and Paris VI).

Deposits of Boussac 1: almost complete left carpometacar-
pus.

Deposits of Boussac 2: distal right ulna and distal left ulna.

Deposits of Escamps 3: distal right humerus.

Deposits of Fonbonne 1: proximal left coracoid.

Deposits of Garrigues; proximal right carpometacarpus
(University of Utrecht).

DESCRIPTION: Archaeotrogon venustus is the smallest
species in the genus. It is also the species most abundantly
represented in the recently collected material and the best
known in regards to the skeleton. All the characters previously
indicated in the description of the genus Archaeotrogon apply
to this species.

MEASUREMENTS: For measurements of this species see
Table 2.

Archaeotrogon zitteli Gaiiiard 1908

Figure 4a-j

1908 Archaeotrogon zitteli Gaiiiard, p. 69, fig. 16; p. 70-72,
fig. 17, pi. 3, figs. 24-25 and 26-27
1933 Archaeotrogon zitteli Gaiiiard, Lambrecht, p. 626
1971 Archaeotrogon zitteli Gaiiiard, Brodkorb, p. 246-247

MATERIAL: Early collections without provenance: almost
complete left humeri, QU 15787, 15790, 15791, 15792a,
15792b, 15795; proximal left humerus, QU 15788; distal left
humeri, QU 15784, 15793, 15947; almost complete right hu-
meri, QU 15783, 15789, 15798, 15801; almost complete left
coracoid, QU 15911; almost complete left carpometacarpi, QU
15647, 15927; almost complete right carpometacarpi, QU
15659, 15662, 15928, 15934, 15942, 15944; proximal right car-
pometacarpus, QU 15946 (Museum of Paris). Complete left
humerus, PQ 1053, cast of no. 128 from the Museum of Mu-
nich, holotvpe; complete right humerus, PQ 1052, cast from
the Museum of Munich (referred to A. venustus by Gaiiiard
(1908), but its size actually corresponds to A. zitteli ); distal
right humerus, PQ 990; 3 left and 2 right tarsometatarsi, 4 of
which are almost complete, PQ 1069 (One of these was figured
by Gaiiiard (1908, fig. 16 and pi. 3, fig. 26-27) and attributed
to A. cayluxensis, but it has suffered a little damage since
then.) (Museum of Lyon).

Deposits of Pech du Fraysse: complete left carpometacarpi,
PFR 11069, 11097; distal left ulna, PFR 11092 (Museum of
Paris).

Deposits of Mas de Got B: complete left ulna, MGB 1548;
complete right ulna, MGB 1555; almost complete left coracoid,
MGB 1553 (University of Montpellier).

Deposits of Belgarite IVa: incomplete right humerus (Uni-
versity of Utrecht).

DESCRIPTION: According to Gaiiiard (1908:70), the hu-
merus of Archaeotrogon zitteli is quite well distinguished an-
atomically from that of A. venustus. He stated that ind. zitteli
the head of the humerus is much more widened transversely,
the tricipital fossa is shallower, and on the anterior face, the
bicipital surface is much reduced. I was able to study a large
number of humeri of both species, and these morphological
differences seem to me to be attributable to individual varia-
tion. The head of the humerus does not appear to be wider
transversely, nor the bicipital surface smaller inT. zitteli. The
tricipital fossa is perhaps slightly shallower in A. zitteli, but
this character is rather variable. Certain specimens such as
QU 15798 (Fig. 4d) have a shallow tricipital fossa, while others
such as QU 15795 (Fig. 4c) have a much deeper tricipital fossa.

It appears to me that the principal character that distin-
guishes A. zitteli from A. venustus is size, the former species

Figure 3. Specimens of Archaeotrogon venustus. Complete right humerus, QU 15782 (formerly QU 3282), Museum of Paris, X3.9, in anconal
(a) and palmar (b) view. Complete right coracoid, PFR 11168, Museum of Paris, X3.6, in anterior (c) and posterior (d) view. Complete right ulna,
PFR 11047, Museum of Paris, X3.7, in external (e) and internal (f) view. Complete right carpometacarpus, MGB 1545, University of Montpellier,
X3.7, in internal (g) and external (h) view. Proximal right scapula, PFR 11254, Museum of Paris, X3. 6, in dorsal (i) view. Distal radius, PFR
11258, Museum of Paris, X3.7, in external (j) and internal (k) view. Complete left femur, PFR 11082, Museum of Paris, X3.7, in posterior (1) and
anterior (m) view. Complete left tarsometatarsus, PFR 11091, Museum of Paris, X3.7, in anterior (o) and posterior (p) view. Distal right tibio-
tarsus, PFR 11132, Museum of Paris, X3.7, in anterior (q) view. Incomplete right tibiotarsus, PFR 11203, Museum of Paris, X3.7, in posterior
(r) view.

24

Mourer-Chauvire: The Archaeotrogonidae

Table 2. Measurements (mm) of A. venustus and A. zitteli bones.

Archaeotrogon venustus Archaeotrogon zitteli

variance

variance

n

min.

max.

mean

s2

n

min.

max.

mean

s2

Humerus

Length

20

25.0

29.7

27.82

1.31

12

30.1

33.4

31.43

0.84

Width head

27

8.4

9.4

8.85

0.07

9

9.4

10.3

9.98

0.06

Width distal end
Width shaft in

56

5.8

7.2

6.30

0.11

16

6.8

7.5

7.13

0.03

the middle

92

2.6

3.3

2.91

0.02

18

3.2

3.7

3.48

0.03

Ulna

Length

4

28.0

29.5

28.80

0.66

2

30.7

30.9

30.80

0.02

Width head

21

4.0

4.7

4.30

0.03

2

4.5

4.9

4.70

0.08

Width distal end

30

3.4

4.0

3.73

0.02

3

4.0

4.4

4.17

0.04

Depth distal end
Width shaft in

30

3.6

4.2

3.84

0.03

3

4.1

4.3

4.20

0.01

the middle

39

2.0

2.4

2.17

0.01

3

2.3

2.4

2.33

0.003

Radius

Width distal end
Width shaft in

6

3.0

3.5

3.18

0.03

the middle

6

1.1

1.2

1.17

0.003

Carpometacarpus

Length

21

17.8

19.8

18.84

0.18

10

19.5

20.9

20.27

0.36

Width head

32

6.2

7.8

7.08

0.20

10

7.5

8.3

7.85

0.08

Width distal end
Width metacarpal 2

25

3.6

4.5

4.01

0.07

7

4.0

4.8

4.46

0.08

in the middle

42

1.5

2.0

1.85

0.02

10

1.9

2.1

1.98

0.006

Coracoid

Length

23

18.8

21.7

20.50

0.42

2

22.7

24.0

23.35

0.85

Width head

38

3.8

4.9

4.30

0.07

2

4.3

4.5

4.40

0.02

Width sternal end
Width shaft in

19

4.7

6.2

5.34

0.16

2

6.2

6.2

6.20

0.00

the middle

46

1.9

2.7

2.22

0.03

3

2.3

2.6

2.43

0.02

Femur

Length

5

22.0

22.4

22.34

0.16

Width head

7

4.4

4.8

4.59

0.03

Width distal end
Width shaft in

7

4.2

4.5

4.36

0.01

the middle

13

1.7

2.0

1.83

0.01

Tibiotarsus

Length

1

~ 29

Width head

2

3.4

3.9

3.65

0.13

Width distal end
Width shaft in

4

3.7

4.1

3.83

0.04

the middle

4

1.6

1.7

1.65

0.003

Tarsometatarsus

Length

2

16.2

16.6

16.40

0.08

5

15.8

16.7

16.26

0.15

Width head

3

4.0

4.4

4.23

0.04

4

4.0

4.2

4.08

0.01

Width distal end
Width shaft in

5

3.8

4.1

3.96

0.02

4

3.8

4.5

4.18

0.09

the middle

7

1.7

1.9

1.79

0.005

5

1.9

2.3

2.04

0.02

Figure 4. Specimens of Archaeotrogon. A. zitteli: Complete right humerus, QU 15801 (formerly QU 3301), Museum of Paris, X1.7, in palmar
(a) and anconal (b) view. Complete left humerus, QU 15795 (formerly QU 3295), Museum of Paris, X1.9, in anconal (c) view. Complete right
humerus, QU 15798 (formerly QU 3298), Museum of Paris, X1.9, in anconal (d) view. Almost complete left carpometacarpus, QU 15927 (formerly
QU 3427), Museum of Paris, xi.8, in internal (e) and external (f) view. Almost complete left tarsometatarsus, PQ 1069, Museum of Lyon, X1.3,
in anterior (g) and posterior (h) view. Complete left coracoid, QU 15910 (formerly QU 3410), Museum of Paris, xl.7, in posterior (i) and anterior
(j) view. A. cayluxensis : Complete right humerus, holotype, PQ 2, Museum of Lyon, xl.7, in palmar (k) and anconal (1) view. Complete left
humerus, QU 15800 (formerly QU 3300), Museum of Paris, X] 9, in anconal (m) view. Distal right ulna, PFR 11054, Museum of Paris, X1.9, in
internal (n) and external (o) view. Complete right carpometacarpus, QU 15949 (formerly QU 3449), Museum of Paris, xl.7, in external (p) and
internal (q) view. Complete left coracoid, QU 15908 (formerly QU 3408), Museum of Paris, xl.7, in posterior (r) and anterior (s) view. A.
hoffstetteri new species: Almost complete right humerus, holotype, QU 15796 (formerly QU 3296), Museum of Paris, xl.7, in anconal (t) and
palmar (w) view. Almost complete right humerus, paratype, QU 15786 (formerly QU 3286), Museum of Paris, xl.7, in anconal (u) and palmar
(v) view.

Mourer-Chauvire: The Archaeotrogonidae

25

26

Mourer-Chauvire: The Archaeotrogonidae

Figure S. Scatter diagram for the humeri of the different species of Archaeotrogon from the Phosphorites du Quercy.

Depth of distal end

Figure 6. Scatter diagram for the distal end of the ulnae of Archaeo-
trogon venustus, A. zitteli, and /l. cayluxensis.

being the larger. In the scatter diagrams (Figs. 5-9), they form
distinct clusters of points. The measurements of the bones at-
tributed to A. zitteli are shown in Table 2. They are slightly
larger than those of A. venustus, and on the whole there is
very little overlap in the measurements of the two species.
This cannot be an example of evolution, i.e., the smaller, 4.
venustus evolving into the larger A. zitteli, as both species
have been discovered together in at least two sites in the new
excavations at Quercy: Mas de Got B, of the lower Oligocene,
and Pech du Fraysse, of the upper Oligocene.

I have referred to A. zitteli five tarsometatarsi from the
Museum of Lyon (PQ 1069), one of which was figured by
Gaillard (1908) and described as A. cayluxensis. The size of
A. cayluxensis is much larger than that of either ,4. venustus
or A. zitteli. Practically all the specimens of Archaeotrogon
found at Pech du Fraysse belong tod. venustus, and it seems
likely that the tarsometatarsi, especially the two complete ones
(PFR 11091, 11175), likewise belong to this species. In the
scatter diagram of the tarsometatarsus (Fig. 9), it is evident
that the specimens numbered PQ 1069 have a total length
comparable to that of thed. venustus from Pech du Fraysse,
but their shafts are much thicker. It seems to me, therefore,
that these tarsometatarsi belong tod. zitteli, all the more so
since d. zitteli is far more numerous in the early collections
than d. venustus. It is logical to assume that if one finds many
humeri one should also have a few foot bones.

Mourer-Chauvire: The Archaeotrogonidae

27

Total length

Figure 7. Scatter diagram for the carpometacarpi of Archaeotrogon venustus, A. zitteli, and A. cayluxensis.

Archaeotrogon cayluxensis Gaillard 1908

Figure 4k-s

1908 Archaeotrogon cayluxensis Gaillard, p. 67-70, fig. 15,
pi. 4, figs. 1-4

1933 Archaeotrogon cayluxensis Gaillard, Lambrecht, p. 625-
626

1939 Archaeotrogon cayluxensis Gaillard, Gaillard, p. 17-18,
fig. 7

1971 Archaeotrogon cayluxensis Gaillard, Brodkorb, p. 246

MATERIAL: Early collections without provenance: almost
complete left humeri, QU 15 778, 15 779, 15800; distal left hu-
meri, QU 15780, 15803, 15806; distal right humerus, QU
15794; complete left coracoid, QU 15908; complete right cor-
acoid, QU 15907; complete right carpometacarpi, QU 15916,
15924, 15948, 15949, 15950; proximal left carpometacarpi, QU
15668, 15944 (Museum of Paris). Complete right humerus, PQ
2 (Holotype of Gaillard); distal left humerus, cast without
number (Museum of Lyon). Almost complete right humerus,
figured in Gaillard (1939, fig. 7) (Department of Earth Sci-
ences, Lyon).

Deposits of Pech du Fraysse: distal right ulna, PFR 11054
(Museum of Paris).

DESCRIPTION: Gaillard (1908:67) says that the humerus
ofd. cayluxensis differs from that ofd. venustus, not only in
size, but also in the following anatomical characters: in A.
cayluxensis, the head of the humerus is thicker anteroposte-
riorlv, the tricipital fossa and the sub-trochanteric fossa are
large and shallow, the pectoral crest is long with a rounded
edge, the bicipital surface is smaller in a vertical direction, the
body of the humerus is more slender and widened toward the
distal end, the epitrochlea and epicondyle are more prominent,
and the inferior groove of the triceps is deeper.

Having been able to examine more material, certain of these
distinctive characters seem valid and others less so. I would
agree that the head of the humerus is thicker anteroposteriorly
in A. cayluxensis, the bicipital surface is proportionately small-
er, the epitrochlea and the epicondyle are more prominent,
and the triceps groove is deeper Further, the radial condyle
is proportionately much more developed anteroposteriorly.

The form of the tricipital fossa is rather variable among
individuals, being very shallow in the holotype, PQ 2 (Fig.
41), but much deeper in other specimens, such as QU 15800
(Fig. 4m). The shape of the sub-trochanteric fossa appears no
different than that of A. venustus, and the pectoral crest is not
especially longer, nor is its border more rounded. The shaft is
slender in the holotype, but it is much heavier in other indi-
viduals, for example, QU 15800 (Fig. 4m). It does not seem
to widen more toward the base than does A. venustus. The
most important distinguishing character is certainly the size,
which is clearly superior to A. venustus and A. zitteli (Figs.
5-8).

Archaeotrogon cayluxensis is known mostly from the early
collections. Only a single bone attributable to this species has
been found in the recent collections from Quercv. It is a distal
ulna from Pech du Fraysse (PFR 11054, Fig. 4n-o). Its mor-
phology corresponds to that of the genus Archaeotrogon, and
its size is very important (Fig. 6).

If one calculates the ratios of the means of the measurements
of all the bones of the two species A. cayluxensis and A. ve-
nustus, the result varies from 1.28 to 1.52. If the same ratios
are taken between A. cayluxensis and A. zitteli, the results
vary from 1.14 to 1.43. This means thatd. cayluxensis is an
average of 1.28 to 1.52 times as large as A. venustus, and 1.14
to 1.43 times as large as A. zitteli. If one takes the only two
measurements possible on the ulna from Pech du Fraysse and

28

Mourer-Chauvire: The Archaeotrogonidae

A. cayluxensis

★ Museum de Paris

A. zitteli

* Museum de Paris

A. venustus
O Pech du Fraysse
+ Pech Desse
v Divers gisements

Total length

Figure 8. Scatter diagram for the coracoids of Archaeotrogon venustus, A. zitteli, and A. cayluxensis.

compares them with the mean values for the corresponding
measurements of the other two species, the following ratios
result: with .4. venustus, 1.58 (depth) and 1.54 (width); with
A. zitteli, 1.41 (depth) and 1.40 (width). The ratios between
the measurements of the ulna from Pech du Fraysse and those
of the other two species are therefore slightly larger than those
generally observed between A. cayluxensis on the one hand,
and 4. venustus andd. zitteli on the other. But the ulna falls
within the range of individual variation. It may belong to a
particularly robust individual of A. cayluxensis.

MEASUREMENTS: For measurements of this species see
Table 3.

Archaeotrogon hoffstetteri new species

Figure 4t-w

HOLOTYPE: Complete right humerus, QU 15796, Nation-
al Museum of Paris.

PARATYPE: Slightly incomplete right humerus, QU 15 786,
National Museum of Paris.

TYPE LOCALITY: Phosphorites du Quercy, France.

TYPE STRATA: Upper Eocene or Oligocene.

DIAGNOSIS: A species of the genus Archaeotrogon, char-
acterized by having a humerus of about the same size as that
of A. venustus or A. zitteli, but with a much more slender

Mourer-Chauvire: The Archaeotrogonidae

29

Figure 10. Diagram of the proximal end of the humerus of Archaeo-
trogon hoffstetteri new species, in anconal view.

shaft and with the proximal end much more recurved inter-
nally.

ETYMOLOGY: This species is named in honor of Dr. Rob-
ert Hoffstetter.

DESCRIPTION: These two humeri are sharply separated
from the other humeri of Archaeotrogon by their slenderness
and by the sigmoid curve of their shafts. Where the other
humeri of Archaeotrogon are massive, these appear much more
slender. Further, the proximal end of the bone is strongly
twisted inwards.

On the posterior side of the bone, the tricipital fossa is very
shallow. The sub-trochanteric fossa, under the internal tro-
chanter, is bordered by two crests, an internal crest and a crest
that Milne-Edwards called the internal trochanteric crest (Fig.
10). In A. hoffstetteri, another crest arises from the base of
this internal trochanteric crest and leads toward the head of
the humerus, crossing the tricipital fossa obliquely.

Below the external trochanter is a muscle insertion surface
that is rather elongate and lies parallel to the long axis of the
bone in A. venustus, A. zitteli, and /l. cayluxensis. Ind. hoff-
stetteri this surface is proportionately shorter and lies more
obliquely.

Ind. hoffstetteri, on the external face of the bone, the pec-
toral crest is very prominent and its upper edge shows a
marked swelling. This pectoral crest is equally as prominent
on the anterior side of the bone, and the bicipital surface is
rather poorly developed.

The distal end of the bone does not show any particular
characters, the more so since it is imperfectly preserved in both
humeri attributed to this species.

RELATIONSHIPS AND DIFFERENCES: This species
can be distinguished from A. cayluxensis by its much smaller
size. The total length of the humerus is comparable to the
largest among d. venustus or the smallest among d. zitteli, yet
though the length is comparable, the shaft is far more slender
in d. hoffstetteri (Fig. 5). Further, the bone has a character-
istically sinuous shape. In addition, there are the other distinct
morphological characters, i.e. , a crest that obliquely crosses
the tricipital fossa, the length and orientation of the muscle
insertion scar below the external trochanter, and the very
strong development of the pectoral crest ind. hoffstetteri (Fig.
10).

MATERIAL AND LOCALITIES: This species is repre-
sented only by the two humeri in the collections of the National
Museum of Natural History in Paris, and is not represented
in the newer collections from the Phosphorites du Quercy. The
original locality is unknown, and it is impossible to assign it
a precise geological age. It is possible that among the skeletal
elements, other than the humeri, at present assigned to d.
venustus and A. zitteli, certain bones may prove to belong to
d. hoffstetteri. There is always the hope that this species may

Table 3. Measurements (mm) of d. cayluxensis and d. hoffstetteri bones.

Archaeotrogon cayluxensis

Archaeotrogon hoffstetteri

variance

variance

n

min.

max.

mean

s2

n

min. max. mean

s 2

Humerus

Length

5

35.6

37.9

36.68

0.87

2

30.0 30.5 30.25

0.13

Width head

3

11.8

12.0

11.90

0.01

1

8.7

Width distal end
Width shaft in

10

7.5

8.5

8.18

0.12

2

6.3 6.8 6.55

0.13

the middle

10

3.7

4.3

4.05

0.04

2

2.9 3.0 2.95

0.005

Ulna

Width distal end

1

5.9

Depth distal end

1

5.9

Carpometacarpus

Length

5

24.7

25.7

25.42

0.24

Width head

6

10.0

10.9

10.58

0.13

Width distal end
Width metacarpal II

4

4.9

5.4

5.13

0.05

in the middle

6

2.4

2.6

2.53

0.01

Coracoid

Length

2

26.6

29.7

28.15

4.81

Width head

2

6.0

6.6

6.30

0.18

Width sternal end
Width shaft in

2

7.4

8.8

8.10

0.98

the middle

2

3.1

3.3

3.20

0.02

30

Mourer-Chauvire: The Archaeotrogonidae

Table 4. Temporal distribution of Trogoniformes in the deposits of the Phosphorites du Quercy. Mammal zones after Fahlbusch

(1975).

-o

c <

c/i -9

<

2 c

«4-i • rj

° c L

iS, JS .™

C aS
§2.
N C £

nj ^2
2 *

Mammal

Zones

Deposits of the
Phosphorites du
Quercy

Species of
Trogoniformes

W

2

w

u

o

o

re

o

w

2:

re

u

o

w

26

32

a

o

■Q

39

2

£

41

03 JJ

re e «

NP 24

NP 23

grande coupure”
36

NP 22

NP 21

NP 20

NP 19

NP 18

NP 17

Boningen

Antoingt

Heimersheim

Montalban

Hoogbutsel

Frohnstetten

Pech du Fraysse

Pech Desse

Itardies

Mounavne

Montmartre — San Cugat

La Debruge

Perriere

Fons 4

Grisolles

Escamps

Perriere

A. venustus
A. zitteli
A. cayluxensis

A. venustus

A. venustus
A. venustus

Mas de Got B

1 A. venustus

\ A. zitteli

Villebramar :

La Plante 2

A. venustus

Roqueprune 2

A. venustus

A. venustus

A. venustus

reappear in the course of new research on the phosphorites,
and that we may then learn more of its skeleton.

MEASUREMENTS: For measurements of this species see
Table 3.

TEMPORAL DISTRIBUTION

The distribution of the different species of the genus Ar-
chaeotrogon in the sites of the Quercy phosphorites is shown
in Table 4. It is evident that A . venustus, which is the species

most numerous in the recent collections, has a very large tem-
poral range. It is already present in the Perriere zone, and
persists at least until the Boningen, stretching across a length
of at least ten mammal zones. In absolute terms this time span
can be evaluated at nearly IS million years.

The species A. zitteli andT. cayluxensis, quite common in
the early collections, have been rediscovered in the course of
the recent excavations in only two sites (Mas du Got B and
Pech du Fraysse). It is therefore not possible to precisely de-
termine their temporal distribution.

Mourer-Chauvire: The Archaeotrogonidae

31

The Archaeotrogonidae are relatively rare, but are nonethe-
less found in beds antedating the “grande coupure,” or “great
change,” such as those of Perriere and Escamps. This great
change is practically on the Eocene-Oligocene boundary and
is characterized by a large-scale replacement among the mam-
malian fauna. In the upper Eocene beds at Quercy, the pre-
dominant forms among the birds belong to the Aegialornithi-
dae. In contrast, after the great change, it is the Archaeotro-
gonidae that become predominant while the genus Aegialornis
disappears. The Aegialornithidae still existed, but they are
represented only by the genus Cypselavus, which is always
rather rare. As for the mammals, the “grande coupure” seems
to correspond to a rather important change in the avian world.

ACKNOWLEDGMENTS

I thank J.P Lehman, Director of the Institute of Paleon-
tology of the National Museum of Natural History of Paris;
L. David, Director of the Natural History Museum of Lyons;
and H. de Bruijn of the University of Utrecht for having kind-
ly loaned me part of the material studied herein. I thank as
well my colleagues from the University of Paris VI and the
University of Montpellier for having kindly entrusted me with
the study of the materials that they collected in their excava-
tions. The Netherlands Organization for Scientific Research
(ZWO) made it possible for me to visit the Rijksmuseum van
Natuurlijke Historic of Leiden in order to compare the fossil
birds from Quercy with living exotic forms. Finally, I thank
K.A. Hiinermann of Zurich for the precise data he gave me
regarding the age of the “Glarner Fischschiefer.”

LITERATURE CITED

Bonis, L. de, J.Y. Crochet, J.C. Rage, B. Sige, J Sudre,
and M. Vianey-Liaud. 1973. Nouvelles faunes de Ver-
tebres oligocenes des Phosphorites du Quercy. Bull. Nat.
Hist., 3° ser., n° 174, Sci. Terre 28:105-113.

Brodkorb, P. 1971. Catalogue of fossil birds, Part 4 (Co-
lumbiformes through Piciformes). Bull. Florida State
Mus., Biol Sci. 15(4): 163-266.

Cavaille, A., B. Geze, L. de Bonis, B, Sige, J. Sudre, L.
Ginsburg, M. Brunet, and J.Y. Crochet. 1974. Ta-
ble ronde sur les Phosphorites du Quercy. Geologie, Car-
nivores, Insectivores, Ongules. Palaeovertebrata 6(1-
2): 1—160.

Collins, C.T. 1976. Two new species of Aegialornis from
France, with comments on the ordinal affinities of the
Aegialornithidae. Smiths. Contrib. Palaeobiol. 27:121-
127.

Crochet, J.Y. 1971. Les Vertebres de l’Oligocene superieur
du Pech du Fraysse poche a phosphate du Quercy (com-
mune de Saint- Project, Tarn-et-Garonne). C.R. Somm.
Soc. Geol. Fr. 197 1(6):3 16— 3 1 7.

. 1978. Les Marsupiaux du Tertiaire d’Europe. These

Sci. Montpellier. 360 pp.

Crochet, J.Y., J.L. Hartenberger, J.A. Remy, B. Sige,
J. Sudre, and M. Vianey-Liaud. 1972. Catalogue des
Mammiferes du Quercy. Pretirage. Table ronde du
CNRS, Montauban. 73 pp.

Fahlbusch, V. 1975. Report of the International Sympo-
sium on mammalian Stratigraphy of the European Ter-
tiary. Newsl. Stratigr. 5(2°): 160—167.

Feduccia, A. 1977. A model for the evolution of perching
birds. System. Zool. 2 6( 1 ): 1 9—3 1 .

Gaillard, C. 1908. Les Oiseaux des Phosphorites du Quer-
cy. Ann. Univ. Lyon, nouv. ser. 23:1-178.

. 1939. Contribution a l’etude des oiseaux fossiles.

Arch. Mus. Hist. Natur. Lyon 1 5(2 ): 1 — 100.

Harrison, C.J.O. 1975. Ordinal affinities of the Aegialor-
nithidae. Ibis 1 17(2): 164—170.

Hartenberger, J.L. 1973. Etude systematique des Theri-
domvoidea (Rodentia) de l’Eocene superieur. Mem. Soc.
Geol. Fr., nouv. ser. 52(117): 1—76.

Hartenberger, J.L., N. Lopez, J.C. Rage, J.A. Remy, B.
Sige, J. Sudre, L. Thaler, and M. Vianey-Liaud.
1974. Table ronde sur les Phosphorites du Quercy. Ta-
phonomie, Squamates, Chiropteres, Rongeurs, Lago-
morphes et Primates. Palaeovertebrata 6(3-4): 161-303.

Lambrecht, K. 1933. Handbuch der Palaeornithologie.
Borntraeger. 1024 pp.

Lydekker, R. 1891. Catalogue of the fossil birds in the Brit-
ish Museum (Natural History). British Museum, London.
368 pp.

Milne-Edwards, A. 1867-1871. Recherches anatomiques
et paleontologiques pour servir a l’histoire des oiseaux fos-
siles de la France. Victor Masson and Sons, Paris 1:1 —
474, 2:1-627, 200 pi.

. 1892. Sur les oiseaux fossiles des depots eocenes de

phosphate de chaux du Sud de la France. C.R. Second
Congr. Ornithol. Intern., Budapest 1891:1-21.

Mourer-Chauvire, C. 1978. La poche a phosphate de Sainte-
Neboule (Lot) et sa faune de Vertebres du Ludien supe-
rieur. 6. Oiseaux. Palaeovertebrata 8( 2 —4 ):217— 229.

Olson, S.L. 1976. Oligocene fossils bearing on the origins of
the Todidae and the Momotidae (Aves: Coraciiformes).
Smiths. Contrib. Paleobiol. 27:111-119.

Peters, J.L. 1945. Check-list of the birds of the world. Vol.
5. Harvard Univ. Press, Cambridge. 306 pp. (reprinted
in 1968).

Peyer, B. 1957. Protornis glaronensis H. v Meyer: Neube-
schreibung des Typusexemplares und eines weiteren
Fundes. Schweiz. Palaeont. Abhandl. 73:3-47.

Sige, B. 1974. Insectivores et Chiropteres de l’Eocene su-
perieur et Oligocene inferieur d’Europe occidentale. Ar-
ticle de synthese. These Sci. Montpellier, n° AO 10537.
85 pp

. 1976. Insectivores primitifs de l’Eocene superieur et

Oligocene inferieur d’Europe occidentale. Nyctitheriides.
Mem. Mus. Nat. Hist. Natur., Nouv. Ser., Ser. C, Sci
Terre 34:1-140.

Sudre, J. 1977. Les Artiodactvles de l’Eocene moven et su-
perieur d’Europe occidentale; systematique et evolution.
These Sci. Montpellier. 257 pp.

Vianey-Liaud, M. 1976. Evolution des Rongeurs a
l’Oligocene en Europe occidentale. Article de synthese.
These Sci. Montpellier, n° AO 12290. 113 pp.

A NEW MIDDLE EOCENE SHOREBIRD
(AVES: CHARADRIIFORMES, CHARADRII)
WITH COLUMBOID FEATURES

By Ella Hoch1

ABSTRACT: Plumumida lutetialis new genus and species is the third fossil bird to be described from

the Messel site in West Germany. It is based upon a fractured postcranial skeleton and is referred to the
Charadrii on the basis of numerous skeletal features. It deviates from all living and known fossil shorebirds
by having a strong perching foot and skeletal structures in the pelvis and hindlimb considered to be
connected with such specialization. By the form and supposed function of the foot the fossil bird shows
similarity to the doves, but other dove apomorphies are lacking. Two skeletal features that may be
autapomorphies for the group to which Plumumida belongs set the bird apart from the doves. Plumumida
is believed to be related to those early shorebirds that are thought to have given rise to, among others,
the doves; the genus is placed incertae sedis in the Charadrii. The depositional environment of the fossil
is reviewed.

ZUSAMMENFASSUNG: Ein neuer Stelzvogel, Plumumida lutetialis gen. et sp. nov., wird als dritter

fossiler Vogel aus dem Messeler Olschiefer beschrieben. Die Art griindet sich auf ein postkraniales Skelett,
das massig verdriickt und unvollstanding ist. Die osteologischen Merkmale deuten auf eine Verwandt-
schaft mit den Charadrii hin. Der Vogel unterscheidet sich jedoch von alien heutigen und fossilen Cha-
radrii durch einen SitzfuB. Zu einer solchen Spezialisierung passen auch die gefundenen Skelettstrukturen
im Bein und Becken. In Form und mutmaBlicher Funktion des FuBes ahnelt Plumumida den Tauben,
doch sind keine anderen Columbiformen Apomorphien am Fossil nachweisbar. Zwei Skelettmerkmale,
die Autapomorphien der taxonomischen Gruppe sein konnten, zu der Plumumida eigentlich gehort, un-
terscheiden den fossilen Vogel von den Columbiformen. Plumumida ist zu den friihen (jungkretazisch-
alttertiaren) Stelzvogeln zu rechnen, aus welchen auch andere Formen, wie allem Anschein nach die
Tauben, abgeleitet werden konnen. Plumumida lutetialis wird incertae sedis den Charadrii eingegliedert.
Ein Uberblick iiber die Lebens- und Einbettungsumstande des Fossils wird gegeben.

Two birds have previously been recorded from the Lutetian
(Middle Eocene) deposits at Messel in the West German Bun-
desstaat Hessen. They are the alleged shorebird Rhynchaeites
messelensis Wittich 1898, and Diatryma cf. steini Matthew
and Granger 1917 (Berg 1965). The purpose of this paper is
to describe a new shorebird from the Messel oilshale that is
about the size of Rhynchaeites messelensis, but differs from
that form in skeletal morphology and relative proportions, and
from living shorebirds by having a perching foot.

The Messel site has yielded large quantities of fossil animal
and plant remains, many of which have only recently been
discovered. The total number of bird fossils so far secured
from the site comes to over one hundred. Among these, more
than fifty were kept by Frau E. Soergel in Freiburg im Breis-
gau (Tobien 1969:165), but these have now been returned to
the Hessisches Landesmuseum in Darmstadt, West Germany,
and about forty are under investigation by Dr. D.S. Peters at

1 Geologisk Museum, University of Copenhagen, 0. Voldgade 5-7,
DK-1350 Copenhagen K, Denmark.

Contrib Sci. Natur. Hist. Mus. Los Angeles County. 1980. 330:33-49.

the Forschungsinstitut Senckenberg, Frankfurt am Main,
West Germany. Such abundance, and the tragic fate now
threatening the Messel pit from local authorities in one of in-
dustrial Europe’s most densely populated areas, justifies a brief
introductory description of the site.

THE MESSEL SITE

The Messel pit is located about 9 km NE of Darmstadt and
22 km S of Frankfurt am Main. It now appears as a 1000 by
700 m crater, 60 to 70 m deep, the bottom of which became
covered by a shallow lake after mining was stopped in 1971
(Fig. 1). The pit was formerly exploited for oilshale, and pro-
duced about 1 million tons of crude oil used as fuel and for
the manufacture of various products for the dyestuffs, elec-
tronics, chemical, and pharmaceutical industries until the site
was closed down.

The oilshale was discovered in the mid-1870’s, and exploi-
tation began in 1886. The shale deposit is known to have had
a maximum thickness of 190 m in the Messel pit, and there
are other, smaller oilshale pits in the area. The shale occurred

34

Hoch: Middle Eocene Shorebird

Figure 1. The Messel site at the end of February 1978. Photograph by J.L. Franzen (Senckenberg Museum, Nos. 296-297).

below Middle Eocene strata containing brown coal lenses up
to 2 m thick (Franzen 1976a). Through the years 1886-1971,
many specimens of fossil leaves, wood, insects, fish, frogs,
tortoises, lizards, snakes, crocodiles, birds, and mammals were
found and recovered by those working in the open air Messel
pit (Tobien 1957), and by the Darmstadt Museum. Fossils
were also obtained from the nearby Prinz von Hessen oilshale
pit in the years 1912-24 (Franzen 1978). Unfortunately, be-
cause of the large water content (up to 40 percent) and the
chemical constituents of the shale (marcasite in particular),
combined with the state of knowledge of preservation tech-
niques at that time, many of these fossils are now in a most
miserable condition.

With the introduction of the “cast resin transfer method”
(see Bornhardt 1975) developed in England in the 1950’s
(Toombs and Rixon 1950; see also Rixon 1976) and first applied
to the Messel fossils by Dr. Walter Kiihne in the early 1960’s
(Kiihne 1961, 1962), a new period of preservation of the oil-
shale specimens began. The resin impregnates and locks the
bones of one level of the fossils, thus hardening and supporting
them sufficiently for further manipulation and studies, as well
as for exhibition purposes. However, the parts of the fossil
specimens that are not impregnated by the resin, or that do
not adhere to it, are often lost.

A veritable fossil-rush followed the closing of the mine
works in 1971. A multitude of collectors, most of them laymen,
searched for traces of ancient life in the pit, not always to the
good of the desired objects. Authorities finally had to take
steps to safeguard, not so much the site, but the lives of the
fossil hunters. At the end of 1974 the admittance of non-au-
thorized persons to the site was prohibited.

Permission to collect fossils in the Messel pit was granted to
the Hessisches Landesmuseum Darmstadt in 1913. This insti-
tution was the sole authorized collector until 1975 when the
Forschungsinstitut Senckenberg, the University of Hamburg,
and the Museum fur Naturkunde Dortmund also obtained the
right to conduct scientific investigations of the site. A large
project is now in progress, financially supported by the foun-
dation “Volkswagenwerk,” involving many scientists and good
amateurs working under the leadership of Dr. Jens Lorenz
Franzen, Forschungsinstitut Senckenberg (Franzen 1978,

1979). More than 18,000 fossils have been recovered to date,
the most conspicuous fossil group being perhaps the perisso-
dactvls, Propalaeotherium spp. (first thoroughly described by
Haupt 1925), although the holosteans Amia and Atractosteus
(Lepisosteidae) and the salmoniform Thaumaturus are the most
numerous vertebrates. Among the tetrapods, birds and bats
dominate (Franzen 1979). A list of fossils is given by Koenigs-
wald (1979). Several forms hint at a NW Europe-North
American land connection that persisted until about the end
of the Early Eocene (West et al. 1977; Berggren et al. 1978).

The fossils of the Messel oilshale are often, except for a
certain flattening, exceedingly well preserved when found. Not
only do major parts of bony skeletons occur, generally in ar-
ticulation, but in many cases the specimens show substantial
“shadows” of the organisms’ soft tissues. Hair, feathers, and
traces of colors in chitinous insect parts may occur at the site
(Haupt 1925; Franzen 1976b, 1979). Even the cuticular struc-
tures of partly digested leaves in the intestines of some fossil
herbivores can be studied in detail. For example, there have
been investigations into the stomach contents of Propalaeothe-
rium messelense (Franzen 1976b, Sturm 1978). These studies
have, in turn, given material support to the hypothesis (first ad-
vanced by Kowalevsky 1873-1874; see also Strelnikov and
Hecker 1968) of a habitat and food change during the evolution
of horses (sensu lato) from pre-Miocene softground forests to
grass plains, and from “omnivorous” to purely graminivorous
equids, respectively.

Presented against this background, recent proposals for the
future of the Messel pit evoke the quotation of a heading in
the catalogue for the Senckenberg Museum Messel exhibition
(Franzen 1977:24): “Wen interessiert eigentlich der Magenin-
halt des Urpferdchens?” — “To whom is the stomach content
of the protohorse of any real interest?” There are very few sites
in the world where remains of an Eocene continental flora and
fauna occur in such abundance and good state of preservation
as at Messel. These fossils permit a multitude of scientific in-
vestigations and conclusions concerning evolution and the eco-
logical, climatological, and physico-chemical aspects of the
paleoenvironment. But, instead of preserving the Messel site
and its fossils for the progress of knowledge, there are now
official plans for the Frankfurt-Darmstadt-Dieburg-Offenbach

Hoch: Middle Eocene Shorebird

35

metropolitan area to use the Messel pit for a large scale garbage
disposal! This would also include so-called “non-dangerous in-
dustrial waste products.” It may be true that large refuse ac-
cumulations are an inevitable result of modern environmental
policy, but it is questionable if it is really necessary to install
a giant garbage dump exactly in the Messel pit! Such short-
sighted solutions to man-made problems are highly tragic, es-
pecially when viewed against the history of life.

AGE AND ORIGIN OF THE
MESSEL OILSHALE

Judged from its fossil content, the bituminous deposit is of
Lutetian age, 43 to 49 million years ago (Berggren 1972, and
following the time scale agreed upon by workers presently
studying the Eocene North Sea tetrapods in the “Projet 124 du
Programme international des correlations geologiques”). To-
bien (1968, 1969:169) refers the oilshale to early Lutetian age
on the basis of contained mammals. This is questioned by
Franzen (1976a:422), however.

During middle to late Eocene times, Tertiary Europe ex-
perienced its maxima of temperature and marine transgres-
sion. Surrounded by the Atlantic Ocean, the North Sea, and
the Tethys Sea, the two latter being connected over present-
day Poland (Russell 1975; Bond 1978; Heissig 1979), the “Mid-
European land” had a humid subtropical-tropical climate
(Nemejc 1970; Buchardt 1978), as is also indicated by the rich
bitumen and lignite deposits formed at that time.

The Lutetian Messel lake and neighboring lakes developed
concurrently with tectonic rifting of the area, as part of a larger
river system (Tobien 1969). Detailed investigations, in partic-
ular of the location and orientation of fish fossils in situ in the
oilshale, reveal the influx of two main water currents in the
Messel lake, one from the northeast and another from the
northwest, and the outflow of a current at the southern edge
of the lake (Franzen 1979). Related studies (ibid.) have also
shed light upon the distribution and probable local geographic
derivation of the fossil organisms in the oilshale. The Messel
flora and fauna were part of a warm, damp Eocene jungle
environment. In its rivers and lakes the near-surface life was
abundant, but the deep bottom waters were quiet and dark
with anaerobic conditions. Thin strata of mineral particles and
organic debris accumulated, embedding those parts of larger
carcasses that were not devoured on their way down through
the waters.

Towards the end of the Middle Eocene the lakes developed
into swamps that were invaded by land plants. These plants
gave rise to the lenses of brown coal that overlie the oilshale.

OTHER SITES IN WESTERN EUROPE
WITH EOCENE BIRD REMAINS

European geography was subjected to great changes during
Paleogene times. The interrelated Alpine folding and Rhine-
graben rifting determined the formation of the central Euro-
pean tectonic features (lilies 1978; lilies and Greiner 1978),
including the primary Messel lake. And North Atlantic rifting
continued the rupture of the northern landbridge between Eu-
rope and North America. Diatryma occurred at Messel (Berg
1965), Geiseltal (Fischer 1967, 1978), and in France (Gaillard
1936) during the Middle Eocene, and is recorded from Upper

Paleocene and Lower Eocene beds in north America and
France (see Brodkorb 1967).

Briefly mentioned below are some Eocene sites in western
Europe of interest to paleornithologists; faunistic or chrono-
logic intercorrelations have not been made, nor have correla-
tions with Eocene sites outside western Europe. Relevant lit-
erature older than that referred to can be found in the cited
works.

British deposits yielding Eocene birds are known from the
southeastern part of England; these fossils have most recently
been discussed by Harrison (1971), and Harrison and Walker
(1977). In France, Eocene bird remains are recorded from the
Paris basin (Louis 1969; Brunet 1970; Hoch 1975); the bone-
fissures of Quercv, or “les Phosphorites de Quercy” (Gaillard
1939; Collins 1976; Mourer-Chauvire this vol.); the Lower
Eocene egg-bearing beds of Provence that are part of an upper
Cretaceous-lower Tertiary series containing bones and egg-
shells of dinosaurs and eggshells of large birds (Touraine 1960,
1961; Dughi and Sirugue 1962); the Lutetian Strait of Carca-
sonne sediments with bird tracks (Plaziat 1964); and the sites
that produced Diatryma referred to above. From Switzerland
are known the Egerkingen siderolithic bone-fissures (Schaub
1940). A reputed bird-yielding locality in East Germany is
Geiseltal, where fossils are occasionally even better preserved
than in Messel (Lambrecht 1935; Fischer 1962). Fissure-fillings
in the Schwabischen and Frankischen Alb in southern West
Germany also contain bird bones (Dehm 1935). Within the
Eocene North Sea area, fossil birds occur, outside England,
at various sites in Denmark and northern Germany (Hoch
1975; the lower Mo-clay and Clay with Tuff deposits with bird
fossils are now assigned to the Upper Paleocene, Hansen 1979).

GENERAL REMARKS ON THE FOSSIL
BIRD AND ITS CLASSIFICATION

The specimen considered here (Fig. 2) is an incomplete, par-
tially articulated skeleton of a small to medium size bird. It is
flattened in an oblique, dorsoventral direction, and is shown
from the ventral side in Figure 2. The dorsal side, the one
exposed when the fossil was excavated, is now embedded in
an artificial matrix (plast resin).

The right fore- and hindlimbs (Fig. 2, r.hu, r.fo, r.h) are
turned out and rest beside the body. The left hindlimb (l.h)
crosses over the abdominal area to lie alongside the right hind-
limb. Soft tissue evidently still remained when the carcass was
embedded in the lake sediments, and it determined the almost
natural position of the limb bones relative to the body. The
toe complex of the left foot (f) apparently became displaced
not too long after deposition, and is now located around the
distal end of the left tibiotarsus. No skull is preserved, which
seems to be fairly usual for bird fossils from the Messel site
(Franzen 1978:126). Other skeletal elements lacking or being
unidentifiable in the fossil include most of the pre- and post-
synsacral sections of the vertebral column, the ribs, the entire
left forelimb, and two toes of the right foot. Generally speak-
ing, the bones are fragmented and more or less incomplete,
and their ends, in particular, are poorly preserved. In some
places molds made of the embedding resin render the missing
bone parts in indistinct contours. The state of preservation of
the fossil seems similar to that described by Russell and Sige
(1970) for bats from the same locality.

36

Hoch: Middle Eocene Shorebird

Hoch: Middle Eocene Shorebird

37

In spite of its incompleteness, sufficient morphologic skeletal
characters can be observed in the fossil for an allocation of the
bird to taxonomic order. The concept presented by Bock
(1974:383) at a symposium on contemporary systematic phi-
losophies in 1973, advocating the presentation of the classifi-
cation in the introduction instead of at the end of taxonomic
papers, will be followed here: . . the reader [then] knows

what statements are available for disproof, what tests will be
attempted and hence why certain empirical evidence is being
presented.”

After comparison with relevant Recent and fossil birds, the
specimen is referred to the Charadriiformes, and to the sub-
order Charadrii therein. Decisive for this taxonomic allocation
is a suite of observable characters, which may or may not be
unique for the Charadrii, but which taken together occur only
within the group of birds traditionally regarded as shorebirds.
I am here following Strauch (1978:270) when he states, refer-
ring to works by D.H. Colless, that one is forced to start with
some sort of phenetic estimate of relationship as a beginning
of a phylogenetic study. The fossil’s state of preservation does
not permit sophisticated morphologic deductions. Rather, it
seems that the present paleontological work is one of those
where, as pointed out by Cracraft (1972a:384), the use of over-
all resemblance is inevitable.

The fossil bird is considered to be related to those early
Cenozoic “lapwings” and “coursors” that, according to Fjeldsa
(1976, pers. comm.), were specializing towards the sandgrouse/
dove complex. In these “basal waders,” the hindtoe had not
been reduced and specialized to a cursorial life, as is the case
in modern shorebirds. The Lutetian bird can be characterized
as a robust member of the Charadrii with specializations sim-
ulating the doves: “a perching shorebird.”

Features that refer the fossil to the Charadrii are:

1. A U-shaped furcula with a sturdy svmphyseal part (Fig.
2, fu; Fig. 3, fu). In general morphology this latter part resem-
bles those depicted by Strauch (1978: Fig. 22b, c) as typical of
shorebirds. In doves, the furcula is weak.

2. A long sternal plate (Fig. 2, from s along the length of
s.c), a good-sized carina with a distinctive pillar-like strength-
ening of the anterior edge (Fig. 2, ca, s.c; Fig. 3, ca, a.c), and
the morphology of the anterior articular area. In ventral view
the sternum shows, as exposed in the fossil, a fragmented
strong anterior medial “lip” (Fig. 3, v.m), believed to be the
base of a large ventral manubrial spine that overhangs the
coracoidal sulcus (Fig. 3, c.s). The observable part of the sul-
cus for the left coracoid suggests a structure that was unbarred
in lateroventral direction, and had a voluminous dorsal lip (see
Description), both features also in harmony with shorebird
morphology. No trace of a dorsal manubrial spine can be dis-
tinguished in the fossil, most probably because it was never
there, which would agree with conditions in the Charadrii. In
doves there are both ventral and dorsal manubrial spines, and
the ventral one is small and does not “overhang” the coracoidal
sulcus.

3. The relative proportions and the morphology of the wing
bones (Fig. 2, r.hu, r.fo), in particular those of the hand. Part
of the proximal end of the right humerus (Fig. 3, h.h, e.t, d.c)
is preserved in the fossil. In the right elbow joint area (as
determined by the position of the ulna and radius), there are
crushed bone material and contours that are difficult to distin-
guish in the figures; these in all probability indicate the location
of the distal end of the humerus. This will permit the statement
that the ulna-radius segment is the longest of the forelimbs,
with the humerus and hand segments about equally long and
somewhat shorter than the ulna and radius (see Measurements
below). Similar relative forelimb proportions are encountered
in a large number of bird taxa, but other taxa differ from the
common pattern, e.g., rails, where the humerus is longer, and
doves, where the hand is longer. In doves the humerus is
noticeably large and “swollen,” but it is by far the shorter of
the three mentioned segments. In the fossil, those traces of the
humerus that are preserved testify to a fairly “ordinary” shape
and to proportions corresponding to humeri in shorebirds (see
Description). The carpometacarpus (Fig. 4, cm), by its long
distal symphysis of metacarpals II and III, its fairly straight
metacarpal III, and other morphologic features, suggests re-
lationship with the Charadriiformes as well as with the An-
seriformes. The strong basal part of metacarpal I (unfortu-
nately its process is not preserved in the fossil), as pointed out
to me by Jon Fjeldsa, could very well indicate that this part
of the carpometacarpus had a spur-like process as is found in,
among others, many lapwings, and, as a potential preadap-
tation, throughout the plovers and allied groups. Such a char-
acter might not, however, exclude the Anseriformes from con-
sideration, among which Anseranas has spurs on metacarpal
I. The character may be primitive within the shorebirds,
ducks, and a few other groups, corresponding to their sup-
posed phvletic relationship and derivation from a common
stock (Fjeldsa pers. comm.). The observable morphology of
the proximal phalanx of digit II (Fig. 4, d. II) corresponds to
that in shorebirds such as Vanellus and Pluvialis. The element
is believed to have been unfenestrated (a small hole (x) in the
preserved lamellar bone in the fossil is a fracture), a state that
excludes the bird from the Lari (Lvdekker 1891), and in part
the Glareolidae, within the Charadriiformes. There is general
agreement that a non-fenestrated proximal phalanx of the
hand-digit II (Strauch 1978:314: “digit III” error for digit II)
is primitive in the Charadriiformes. The same element in doves
shows a certain morphologic similarity to that in shorebirds
but seems more “elaborated,” with, e.g., the internus indicis
process (corresponding to p, Fig. 4) much stronger than in
unspecialized shorebirds, as is also indicated by Stegmann
(1969:10 and Fig. 8). In doves the bone is also non-fenestrated.

The coracoids, pelvis, and long bones of the hindlimbs are
reminiscent of those in shorebirds, but may at first sight be
ascribed some dove traits. The coracoids (Fig. 2, l.c, r.c) are
fairly long and straight with a short external lip of the glenoid
facet (Fig. 3, g.l). Both in relative proportions and in detailed

Figure 2. Plumumida lutetialis new genus and species. Holotype. Middle Eocene oilshale at Messel, Hessen, West Germany. S.G.P.I. Kat. Nr.
2183, Hamburger Geologisches Institut (Section of Palaeontology). The specimen is mounted on a slab of plast resin with a thin sediment coating
ca, sternal carina; f, toe complex left foot; fu, furcula; l.c, left coracoid; l.h, left hindlimb; pe, pelvis; r.c, right coracoid; r.fo, right forelimb, sub-
elbow part; r.h, right hindlimb; r.hu, right humerus; s, sternum: anterior end; s.c, sternum: broken base of the sternal carina.

38

Hoch: Middle Eocene Shorebird

Figure 3. Shoulder girdle of the holotype of Plumumida lutetialis. a.c, anterior carinal margin of sternum; ar, articular markings at sternal edge
of coracoid; ca, sternal carina; c.h, coracohumeral surface; c.s, coracoidal sulcus; d.c, deltoid crest of right humerus; d.l, dorsal lip of coracoidal
sulcus; e.t, external tuberosity of right humerus; fu, furcula; g.f, glenoid facet; g.i, glenoid lip; h.h, humeral head; l.c, left coracoid; q, Pprocoracoid;
r.c, right coracoid; v.l, ventral lip of coracoidal sulcus; v.m, base of ventral manubrial spine; x-x, natural depression in coracoidal bone wall.

Figure 4. Right forelimb of the holotype of Plumumida lutetialis. cm, carpometacarpus; c.t, carpal tuberosity; cu, cuneiform; d.l, digit I; d. II,
digit II; i.c, internal condyle; l.p, ligamental prominence; p, internus indicis process; p.f, process pertaining to pollical facet; p.p, pisiform process;
ra, radius; sc, scapholunar; u, ulna; x, fracture in lamellar part of proximal phalanx of digit II.

Hoch: Middle Eocene Shorebird

39

40

Hoch: Middle Eocene Shorebird

morphology, as far as they can be discerned in the fossil, they
are closer to those in shorebirds, e.g., Charadrius, than to
those of doves (see Description). The pelvis (Fig. 2, pe) is a
robust, short and wide element with distinct and fairly large
iliac and ischiatic posterior processes. The robustness of the
pelvis corresponds to that of the bones of the fossil in general.
Robustness characterizes the pelvis, as well as the entire skel-
eton, of many doves, but such shorebirds as the oystercatchers,
which some authors (e.g., Fjeldsa pers. comm.) regard as
primitive among Recent Charadrii, also have fairly robust
skeletons. The visual impression of wideness of the fossil pelvis
is exaggerated because of the flattened condition of the speci-
men. The shape of the posterior pelvic processes are, in fact,
dove-like, as are some other details of the pelvis. The hind-
limbs are dove-like, although some fine morphological features
in the long hindlimb bones are shorebird-like. The relative
proportions of the bones and the shape of the foot show sim-
ilarity to the condition in doves.

The taxonomic status of the fossil at the family level has
been subject to considerable circumstantial consideration.
Judged from observed characters the bird does not fit into any
of the generally acknowledged shorebird or dove families, and
no living birds are found that exhibit intermediate shorebird-
dove morphology (Stegmann 1969). Some paleontologists pre-
fer the erection of a separate family for such a form, in order
to (1) indicate that the bird is an aberrant but valid member
of the order and suborder to which it is referred, and (2) secure
it from being consigned to oblivion by being placed incertae
sedis somewhere in the system when future reorganizers of the
avian hierarchy find it a burdensome misfit. Another point of
view is that such a monotypic bird family, based upon a single
incomplete fossil exhibiting aberrant features, will be a nui-
sance to anyone who wants to make practical use of the bird
system, and consequently that the bird should be included in
one of the existing families.

According to Fjeldsa (1976, pers. comm.) the early shore-
birds, from which a coursor/sandgrouse/dove line can be de-
rived, constituted the basic charadriiform group, together with
the stone curlews that are close to the lapwings (compare Cra-
craft 1972b). Fjeldsa groups some Recent Old World species
(generally regarded as lapwings), which “must be very close
to the basal late Cretaceous wader stock,” in the genus Xiphid-
iopterus, while even the living stone curlews, oystercatchers,
avocets, and certain others are “basal” in several aspects, al-
though each group has its own specializations. Fjeldsa’s system
is based on studies of Recent birds. It seems unjustifiable, on
the basis of present evidence, to refer the Lutetian bird to any
of Fjeldsa’s charadriiform taxa. The bird, by the structure of
its foot, is evidently different from known shorebirds and their
allies, the foot resembling that of doves. But a number of
clearly apomorphic traits of doves (and sandgrouse and cour-
sors) are missing. Reduction of the hindtoe is apparently ha-
bitually conditioned in plovers and other shorebirds: it tends
to disappear in cursorial inhabitants of arid plains, whereas
a small hindtoe remains in inhabitants of marshy environments
(Fjeldsa 1976). A well-developed hindtoe is considered a ple-
siomorphic character within the Charadriiformes, in accor-
dance with Strauch’s view (1978:320). The strong development
of the hindtoe in the Lutetian bird, as part of a perching foot
type, may be a plesiomorphic state. But rather, the whole foot
structure should be regarded a specialization, paralleling that

in doves, that sets the fossil bird apart from the Charadriidae
sensu Fjeldsa (including oystercatchers, stilts, and avocets).
The fossil bird will be placed incertae sedis in the suborder
Charadrii, a provisional status open to reconsideration when
further evidence becomes available through new finds of fossil
material. There are good possibilities for this, if paleontolog-
ical work can be continued at the Messel site.

SYSTEMATICS

Order Charadriiformes
Suborder Charadrii — Shorebirds
Family incertae sedis

Plumumida new genus

TYPE SPECIES: Plumumida lutetialis new species.

INCLUDED SPECIES: Type species only.

DIAGNOSIS: Shorebirds with a robust skeleton, the fore-
limbs longer than the hindlimbs. The coracoid has a long and
distinct glenoid facet with a glenoid lip that is short in basal
extension. The carpometacarpus has a distal metacarpal sym-
physis that is about one-fifth its length. The pelvis and hind-
limbs show features reminiscent of doves. The synsacrum has
the anterior part of the caudal section broad and flat ventrally,
with two pairs of ventral parapophvses. The foot is a strong
perching foot.

DISTRIBUTION: So far confined to the Middle Eocene
(Lutetian) of Germany (BRD).

ETYMOLOGY: Latin, pluma, feather; and Latin, (h)umida,
moist. The name alludes to the only true story we know of the
bird, i.e., that it was post-mortem soaked in the Eocene Messel
waters. The sound of the name may invoke the impression of
the plumpness that characterizes the fossil bird skeleton, but
it also conveys the feeling of the deep, dark mud that softly
veiled the dead bird on the lake bottom and held it for over
40 million years.

Plumumida lutetialis new species

HOLOTYPE: An incomplete, articulated, obliquely dor-
soventrally flattened and partially crushed skeleton now pre-
served on a slab of artificial resin. Hamburger Geologischen
Institut (Section of Paleontology), S.G.P.I. Kat. Nr. 2183; col-
lected by Mr. Hans-Peter Schierning, Hamburg, West Ger-
many, in 1972.

TYPE LOCALITY: The Messel oilshale pit, West Ger-
many. According to the finder (letter of 24 November 1978):
“Die Fundstelle liegt am westlichen Hang der Grube Messel
auf der vierten Sohle des friiheren Tagebaus. Die Schichten
sind nicht bestimmbar, da eine Festlegung wegen fehlender
Merkmale bisher nicht erfolgen konnte. Alle Ablagerungen
fallen stark nach Osten ein.” — “The fossil was found on the
western slope of the Messel pit at the fourth exploitation level
of the ancient open mine. The layers [with the fossil bird]
cannot be correlated with others in the pit because stratigraph-
ic indicators are lacking. All strata dip strongly towards the
east.”

DIAGNOSIS. As for genus.

ETYMOLOGY: Latin, Lutetia, Paris (that gave name to
the Middle Eocene stage, the Lutetian, of western Europe);

Hoch: Middle Eocene Shorebird

41

and Latin, -alls, belonging to. In reference to the time when
the bird lived.

MEASUREMENTS (in mm): Since all bones measured in
the fossil are in some state of fragmentation and/or incom-
pleteness, the measurements given are approximate. Shoulder
girdle and forelimb: coracoid (shortest distance from top of
bone to edge with sternal facet) 30; humerus (from top of head
to elbow joint as determined by the proximal ends of ulna and
radius) 58; ulna (maximum extension) 61; radius (maximum
extension) 59; carpometacarpus (maximum extension) 34; pha-
lanx 1 in digit II (length between metacarpal and digital facets)
15; total hand (from top of carpometacarpus to end of digit II)
58. Hindlimb: femur (see Description) 36; tibiotarsus (from
proximal articular surface to, and including, distal condyles)
56; tarsometatarsus (maximum extension) 36; phalanx 1 in dig-
it I 9.

DESCRIPTION: The bone terminology used is primarily
that of Howard (1929).

The Shoulder Girdle. In this section of the holotvpe (Fig. 3),
the furcula (fu) is seen displaced a little toward the right side
of the bird relative to the coracoids, so that the right and
median parts of the furcula are now resting on top of the right
coracoid (r.c). The left furcular branch is located alongside the
medial edge of the left coracoid (l.c). In the symphyseal region
lies the anterior and stronger part of the sternum, turned out
of its original position and partly crushed. The thickened an-
terior carinal margin (a.c) occupies a more or less transverse
position in the figure, corresponding to an overturning of the
carina toward the right side in the fossil. The left coracoidal
sulcus (c.s) is exposed and can be followed into the matrix
above the anterior part of the ventral lip (v.l). Mediad to the
latter is the broken base of the ventral manubrial spine (v.m),
as referred to above. The remains of the left dorsal lip (d.l)
show that this structure attained its largest size toward the
lateral termination of the coracoidal sulcus. The anteroventral
termination of the carina is difficult to determine. Lamellar
bone now adhering to the inner side of the deltoid crest (d.c)
of the right humerus and covering the area between this and
the anterior carinal margin is believed to belong to the carina
(ca). Its natural edge is not preserved. Most of the carinal base
(Fig. 2, s.c), including its triangular xiphial part (Fig. 2, lower
s.c; Fig. 6, s.c), can be seen. Lamellar bone alongside the
carinal base is either parts of the sternal plate or, as may be
the case along its right side, remains of the overturned carina.

The coracoids are strong elements. Compared with cora-
coids in Recent shorebirds they seem little specialized in their
sternal ends, where they show a similarity to the coracoids of
such forms as Haematopus, Pluvialis and Charadrius. As seen
in ventral view, their broader sternal part (not naturally de-
limited in the fossil in its present state of preservation) appears
to have been moderately rounded from side to side, with a
stronger rounding toward the medial edge of the bone. This
is visible on the left coracoid, whereas the right one has suf-
fered flattening in this area. The articular markings (Fig. 3,
ar), visible in the right coracoid, are relatively small and al-
most identical in shape with those in Haematopus ostralegus.
Toward the middle of the shaft, as seen in the left coracoid
(Fig. 3, x-x), is a natural depression, also a morphologic cor-
respondence with Haematopus ostralegus. An intermuscular
line extends along the length of the bone (distinct in the left
coracoid, Figs. 2 and 3). The humeral end of the left coracoid

is comparatively well preserved, whereas in the right coracoid
the humeral end is mainly in replica and shows no fine mor-
phologic details of the original bone. In the left coracoid the
glenoid lip (Fig. 3, g.l), of which only the base remains, is
short in basal extension, and the neck of the bone, which is
broken, diverges from the mainstem of the coracoid very close
to its upper termination. The coracohumeral surface (c.h, only
partially visible in the figure) is broad. The shortness of these
structures is in contrast to the remarkably long glenoid facet
(g.f) sharply set off from the surrounding bone wall. Bordering
it sternallv is a distinct, oblong-triangular sidewall of the cor-
acoid. The structure indicated “q” in Figure 3 may be the
distal part of the procoracoid.

The U-shaped furcula is very stout, or robust, judging from
the preserved right side of its middle part and the remains of
its branches. A ridge can be followed on the exposed surface
of the bone, extending to the lower part of the symphyseal
area. The ridge from right and left sides delimits a shallow
symphyseal depression in the anterior wall of the furcula. No
distinct furcular process can be observed.

The Forelimb. Part of the proximal end of the right humerus
can be discerned (Fig. 2, r.hu; Fig. 3, h.h, e.t, d.c). And of
the sub-elbow section (Fig. 2, r.fo; Fig. 4), the radius (Fig. 4,
ra), ulna (u), cuneiform (cu), scapholunar (sc), carpometacar-
pus (cm), and digit I (d.l) and digit II (d.II) are comparatively
well preserved. No trace is left of digit III.

The proximal end of the humerus is represented by the in-
complete head (Fig. 3, h.h), external tuberosity (e.t), and del-
toid crest (d.c). The external tuberosity is protruding, forming
a “corner” or “shoulder” of the bone; in this character, the
bone is similar to the humeri in shorebirds, but differs from
those in doves. The base of the deltoid crest, as preserved in
the fossil, appears broad and strong. It is fractured and ap-
parently artificially widened, thus seeming broader than it
originally was. It is about the same relative length as the del-
toid crest in shorebirds. Possible crushed remains and vague
imprints of the distal end of the humerus in the elbow area
have been mentioned above. No bone fragments or structures
in the area between the proximal and the presumed distal end
of the humerus can be reliably referred to that bone.

The radius (Fig. 4, ra) and ulna (u) are fairly complete in
outline, although their bone walls are fragmented and, espe-
cially in the radius, partially in replica. A fracture zone cuts
across the middle of the radius and ulna, and disturbs the
impression of the form of the bones. The radius was an almost
straight bone with a swollen ligamental prominence (l.p). The
ulna is crushed in its proximal end, where its limits can be
traced against the background only with difficulty. In the dis-
tal end of the ulna, the carpal tuberosity (e.t) and internal
condyle (i.c) are visible. The carpal tuberosity is a bulging
structure with a large terminal ligamental attachment. The
fossil ulna exhibits no papillar markings. The anconal papillae,
if present, will be situated beneath the plast matrix.

The two free carpals, the cuneiform (cu) and the scapholu-
nar (sc), are incompletely preserved. Compression of the spec-
imen has caused a combined crushing and plastic deformation;
where several bones were originally situated one on top of the
other, their individual surfaces and limitations may now be
difficult to trace in the fossil. This is evident in the wrist. The
cuneiform in situ in the living bird is a U-shaped bone that

42

Hoch: Middle Eocene Shorebird

Figure 5. Pelvis of the holotype of Plumumida lutetialis. ?c.v, Pcaudal vertebrae; d.p, dorsal parapophyses; f.h, head of right femur; fi, left
fibula; l.f, left femur; l.il, left ilium; l.ti, left tibiotarsus; m, replica structure representing posterior svnsacral vertebra(e); p. 1-3, lumbar par-
apophyses; r.il, right ilium; r.is, right ischium; r.pu, right pubis; 3.s, bases of third pair of single parapophyses in the caudal section of the synsacrum;
s.c, xiphial part of sternum with broken base of sternal carina; s.p, sternal plate; sy.v, anterior synsacral vertebra; t.v, thoracic vertebra; v.p,
ventral parapophyses; y, posterior iliac crest; z, interior pelvic ridge.

embraces part of the carpal trochlea, with one branch lying on
the underside of the wing. This branch can be seen in the
fossil, a little distal to its original position because of a 90
degree turning over of the bone. The other branch, beneath
part of the carpal trochlea, supposedly together with the ex-
ternal condyle of the ulna, has been pressed into the trochlear
bone material, causing a swelling proximal to the visible
branch of the cuneiform. The deformed trochlear bone mate-

rial is difficult to delimit from the fractured part of the cunei-
form bone. Where unfragmented, the cuneiform shows a bulky
shape with a median furrow, clearly visible in Figure 4, where
may have lodged, judging from conditions in Haematopus os-
tralegus, the tendon that passes along the length of metacarpal
II and fastens at the medial edge of digit II, together with (or
forming part of) the exterior indicus longus tendon.

The carpometacarpus (cm), although easily identifiable in

Hoch: Middle Eocene Shorebird

43

the fossil, is not too well preserved with respect to details. The
trochlear section, as stated above, has suffered a good deal of
deformation; the terminal parts of ridges and processes are
lacking, as can be observed in the trochlear ridge, the process
of metacarpal I, the pisiform process (p.p) and the process
pertaining to the pollical facet (p.f). The shaft of metacarpal

II is fractured, with part of the bone wall lacking or covered
by plast matrix. The proximal part of the shaft of metacarpal

III is in replica. And, in the distal end of the carpometacarpus,
part of the bone wall is lost, and the tuberosity of metacarpal
II is either lost or, as it appears, is covered by plast. The distal
metacarpal symphysis is fairly long, surpassing in length that
in some Recent shorebirds, including Haematopus ostralegus
and Charadrius hiaticula. Scolopax rusticola has a similarly
long distal metacarpal symphysis that might indicate that this
condition is a specialized character within the shorebirds. In
doves the distal metacarpal symphysis is short.

Digit I (d.I), the pollex, is recognizable in the fossil, but is
incomplete. It has not been possible to free its distal end (if
preserved) from the resin.

Both phalanges of the second digit (d. II) are relatively well
preserved. The proximal phalanx, as mentioned earlier, is of
some diagnostic importance within the Charadriiformes. The
medial part of the bone is strong and forms the axis of the
element, with strong and elaborate metacarpal and digital fac-
ets. The lateral part consists of a thinner lamellar section,
laterally bordered by a thicker edge or rim. In the fossil, the
proximal lateral part of the bone is lacking. Distallv, the lateral
bone section terminates in the internus indicis process (p) that
in the living bird was connected by ligamental tissue to the
distal end of the second phalanx, thus, together with the latter,
forming a good support for the base of the distal remex. The
chevron-formed mark close to the interphalangeal joint was
the base for a strong tendon to the proximal edge of phalanx
2. This bone in the fossil is partly covered by resin and cannot
be seen in its full extent. Remiges were also fastened to the
upper side of phalanx 1 (and to the first finger, carpometacar-
pus, and ulna). Its elaborated and strong articular ends, well
developed tendon mark, and entire morphology, reminiscent
of that in shorebirds, together with the morphology of the
other preserved bones of the wing and shoulder girdle, indicate
that Plumumida lutetialis was a capable flyer.

A “stray” bone fragment occupies the original position of
the third finger.

The Pelvic Girdle. A good deal of the pelvis (Fig. 5), including
the synsacral (sy.v- — m) and post-acetabular right portions (r.il,
r.is, r.pu), is comparatively well preserved.

In the synsacrum, the sacral section is covered by part of
the shaft of the left femur (l.f). A structure regarded to be the
broken base of the posterior right parapophysis (p.l) of the
lumbar section can be identified anterior to it (above it in the
figure), together with the bases of the second-to-last (p. 2 ) and
third-to-last (p. 3) right lumbar parapophvses. The bone wall
of the right side of the corresponding part of the synsacral
body is preserved, although fragmented, whereas most of the
wall in the left side is in replica. The structure marked “sy.v”
is considered to be the anterior synsacral vertebral element (or
it may be the posterior free thoracic vertebra). In spite of frag-
mentation, its posterior limitation can be perceived; modern
birds may also have a terminal marking of the first synsacral
vertebra. Between this and the three lumbar synsacral ele-

ments, represented by their right parapophyses, are the re-
mains of still another element. Thus intepreted, the number
of presacral vertebrae included in the synsacrum is five. A
fairly well preserved vertebral body (t.v) is closely attached to
the anterior end of the structure marked “sy.v,” but not fused
with it as is seen from its position a little out of line with the
synsacral axis, and from the presence of terminal bone walls
in the intervertebral joint area. It has preserved a large right
anterior projection for the support of the rib.

The pelvis in birds exhibits some individual variation in
morphologic details (see also Boas 1933). Thus the number of
synsacral thoracic vertebrae may vary within a species. One
available pelvis of Haematopus ostralegus has five fused pre-
sacral vertebral elements, whereas another pelvis of the same
species has four fused presacral elements and one vertebra
whose body is not fused with the synsacrum, but whose trans-
verse processes and ribs support the anterior parts of the ilia.
The vertebral body structure marked “sy.v” in Figure 5 may
not have been completely fused with the axial structures be-
hind it, but its position suggests that it was part of the pelvis.
Fragments of lamellar bone (s.p) situated along the right side
of the axial structure in the fossil are most probably parts of
the sternal plate and do not belong to the ilium, thus giving
no indication of the anterior extension of the pelvis. The mor-
phology of the presacral part of the synsacrum, as far as it can
be observed, is reminiscent of that in shorebirds, although the
elements are more robust in the fossil than in shorebirds. As
stated for Haematopus ostralegus, the synsacrum in modern
shorebirds may include five presacral vertebral elements. In
some shorebirds, such as Vanellus vanellus and Calidris al-
pina, the number is generally four. The corresponding part of
the synsacrum in doves is shorter and more compact than that
in shorebirds, with only three, or in some cases four, included
vertebrae.

Posterior to the superimposed left femur is the caudal section
of the synsacrum (sensu Howard 1929). Anteriorly, in the fos-
sil’s right side, remains of two pairs of dorsal (d.p) and ventral
(right v.p) parapophyses, and in the left side, the undisturbed
proximal parts of two ventral parapophyses (left v.p), can be
distinguished. The presence of both dorsal and ventral par-
apophyses shows that the corresponding two vertebral elements
are the first and second synsacral caudal vertebrae. The num-
ber of ventral parapophyses in this region in birds is also sub-
ject to some intraspecific variation, so that, e.g., some mem-
bers of a species that generally has only one pair of ventral
parapophyses may have two pairs, or one pair and a right or
left ventral parapophysis in front or behind it. The two pairs
of ventral parapophyses in the fossil were strong structures,
which indicates that they represent a normal condition of two
pairs for this species. Fragmented remains of their distal ends
in the right side show that the two ventral parapophyses of
each side converged distallv, where, as known from Recent
birds, they joined the inside of the ilium, thus forming struts
from the synsacral body to the acetabular region. The ventral
wall of the anterior part of the caudal section of the synsacral
body is broad and flat. It is only slightly disturbed in the fossil
and thus gives a fairly correct idea of the morphology of that
part in the living bird also. Posteriorly, remains of three pairs
of “single” (i.e., not “split into” dorsal and ventral) par-
apophyses can be seen. No synsacral bone material was pre-
served posterior to the third pair of single parapophyses (3 . s),

44

Hoch: Middle Eocene Shorebird

and an indistinct replica contour (m) gives no morphologic
details of the skeletal structure. But, judging from the shape
of the neighboring part of the right ilium, which in the living
bird was attached to the svnsacrum, hardly more than one
posterior vertebral element is lacking in the synsacrum. This
would make six the number of caudal vertebrae included in
the synsacrum. Fragmented bone walls remain in the two
hindmost preserved interparapophvseal spaces (right side in
the fossil), showing that these spaces in all probability were
closed or nearly closed in the living bird. This also seems to
have been the case in the most anteriorly preserved interpar-
apophyseal space (between the double parapophvses). Matrix
fills out the “background” in the intervening space, but la-
mellar bone fragments in the left side may also indicate the
original presence of a bone wall in this interparapophyseal
space. A structure in the fossil (?c.v), discernible in Figure 5
below and to the left of the caudal section of the synsacrum,
is presumed to be the remains of the caudal vertebrae.

Most modern shorebirds have one pair of strong synsacral
struts in the acetabular region of the pelvis, although two pairs
of struts may occur. Doves have one, rarely two, pair(s) of
comparatively weak synsacral struts. The number of synsacral
vertebral elements posterior to the struts is generally four to
five in shorebirds and five to six in doves. The interparapoph-
yseal bone walls are usually fenestrated in the shorebirds,
whereas in doves they tend to be closed.

The ilium, ischium, and pubis of each side are fused in
Plumumida lutetialis, as are the corresponding bones in Re-
cent birds. This leaves an ilio-ischiatic fenestra and a posterior
incisure between the ilium and ischium, and an ischio-pubic
fenestra or incisure between the ischium and pubis. Only frag-
ments of the ilium (l.il) occur in the left side of the fossil. In
the right side, because of compression, there is a large opening
between the posterior part of the ilium and the synsacrum,
which in many birds, including the shorebirds and (all?) doves,
are not completely fused. The ilio-ischiatic fenestra has been
closed by the ischium being pressed onto the lateral edge of
the ilium (the presence of plast matrix obscures the details
here). Most of the bone wall of the postacetabular part of the
ilium is preserved (although fragmented), including a large
posterior iliac process that in the present flattened condition
exhibits transverse ripples, testifying to a certain degree of
introflexion of the process in the living bird. The acetabulum
has been protruded by the right femoral head (f.h), and the
proximal parts of the right femur and of the left tibiotarsus
(l.ti) and fibula (fi) have been pressed into the pelvic bone in
the acetabular region causing fragmentation. The iliac bone
wall between the distal ends of the right synsacral struts and
the acetabulum is much disturbed and gives no information
as to the original position of the struts relative to the acetab-
ulum (see Strauch 1978:315). The interior pelvic ridge (z) has
also been damaged in the acetabular region. The postacetab-
ular part of this ridge, as far as it is preserved, is a strong and
conspicuous structure in the fossil. It is almost straight, and
continues posteriorly, where it is now broken, into the poste-
rior ischial process. A bone structure beneath the posterior
ridge fracture in the fossil may be the displaced fractured end
of this process. Remains of a posterior iliac crest (y), which is
the posterior limitation of the renal depression, can be ob-
served connecting with the medial wall of the interior pelvic
ridge. Along the lateral side of this latter structure is a fairly

large lamellar bone wall of the ischium, somewhat disturbed
by the xiphial part of the sternum (s.c). The preserved section
of the right pubis (r.pu) indicates by its strength that the pubes
were well developed, and apparently extended a good distance
behind the medial part of the pelvis.

In many pelvic features Plumumida lutetialis shows good
correspondence with conditions in doves. The innominate
bone, as far as it can be studied, is close in morphology to that
bone in, e.g., Columba loricata. The synsacrum, by its broad
ventral wall of the anterior caudal section and apparently
closed interparapophyseal walls, is reminiscent of the synsa-
crum in, e.g., Columba palumbus. The presence of two pairs
of strong synsacral struts in the acetabular region is, however,
atypical of doves. In some modern columbiforms the pair of
main synsacral struts extends from the vertebral element an-
terior to that corresponding to the main strut element in shore-
birds. Following Strauch (1978:314), the condition in these
doves is the relatively derived of the two. He states (ibid.)
that, in the most widely distributed and presumably primitive
state for the Charadriiformes, the pair of main struts arises
from the fifth vertebra from the posterior end of the synsa-
crum. The interpretation given above of the number of caudal
synsacral elements in the fossil would permit the conclusion
that the two pairs of synsacral struts in Plumumida lutetialis
correspond to a combination of those of the shorebirds and
those of the advanced doves. Thus, in this particular feature,
Plumumida lutetialis may represent a form intermediate be-
tween shorebirds and advanced doves. Another interpretation
is that both pairs of struts that may occur in specimens of
modern shorebirds were strongly developed in Plumumida lu-
tetialis as a consequence of the general robustness of the pelvis,
which might also be said for other features that give the
impression of strength. Recent shorebirds may have more than
four, and doves more than five, caudal synsacral elements
posterior to the strut element, suggesting that the number of
caudal vertebrae included in the synsacrum is of limited di-
agnostic significance. Unfortunately, the sacral section, which
would permit a correct determination of the homology of the
vertebral elements in this region, is covered in the holotype.

The Hindlimbs. The greater portions of both hindlimbs (Figs.
5 and 6) are visible in the fossil, but their state of preservation
permits few fine details to be studied. The right hindlimb (Fig.
6) is positioned in articulation with the pelvis, with the femoral
head (f.h) located in the acetabulum as described above. Sec-
tions of the right femur (r.f) are seen as bone fragments oi
impressions in the matrix to the left of the superimposed left
tibiotarsus (l.ti). The distal condylar part of the femur is
crushed into more or less complete fusion with the crushed
proximal articular part of the right tibiotarsus (r.ti). The femur
length of 36 mm is measured between the visible extremes of
the femur in the fossil, indicated in Figure 6 as the end points
of the white lines “r.f.” The original femur may have been
slightly longer. In the proximal part of the right tibiotarsus,
which is seen in anterior view, a distinct mark for the attach-
ment of the M. flexor cruris medialis (f.a) can be observed. A
ridge (on various fragments of bone), the intermuscular line
(i.m), terminates proximally in the base of the (broken) inner
cnemial crest (c.c). Because of the serious crushing of the prox-
imal end of the tibiotarsus, the proximal continuation of the
inner cnemial crest cannot be reliably indicated. A replica con-

Hoch: Middle Eocene Shorebird

45

Figure 6. Hindlimbs of the holotype of Plumumida lutetialis. c.c,
inner cnemial crest; e.o, external oblique ligament apophysis; f.a, flexor
cruris medialis attachment; f.h, femoral head; fi, fibula; i.m, intermus-
cular line; 1.1, left hindtoe; l.e, linea extensoris; l.f, left femur; l.ta, left
tarsometatarsus; l.ti, left tibiotarsus; mt, metatarsal I; o.c, outer
cnemial crest (in the right tibiotarsus the interpretation is based upon a
vague replica structure); r.I, right hindtoe; r.f, right femur; r.t, right toe;
r.ta, right tarsometatarsus; r.ti, right tibiotarsus; s.b, supratendinal
bridge; t.II-III, trochleae II and III; II— III— IV, left toes II, III, and IV.

46

Hoch: Middle Eocene Shorebird

tour is tentatively interpreted as the imprint of the outer cne-
mial crest (?o.c). The shaft of the right tibiotarsus has under-
gone a special fragmentation because of lateral pressure from
bones of the left hindlimb, resulting in a slight bending and
some extension in a zone close to the present location of the
proximal end of the left tarsometatarsus (l.ta), where little bone
material is preserved. Part of the original bone wall remains
near the distal end of the tibiotarsus, but around the intertarsal
joint, most morphological details are obscured by intense
crushing of the opposing ends of the tibiotarsus and the tar-
sometatarsus. One bone structure, the supratendinal bridge
(s.b), appears strangely unaffected by the crushing. Rounded
structures (set off in the figure) apparently represent the con-
dyles. No trace of a right fibula can be seen. The remains of
the right tarsometatarsus (r.ta) show an element with a wide
median furrow. Some bone material is preserved, but most of
the element appears in replica. Distally, the trochleae have not
been preserved. Their termination was originally located a
little distal to the base of the hallux (r.I). A fracture now occurs
in the tarsometatarsus above the hallux articulation. The hal-
lux, also in replica, has been determined on the basis of the
presence of a small body (mt) interpreted as metatarsal I, and
on the structure of the toe that seems to consist of one larger
element and a curved and pointed distal segment, correspond-
ing to the first and the ungual phalanges, respectively. The
terminal structure (r.t), ending in a “claw,” is also a toe replica,
but the number of the toe cannot be determined since no re-
liable traces of articulation can be discerned.

In the left hindlimb (Figs. 5 and 6), the femur (l.f) is incom-
pletely preserved, with the proximal part lacking, and the bone
wall in the remaining part strongly fragmented. The left tib-
iotarsus (l.ti) is seen in its full length. Proximally, the base of
the inner cnemial crest (c.c) can be traced, together with parts
of the bone wall, medial and lateral to the crest. A broken
structure (o.c) represents the outer cnemial crest. The bases of
these two crests extend about similar lengths distad in the
tibiotarsus. An intermuscular line (i.m) can be followed as a
distal continuation of the inner cnemial crest for some length
of the bone. The middle part of the shaft is much fragmented,
with pieces of the bone wall lacking. Towards the distal end
of the bone, a very clear linea extensoris (he) can be traced in
three fragments, terminating distally in the broken internal
oblique ligament apophysis. The external oblique ligament
apophysis (e.o) is preserved in the fossil, as well as the supra-
tendinal bridge (s.b). The distal condyles, although incom-
plete, are recognizable, as is the condylar fossa between them.
Both condyles have suffered some deformation because of
pressure against the surrounding bones. Sections of the fibula
(fi) are preserved in what appears to be the complete original
extension of the bone. Its proximal “head” is rather large.

The left tarsometatarsus (l.ta), in its exposed anterior side,
has much bone material preserved. Proximally, the external
cotvla has become somewhat deformed by the external condyle
of the tibiotarsus being pressed into it. The internal cotvla is
little disturbed. The depth and length of the anterior median
furrow was artificially increased during the fragmentation of
the bone by the left lateral wall being turned into the furrow.
There was probably a median furrow in the upper part of the
undisturbed tarsometatarsus, perhaps as wide as that seen in
the remains of the right tarsometatarsus. But in the distal end
of the bone, metatarsal III protrudes, thus making the front

side of the bone convex. The remains of trochlea II (t. II) and
trochlea III (t. Ill) show that the latter was the distalmost of
the two, and presumably of all three trochleae. Trochlea IV
is not preserved. No digits are found in natural articulation
with the left tarsometatarsus, but partially articulated toe
bones located around the distal end of the left tibiotarsus are
considered to belong to the left foot. A long bone (left portion
of 1.1), situated transversely between the right and left tibio-
tarsi, is believed to be the proximal phalanx of the left hallux,
corresponding in size to that in the right hallux (r.I). Its ungual
phalanx may be the one exposing its hemispheric flexor tuber-
cle (1.1), with a central foramen, close to the lateral side of the
left tibiotarsus and pointing its distal tip along this bone in its
proximal direction. Below it are two articulated toe structures
that show strong, curved, and pointed ungual phalanges with
transversely rounded undersides and deep lateral furrows.
Four phalanges, of which the two middle ones are compara-
tively short, can be seen in one, and, in the other, two pha-
langes and the ungual phalanx are visible. In birds, four bones
may occur in toes III and IV, and three bones may also occur
in toe II. The interpretation of the two articulated toes shown
in Figure 6 as toes II and IV, and the free phalanx representing
toe III (or with numbers II and III interchanged) would cor-
respond to conditions in, among others, doves.

No morphologic peculiarities of diagnostic interest can be
studied in the fossil femora. The tibiae exhibit features that
show similarity to the shorebirds and, for some of them, dis-
tinguish the fossil from the doves. In shorebirds, the M. flexor
cruris medialis attachment is primarily a distinct oblong mark-
ing situated on the bone as it is in the fossil, whereas in doves
the marking is different in shape, and is often more diffuse in
outline than in shorebirds. Some shorebirds, such as Vanellus
vanellus and Calidris alpina, have the bases of the inner and
outer cnemial crests about equally long; this may, however,
also be the case in some doves. The internal oblique ligament
apophysis is in Plumumida lutetialis and Recent shorebirds
situated rather closely above the level of the supratendinal
bridge; in doves it is located farther proximad on the shaft of
the bone. It appears in the fossil that the external condyle of
the tibiotarsus is larger than the internal condyle; such is the
case in shorebirds, but not in doves, where the internal condyle
tends to be the larger.

The foot in Plumumida lutetialis is reminiscent of the foot
in doves. Characters encountered in a moderately advanced
perching foot include a relatively short and strong tarsometa-
tarsus, and robust digits consisting of a long hallux and, as it
also appears, three anterior toes of medium lengths, all with
strong, curved, and pointed (although not to the extent seen
in passeriforms or birds of prey) claws. Specialized perchers,
as the Passeriformes, have all three trochleae of the tarso-
metatarsus in one line, whereas groundbirds, as the Charadrii,
have trochlea III longer and more anteriorly protruding than
trochleae II and IV (compare Stegmann 1969:25 ff). The dove
foot seems intermediate in morphology and evolutionary stage
between that in groundbirds of the shorebird type and the
advanced perching foot.

The tarsometatarsus in Recent doves has no well developed
anterior median furrow as has the tarsometatarsus in Recent
shorebirds and, apparently, in Plumumida lutetialis. In doves
the tarsometatarsus is shorter than the femur, whereas in
shorebirds it is longer, in some cases much longer, than the

Hoch: Middle Eocene Shorebird

47

femur. In Plumumida lutetialis the two bones are of about
equal length. As stated above, Recent shorebirds exhibit var-
ious degrees of reduction of the hallux, showing that the ple-
siomorphic state in this group of birds is the presence of a
hallux. In skeletal morphology and proportions the foot of
Plumumida lutetialis appears to be intermediate between that
of shorebirds and doves.

CONCLUSIONS

There seems to be fairly wide agreement that the Colum-
biformes developed from early Charadriiformes, perhaps dur-
ing late Cretaceous times (for various elaborations of the idea,
see cited works by Fjeldsa and Stegmann, and works referred
to therein). It is tempting, on the basis of the above descrip-
tion, to see in Plumumida lutetialis a form “on the line” from
early shorebirds to doves. Only one pre-Lutetian columbiform,
Microena goodwini Harrison and Walker 1977, described from
the British Lower Eocene on the basis of a left tarsometatarsus
lacking the trochlea for the 4th digit, has been recorded. Shore-
birds are known as far back as the late Cretaceous (see Brod-
korb 1967 and later works such as Brunet 1970 and Harrison
and Walker 1976, 1977). Rhynchae'ites messelensis, stated by
Wittich (1898:144) to be intermediate between shorebirds and
rails, is referred to the Scolopacidae by Brodkorb (1967) (on
the basis of Wittich’s investigations), an allocation that is,
however, questioned by some workers.

Most features in the fossil specimen used to refer it to the
shorebirds are believed to be plesiomorphic characters within
the shorebird complex (i. e. , the shorebirds and groups derived
from them). A perching foot is a relatively apomorphic char-
acter therein.

Perching birds are found in various taxa of typically non-
perching groups, such as the Cairinini within the Anseri-
formes, and Anous spp. within the Lari. In both the perching
ducks and geese and the noddies, the hindtoe is markedly long
and the claws are more robust than in other ducks and terns.
The whole foot, however, is not very different from the foot
in their close relatives. Yet it is said about the young of these
birds that they have very strong claws and are good at climb-
ing. Very advanced perching feet are encountered in the Pas-
seriformes, less advanced in the Columbiformes. A perching
foot with corresponding structural modifications in the hind-
limb and pelvic girdle is a specialization that evidently devel-
oped more than once in bird history.

One feature in the fossil, the short glenoid lip-long glenoid
facet structure in the coracoid, does not have a morphologic
parallel in the investigated shorebirds and doves. Although
encountered in other bird taxa, it seems to be unique for Plu-
mumida lutetialis among known shorebirds and doves. The
character is considered an autapomorphv for the taxon to
which the fossil belongs.

The long distal metacarpal symphysis distinguishes Plu-
mumida both from doves and from those shorebirds that are
generally regarded as unspecialized. Specialized shorebirds,
such as Scolopax, also have a long symphysis of the distal ends
of metacarpals II and III. The character, presumably, is rel-
atively apomorphic within the shorebirds, but is encountered
within other bird taxa also.

Formally, following the above given statements, Plumu-
mida lutetialis should be placed within the shorebird-complex

as a separate taxon (because of autapomorphv) with sister-
group status relative to either (1) the doves (svnapomorphy:
the perching foot), or (2) the Scolopax group (svnapomorphy:
the long distal metacarpal symphysis). For the present, how-
ever, acknowledging the incompleteness of the fossil, Plumu-
mida lutetialis will be placed incertae sedis in the Charadrii.

ACKNOWLEDGMENTS

I express my gratitude to the collector of the fossil, HP.
Schierning, Hamburg, who permitted me to study and de-
scribe the specimen, and to J.L. Franzen and D.S. Peters of
the Forschungsinstitut Senckenberg, and O. Feist, Miihltal,
who readily undertook correspondence and co-work with me
on the Messel fossils.

The following individuals are also heartily thanked: Joel
Cracraft for inspiring me to undertake the present work. Jon
Fjeldsa who discussed his ideas of charadriiform phvlogenv
with me, and let me read an unpublished manuscript. C.A.
Walker of the British Museum (Natural History) for showing
me the collections of fossil birds, and L. Ginsburg of the Mu-
seum National d’Histoire Naturelle for placing fossil and Re-
cent bird material at my disposal. G.S. Cowles and C.J.O.
Harrison for permitting me to study Recent bird material at
the British Museum Ornithological Department, Tring, and
J. Fjeldsa and U. Mphl for doing the same at the Zoological
Museum of Copenhagen. C.J.O. Harrison is thanked for his
kind interest in discussing my ideas of the Messel bird. S.E.
Bendix-Almgreen, Niels Bonde, Ulrik Mphl, and Kaj Strand
Petersen have critically commented on the text. Bente Bang,
O. Bang Berthelsen, and P Nielsen carried out the photo-
graphic work connected with the study of the fossil, and Chr.
Rasmussen carefully made the important details of the fossil
stand out clearly in the illustrations. John Bailey helped with
the English language, and Annemarie K. Brantsen carefully
typed the manuscript.

A grant from Statens Naturvidenskabelige Forskningsrad
(J. Nr. 511-10475) allowed me to visit the British and French
museums.

LITERATURE CITED

Berg, D.E. 1965. Nachweis des Riesenlaufvogels Diatryma
im Eozan von Messel bei Darmstadt/Hessen. Notizbl.
hess. L.-Amt Bodenforsch 93:68-7 2.

Berggren, Wt.A. 1972. A Cenozoic time-scale — some im-
plications for regional geology and paleobiogeography.
Lethaia 5:195-215.

Berggren, W.A., M.C. McKenna, J Hardenbol, and
J.D. Obradovich. 1978. Revised Paleogene polarity
time scale. J. Geol. 86:67-81.

Boas, J.E.V. 1933. Kreuzbein, Becken und Plexus lumbo-
sacralis der Vogel. Danske Videnskabsselsk. Skrifter,
Naturvid. Ser., 9. Rk. 5(1): 1—77.

Bock, W.J. 1974. Philosophical Foundations of Classical
Evolutionary Classification. Syst. Zool. 22:375-392.
Bond, G. 1978. Evidence for Late Tertiary uplift of Africa
relative to North America, South America, Australia and
Europe. J. Geol. 86:47-65.

Bornhardt, J.F. 1975. Neue Fossilfunde aus der Grube
Messel und ihre Preparation. Aufschluss 12:453-473.
Brodkorb, P. 1967. Catalogue of fossil birds: Part 3 (Ralli-

48

Hoch: Middle Eocene Shorebird

formes, Ichthyornithiformes, Charadriiformes). Bull.
Florida State Mus., Biol. Sci. 11:99-220.

Brunet, J. 1970. Oiseaux de l’eocene superieur du Bassin de
Paris. Arm. Paleont. (Vertebres) 5 6( 1 ): 1—5 7 .

Buchardt, B. 1978. Oxygen isotope palaeotemperatures
from the Tertiary period in the North Sea area. Nature
275:121-123.

Collins, C.T. 1976. Two New Species of Aegialornis from
France, with Comments on the Ordinal Affinities of the
Aegialornithidae. Smithson. Contr. Paleobiology 27:12 1-
127.

Cracraft, J. 1972a. The relationships of the higher taxa of
birds: Problems in phylogenetic reasoning. Condor
74(4):3 7 9 — 392 .

. 1972b. A New Cretaceous Charadriiform Family.

Auk 89( 1 ):36— 46.

Dehm, R 1935. Uber tertiare Spaltenfiillungen im Frank-
ischen und Schwabischen Jura. Abh. baver. Akad. Wiss.
29:1-86.

Dughi, R., and F. Sirugue. 1962. Distribution verticale
des oeufs d’oiseaux fossiles de 1’Eocene de Basse-Pro-
vence. Bull. Soc. geol. Fr., 7C ser. 4:69-78.

Feist, O. 1978. Fossilien der Grube Messel. — Eigenverlag,
Katalog von 40 verfiigbaren Fossil-Kopien, 37 Abbildun-
gen in Fotokopien Din A4 mit kurzen Untertiteln. Miihl-
tal.

Fischer, K. 1967. Neufunde von Tapiroideen und Diatryma.
Ber. deutsch. Ges. geol. Wiss., A, Geol. Palaont.
12(6):655.

. 1978. Neue Reste des Riesenlaufvogels Diatryma

aus dem Eozan des Geiseltales bei Halle (DDR). Mitt.
Zool. Mus. Berlin, Suppl. Band 54, Ann. Orn. 2:133-144.

FjeldsA, J. 1976. The systematic affinities of sandgrouses,
Pteroclididae. Vidensk. Meddr dansk naturh Foren.
139:179-246.

Franzen, J.L. 1976a. Die Fossilfundstelle Messel Ihre Be-
deutung fur die palaontologische Wissenschaft. Natur-
wissenschaften 63:418-425.

— . 1976b. Senckenbergs Grabungskampagne 1975 in

Messel: Erste Ergebnisse und Ausblick. Natur u. Museum
1 06(7 ): 2 1 7— 2 23 .

. 1977. Urpferdchen und Krokodile. Messel vor 50

Millionen Jahren. Kl. Senckenberg-Reihe Nr. 7:1-36.

. 1978. Senckenberg-Grabungen in der Grube Messel

bei Darmstadt. 1. Problem, Methoden, Ergebnisse 1 976—
77. Cour. Forsch.-Inst. Senckenberg 27:1-135.

. 1979. Senckenberg-Grabungen in der Grube Messel

bei Darmstadt. 2. Ergebnisse 1978. Cour. Forsch.-Inst.
Senckenberg 36:1-144.

Gaillard, C. 1936. Un oiseau geant dans les depots eocenes
du Mont-d’Or Lyonnais. Annls. Soc. Linn. Lyon 80:111-

126.

. 1939. Contribution a l’etude des Oiseaux fossiles.

Archs. Mus. Hist. nat. Lyon 15,11:1-100.

Hansen, J.M. 1979. Age of the Mo-Clav Formation. Bull,
geol Soc. Denmark 27:89-91.

Harrison, C.J.O. 1971. Flamingo (Phoenicopteridae) re-
mains from the British Upper Eocene. Bull. Br. Orn.
Club 91:36-39.

Harrison, C.J.O. , and C.A. Walker. 1976. Birds of the
British Upper Eocene. J. Linn. Soc. (Zool.) 59:323-351.

. 1977. Birds of the British Lower Eocene. Tertiary

Res. Sp. Pap. 3:1-52.

Haupt, O. 1925. Die Palaohippiden der eocanen
SiiBwasserablagerungen von Messel bei Darmstadt. Abh.
hess. geol. Landesamt. 6:239-398.

Heissig, K. 1979. Die hvpothetische Rolle Siideuropas bei
den Saugetierwanderungen im Eozan und Oligozan. —
Hypothetical land connections in south east Europe and
the Eocene and Oligocene migrations of mammals. N. Jb.
Geol. Palaont. Mh.:83-96.

Hoch, E. 1975. Amniote remnants from the eastern part of
the Lower Eocene North Sea basin. Colloques int. Cent,
natn. Rech scient., no. 218. Problemes actuels de pa-
leontologie — Evolution des Vertebres. :543— 562 .

Howard, H. 1929. The Avifauna of Emeryville Shellmound.
Univ. California Publ. Zool. 32:301-394.

Illies, J.H. 1978. Two stages Rhinegraben rifting. Pp. 63-
71 in Tectonics and Geophysics of Continental Rifts (I.B.
Ramberg and E.R. Neumann, Eds.). Dordrecht, Holland.

Illies, J.H., and G. Greiner. 1978. Rhinegraben and the
Alpine system. Bull. Geol. Soc. Am. 89:770-782.

Koenigswald, W.v. 1979. Die Erforschung der Messeler
Schichten. Pp. 27-38 in Fossilien der Messeler Schichten
(R Heil, W.v. Koenigswald and H.E. Lippmann, Eds.).
Darmstadt.

Kowalevsky, W. 1873-74. Monographic der Gattung An-
thracotherium Cuv. und Versuch einer naturlichen Clas-
sifikation der fossilen Hufthiere. Palaeontographica.

35(3/4):25 1-252 .

Kuhne, W.G. 1961. Preparation von flachen Wirbeltierfos-
silien auf kiinstlicher Matrix. Palaontol. Z. 35(3/4): 2 51 —
252, 1 pi.

. 1962. Praparation von Wirbeltierfossilien aus kolloi-

dalem Gestein. Ein Beheiter fur kleine Fossilien. Palaon-
tol. Z. 36(3/4):285— 286.

Lambrecht, K. 1935. Drei neue Vogelformen aus dem Lute-
tian des Geiseltales. Nova Acta Leopoldina (N.S.) 3:361-
367.

Louis, P. 1969. Les environs de Reims a l’Eocene Inferieur.
Soc. Et. sci. nat. Reims 136:3-11.

Lydekker, R. 1891. Catalogue of the Fossil Birds in the
British Museum (Natural History). London.

Mourer-Chauvire, C. 1980. The Archaeotrogonidae of the
Eocene and Oligocene Phosphorites du Quercy (France)
(this vol.).

Nemejc, F. 1970. The Origin of the European Vegetation
and its main Alterations during the Tertiary. Palaont.
Abh. B, III (3/4):553— 562 .

Plaziat, J.-C. 1964. Piste d’Oiseaux et remaniements svn-
sedimentaires dans le lutetien du Detroit de Carcasonne
(Aude). Bull. Soc. geol. Fr., 7e ser. 6:289-293.

Rixon, A.E. 1976. Fossil Animal Remains, Their Prepara-
tion and Conservation. London.

Russell, D.E. 1975. Paleoecology of the Paleocene-Eocene
Transition in Europe. Contrib. Primat. 5:28-61.

Russell, D.E., and B. Sige. 1970. Revision des chiropteres
lutetiens de Messel (Hesse, Allemagne). Palaeovertebrata
3(4):83 — 182 , 6 pis.

Schaub, S. 1940. Ein Ratitenbecken aus dem Bohnerz von
Egerkingen. Eclogae geol. Helv. 33(2): 2 74—2 84.

Hoch: Middle Eocene Shorebird

49

Stegmann, B. 1969. Uber die systematische Stellung der
Tauben und Flughiihner. Zool Jb. Syst. 96:1-51.

Strauch, J.G. 1978. The phylogeny of the Charadriiformes
(Aves): a new estimate using the method of character com-
patibility analysis. Trans. Zool. Soc. Lond. 34:263-345.

Strelnikov, I., and R. Hecker. 1968. Wladimir Kowalev-
sky’s sources of ideas and their importance for his work
and for Russian evolutionary palaeontology. Lethaia
1:219-229.

Sturm, M. 1978. Maw contents of an Eocene horse (Pro-
palaeotherium) out of the oil shale of Messel near Darm-
stadt. (Preliminary report.) Cour. Forsch -Inst. Senck-
enberg 30:120-122.

Tobien, H. 1957. Zur Palaontologie des mitteleozanen Ol-
schiefer-Vorkommens von Messel bei Darmstadt. Z. dt.
geol. Ges. 1 09(2 ):665 — 666.

. 1968. Mammiferes eocenes du Bassin de Mayence et

de la partie orientale du Fosse Rhenan. Colloque sur
l’eocene. Paris, mai 1968. B.R.G.M. 58:297-307.

. 1969. Die alttertiare (mitteleozane) Fossilfundstatte

Messel bei Darmstadt (Hessen). Mainzer Naturwiss.
Arch. 8:149-180.

Toombs, H.A., and A.E. Rixon. 1950. The use of plastics
in the “Transfer Method” of preparing fossils. The Mu-
seum’s J. 50(5): 105—107.

Touraine, F. 1960. Oeufs d’oiseaux de tres grande taille
dans l’Eocene inferieur de Provence. Bull Soc. geol. Fr.,
T ser. 2:783-789.

. 1961. Les oeufs d’Oiseaux eocenes et la stratigraphie

du Cengle (Bouches-du-Rhone). C.R. Soc. geol. Fr.:90-
91.

West, R.M., M.R. Dawson, and J.H. Hutchison. 1977.
Fossils from the Paleogene Eureka Sound formation,
N.W.T., Canada: Occurrence, climatic and paleogeo-
graphic implications. Pubis. Biol Geol. Milwaukee publ.

Mus. 2:77-93.

Wittich, E. 1898. Beitrage zur Kenntnis der Messeler
Braunkohle und ihrer Fauna. Abh. hess. geol Lande-

sanst. 3(3): 7 9—147 .

A NEW GENUS OF PENGUIN-LIKE PELECANIFORM BIRD
FROM THE OLIGOCENE OF WASHINGTON
(PELECANIFORMES: PLOTOPTERIDAE)

By Storrs L. Olson1

ABSTRACT: New specimens from the state of Washington, USA, anci from Japan show that the

family Plotopteridae Howard, previously known only from a portion of a coracoid from the early Miocene
of California, consists of flightless Pelecaniformes, with the wing modified as a paddle remarkably con-
vergent towards that of penguins and flightless members of the Alcidae. The Plotopteridae is rediagnosed
and a new genus and species, Tonsala hildegardae, is described from a partial associated skeleton from
the late Oligocene of Washington. Postcranial morphology shows the Plotopteridae to be closest to the
Anhingidae, although the specialized spearing apparatus of anhingas is lacking. Plotopterids are known
only from the North Pacific and only from deposits of late Oligocene to early Miocene age. The apparently
simultaneous disappearance of the Plotopteridae in the Northern Hemisphere and the giant penguins in
the Southern Hemisphere may be correlated with the rise of seals and porpoises. Brief comments are
appended on convergence in the evolution of diving birds.

A little more than ten years ago, Hildegarde Howard (1969),
in a brief and succinct note, introduced to science a new genus
and species of bird, Plotopterum joaquinensis, based on the
humeral end of a coracoid from an early Miocene deposit in
Kern County, California. From this single specimen she con-
cluded that Plotopterum should be made the type of a new
family of Pelecaniformes, the Plotopteridae, related to anhin-
gas and cormorants but with convergent similarities to pen-
guins and alcids that suggested Plotopterum was a wing-pro-
pelled diver with a paddle-like forelimb. Although having no
further information, Brodkorb (1971) assigned Plotopterum to
a separate subfamily in the Phalacrocoracidae. However, sub-
sequent discoveries of fossils from Japan and Washington have
fully substantiated Dr. Howard’s extraordinary perspicacity in
recognizing the affinities and adaptations represented by the
original fossil fragment.

Most of the new material of Plotopteridae, and also the best
preserved, comes from several late Oligocene and early Mio-
cene localities in Kyushu and Honshu, Japan, which I am
studying in collaboration with Dr. Yoshikazu Hasegawa of the
National Science Museum, Tokyo. We have summarized else-
where some of our overall findings (Olson and Hasegawa
1979). The general nature of the Japanese specimens, with
details of locality and stratigraphy, have been documented by
Hasegawa et al. (1979). A more complete description of the
Japanese material awaits preparation and study of recently
discovered specimens. In the present paper I shall concentrate
on the only specimen of Plotopteridae yet known from the
eastern side of the Pacific, apart from the original fossil de-
scribed by Howard.

1 Department of Vertebrate Zoology, National Museum of Natural
History, Smithsonian Institution, Washington, D.C. 20560.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 330:51-57.

The following diagnosis of the family Plotopteridae is based
partially on characters ascertained from the as yet unnamed
Japanese specimens, while that for the new genus is based on
characters of the coracoid, the only element known for the sole
taxon of the family hitherto named.

SYSTEMATICS

Order Pelecaniformes Sharpe 1891

In the following characters the Plotopteridae resemble the
Pelecaniformes and differ from the Sphenisciformes and Cha-
radriiformes: (1) absence of supraorbital furrows for salt
glands; (2) deep transverse naso-frontal hinge; (3) sternum with
large, pointed carina projecting far anterior to coracoidal sulci;
(4) furcula articulating solidly by a large rounded facet with
apex of carina (Fig. 1); (S) scapula with very large acromion
projecting anteriorly well beyond coracoidal articulation; (6)
coracoid with large flat furcular facet; (7) procoracoid process
simple, without foramen (foramen lacking in certain alcids and
incomplete in some penguins); (8) femur with proximal and
distal ends proportionately broader, neck elongate; (9) internal
condyle of tibiotarsus with marked medial deflection, and ten-
dinal groove and openings wide; (10) tarsometatarsus with fol-
lowing combination of characters — metatarsals completely
fused, hvpotarsus with large medial crest, outer trochlea ele-
vated well above others, inner trochlea elongate and at same
level as middle trochlea.

Suborder Sulae Sharpe 1891
Family Plotopteridae Howard 1969

INCLUDED GENERA: Plotopterum Howard (1969); Ton-
sala, new genus; genus or genera unnamed (Japanese speci-
mens).

52

Olson: Oligocene Plotopteridae

Figure 1. Right lateral view of the sternum (s) and furcula (f) in a plotopterid (specimen from Ainoshima Island, Japan), showing the far anterior
projection of the carina (c) and its solid articulation (a) with the furcula, a characteristic of the Pelecaniformes.

DIAGNOSIS: Medium to extremely large, flightless, wing-
propelled diving Pelecaniformes with forelimbs modified into
penguin-like paddles. Humerus with shaft greatly flattened
and curved, proximal end very heavy and rounded as in
Spheniscidae, distal end appearing more similar to certain
Alcidae. Ulna shortened, with row of distinct pits for attach-
ment of secondaries. Radius flattened and expanded. Carpo-
metacarpus short and flattened, with metacarpal I extending
nearly half the length of the bone. Coracoid very straight and
elongate; furcular facet projecting far ventrad; triosseal canal
with lower part markedly convex, separated from glenoid facet
by distinct longitudinal groove; procoracoid process long and
acuminate. Scapula with blade thin and greatly expanded,
somewhat as in Spheniscidae but acromion greatly elongated
and narrow. Skull and cervical vertebrae not greatly narrowed
and elongate as in Anhingidae; temporal fossae deep and post-
orbital processes large as in Sulidae. Caudal vertebrae very
large. Pelvis broad and shallow with anterior portions of ilia
expanded as in Anhingidae and Phalacrocoracidae; acetabu-
lum lying entirely anterior to obturator foramen, unlike other
Pelecaniformes. Femur and tibiotarsus most similar to Anhin-
gidae. Tarsometatarsus somewhat similar to Anhingidae, but
much heavier, not as excavated anteriorly, and with distal
foramen continuous with intertrochlear notch.

TEMPORAL AND GEOGRAPHIC DISTRIBUTION:
Known only from late Oligocene and early Miocene deposits
of the North Pacific: Kyushu and Honshu in Japan; Washing-
ton and southern California, in the United States.

Tonsala new genus

TYPE SPECIES: Tonsala hildegardae new species.

DIAGNOSIS: Distinguished from Plotopterum by having
coracoid with (1) glenoid facet more elongate, with margins
not as distinctly raised above shaft, and (2) sternal margin not
sinuate; (3) furcular facet projecting farther ventrad; (4) cor-
acohumeral surface relatively longer and narrower.

ETYMOLOGY: Latin, tonsa, oar, and ala, wing, feminine;
so named for the paddle-like development of the forelimb.

Tonsala hildegardae new species

Figures 2a-f, 3a-h, 4a-c, 5b

HOLOTYPE: Partial associated skeleton, vertebrate pa-
leontological collections, USNM 256518. The specimen con-
sists of the distal two-thirds of a right humerus, right ulna,
proximal and distal ends of right radius, right ulnare and ra-
diale, right carpometacarpus lacking most of the proximal end,
proximal and distal (pathological) portions of left humerus and
shaft of left ulna (pathological), humeral ends of right and left
coracoids (both worn), right scapula, anterior portion of syn-
sacrum, and right patella; also, several vertebrae, ribs, and
unidentified bone fragments still in matrix. Collected 1 Janu-
ary 1977 by Douglas Emlong (field number E-77-1). The con-
dition of the holotype suggests considerable predepositional
breakage and wear of the bones, although some of the elements
remained nearly in articulation. The left humerus is in two
pieces, possibly due to a premortem break as the distal end is
grossly pathological and so grown over with spongy bone as
to be almost unrecognizable. Likewise, the left ulna appears
to be atrophied. The specimen was preserved in an excessively
refractory sandstone, necessitating laborious preparation by
grinding.

DIAGNOSIS: As for the genus. Much larger than Plotop-
terum joaquinensis.

TYPE LOCALITY: Washington, Clallam County, Olym-
pic Peninsula, south side of Strait of Juan de Fuca. On Disque
Quadrangle, U.S. Geological Survey 7.5-minute series to-
pographic map, 1950 edition, the locality is about 0.4 km E of
first point of land extending into strait, slightly more than 3.2
km W of mouth of Lyre River, and immediately W of the
mouth of Murdock Creek.

HORIZON: Late Oligocene, Pvsht Formation of Twin Riv-
er Group (see correlation chart in Snavely et al. 1978). Spec-

Olson: Oligocene Plotopteridae

S3

Figure 2. Humerus of Tonsala hildegardae, holotype (USNM 256518), and a late Eocene penguin, a, proximal end of left humerus of Tonsala
hildegardae, external view; b, same, internal view; c, same, proximal view; d, distal portion of right humerus of Tonsala hildegardae , external
view; e, same, internal view; f, same, distal view; g, internal view of proximal end of left humerus of a late Eocene penguin (gen. and sp. indet.,
USNM 244144) from Seymour Island, Antarctica, to show overall similarity in morphology to Tonsala . All figures xl except g, which is about
x'/2. The specimens are actually dark, but in this and the following two figures they have been coated with ammonium chloride to enhance detail.

imen found in float about 10 m from bank. Matrix barren of
microfossils (C.A. Repenning pers. comm ). The locality is in
the reference section of the “upper member” of the Twin River
“Formation” in the terminology of Brown and Gower (1958).
It is close to or at “locality A3690” of Durham (1944) and is
in his Echinophoria rex zone. It is also very near or at “locality
f 11810” of Rau (1964), regarded as upper Zemorrian in the
California benthic foraminiferal stages. The most recent data
would make the age of this deposit greater than 30 million
years (Addicott 1977).

ETYMOLOGY: In honor of Dr. Hildegarde Howard, in
recognition of her many contributions to the study of fossil

birds, but more particularly of her correct diagnosis of an en-
tirely new family from a single fragment of bone.

MEASUREMENTS OF HOLOTYPE (in mm): Humerus:
proximal width 27.9, proximal depth 19.0, distal width
(through external condyle) 22.7, distal depth (through internal
condyle) 13.3, shaft width just distal to palmar crest 16.8,
shaft depth at same point 7.9. Coracoid: distance from head
to distal extent of glenoid facet 41.8, length of glenoid facet

24.6, breadth below head across triosseal canal 12.7. Scapula:
total length (as preserved) 141.1, width at narrowest point

10.7. Ulna: length 72.5, proximal depth 18.7, proximal width
12.5. Carpometacarpus: distance from distal end of metacarpal

54

Olson: Oligocene Plotopteridae

Figure 3. Distal wing elements and patella oiTonsala hildegardae, holotvpe (USNM 2S6518). a, right ulna, internal view (the lunate incision in
the middle of the shaft is an artifact of preparation); b, same, proximal view; c, same, external view (note the pits for the attachment of the
secondaries — these are obscured distally by breakage); d, right radius, external view (part of shaft missing); e, right carpometacarpus, lacking
proximal end, internal view (arrow indicates distalmost portion of metacarpal I; retouched to eliminate matrix in intermetacarpal space); f, same,
external view; g, right ulnare, ventral view; h, right patella, anterior view. All figures XI except g, which is X2.

I to distal end of metacarpal II 24.8, distal depth 14.4, distal
width 6.0, length of intermetacarpal space 25.9. Radius: great-
est diameter of proximal articulation 8.6. Radiale: greatest
diameter 13.2. Patella: greatest diameter 25.5.

DESCRIPTION: Through modification for underwater lo-
comotion, the wing elements of Tonsala hildegardae have lost
all resemblance to those of any other pelecaniform birds and
have become remarkably similar to those of other wing-pro-
pelled divers, i.e., penguins and auks (Sphenisciformes:
Spheniscidae and Charadriiformes: Alcidae). This is most ap-

parent in the proximal end of the humerus (Fig. 2a-c), which
is very heavy, with a deep, rounded head having a resem-
blance among known birds only to penguins. The ligamental
furrow is very deep and diagonally oriented and the bicipital
surface is well demarcated — in contrast to Recent penguins
but quite similar to the condition seen in certain late Eocene
penguins (Fig. 2g). The capital groove is better developed than
in living or fossil penguins, but the tricipital fossa is consid-
erably smaller and shallower.

The shaft of the humerus is sigmoid and very compressed

Olson: Oligocene Plotopteridae

55

dorsoventrally, having a large anterior crest most similar to
that seen in the humerus of the flightless alcids of the fossil
genus Mancalla, but better developed (Fig. 2d, e; see Miller
and Howard 1949, for illustrations of Mancalla). At the distal
end, the tricipital grooves are very deep, as in penguins and
alcids. The entepicondvle extends slightly more distad than
the ridge between the tricipital grooves, Tonsala being inter-
mediate in this respect between the flightless alcid Pinguinus
and the more specialized flightless alcid Mancalla. The bra-
chial depression is still fairly well developed, as in Pinguinus
and unlike either Mancalla or penguins.

Overall, the humerus is more specialized for wing-propelled
diving than in any birds except penguins, although the distal
end is slightly less modified in this direction than in Mancalla.

The ulna is quite distinctive (Fig. 3a-c). The shaft is not
curved, and it tapers evenly from a broad proximal end to a
relatively small distal articulation. Along the dorsal surface is
a row of approximately 13 deep circular pits for the attachment
of secondaries. This is a unique condition in birds, and quite
unlike that in penguins, in which the remiges can no longer
be differentiated morphologically from other feathers of the
wing. The olecranon is reduced. The internal cotvla is quite
large and deep, but the external cotvla is so modified as to be
convex in proximal view.

The radius of the holotype (Fig. 3d) lacks a section of shaft,
so that it is not possible to determine its exact shape. It is
highly distinctive in being flattened and in having both a prox-
imal and a distal expansion on the anterior edge. This is much
more modified than the radius in Pinguinus, but somewhat
similar to, though still more specialized than, that of Mancalla.
In the latter the radius is also expanded, but by a single crest
located farther distally than the proximal crest in Tonsala.

The radius and ulna in Tonsala are not nearly as modified
as in the flattened, almost unrecognizable structures of pen-
guins, nor are they quite as foreshortened as in Mancalla,
although in other respects they are perhaps more modified than
in that genus.

The ulnare (Fig. 3g) is flattened, with a large posterior ex-
pansion, foreshadowing the even more specialized structure of
penguins. The ulnar condyle is a distinct lunate incision that
is on a line with the notch for the carpal trochlea, quite in
contrast to typical birds in which these two articulating sur-
faces are nearly perpendicular to each other. This indicates
that the distal portion of the wing in Tonsala was held parallel
to the ulna and was probably capable of very little flexion.

The carpometacarpus of the holotype lacks most of the prox-
imal end, but the distalmost part of metacarpal I is preserved
(Fig. 3e, f). From this, and one Japanese specimen in which the
carpometacarpus is complete, it is seen that metacarpal I was
greatly elongated — to about the same extent as in Mancalla.
The entire metacarpus is flattened and the distal end, partic-
ularly the tuberosity of metacarpal II, is much more expanded
than in Mancalla.

The scapula of Tonsala has a very thin, sheetlike, greatly
expanded blade (Figs. 4a, 5b), unlike that of any other birds
except penguins. This is also an adaptation for wing-propelled
diving; the dorsal elevators of the wing arise mainly from the
scapula and are enlarged because the upstroke must be made
against water and also provides propulsive force. The
very long, narrow acromion in Tonsala is a pelecaniform
feature, this process being small and poorly developed in both
penguins and alcids. The acromion in Tonsala is longer and

Figure 4. Pectoral girdle of Tonsala hildegardae , holotype (USNM
256518). a, right scapula, dorsal view (most irregularities in margins
of blade are probably artifacts of preservation); b, humeral end of left
coracoid, dorsal view; c, same, lateral view. All figures xi.

more slender than in any of the other members of the Pele-
caniformes.

Both coracoids of the holotype of Tonsala are poorly pre-
served and lack the sternal two-thirds. The left coracoid, how-
ever, shows the distinctive features of the Plotopteridae as
discussed by Howard (1969) and in the familial and generic
diagnoses above. In this specimen (Fig. 4b-c), the procoracoid
process is preserved. It is a large, anteriorly curved spine,
much larger than in the Anhingidae or Phalacrocoracidae, but
not too unlike that in the Sulidae.

The patella was the only element of the hindlimb preserved
in the holotype of Tonsala hildegardae . It is a large and heavily
ossified bone with a distinct small transverse perforation for
the tendon of the ambiens muscle (Fig. 3h). An ossified patella
occurs only in the Sulidae, Anhingidae, and Phalacrocoraci-
dae, among the Pelecaniformes. The form of the patella of
Tonsala is more similar to that in the Anhingidae, differing
mainly in being a heavier bone and in having the dorsal por-

56

Olson: Oligocene Plotopteridae

Figure 5. Ventral view of right scapula of (A) Anhinga anhinga, An-
hingidae, Pelecaniformes; (B) Tonsala hildegardae, Plotopteridae, Pel-
ecaniformes; (C) Eudyptes chrysolophus, Spheniscidae, Sphenisci-
formes. The acromion (a) is well developed in the two Pelecaniformes,
in contrast to the penguin; however, the very broad but thin blade
occurs only in penguins and the convergentlv similar Plotopteridae.
Not to scale.

tion projecting anteriad as a distinct knob. In the Sulidae the
patella is a more flattened, simpler structure, lacking the en-
closed canal for M. ambiens. The patella in the Phalacroco-
racidae, while varying within the family, is quite different,
taking the form of a pyramid with a tetragonal base and pro-
jecting much farther anteriad than in Tonsala.

DISCUSSION

Tonsala hildegardae was a much larger bird than Plotop-
terum joaquinensis and also exceeded in size any of the living
penguins except the two species of Aptenodytes. It is generi-
cally distinct not only from Plotopterum, but also from a much
larger and as yet unnamed Japanese species for which com-
parable elements are known. The holotvpe of Tonsala hilde-
gardae is the only specimen of bird yet known from the Oli-
gocene marine deposits of the eastern Pacific. It is somewhat
older than Plotopterum joaquinensis, but probably nearly con-
temporaneous with most of the plotopterids from Japan. In
the deposits in which they occur, plotopterids are the only
birds so far known. Yet they are absent from later deposits
and thus evidently became extinct toward the end of the early
Miocene. The giant penguins in the Southern Hemisphere died
out at the same time. There is a strong possibility that the
disappearance of these two unrelated groups in different hemi-
spheres is linked with the contemporaneous ascendency of
seals and porpoises (Simpson 1974; Olson and Hasegawa
1979).

The Plotopteridae not only belong in the Pelecaniformes,
but are clearly derived from members of the suborder Sulae,
which includes the Sulidae, Anhingidae, and Phalacrocoraci-
dae. The species in the latter two families are entirely foot-
propelled divers, but at least some of the Sulidae, all of which
are plunge divers, are known to use the wings occasionally
underwater to extend the depth of their dives (Thomas R.
Howell pers. comm.). Increased specialization for such loco-
motion in some early pelecaniform group led ultimately to the
development of the Plotopteridae.

In the course of modifying the forelimb into a paddle-like
propulsive organ, plotopterids, penguins, and alcids have
evolved numerous “shared derived character states,” but only
by blind adherence to cladistic methodology could these three
families be classified as a monophvletic group. The profound
differences between plotopterids and penguins or alcids and
the many characters, including presumably derived ones, that
link the Plotopteridae and the Pelecaniformes have been out-
lined above. To ignore such differences in favor of emphasiz-
ing similarities in what are clearly locomotor adaptations is to
disregard the very information that leads to a true understand-
ing of the evolutionary history of these taxa. This is neverthe-
less what Cracraft (1972:387) has done in attempting to res-
urrect the hypothesis that foot-propelled diving birds of the
orders Gaviiformes, Podicipediformes, and Hesperornithi-
formes “evolved from a common ancestor” that was also a
foot-propelled diver.

The physical constraints of extreme specialization of one or
the other set of limbs for underwater propulsion obviously
impose a certain morphological uniformity on those organs in
the bird that happens to adopt such a mode of locomotion,
regardless of relationships. Storer’s (1960) analysis of evolution
in diving birds, which does not ignore differences and which
requires independent development of similarities in locomotor
adaptations, is logical and in full accordance with observed
facts. To this the Plotopteridae add a striking new example of
the significance of convergence.

ACKNOWLEDGMENTS

Douglas Emlong collected, and G.B. Sullivan prepared, the
holotype oiTonsala hildegardae . Emlong’s field work was sup-
ported by a grant from the Smithsonian Research Foundation.
Clayton E. Ray kindly provided detailed stratigraphic infor-
mation. I gratefully acknowledge the cooperation of my col-
league Yoshikazu Hasegawa, National Science Museum, To-
kyo, who has supplied information, casts, and specimens of
the Japanese plotopterids as well as much generous hospitality
during my visit to Japan in 1976. Robert McKenzie, Natural
History Museum of Los Angeles County, supplied a cast of
Plotopterum. For their comments on the manuscript I thank
John Farrand, Jr., Clayton E. Ray, and David W. Steadman.
The line drawings are by Bonnie Dalzell. I am particularly
indebted to Victor E. Krantz for his skillful photography of
specimens.

LITERATURE CITED

Addicott, W.O. 1977. Neogene chronostratigraphv of near-
shore marine basins of the Eastern North Pacific. Proc.
First Intern. Congr. Pacific Neogene Stratigraphy, To-
kyo, 1976. Pp. 151-175.

Olson: Oligocene Plotopteridae

57

Brodkorb, P. 1971. Catalogue of fossil birds: Part 4 (Co-
lumbiformes through Piciformes). Bull. Florida State
Mus., Biol. Sci. 15(4): 163-266.

Brown, R.D., Jr., and H.D. Gower. 1958. Twin River
Formation (redefinition), Northern Olympic Peninsula,
Washington. Bull. Amer. Assoc. Petrol. Geol. 42(10):2492-
2512.

Cracraft, J. 1972. The relationships of the higher taxa of
birds: problems in phylogenetic reasoning. Condor
74(4):3 7 9—392 .

Durham, J.W. 1944. Megafaunal zones of the Oligocene of
Northwestern Washington. Univ. Calif. Publ., Bull
Dept. Geol. Sci. 25(5): 101-212.

Hasegawa, Y., S. Isotani, K. Nagai, K. Seki, T. Suzuki,
H. Otsuka, M. Ota, and K. Ono. 1979. Preliminary
notes on the Oligo-Miocene penguin-like birds from Ja-
pan. Bull. Kitakyshu Mus. Nat. Hist. 1:41-60. [In Jap-
anese.]

Howard, H. 1969. A new avian fossil from Kern County,
California. Condor 7 1( 1):68— 69.

Miller, L., and H. Howard. 1949. The flightless Pliocene
bird Mancalla. Carnegie Inst. Washington Publ. 584:201-
228.

Olson, S.L., and Y. Hasegawa. 1979. Fossil counterparts
of giant penguins from the North Pacific. Science
206(44 19):688-689.

Rau, W.W. 1964. Foraminifera from the Northern Olympic
Peninsula, Washington. U.S. Geol Surv. Prof. Paper
374G:Gl-G33.

Simpson, G.G. 1974. Fossil penguins. Pp. 19-41 in The Bi-
ology of Penguins (B. Stonehouse, Ed.). Macmillan, Lon-
don.

Snavely, P.D., Jr., A.R. Niem, and J.E. Pearl. 1978.
Twin River Group (Upper Eocene to Lower Miocene) —
Defined to include the Hoko River, Makah, and Pvsht
Formations, Clallam County, Washington. Pp. A 1 1 1 —
A120 in Changes in stratigraphic nomenclature by the
U.S. Geological Survey, 1977 (N.F. Sohl and W.B.
Wright, Eds.). Geol. Surv. Bull. 1457-A.

Storer, R.W. 1960. Evolution in the diving birds. Proc.
Xllth Intern. Orn. Congr., Helsinki 1958:694-707.

A NEW GENUS OF TERATORN FROM THE HUAYQUERIAN OF
ARGENTINA (AVES: TERATORNITHIDAE)

By Kenneth E. Campbell, Jr.,1 and Eduardo P. Tonni2

ABSTRACT: A review of the family Teratornithidae, heretofore known only from two genera and

three species restricted to North America, is followed by the description of a new genus and species,
Argentavis magnificens, from the Huayquerian (late Miocene) of Argentina. The new teratorn possessed
cranial adaptations similar to those of Teratornis merriami L. Miller. It was approximately twice as large
as T. merriami, with a probable wingspan of 6.5 to 7.5 m, the largest flying bird known to science. A
possible second occurrence of a teratorn in late Pleistocene deposits of South America (La Carolina,
Ecuador) is noted.

RESUMEN: Se realiza una revision de la familia Teratornithidae, solo conocida hasta el momento

a traves de dos generos y tres especies restringidas a America del Norte. Se describe un nuevo genero v
especie, Argentavis magnificens , procedente de sedimentos de Edad Huavqueriense (Mioceno tardio) de
la Argentina. Este nuevo teratorno poseia adaptaciones craneanas similares a aquellas de Teratornis
merriami L. Miller, siendo su tamano aproximadamente el doble que el de esta ultima especie. Argentavis
magnificens tenia una envergadura probable de 6. 5-7. 5 m, por lo que representa el ave voladora de mayor
tamano conocida hasta ahora. Se hace referencia tambien a otro posible registro para un teratorno en el
Pleistoceno tardio de America del Sur (La Carolina, Ecuador).

The teratorns are members of an extinct avian family, the
Teratornithidae Miller 1925, long considered to be related to
the New World vultures of the family Vulturidae. This rela-
tionship was based primarily on the raptorial appearance of
the beak and certain parts of the postcranial skeleton, although
it was questioned even as it was originally proposed (Miller
1909). All known species of the family were very large to gi-
gantic birds, a fact that led many people to consider the ter-
atorns as necessarily having a condor-like style of flying.

To date, the family Teratornithidae has been composed of
only two genera, Teratornis and Cathartornis . The former
contains two species, Teratornis merriami L. Miller 1909 and
T. incredibilis Howard 1952. Teratornis merriami was the first
described and is the best known species of the family, being
represented by hundreds of specimens recovered from the as-
phalt deposits at Rancho La Brea, California, as well as spec-
imens from other late Pleistocene localities in California, Flor-
ida, and Nuevo Leon, Mexico (Brodkorb 1964).

In his original description of Teratornis merriami, Miller
(1909:315) stated: “ Teratornis , if it be considered raptorial,
displays characters more or less distinctive of each of these
groups [other families of the order Accipitriformes], though a
preponderance of cathartid affinities is evident.” While even
then believing that Teratornis should be placed in its own

1 Natural History Museum of Los Angeles County, 900 Exposition
Boulevard, Los Angeles, CA 90007.

2 Division Paleontologia Vertebrados. Facultad de Ciencias Naturales
y Museo, 1900 - La Plata, Argentina.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County . 1980. 330:59-68.

family, he hesitated to take that step because of the lack of
any hindlimb elements assignable to T. merriami. The follow-
ing year, Miller (1910) described a new genus and species,
Pleistogyps rex, based upon the hindlimb elements of T. mer-
riami, an error he later recognized and corrected (Miller
1925:92). At that time, he established the family Teratornith-
idae, stating that “ Teratornis . . . shows very bold divergence
in its osteology from the closely knit family of the Cathartidae
[ = VulturidaeJ, the divergence taking a number of different
pathways. The degree of divergence is in excess of those os-
teological differences to be noted between most families of
living birds classified under one order” (Miller 1925:94).

Teratornis merriami was a very large bird, standing about
0.75 m tall, with a wingspan of 3.5 to 3.8 m. Early estimates
(Fisher 1945; Stock 1956; Howard 1972) placed its weight at
about 23 kg, but new data and calculations (John Anderson
pers. comm.) indicate that 15 kg is a more accurate estimate.
The California Condor, Gymnogyps californianus (Shaw),
reaches a wingspan of 2.75 to 3. 1 m and a weight of 9 to 10.5
kg (Koford 1953). Because of its size, it was long thought that
T. merriami must have been a soaring bird, using wind cur-
rents and updrafts to maintain flight, much as the condors do.
The tendency to equate large size with soaring flight probably
played a significant role in maintaining the concept of Tera-
tornis as a condor-like bird. After a study of the postcranial
osteology, Fisher (1945) concluded that T. merriami was better
adapted for flapping flight than condors. He suggested that
the type of flight of T. merriami may have been similar to that
in modern herons and pelicans, and also that it was not ca-

60

Campbell and Tonni: Huayquerian Teratorn

pable of soaring under conditions that would keep Gymnogyps
in the air indefinitely.

With this background, the discovery of the even larger Ter-
atornis incredibilis was quite astounding. Unfortunately, T.
incredibilis is known from only three specimens, none of
which is particularly diagnostic. The species was named on
the basis of a complete cuneiform bone from Smith Creek
Cave, Nevada, a site that is “certainly no older than late Pleis-
tocene” (Howard 1972:343). The second specimen referred to
the species came from Irvingtonian deposits in the Vallecito
Creek valley of the Anza-Borrego Desert, San Diego County,
California. This specimen, a distal end of a radius, was re-
ferred to T. incredibilis “based on its general resemblance to
that of Teratornis merriami and its tremendous size” (Howard
1963:16). The third specimen, the anterior portion of a beak,
came from Blancan deposits in the Fish Creek beds of the
Anza-Borrego Desert. This specimen was also referred to T.
incredibilis on the basis of its general resemblance to T. mer-
riami and its large size (Howard 1972).

Whether all of the three specimens referred to T. incredibilis
are actually from the same species is problematical. As dis-
cussed by Howard (1972:343), if the three specimens are from
the same species, its longevity would be in excess of three
million years. However, these specimens are so undiagnostic
that they may not even all belong to the same genus, much
less the same species, and if they are all of the same species
they may belong to a genus other than Teratornis (see Howard
1972:343). We hasten to add that we believe Howard’s method
of describing the specimens was most appropriate; she brought
their existence to the attention of the scientific community,
while at the same time leaving the resolution of higher level
taxonomic categories until the discovery of more diagnostic
material.

The three specimens referred to T. incredibilis are each ap-
proximately 40 percent larger than corresponding specimens
of T. merriami. The large size of the cuneiform and radius
indicates that T. incredibilis was a flying bird, and Howard
(1952:52) has suggested that it had a wingspan of about 4.9 to
5.2 m, an estimate based upon the size of its cuneiform relative
to that of T. merriami. Teratornis incredibilis, then, was ri-
valed only by Osteodontornis orri Howard 1957, a gigantic
marine bird from the Miocene of California, for the title of the
world’s largest flying bird. Howard (1957:15) suggested that
0. orri may have had a wingspan near 5 m.

The genus Cathartornis is composed of only one species, C.
gracilis Miller 1910, a taxon based upon two tarsometatarsi
from Rancho La Brea, California. In a reevaluation of C.
gracilis, Miller and Howard (1938) considered it to be gener-
icallv distinct from Teratornis. They also considered Terator-
nis and Cathartornis to be sufficiently similar to warrant the
transfer of the latter from the Vulturidae, wherein it was orig-
inally placed, to the Teratornithidae. Based upon the size of
the tarsometatarsus, which is as long as but more slender than
that of Gymnogyps calif or nianus, C. gracilis is the smallest of
the known teratorns.

Brodkorb (1964) reduced the Teratornithidae to subfamilial
rank within the Vulturidae. On the other hand, Jollie
(1977:111) considered T. merriami to be “the most extreme
cathartic! in some respects” and the teratorns to be distinct at
the familial level within the Accipitriformes. Olson (1978:168),
however, has suggested that the teratorns may be a pelecani-

form group. This suggestion was based in part upon the shape
of the sternum of T. merriami, about which Fisher (1945:727)
noted, “There is nothing cathartid about this bony element
...” The senior author of the present paper recently initiated
detailed studies of the osteology of T. merriami, with the in-
tended goal of further determining its functional morphology
and phylogenetic relationships. Preliminary results indicate
that T. merriami was condor-like in its locomotory but not
its feeding behavior, and that the teratorns may not be related
to any of the families of Accipitriformes. In fact, T. merriami
does have many structural similarities to pelecaniform birds,
both in its cranial (as noted below) and postcranial osteology.
However, these similarities appear to be a result of conver-
gence and probably do not reflect phylogenetic relationships.

In summary, the Teratornithidae has been comprised of
three species of very large to gigantic flying birds placed in
two genera, all known from North America. Two of the
species, Teratornis incredibilis and Cathartornis gracilis, are
known from only a few specimens, and may or may not be
related to T. merriami. The latter is known from hundreds of
specimens, but its physical characteristics and relationships
with other avian groups are still poorly understood.

To the Teratornithidae we can now add a new genus and
species of such staggering proportions that one can only marvel
that such a bird could have existed, and at the good fortune
of finding a fragmented associated skeleton of it.

SYSTEMATICS

Order Accipitriformes (Vieillot 1816)

Family Teratornithidae L. Miller 1925

DESCRIPTION: Family characters listed by Miller
(1925:94) include: (1) lateral and backward extension of post-
auditorv prominences; (2) close approximation of maxillopal-
atines; (3) reduction of cerebellar region; (4) compression and
vaulting of beak; (5) elliptical foramen magnum; (6) broaden-
ing and shortening of sternum; (7) weakness and openness of
furcula; (8) ruggedness of humeral head; (9) elongation and
attenuation of ulna and metacarpus; (10) relative weakness of
posterior limbs; (11) reduction of trochanter of femur; (12) re-
duction of tibial crests; (13) columnar character of tarsometa-
tarsus. Additional characters not listed by Miller include (14)
skull broad and dorsoventrallv flattened; and (15) quadrate
with an L-shaped mandibular articulation extending without
break from quadratojugal socket to anteromost point of ven-
tral surface.

Argentavis new genus

TYPE SPECIES: Argentavis magnificens new species.

DIAGNOSIS: Differs from Teratornis L. Miller 1909 by
having skull (Fig. la, b) (1) broader, more flattened dorsoven-
trally, with greater posterolateral extension of postauditory
prominences; with (2) foramen magnum lying in a plane facing
more posteriad, i.e., more vertical; (3) foraminal openings im-
mediately anterolateral to occipital condyle large, but possibly
enlarged by breakage (very small in Teratornis)', (4) occipital
condyle as wide as widest portion of foramen magnum (about

Campbell and Tonni: Huayquerian Teratorn

61

Table 1. Measurements (in mm) of Argentavis magnificens new genus new species, Teratornis merriami L. Miller,1 and Gymnogyps californianus
(Shaw) (n = 1).

Argentavis

magnificens

Teratornis

merriami

Gymnogyps

californianus

Skull

Length

435 ± 20

222.0

158.0

Maximum width through postauditory prominences

150 ± 10

,86.7

50.0

Top of cranium through ventral tip of

66 ± 5

55.7

45.0

occipital condyle

Maximum width of foramen magnum

15.5 ± 1

12.2

11.4

Height of foramen magnum

17.5 ± 1

13.4

12.5

Width of occipital condyle

15.0

9.5

6.1

Height of occipital condyle

11.0

6.1

5.0

Quadrate

Maximum distance from squamosal articulation
to tip of mandibular articulation

66 ± 2

36.7-39.2

38.3

27.8

Anteroposterior ventral length

46 ±3

24.2-28.5

26.4

18.0

Center of socket for quadratojugal to anterior
end of mandibular articulation

53 ± 2

25.5-28.3

26.8

15.3

Humerus

Length

570 ± 10

310.0-330.0

318.2

271.0

Least width of shaft

49.0

22.9-26.7

24.6

21.0

Depth of shaft at point of least width

35.0

17.6-20.5

19.5

16.0

Coracoid

Head to internal distal angle

325 (est.)

(as preserved, 205)

151.3-163.5

156.5

98.0

Head to medial opening of coracoidal fenestra

125.0

70.1-77.7

74.4

53.2

Maximum width of glenoid facet

31.0

17.9-18.8

18.3

13.4

Dorsal end of glenoid facet to ventral end
of procoracoid

78.0

39.1-42.6

40.5

35.7

T arsometatarsus

Length

240 (est.)

(as preserved, 133)

130.4-145.8

139.8

121.5

Width at distal end of distal foramen

42.0

20.8-23.4

22.0

22.5

1 Measurements for the skull were taken from specimen No. LACM HCB1381. For the other elements, measurements were taken from five
complete specimens of each from the collections in the George C. Page Museum, Natural History Museum of Los Angeles County. This group of
measurements is not intended to be definitive for the species, but only to demonstrate its general size.

one-third narrower than foramen magnum in Teratornis)', (5)
transverse ridge connecting the postauditorv prominences ab-
sent; (6) postauditorv prominence with posterolateral corner
less angular, not projecting ventral to occipital condyle in pos-
terior view; (7) quadratojugal with quadrate articulation pro-
jecting much less sharply ventrad.

Tarsometatarsus (Fig. 4l-m) with (1) center of shaft in an-
terior view distinctly elevated above those portions of shaft
leading to internal and external trochleae, resulting in the dis-
tal foramen lying well below the elevation of the center of the
shaft (in Teratornis the shaft is well rounded in this area, with
opening for distal foramen lying at same level as anterior edge
of center of shaft); (2) distal foramen of uniform width through-
out its length, with outer extensor groove leading to it restrict-
ed in width by elevated center of shaft (distal foramen wider
proximally than distally in Teratornis, with outer extensor
groove wide proximally, narrowing significantly at distal fo-

ramen); (3) shaft with anterior half at most proximal preserved
point quite convex, with medial side extending farthest ante-
riad (in Teratornis, anterior metatarsal groove extends distad
to become outer extensor groove, so anterior half of shaft is
not convex at any point proximal to distal foramen); (4) shaft
appears elliptical in cross section at most proximal point pre-
served, with long axis of ellipse running anteromedially-pos-
terolaterallv (roughly rectangular in Teratornis, being wider
than deep); (5) shaft edge external to distal foramen more con-
vex.

Differs from Catharthornis Miller 1910 by having tarso-
metatarsus with anterior surface of shaft convex (strongly
grooved, or channeled, throughout length in Cathartornis).

ETYMOLOGY: Latin, argentum, silver; avis, feminine,
bird. In reference to Argentina, the country of origin.

MEASUREMENTS: For measurements of the holotvpe see
Table 1.

62

Campbell and Tonni: Huayquerian Teratorn

0 CM 10

B

Figure 1. Holotvpe (Museo de la Plata No. 65-VII-29-49) skull of Argentavis magnificens new genus new species in lateral (a) and posterior (b)
view. xO.30. In this and all other figures the hatched areas represent portions of the specimen where the bone has flaked away, but the matrix
remains to show form; the dotted lines show estimated outline of bone where missing, based upon corresponding bones of Teratornis merriami.

Argentavis magnificens new species

Figures 1, 2a-c, 3, 4

HOLOTYPE: Associated partial skeleton, consisting of por-
tions of skull, right quadrate, humeral end and shaft of right
coracoid, left humerus with badly damaged proximal and dis-
tal ends, portion of shaft of left(?) ulna, portion of shaft of
right radius, distal end of left metacarpal II, midportion of left
metacarpal III, shaft of right tibiotarsus, shaft of right tarso-
metatarsus. Original in the Division Paleontologia Vertebrados
del Museo de La Plata, No. 65-VII-29-49; cast in Natural
History Museum of Los Angeles County, LACM 120074. Col-
lected by Rosendo Pascual and Eduardo Tonni.

TYPE LOCALITY: Salinas Grandes de Hidalgo, Depar-
tamento Atreuco, La Pampa Province, Argentina. Located
about 15 km south of the Hidalgo station on the railroad con-
necting Carhue (Buenos Aires Province) with Doblas (La Pam-
pa Province), approximately 37°14'S, 63°36'W; see Figure 5.

HORIZON AND AGE: Epecuen Formation (fide Pascual
1961) (lowest level outcropping at locality). Huayquerian (late
Miocene).

DIAGNOSIS: As for genus. For measurements see Table 1.

ETYMOLOGY: Latin, magnificens, magnificent.

DESCRIPTION: All of the bones have been severely frac-
tured, but, except for the skull, crushing has been minimal.
The fracture lines have been omitted from the illustrations. In
some places the bone has flaked away, leaving only a replica
in matrix to indicate its general form. Where this has happened
in areas without diagnostic characters, the illustrations were
prepared as if the bone were still present. Hatching indicates
where bone has broken away in diagnostic areas, leaving only
the general form. Unfortunately, all the bones of the postcra-

nial skeleton lack their most diagnostic portions. Were it not
for the partial skull and quadrate, the specimen would have
to be considered indeterminate; but these two elements provide
strong evidence that relates Argentavis to Teratornis.

The quadrate (Fig. 2a-c) of Argentavis differs from that of
Teratornis by having (1) quadratojugal socket positioned far-
ther from main body of quadrate, i.e., with short leg of
L-shape proportionately longer, giving appearance of having
a “neck;” (2) mandibular articulation extending farther antero-
ventrad, but not as far anteriad proportionately, giving greater
degree of curvature to ventral edge in medial view; (3) ptery-
goid articulation positioned more laterally; (4) squamosal ar-
ticulation with medial portion hemispheric, mounted on co-
lumnar-like structure (medial portion elongated, positioned on
more massive extension of main body of quadrate in Terator-
nis)', (5) mandibular articulation with anterior one-half of me-
dial portion, i.e., its long leg, proportionately much larger,
lying at less of an angle to horizontal.

The coracoid of Argentavis (Fig. 4a-d) is characterized by
having (1) shaft laterally compressed at humeral end, nearly
fiat anterior to glenoid facet (not compressed, and well rounded
anterior to glenoid facet in Teratornis ); (2) procoracoid re-
duced, with ventral margin lying at about 45 degrees to main
axis of shaft (not reduced, with ventral margin straight and
lying at 90 degrees to main axis of shaft in Teratornis)', (3)
glenoid facet deeply concave in lateral view, with deepest point
lying just ventral to horizontal midline of facet (slightly con-
cave in lateral view, with deepest point lying near ventral end
in Teratornis)', (4) glenoid facet in posterior view with medial
edge roughly vertical and in line with coracoidal fenestra, and
parallel to main axis of shaft (sloping significantly mediad from
dorsal to ventral points in posterior view, not in line with

Campbell and Tonni: Huayquerian Teratorn

63

Figure 2. Holotype (Museo de la Plata No. 65-VII-29-49) right quadrate of Argentavis magnificens new genus new species in posterolateral (a),
lateral (b), and ventral (c) view; quadrate of Teratornis merriami L. Miller (LACM HCB747) in lateral (d) and ventral (e) view; quadrate of
Gymnogyps californianus (Shaw) (LACM Bil800) in lateral (f) and ventral (g) view. All xl.

coracoidal fenestra or main axis of shaft in Teratornis)', (5)
coracoidal fenestra lying much nearer procoracoid, and open-
ing mediae! more posteriorly; (6) ridge leading ventrad from
procoracoid toward internal distal angle small, but distinct
(absent in Teratornis).

The humerus of Argentavis (Fig. 3a-b) differs from that of
Teratornis by having (1) shaft in anterior view with proximal
two-thirds relatively straighter and distal one-third curving
more sharply dorsad; (2) shaft in dorsal view appearing more
strongly sigmoid; (3) external tricipital groove appearing to
extend proximad to ectepicondvlar prominence, which is bro-
ken away (does not extend proximad to ectepicondvlar prom-
inence in Teratornis)', (4) deltoid crest with very pronounced
knob, the distal portion of which is broken away (similar, but
with knob less elevated above and less sharply demarcated
from shaft proximallv in Teratornis ); (5) shaft slightly less but
still deeply convex between deltoid crest and bicipital crest.

The preserved portion of the ulna of A rgentavis (Fig. 4i) has
no diagnostic characters, displaying only three papillae of the
secondaries spaced about 30 mm apart (spaced about IS to 18
mm apart in Teratornis merriami) .

The carpometacarpus of Argentavis (Fig. 4e-h) differs from
that of Teratornis by having metacarpal II with (1) tendinal
groove deeper, bordered by more pronounced ridges, and lying

more anteriorly on external side of shaft; (2) shaft with pos-
terior half more rounded, with a small ridge lying on posterior
side and extending a short distance proximad to most proximal
point preserved (ridge absent in Teratornis ); (3) distal meta-
carpal symphysis lies closer to center of shaft proximallv; (4)
facet for digit II with that portion preserved having anterior
end extending farther posteriad at a greater angle. Metacarpal
III has (1) shaft more triangular in cross section; (2) anterior
surface more excavated, bordered externally by more pro-
nounced ridge).

The tibiotarsus of Argentavis (Fig. 4j-k) lacks any diagnos-
tic characters, but can be seen to differ from that of Teratornis
by having (1) shaft slightly curved in anterior view, although
some curvature seen in Figure 4j-k may be a result of breakage
(essentially straight in Teratornis)', (2) fibular crest much less
developed, although this may be a result of breakage; (3) ten-
dinal groove with proximal end more symmetrical and lying
near center of shaft rather than near internal edge of shaft.

AGE AND ASSOCIATED FAUNA

The holotype of Argentavis magnificens was collected from
the brownish to reddish terrestrial sediments of the late Mio-
cene Epecuen Formation (fide Pascual 1961). This formation

64

Campbell and Tonni: Huayquerian Teratorn

C M

Figure 3. Holotype (Museo de la Plata No. 65-VU-29-49) left humerus of Argent avis magnificens new genus new species in anconal (a) and palmar
(b) view. X0.30.

is composed primarily of fine sand with minor amounts of silt
and rare lenses of clay. Irregular thicknesses of caliche-like
concretions occur at several levels; isolated concretions may
also occur.

The late Miocene age assignment of the Epecuen Formation
is based on the following mammalian fauna reported for the
deposits of Salinas Grandes de Hidalgo by Zetti (1972): Order
Marsupialia, Family Borhvaenidae: Borhyaenidium muste-
loides Pascual and Bocchino, Thylacosmilus aff. atrox Riggs;

Order Carnivora, Family Procyonidae: Cyonasua brevirostris
Moreno and Mercerat; Order Notoungulata, Family Toxodon-
tidae: Pisanodon n. sp.; Family Hegetotheriidae: Hemihege-
totherium n. sp., Paedotherium borrelloi Zetti; Order Litop-
terna, Family Macraucheniidae: ?Promacrauchenia sp.; Order
Edentata, Family Mvlodontidae: Elassotherium altirostre Ca-
brera; Family Dasypodidae: Proeuphractus sp., Macroeu-
phractus sp.; Family Glyptodontidae: Sclerocalvptinae gen. et
sp. indet.; Order Rodentia, Family Caviidae: Orthomyctera

Campbell and Tonni: Huayquerian Teratorn

65

J K

Figure 4. Holotype (Museo de la Plata No. 65-VII-29-49) of Argentavis magnificens new genus new species: right coracoid in anterior (a), lateral
(b), posterior (c), and medial (d) view; distal end of left metacarpal II in internal (e) and external (f) view; medial portion of left metacarpal III in
lateral (g) and medial (h) view; portion of shaft of left(?) ulna in anconal (i) view; shaft of right tibiotarsus in anterior (j) and posterior (k) view;
shaft of right tarsometatarsus in anterior (1) and posterior (m) view. X0.30.

sp., Paleocavia sp.; Family Hydrochoeridae: PProtohvdro-
choerinae gen. et sp. indet.; Family Chinchillidae: Lagosto-
mopsis sp.; Family Octodontidae: Phtoramys sp., Pseudopla-
taeomys sp.; Family Echimyidae: ?Eumysops sp.

This assemblage of mammalian taxa is characteristic of the

Huayquerian (sensu Pascual et al. 1965), a South American
land mammal age conventionally referred to the late Miocene
(Marshall et al. 1979). In addition to Argentavis magnificens
and the mammalian fauna, reptiles and other birds are known
from the deposits, but have yet to be described.

66

Campbell and Tonni: Huayquerian Teratorn

Figure 5. Map showing location of type locality, Salinas Grandes,
in Argentina.

DISCUSSION

The similarities of the skull and quadrate of Argent avis to
those of Teratornis are very striking when characters of these
two genera are contrasted with those of genera of the other
accipitriform families. Although the unique structure of the
teratorn skull has been commented on since its description
(Miller 1909), there has been no attempt to analyze it as there
has been for its postcranial skeleton (Fisher 1945). The studies
now in progress on Teratornis merriami will attempt to fill this
void. A few preliminary comments about functional morphol-
ogy that apply to both Argentavis and Teratornis are presented
here.

Teratornis appears to be more specialized than Argentavis,
e.g. , by having the postauditory prominences ending in a more
angular corner that projects ventral to the occipital condyle
and a prominent transverse ridge connecting the postauditory
prominences. Teratornis also has the posterior portion of the
skull much more rounded, in both lateral and posterior view
(for illustrations of T. merriami see Miller 1909, 1925; Jollie
1978). It is not possible to make additional cranial comparisons
because of the damaged nature of the holotvpe skull of Argen-
tavis magnificens .

The posterior extension of the postauditory prominences is
an adaptation to increase the gape of the mouth by moving
the hinge line of the jaw posteriad. In both Argentavis and
Teratornis the quadrate articulates with the squamosal pos-
terior to the occipital condyle, giving the maximum possible
gape without actually having the squamosal lying farther pos-
teriad than the parietal or supraoccipital.

The articulation of the quadrate with the squamosal is such
that, when the ventral end is swung through its arc, it moves

posterolaterallv at an angle of about 45 degrees to the long axis
of the skull. This contrasts with the condition found in other
accipitriform families where the quadrate movement is almost
parallel to the long axis of the skull, and is far more restricted.
By rotating the quadrate so that the ventral end moves laterad
as much as it moves posteriad, pressure is exerted on the ar-
ticular of the lower jaw, forcing the rami of the lower jaws
apart posteriorly. A similar, but less developed, condition is
found in pelicans (Pelecaniformes: Pelecanidae), and the peli-
can quadrate bears a strong superficial resemblance to the ter-
atorn quadrate. The Frigatebird, Fregata magnificens (Pele-
caniformes: Fregatidae), and albatrosses (Procellariidae: Dio-
medeidae) also have a similar condition.

As illustrated by Gymnogyps (Fig. 2f— g), in the family Vul-
turidae the mandibular articulation is not “L-shaped” or con-
tinuous, and the two portions do not lie perpendicular to each
other. All genera of vulturids have a distinct shelf on the me-
dial side of the anterior portion of the mandibular articulation,
a character limited to that family within the Accipitriformes.
The lateral component of the articular movement on the quad-
rate, and of the quadrate on the squamosal, in Gymnogyps and
other vulturids is minimal.

In the teratorn quadrate, the quadratojugal socket is much
less restrictive than in vulturids, an adaptation that assists the
lateral movement of the quadrate. A similar condition exists
in frigatebirds and albatrosses; in the pelicans there is no sock-
et present, only a flat or convex articular surface.

The lower jaw of Argentavis is unknown, which is perhaps
to be expected if it resembled that of Teratornis. The lower
jaw of Teratornis merriami is very weak, as noted by Howard
(1950), and even at Rancho La Brea no complete specimens
are known; the portion immediately anterior to the mandibular
foramen was apparently such a thin sheet of bone that it was
never preserved, or it was lost in collection and preparation.
This character is also an adaptation for lateral movement of the
posterior portion of the lower jaw; it provides a weak spot
where the jaw can flex without having a weak symphysis. This
condition is also present in frigatebirds and albatrosses. The
exact function of this character complex in feeding remains to
be worked out, but it appears very unlikely that teratorns fed
in a manner similar to any other accipitriform.

A comparison of the measurements of Argentavis magnifi-
cens and Teratornis merriami reveals that the former is almost
twice the size of the latter in almost all measurements. If we
were to assume that it is reasonable to extrapolate directly
from the estimated size of T. merriami (isometric scaling), we
could say thatT. magnificens had a wingspan of 7 to 7.6 m,
a height of 1.5 m, and a weight of 120 kg. Of course, there is
the possibility that isometric scaling may not be applicable in
this case. Also, because the size of T. merriami was calculated
with the consideration in mind that it was a condor-like bird,
its estimated wingspan may be quite erroneous; and the esti-
mate may as well be too small as too large. The estimate of
the height and the new weight estimate of T. merriami are
probably much more accurate. In spite of these qualifications,
A. magnificens is certainly the largest flying bird known to
have existed.

The question as to how such a tremendously large bird like
Argentavis magnificens could fly remains unanswered. It is
often believed that very large flying birds must depend on
wind currents to become airborne and remain aloft, and that

Campbell and Tonni: Huayquerian Teratorn

67

“the maximum size attainable by Hying birds is limited by
surface-volume ratio and the speed of flight” (Storer 1971:152).
Or, “The larger the bird, the faster it must fly to stay airborne”
(Pettingill 1970:2). As noted above, however, Fisher (1945)
suggested that T. merriami was capable of flapping flight, pos-
sibly similar to that of herons and pelicans, both of which may
fly at speeds considerably slower than that observed for many
smaller species. Storer (1971:153) commented that “Under
present conditions, the larger albatrosses, pelicans, storks,
swans, condors, turkeys,, and bustards must represent about
the largest size to which flying birds can evolve.” While it is
certainly true that environmental conditions in La Pampa
Province of Argentina were very different in the Huayquerian
than they are today, it is questionable whether the mechanics
of avian flight have changed. Rather, there is a greater prob-
ability that our understanding of avian flight is still very in-
complete.

The presence of a teratorn in South America should not be
considered too surprising. Campbell (1979), in a study of the
late Pleistocene avifauna of the Talara Tar Seeps of north-
western Peru, described a new species of Gymnogyps and a
new genus and species of large eagle, Amplibuteo hibbardi ;
both genera were previously known only from North America
(G. ampins and G. calif ornianus\ Amplibuteo ( =Morphnus )
woodwardi). Many Recent species previously reported as fos-
sils only from North America were also reported from the
Talara Tar Seeps. Earlier, Campbell (1976) reported an in-
determinate fragmentary vulturid tarsometatarsus from La
Carolina, Ecuador, that differed markedly from the three gen-
era of condors later reported from the Talara Tar Seeps. A
recent comparison of this specimen with tarsometatarsi of
T. merriami from Rancho La Brea, California, shows that al-
though it is not referable to Teratornis merriami, there is a
very good possibility that it is from a different species of Ter-
atornis. As collections of avian fossils, particularly those from
South America, increase, we can expect to find many more
examples of what have been considered North American
groups appearing in South America, and vice versa (e.g., see
Campbell this vol.).

Although there is a good possibility that the Teratornithidae
should not be placed within the Accipitriformes, it is prudent
at the present time to leave it there pending completion of
more detailed studies. It can be stated that there are almost
no points of similarity between the cranial osteology of tera-
torns and that of the members of the Falconidae, Accipitridae,
Serpentariidae, or Vulturidae. And, although there are simi-
larities between the postcranial skeleton of teratorns and those
of the other families of Accipitriformes, there are many more
striking differences.

ACKNOWLEDGMENTS

We express our deepest gratitude to Jill Littlewood, who
gave so freely of her time and talent to produce the drawings
for this paper. Without her enthusiastic effort this work would
not have been possible at this time.

William Akersten read an early draft of this paper and dis-
cussed various points of osteology. Lidia Lustig assisted the
senior author by providing valuable information and discus-
sions about Argentina.

A National Geographic Society grant to the senior author

made possible, in part, his trip to Argentina, which resulted

in this study.

LITERATURE CITED

Brodkorb, P. 1964. Catalogue of Fossil Birds. Part 2 (An-
seriformes through Galliformes). Bull. Florida State
Mus., Biol. Sci. 8(3): 195-335.

Campbell, K.E., Jr. 1976. The Late Pleistocene Avifauna
of La Carolina, Ecuador. Smith. Contrib. Paleobiology
27:155-168.

. 1979. The Non-Passerine Pleistocene Avifauna of the

Talara Tar Seeps, Northwestern Peru. Royal Ontario
Mus., Life Sci. Contrib. 118:1-203.

. 1980. A Review of the Rancholabrean Itchtucknee

River Avifauna, Florida, (this vol.)

Fisher, H. 1945. Locomotion in the Fossil Vulture Terator-
nis. Amer. Mid. Nat. 33:725-742.

Howard, H. 1950. Wonder Bird of the Ice Age. Los Angeles
County Museum Leaflet Series. Science No. 3:1-3.

. 1952. The prehistoric avifauna of Smith Creek Cave,

Nevada, with a description of a new gigantic raptor. Bull.
So. Calif. Acad. Sci. 51:50-54.

. 1957. A Gigantic “toothed” Marine Bird from the

Miocene of California. Santa Barbara Mus. Nat. Hist.,
Bull. Dept. Geol. 1:1-23.

. 1963. Fossil birds from the Anza-Borrego Desert.

Nat. Hist. Mus. Los Angeles Co., Contrib. Sci. 75:1-33.

. 1972. The incredible teratorn again. Condor

74(3):341-344.

Jollie, M. 1977. A Contribution to the morphology and phv-
logeny of the Falconiformes (Parts 2, 3). Evol. Theory
2(4,5): 115-300.

. 1978. A Contribution to the morphology and phv-

logenv of the Falconiformes (Part 4). Evol. Theory 3( 1): 1 —
142.

Koford, C. 1953. The California Condor. National Audu-
bon Society. Reprinted by Dover Publ., New York. 154
PP

Marshall, L.G., R.F. Butler, R E. Drake, G.H. Curtis,
and R.H. Curtis. 1979. Calibration of the great Amer-
ican interchange. Science 204(4390):272-279.

Miller, L. 1909. Teratornis, a new avian genus from Ran-
cho La Brea. LTniv. Calif. Publ., Bull. Dept. Geol.
5(2 1):305— 317.

. 1910. The Condor-like Vultures of Rancho La Brea.

Univ. Calif. Publ., Bull. Dept. Geol. 6(1): 1-19.

. 1925. The Birds of Rancho La Brea. Carnegie Instn.

Washington Publ. 349:63-106.

Miller, L., and H. Howard. 1938. The status of the ex-
tinct Condor-like birds of the Rancho La Brea Pleisto-
cene. Publ. Univ. Calif. Los Angeles, Biol. Sci. 1:169-
176.

Olson, S.L. 1978. Multiple origins of the Ciconiiformes.
Proc. Colonial Waterbird Group 1978:165-170.

Pascual, R. 1961. Un nuevo Cardiomvinae (Rodentia, Ca-
viidae) de la Formacion Arroyo Chasico (Plioceno inferior)
de la provincia de Buenos Aires. Ameghiniana 2(4): 101—
132.

Pascual, R., and A. Bocchino. 1963. Un nuevo Borhvae-

68

Campbell and Tonni: Huayquerian Teratorn

ninae (Marsupialia) del Plioceno medio de Hidalgo (La
Pampa). Ameghiniana 3(4):97— 107.

Pascual, R., E.J. Ortega, D. Gondar, and E. Tonni.
1965. Las Edades del Cenozoico mamalifero de la Ar-
gentina con especial atencion a aquellas del territorio bon-
aerense. An. Com. Inv. Cient. Prov. Buenos Aires 6:165-
193.

Pettingill, O.S. 1970. Ornithology in Laboratory and
Field. 4th ed. Burgess Publ. Co., Minneapolis. 524 pp.

Stock, C. 1956. Rancho La Brea. A Record of Pleistocene

Life in California. 6th ed. Nat. Hist. Mus. of Los Angeles
County, Sci. Series No. 20; Paleontology, No. 11:1-81.
Storer, R.W. 1971. Adaptive Radiation of Birds. Pp. ISO-
188 in Avian Biology (D.S. Farner and JR King, Eds.).
Vol. 1. Academic Press, New York.

Zetti, J. 1972. Los mamiferos fosiles de Edad Huayqueri-
ense (Plioceno medio) de la Region Pampeana. Thesis No.
240, Fac. Cs. Nat. and Museo de la Plata. 86 pp. (un-
published)

MIDDLE PLIOCENE RAILS FROM WESTERN MONGOLIA

By E.N. Kurochkin1

ABSTRACT: The fossil remains of three new species of rails from three Middle Pliocene localities in

the Ich Nuuryn Tochom of Western Mongolia are described. These include Palaeoaramides tugarinovi
new species, Rallns risillus new species, and Crex zazhigini new species. Rails are practically absent from
Ich Nuuryn Tochom today, and the presence of three species of rails in Western Mongolia during the
Middle Pliocene indicates that there has been a change in the climatic and ecological conditions found
there since that time.

Soviet and Mongolian paleontologists and geologists have
discovered many fossil localities (Chirgis Nuur II, Chono Ha-
riagh, Dzavchan, “point 1080 m” in Sargyn Gov’ Desert, Javor

l, and others) in the western part of the Mongolian People’s
Republic (MPR) in the Ich Nuuryn Tochom (The Great Lakes
Depression) in the past few years. These workers have re-
covered numerous fragmentary remains of Middle Pliocene
vertebrates.

The fossil localities are located on the eastern border of the
Tochom, and run in a line from north to south for almost 400
km. The vertebrate remains occur in the Middle Pliocene
(Devjatkin and Zhegallo 1974) sand and aleurite sediments of
the lacustrine and nearshore-lacustrine facies. These deposits
are stratigraphicallv apportioned by Devjatkin (1970) to the
Chirgis Nuur series.

The majority of the vertebrate remains from these localities
are mammalian, but fossils of hsh, reptiles, amphibians, and
ostracods, as well as a considerable number of birds were also
collected. The birds are represented by approximately 200
fragments of postcranial bones, as well as by numerous ratite
egg shell fragments. The total number of birds identified from
the avifauna include 55 species belonging to 11 orders and 15
families (Phalacrocoracidae, Ardeidae, Ciconiidae, Anatidae,
Phasianidae, Gruidae, Ergilornithidae, Rallidae, Scolopaci-
dae, Phalaropodidae, Pteroclidae, Strigidae, Psittacidae, Cor-
vidae, and Turdidae). Water birds and shorebirds are predom-
inant in the collection, and the waterfowl are most numerous,
with 14 species. Part of the paleornithological material has
been described previously (Kurochkin 1971, 1976), and a de-
scription of all of the material is now being prepared for pub-
lication as a monograph. The present paper contains the de-
scription of the rallid remains from three localities: “point 1080

m, ” located in the central region of the Sargyn Gov’ Desert in
the south of the Tochom; “Chono Hariagh,” located on the
northern shore of the river with the same name between Chovd
Dalaj Nuur and Char Nuur Lakes; and “Chirgis Nuur II,”
located on the northern shore of Chirgis Nuur Lake.

1 Paleontological Museum of the USSR Academy of Sciences, Lenin-
sky Avenue, 16, 117071, Moscow, USSR.

SYSTEMATICS

Order Ralliformes
Suborder Ralli
Family Rallidae
Subfamily Rallinae

Genus Palaeoaramides Lambrecht 1933

Palaeoaramides tugarinovi new species

Figures 1, 5a

HOLOTYPE: Distal end of right humerus, No. 2614-121,
Collection of the Paleontological Institute of the USSR Acad-
emy of Sciences (PIN).

LOCALITY: “point 1080 m” in Sargyn Gov’, the Gov’ Altaj
ajmak, MPR; Middle Pliocene.

DIAGNOSIS: Humerus with (1) sulcus anconeus externus
shallow; (2) processus supracondylus externus well devel-
oped, forming prominent transverse step; (3) attachment of M.
pronator brevis distinctly separated.

MEASUREMENTS (in mm): Greatest width of distal end
5.5; anteroposterior depth of condylus radialis 3.3; anteropos-
terior depth of condylus ulnaris 1.9; distance from top of facies
ligamenti interni to distal edge of condylus ulnaris 3.2; least
depth of distalmost portion 2.1.

ETYMOLOGY: This species is named in honor of the mem-
ory of Professor A.Y. Tugarinov.

COMPARISON: Four species of Palaeoaramides are known
from the Lower (Aquitanian) and Upper Miocene of Europe
(Olson 1977). Three of these have been described and com-
pared on the basis of tibiotarsi and tarsometatarsi, but P.
beaumontii (Milne-Edwards 1869) was described from a hu-
merus from the LTpper Miocene (Helvetian) of France (Sansan
locality in the Gers Department). Illustrations of the humerus
of P. beaumontii are given in the Atlas by Milne-Edwards
(1869-1871), as well as by Cracraft (1973) in stereophoto-
graphs. These illustrations proved to be sufficient for the de-
termination and comparison of the rallid humerus from Sargyn
Gov’.

The humerus of P. tugarinovi new species and P. beau-

Contrib. Sci. Natur. Hist. Mas. Los Angeles County . 1980. 330:69-73.

70

Kurochkin: Pliocene Mongolian Rails

Figure 1. Palaeoaramides tugarinovi new species, holotype, distal
end of right humerus, No. 2614-121 PIN; locality “point 1080 m,”
Sargyn Gov’, Mongolia, in dorsal (a), palmar (b), ventral (c), and distal
(d) view. 1, incisura intercondylaris; 2, processus supracondylus ex-
ternus; 3, eminentia M. pronator brevis.

montii are very similar in the structure and disposition of both
condyles, as well as in the structure of the epicondvlus ulnaris
(or processus flexoris). The latter is notably elongated distally
and salient on the internal surface of the specimen. The facies
ligamenti interni is similar in both. It is oval in outline, extends
high externally, with its plane directed laterad and dorsad.
The impression of M. brachialis inferioris also has the same
outline and dimensions in both. The condylus ulnaris and epi-
condylus ulnaris in both species are separated by a distinct
groove that is very characteristic of the genus.

Structural differences in the distal end of the humeri be-
tween P. tugarinovi and P. beaumontii were presented in the
diagnosis. The sulcus anconeus externus is notably smaller in
P. tugarinovi than in P. beaumontii. The processus supra-
condylus externus and eminentia M. pronator brevis are more

Figure 3. Rallus risillus new species, holotype, proximal portion of
left carpometacarpus, No. 2614-100 PIN; locality “point 1080 m” in
Sargyn Gov’, Mongolia, in proximal (a), internal (b), and posterior (c)
view. 1, facies articularis pollicis; 2, fossa carpalis interna; 3, fossa
carpalis posterior.

developed in P. tugarinovi, as compared with P. beaumontii.
Palaeoaramides tugarinovi was smaller than P. beaumontii
(width of distal epiphysis 6.2; anteroposterior depth of the con-
dylus radialis 3.6; anteroposterior depth of the condylus ulnaris
1.9 (from Cracraft 1973).

DISCUSSION: Cracraft (1973) pointed out the general sim-
ilarity of Palaeoaramides and Recent Rallus Linnaeus. This
position is confirmed with this specimen. Of all modern species
of the Rallinae, Palaeoaramides is most similar to Rallus, as
concluded from the general proportions of the condyles, from
the outline and dimensions of the fossa olecrani, from the cur-
vature of the distal part of the diaphysis, and from the outline
of the facies ligamenti interni. But these two genera can well
be distinguished by the structure of the epicondylus ulnaris,
which is narrower and elongated internally in Palaeoaramides
and weakened in Rallus. In Palaeoaramides the visible depres-
sion lies between the epicondylus ulnaris and condylus ulnaris.
Rallus does not have such a depression, which results from
the distal prolongation of the ventral edge of the condylus
ulnaris. The impression of M. brachialis inferioris in Palaeoar-
amides is shallower and broader than in Rallus.

Figure 2. Cf. Palaeoaramides tugarinovi, referred humeral end of
right coracoid, No. 3222-55 PIN; locality Chirgis Nuur II on the shore
of Chirgis Nuur Lake, Mongolia, in internal (a), anterior (b), and
posterior (c) view. 1, foramen supracoracoideum; 2, processus procor-
acoideus; 3, facies glenoidalis; 4, cotyla scapularis; 5, tuber brachialis.

Figure 4. Crex zazhigini new species, holotype, distal end of left
humerus, No. 2614-90 PIN; locality Chono Hariagh in Chovd ajmak,
Mongolia, in ventral (a), palmar (b), dorsal (c), and distal (d) view. 1,
entepicondylus; 2, ectepicondylus; 3, processus supracondylus exter-
nus; 4, transversal line tuberosity.

f

Figure 5. a. Palaeoaramides tugarinovi new species, holotype, No. 2614-121 PIN, palmar view; b-c. cf. Palaeoaramides tugarinovi, referred
coracoid, No. 3222-55 PIN, in internal (b) and posterior (c) view; d-e. Rallus risillus new species, holotype, No. 2614-100 PIN, in internal (d) and
external (e) view; f. Crex zazhigini new species, holotype, No. 2614-90 PIN, palmar view, (all X4).

cf. Palaeoaramides tugarinovi

Figure 2, 5b, c

MATERIAL: Humeral end of left coracoid, No. 3222-55

(PIN).

LOCALITY: Chirgis Nuur II, MPR; Middle Pliocene.

DISCUSSION: The coracoids of the four described fossil
species of Palaeoaramides remain unknown. The specimen
here referred to P. tugarinovi has its most pronounced struc-
tural similarity with Rallus, but it still differs in certain mor-
phological characters from that genus. The similarities include
( 1) the identical structure of the dorsal portion of the diaphysis,
with the same localization and form for foramen supracora-
coideum; (2) the same degree of development and form for
processus procoracoideus; (3) the same form for cotyla scapu-
laris; and (4) the same proportions of the acrocoracoideum.
However, the details of the acrocoracoid are different: (1) The
facet on the external side of the acrocoracoid is narrower and
more extended in P. tugarinovi. (2) This is also the case with
the facies glenoidalis, which is more extended distally over the

level of processus procoracoideus in P. tugarinovi, whereas in
Rallus aquaticus Linnaeus the facies glenoidalis and processus
procoracoideus are positioned at one transverse level. (3) The
tuber brachialis in P . tugarinovi is smaller and more extended
along the diagonal. It is more elongated internally and projects
somewhat over the foramen triosseum, as compared with that
of R. aquaticus.

This specimen is referred to P. tugarinovi on the basis of a
unique combination of morphological characters that occur in
the coracoid and humerus of modern rails. This conclusion
results from similar comparisons, taking into account the ap-
propriate relative measurements of the holotype of P. tugari-
novi and the referred coracoid.

MEASUREMENTS (in mm): Transverse width of di-
aphysis 2.4; length of dorsal epiphysis (from ventral edge of
cotyla scapularis) 5.6; width of facies articularis scapularis 2.4.
On the basis of measurements, this specimen appears to have
come from a bird between the size of R aquaticus and R.
longirostris Boddaert, being slightly closer to the former. This
is also true for the holotype humerus of P. tugarinovi .

72

Kurochkin: Pliocene Mongolian Rails

Rallus Linnaeus 1758
Rallus risillus new species

Figures 3, 5d-e

HOLOTYPE: Proximal end of left carpometacarpus, No.
2614-100, Collection of the Paleontological Institute of the
USSR Academy of Sciences.

LOCALITY: “point 1080 m” in the central region of the
Sargvn Gov’ Desert, the Gov’ Altaj ajmak, MPR; Middle Plio-
cene.

DIAGNOSIS: Carpometacarpus with (1) facies articularis
pollicis appearing as small step, not sharply set off from meta-
carpal II; (2) fossa carpalis posterior lengthened and shallow;
(3) fossa carpalis interna small; (4) anteroproximal end of
trochlea radialis lying on same longitudinal axis as apophysis
pisiformis; (5) size very small.

MEASUREMENTS (in mm): Transverse width of trochlea
carpalis 1.5; anteroposterior width of trochlea carpalis (with
processus metacarpalis I) 4.0.

ETYMOLOGY: From Latin, risillus, masculine, very
small.

COMPARISON: The fossil rails are one of the best known
groups of fossil birds (Feduccia 1968; Olson 1973, 1974, 1977).
Unfortunately, no carpometacarpi of described fossil rails are
available for comparison with Rallus risillus. We compared
it with Recent R. elegans Audubon, R. longirostris , R. aqua-
ticus, and!?, limicola Vieillot. Rallus risillus differs from these
species in the details of the carpometacarpus listed in the di-
agnosis. Size is a very important character of R. risillus, it
being 1.5 times smaller than the American R. limicola, the
smallest modern representative of the genus. Measurements
(in mm) of the carpometacarpus of the four species of Recent
rails are as follows. Transverse width of trochlea carpalis: R.
elegans 3.1; R. longirostris 2.8; R. aquaticus 2.2; R. limicola
2.1. Anteroposterior width of trochlea carpalis (with processus
metacarpalis I): R. elegans 7.1; R. longirostris 6.2; R. aqua-
ticus 4.6; R. limicola 4.3.

In Recent Rallus the articulating surface of facies articularis
pollicis is widened on each side, with the surface of metacarpal
II sharply set off at almost a right angle to it. This contrasts
with the narrow surface in R. risillus that is not set off from
the surface of metacarpal II by a sharp angle. But a small R.
limicola has this angle somewhat blunted. The fossa carpalis
posterior, located on the interior side of trochlea radialis, is
much shorter and deeper in Recent Rallus than in R. risillus.
Only in R. aquaticus is it slightly elongated, tending toward
that of R. risillus. The fossa carpalis interior, on the interior
face of trochlea carpalis, of modern Rallus is deeper, larger,
and farther from the apophysis pisiformis than that of R. ris-
illus. In R. risillus the anteroproximal angle of the trochlea
carpalis lies on the same longitudinal axis as the apophysis
pisiformis, approximately the same position found in/?, aqua-
ticus. The three other species of Rallus have this angle shifted
more caudad.

REMARKS: The comparative material of the modern Ral-
linae used for the description of R. risillus was naturally in-
sufficient. Most of the modern tropical genera of this subfamily
were not represented in the comparative series. However, I
am quite certain that R. risillus is closest to the genera Rallus
and Porzana Vieillot. Rallus risillus resembles Porzana, as
indicated by comparison with P. porzana Linnaeus, P. parva

(Scopoli), P. Carolina Linnaeus, and P. flaviventer (Boddaert),
by having (1) the same structure of metacarpal II, which also
rises gradually to the facies articularis pollicis, (2) relatively
similar dimensions, and (3) the fossae carpalis posterior et in-
terior similar in form. However, the relative dimensions of
metacarpal I and metacarpal II, and their position with respect
to the carpal trochlea, indicate that R. risillus should be re-
ferred to Rallus. In addition, the groove running between
metacarpal I and metacarpal II begins at approximately the
same position in modern Rallus as it does in R risillus, but
in Rallus it begins notably more proximal than in Porzana.
The proximal articulating surface of trochlea carpalis is divid-
ed on its sides in modern Rallus, as in R. risillus, and it is
relatively wider than in Porzana.

Crex Bechstein 1803
Crex zazhigini new species

Figures 4, 5f

HOLOTYPE: Distal end of left humerus, No. 2614-90, Col-
lection of the Paleontological Institute of the USSR Academy
of Sciences.

LOCALITY: Chono Hariagh, on the northern shore of the
Chono Hariagh River in Ich Nuurvn Tochbm, Chovd ajmak,
MPR; Middle Pliocene.

DIAGNOSIS: Humerus with (1) impression of M. brachialis
inferioris deep and clearly outlined; (2) ectepicondylus short-
ened; (3) entepicondylus expanded and protruding externad;
(4) processus supracondylus externus obtuse and broad; (5)
transversal line tuberosity lying proximal from processus su-
pracondylus externus extends across approximately one-third
of the shaft at that point.

MEASUREMENTS (in mm): Greatest distal width 6.2;
anteroposterior depth of condylus radialis 3.1; anteroposterior
depth of condylus ulnaris 1.8.

COMPARISON: Crex zazhigini closely resembles Recent C.
crex (Linnaeus), but differs by having (1) impression of M.
brachialis inferioris deep, with clearly marked borders (shal-
lower, without clearly marked borders in C . crex); (2) entepi-
condylus elongated and produced (shortened and not produced
in C. crex); (3) ectepicondylus shortened (elongated and narrow
in C. crex); (4) processus supracondylus forming step, wid-
ened and blunted medially (pointed in C. crex); (5) transversal
line tuberosity extending across one-third of the shaft, ending
internally at impression of M. brachialis inferioris, and exter-
nally at edge of shaft (that tuberosity is lower and narrower
in C. crex); (6) somewhat larger size (transverse width of distal
end of humerus in C. crex, 5.3 to 5.5 mm). This is the first
record of the genus Crex from Neogene deposits.

ETYMOLOGY: This species is named in honor of paleo-
mammalogist V.S. Zazhigin in recognition of his contributions
to collecting of Neogene birds in Mongolia.

CONCLUSIONS

Porzana pusilla (Pallas) is the only rail inhabiting western
Mongolia today. The presence of several specimens of species
of the subfamily Rallinae in upper Middle Pliocene deposits
of the Ich Nuurvn Tochom indicates that the ecological and
climatic conditions, and the zoogeographical character, of this
region may have been quite different than now. It appears

Kurochkin: Pliocene Mongolian Rails

73

that at the end of the Middle Pliocene the climate of Western
Mongolia was not as continental, with milder winters than
occur there today. The lakes were probably not as salty, and
their shores and the valleys of rivers and streams were covered
with rich grass and bush vegetation. Additional evidence of
such an environment are the large numbers of waterfowl and
gallinaceous birds found in the same deposits.

ACKNOWLEDGMENTS

I extend thanks to E.V. Devjatkin and V.S. Zazhigin of the
Geological Institute of the USSR Academy of Sciences, and to
V I. Zhegallo of the Paleontological Institute of the USSR
Academy of Sciences, who contributed the fossil material for
the study. I am also grateful to T.D. Rakova and V.A. Pres-
njakov for preparing the illustrations, and to B.D. Delazari
for helping with the English translation.

LITERATURE CITED

Devjatkin, E.V. 1970. Geologia kainozoja Zapadnoi Mon-
golii. Trudy Sovmestnoi Sovetsko-Mongolskoi nauchno-
issledovatelskoi geologicheskoi ekspeditzii. 2:44-102.
Devjatkin, E.V., and V.I. Zhegallo. 1974. Novie dannie
o mestonahozhdenijah neogenovih faun severo-zapadnoi

Mongolii. Trudy Sovmestnoi Sovetsko-Mongolskoi pa-
leontologicheskoi ekspeditzii. 1:330-356.

Cracraft, J. 1973. Systematics and evolution of the Grui-
formes (Class Aves). 3. Phylogeny of the suborder Grues.
Bull. American Mus. Nat. Hist. 15 1(1): 1-127.

Feduccia, J. A. 1968. The Pliocene rails of North America.
Auk 85(3):441-453.

Kurochkin, E.N. 1971. K avifaune pliotzena Mongolii.
Trudy Sovmestnoi Sovetsko-Mongolskoi nauchno-issle-
dovatelskoi geologicheskoi ekspeditzii. 3:58-67.

. 1976. Novie dannie o ptitzah pliotzena Zapadnoi

Mongolii. Trudy Sovmestnoi Sovetsko-Mongolskoi pa-
leontologicheskoi ekspeditzii. 3:51-67.

Milne-Edwards, A.M. 1869-1871. Recherches Anato-
miques et Paleontologiques pour Servir a l’Histoire des
Oiseaux Fossiles de la France. T. 2, 632 ss., Atlas t. 2,
pi. 97-200. Paris, Librairie de G. Masson.

Olson, S.L. 1973. A classification of the Rallidae. Wilson
Bull. 85(4):38 1-4 16.

. 1974. The Pleistocene rails of North America. Con-
dor 76(2): 1 69— 175.

. 1977. A synopsis of the fossil Rallidae. Pp. 339-373

in Rails of the World by S.D. Ripley. D.R. Godine, Bos-
ton.

A NEW GOOSE FROM THE LATE PLIOCENE OF NEBRASKA
WITH NOTES ON VARIABILITY AND PROPORTIONS
IN SOME RECENT GEESE

By Larry D. Martin1 and Robert M. Mengel2

ABSTRACT: A nearly complete but moderately crushed skeleton of a late Pliocene (Blancan) goose

from the Broadwater Local Fauna of western Nebraska is described and named as Anser thompsoni new
species. This rather large goose was larger than Recent Anser caerulescens (Linnaeus) and smaller than
the larger races of Recent Branta canadensis (Linnaeus). Particularly distinctive characters include a
relatively short bill and a relatively very small furcula. The wing resembled A nser caerulescens and Anser
rossii Cassin in its relatively long ulna. The leg was relatively short as in Recent Anser albifrons (Scopoli)
and Branta canadensis.

Among Hildegarde Howard’s many contributions to avian
paleontology is her excellent review (1964) of the fossil anser-
iforms, in which she lucidly considered the strengths and
weaknesses of the fossil record of waterfowl and summarized
all that was then known. This has been helpfully supple-
mented by Woolfenden’s (1961) thorough study of the quali-
tative osteology of modern anseriforms.

We have recently begun to study the fossil birds of the Blan-
can Broadwater Local Fauna of western Nebraska. Most of
these are water birds, including a remarkably complete skel-
eton of a goose that appears to represent a new species of the
widespread Holarctic genus Anser Brisson.

The completeness of this specimen provided the rare op-
portunity for a fairly accurate reconstruction of the body pro-
portions of a fossil species. Comparing these with those of
Recent geese, however, posed problems. Although Verheyen
(1955a, 1955b) has extensively surveyed the rather variable
proportions of Recent geese, his sample sizes (one or two, rare-
ly up to four specimens per species) provide little indication of
the limits and nature of variation. For comparative purposes
we have therefore been obliged to undertake a limited analysis
of the relevant aspects of variation in several Recent geese
presently represented by major populations in interior North
America.

The description of this unusually complete fossil seemed es-
pecially appropriate for the present volume in recognition of
Howard’s long interest in and study of fossil waterfowl and in
appreciation of her often stated and intelligently tempered con-
cern (e.g., Howard 1964:235-237) about the problems posed
in study of often fragmentary individual specimens.

Only two Blancan local faunas have extensively studied avi-

1 Associate Curator of Vertebrate Paleontology, Museum of Natural
History, and Associate Professor of Systematics and Ecology; The
University of Kansas, Lawrence, Kansas 66045

2 Curator of Birds, Museum of Natural History, and Professor of Sys-
tematics and Ecology; the University of Kansas, Lawrence, Kansas
66045.

faunas: the Rexroad Local Fauna ( sensu lato ) and the Hag-
erman Local Fauna. Feduccia (1975) has summarized the in-
formation pertaining to birds for these two faunas. Other
Blancan sites are probably also rich in bird material; this is
certainly true of the Broadwater Local Fauna (for correlation
chart see Schultz and Martin 1977), which is one of the least
studied.

In addition to the birds, the Broadwater Local Fauna has
a large mammalian component (list modified from Schultz and
Stout 1948:563-564): Sorex sp. (shrew), Paramylodon sp.
(ground sloth), Megalonyx sp. (ground sloth), Hypolagus sp.
(rabbit), Spermophilus sp. (ground squirrel), P aenemarmota
barbouri Hibbard and Schultz (giant ground squirrel), Geomys
sp. (pocket gophers), Procastoroides sweeti Barbour and
Schultz (giant beaver), Castor sp. (beaver), Peromyscus sp.
(white-footed mouse), Neotoma sp. (wood rat), Pliopotamys
meadensis Hibbard (Extinct muskrat), Pliophenacomys sp.
(extinct vole), Pliozapus ? sp. (jumping mouse), Canis lepopha-
gus Johnston (extinct coyote), Borophagus diversideus (Cope)
(extinct canid), Satherium piscinaria middleswarti Barbour
and Schultz (extinct otter), Lutravus sp. (extinct mustelid), Is-
chorosmilus crusifonti Schultz and Martin (scimitar-toothed
cat), Stegomastodon mirificus (Leidy) (short-jawed mastodon),
Pliomastodon sp. (ancestral American mastodon), Equus (Dol-
ichohippus) simplicidens (Cope) (extinct horse), N annippus sp.
(extinct horse), Platygonus sp. (peccary), Camelops sp. (extinct
camel), Tanupolama sp. (extinct camel), Titanotylopus spatulus
(Cope) (giant camel), Capromeryx arizonensis schultzi Skinner
(ancestral pronghorn).

The presence of the more advanced muskrat Pliopotamys
meadensis in the Broadwater Local Fauna and of the less ad-
vanced P. minor in the Hagerman Local Fauna suggests that
the Broadwater is the younger of the two. These faunas are
near the age of, or somewhat younger than, the earliest evi-
dence of extensive continental glaciation (Boellstorff 1976;
Mercer 1978), approximately 3.5 million years ago.

The Broadwater fossils come from unconsolidated sands and

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 330:75-85.

76

Martin and Mengel: Pliocene Goose

fine-grained silts that were deposited in or adjacent to a large
Pliocene river system. The goose skeleton was found in the
“marly” facies, suggesting an area of local ponding. In asso-
ciation with it were many specimens of the extinct muskrat
Pliopotamys Hibbard, and it probably also shared its habitat
with the giant aquatic beaver Procastoroides Barbour and
Schultz and the extinct river otter Satherium Gazin.

MATERIAL

In addition to the fossil (see Systematics section), the ma-
terial studied consisted of relevant specimens from the neo-
osteological collections of the University of Kansas Museum
of Natural History, Division of Birds. Available was a series
of 17 male and 12 female “lesser” Snow Geese from Kansas
and Missouri (Anser caevulescens caerulescens (Linnaeus)),
migrant representatives of the population breeding on the
southwestern shore of Hudson’s Bay (Palmer 1976:137, 140).
These provided a comparatively good picture of variability in
a fairly homogeneous population of modern geese. Also avail-
able were a good series (7 males, 8 females) of Ross’s Goose
(Anser rossii Cassin), a heterogeneous assemblage of Canada
Geese (Branta canadensis (Linnaeus)), including 8 sexed ex-
amples of several of the larger subspecies and several addi-
tional specimens of the smaller ones, 3 Brant (B. bernicla
(Linnaeus)), and 2 White-fronted Geese (A. albifrons (Scopoli)).
These were supplemented by a loan of selected elements of a
Greylag (A. anser (Linnaeus)) and a Bean Goose (A. fabalis
(Latham)) from the National Museum of Natural History,
Smithsonian Institution, and measurements taken for us in
that museum of 3 additional Canada Geese and 3 more White-
fronted Geese.

Because of the nature and preservation of the fossil’s ele-
ments, the literature proved adequate for consideration of the
fossil forms that seemed relevant, primary sources being al-
most all at hand, as well as having been conveniently sum-
marized by Howard (1964).

Terminology is that of Howard (1929). All measurements
were taken to the nearest 0.1 mm with dial calipers.

SYSTEMATICS
Order Anseriformes

Family Anatidae — Ducks, Geese, and Swans
Genus Anser Brisson 1760

Anser thompsoni new species

Figures 1-4

HOLOTYPE: Most of a skeleton UNSM (University of
Nebraska State Museum) 1110, including skull, furcula, cor-
acoid, humerus, carpometacarpus, tibiotarsus, and other ele-
ments. Collected in 1939 by Joseph Johnson.

LOCALITY AND HORIZON: Broadwater Quarry 4 (NE
14, Sec. 20, T.19N., R.47W. — on the Dan Bowman ranch,
8.4 km E and 1.2 km N of Broadwater, Morrill County, Ne-
braska), from the Lisco Member, Broadwater Formation, late
Pliocene (Blancan). A cast of this specimen is on deposit at the
University of Kansas Museum of Natural History (KU 24669).

DIAGNOSIS: A moderately large goose, smaller overall
than the larger subspecies of modern Branta canadensis but
larger than any known subspecies of Anser caerulescens or A.
albifrons.

Compared with all other geese examined, Anser thompsoni
is characterized by having bill from fronto-nasal hinge to tip
shorter in relation to skull length (total length/bill length 2.13;
1.87-1.97 in Recent geese seen). Posterior rami of lower jaw
deeper.

Furcula relatively very small.

Humerus with impression of M. brachialis anticus more
elongate and more nearly parallel with the shaft; attachment
of anterior articular ligament nearly circular (oval in all other
geese examined); pectoral attachment not extending as far an-
conad.

Carpometacarpus with the process of metacarpal I longer
and narrower, with its proximal edge angling proximad (as
opposed to vertically or distally); distal metacarpal symphysis
relatively and absolutely longer.

Tibiotarsus with the tendinal foramen beneath the supra-
tendinal bridge circular rather than oval, and the groove for
M. peroneus profundus relatively broad, shallow, and short.

MEASUREMENTS (in mm): Skull: from posteromost
point of supraoccipital to fronto-nasal hinge (articulation with
beak) 63.9; beak from this point to tip 56.4 (total 120.3). Lower
jaw: depth of right mandibular ramus at posterior margin of
coronoid process 14.4. Coracoid (left): anterior end of brachial
tuberosity to middle of furcular facet 68.4. Humerus: total
length 167.0 (left); greatest width near head, at right angles to
shaft 36.3 (right). Ulna (left, on slab): estimated length 166 ±
1 (the olecranon process and the condyles are somewhat erod-
ed). Carpometacarpus (left): total length 97.8; apex of process
of metacarpal I to posterior margin of internal carpal trochlea
23.7. Proximal phalanx of digit II: length 40.0. Second phalanx
of digit II: length 27.2. Femur (left, on slab): estimated length
83 ± 1. Tibiotarsus (left): length from external condyle to prox-
imal articulating surface (i.e. , exclusive of cnemial crests)
133.9; width across condyles 15.5; greatest anteroposterior
depth between condyles 11.0. Tarsometatarsus (right, on slab):
length 90.3.

ETYMOLOGY: Anser thompsoni is named for Max Clyde
Thompson, Southwestern College, Winfield, Kansas, in grate-
ful recognition of his many contributions to the bird collection
of the University of Kansas Museum of Natural History, in-
cluding its extensive osteological component.

DESCRIPTION: UNSM 1110 is a nearly complete, but
moderately crushed skeleton (Fig. 1). Crushing, while not af-
fecting most of the articular surfaces, has generally flattened
the shafts of the long bones, precluding meaningful measure-
ments of their diameters. The bones, while associated, are
generally not in precise articulation. The general orientation
is with the left side up. The head and neck are extended and
the limbs are folded into a mass, rendering their removal and
study difficult. The humeral end of the right coracoid and a
fragment of the synsacrum have been displaced to the region
of the head. While the sternum is missing, the undamaged
furcula is nearly in correct anatomical position. The relation-
ship of the elements is not inconsistent with the notion that
the breast, viscera, and hip region were removed by some
predator prior to burial.

Present in some form are all of the long bones; the shoulder
girdle lacking the sternum; and the skull, lower jaw, and many
of the cervical vertebrae. The synsacrum is missing, except for
the above-mentioned fragment (which was removed from the
slab (Fig. 1) before the photograph was taken), as are most of
the ribs. The following elements, at least one of nearly every

Martin and Mengel: Pliocene Goose

77

Figure 1. The holotype, UNSM 1110, of Anser thompsoni new species approximately as it appeared in situ

one successfully removed from the matrix and associated
bones, are sufficiently well preserved to permit meaningful
comparisons.

Skull. This is crushed and flattened in a plane between ver-
tical and horizontal, but considerable detail is preserved, es-
pecially on the under (right) side (Fig. 2). In every respect it
resembles Anser rather than Branta. The relative size is com-
parable to the present-day Snow Goose, but the bill is decid-
edly short relative to total length. Nareal opening ovoid, rel-
atively short and broad, like A. caerulescens rather than B.
canadensis. Profile of forehead and proximal bill straight as
in living species of Anser (concave in B. canadensis ); supraor-
bital region not noticeably excavated as in Anser (in B. can-
adensis this is moderately excavated, presumably for a nasal
gland); orbit large and lachrymal short and broad as in Anser ;
lateroventral margin of maxilla moderately arched as in A.
caerulescens andd. albifrons (it is nearly straight in B. can-
adensis). (The so-called “grin patch” on the bill of the Recent
Snow Goose appears to be a ramphothecal feature and not an
osteological one.)

Lower jaw. Rami short and massive (Fig. 2) with the pos-
terior portion deeper than in other geese examined. Dentary

curved as in Anser caerulescens. Coronoid process with a
straight anterior margin as in A . caerulescens andd. albifrons,
and a gently sloping posterior margin as in Branta canadensis
andd. caerulescens (A. albifrons has a more abruptly curving
posterior coronoid margin). Posterior margin of dentarv below
the lateral process of the surangular as in Anser (anterior to
the lateral process in Branta).

Cervical vertebrae. These suggest a neck of average length
for a species of Anser of comparable size. A number are miss-
ing; only 10 are visible on the slab. Recent species of Anser
have 18, 19, or 20 (Verheyen 1955b: 10-1 1).

Furcula. Somewhat warped and distorted, the furcular rami
of Anser thompsoni are relatively thick, relatively short, and
relatively uncurved anteroposteriorlv. Somewhat surprisingly,
the wishbone was clearly smaller than those of Recent d. cae-
rulescens, which is considerably smaller than the fossil in other
measurements; it is little larger than that of a large A. rossii.
In short, the fossil seems to have had a remarkably small
furcula, indeed, relative to Recent geese examined. Qualita-
tively the element seems closest to that ofd. albifrons. (The
furcula of Branta canadensis differs from all of those discussed

78

Martin and Mengel: Pliocene Goose

Figure 2. Upper, restoration of the skull and lower jaw of Anser
thompsoni new species (x%). Lower, right side of the skull of the
holotvpe of Anser thompsoni new species.

in having the rami relatively long and set comparatively close
together.)

Scapula. The one scapula present is robust in comparison
with those of other geese examined and its pneumatic foramen
is relatively very small. The posterior end is missing.

Coracoid. As do all of the living species of Anser, A. thomp-
soni has, just posterior to the brachial tuberosity, a distinct
fossa that contains pneumatic foramina (Fig. 3b). However,
the foramina are not as large and numerous as they are in
modern species of Anser. In Brant a canadensis (and its dimin-
utive relative B. bernicla), the size and number of pneumatic
foramina reaches an extreme; the whole furcular facet is gen-
erally undercut with pneumatic foramina along its entire pos-
terior margin (noted by Woolfenden 1961:49). In .4. thompsoni
the furcular facet is less depressed than in A. caerulescens.
The procoracoid is short and blunt, and the glenoid facet is
narrow and AnserAWse (less nearly circular than in Branta ).
Breakage precludes measurement of its greatest length, but
the coracoid is relatively long, approximately equal to that of
B. canadensis specimens that are considerably larger in most
other dimensions.

Humerus. Humeri badly crushed but entire (Fig. 3c). They
appear to have been robust. Impression of M. brachialis an-
ticus more elongate and more nearly parallel to the shaft than
in other geese examined (one specimen of Old World Anser

anser approaches it in this respect); attachment of anterior
articular ligament nearly circular (oval in other geese, so that
its proximal border extends well proximad to the external con-
dyle); ectepicondyle narrow as in A. caerulescens', pectoral at-
tachment does not project as far anconad as in other geese;
pneumatic foramen (proximal end) small and capital groove
relatively small.

Ulna. The left ulna, exposed on the slab, possesses no rel-
evant features other than length sufficiently intact for com-
parisons.

C arpometacarpus . Relatively rather long (Fig. 3a). Internal
carpal trochlea does not project as far beyond the plane of
metacarpal III as in Recent geese; internal ligamental fossa
more nearly circular (less oval) than in other geese studied;
process of metacarpal I longer and narrower than in modern
geese, its proximal profile angling proximad (it angles verti-
cally or posteriad in living geese); extensor attachment rela-
tively large (resembling Branta esrneralda Burt in these par-
ticulars); distal metacarpal symphysis longer, absolutely and
relatively, than in other geese examined.

Proximal phalanx, digit 11 Comparatively large and broad,
not as elongate, relatively, as in Branta, but too badly crushed
to make detailed comparison useful.

Distal phalanx, digit II. This element resembles the com-
parable one in Anser caerulescens.

Femur. Still attached to slab. Imperfect and badly crushed
but comparatively large and (though not precisely measurable)
at least as long as that of a fairly large specimen of Branta
canadensis, which is considerably larger in most other dimen-
sions.

Tibiotarsus. Distal end and about two-thirds of the shaft of
the left tibiotarsus present (Fig. 4a). The proximal end, badly
crushed, could not be saved during removal; however, the
total length was ascertained. The element is relatively very
short. Also of interest is the shape of the distal articular surface
when viewed end-on. In Anser thompsoni andT. caerulescens
it is more compressed than in Branta canadensis and A. al-
bifrons (the width across the condyles divided by the depth of
the shaft between them in randomly selected specimens gives
ratios of 0.61 in B. canadensis, 0.63 inT. albifrons, and 0.72
in both A. caerulescens and A. thompsoni). That this feature
is rather variable among living geese, however, is indicated by
ratios of 0.69 and 0.79 respectively for single specimens of A.
fabalis and .4. anser. The tendinal foramen beneath the su-
pratendinal bridge is circular in A. thompsoni, rather than
oval as in other geese. The external condyle is relatively round-
ed, and the groove for M. peroneus profundus is relatively
broad, shallow, and short.

Tarsometatarsus. Relatively elongate as in Anser caerules-
cens (Fig. 4b). Detailed characters not well preserved.

DISCUSSION

Comparisons with Extinct Taxa

No extinct goose seems particularly close to Anser thompsoni
on the basis of available evidence. The fossil species of greatest
chronological and systematic relevance are represented by only
one or a few elements (some of them referred to these species
by later workers), most of which are not directly comparable.

Nearest in age is Anser pressus Wetmore (1933), known by
a femur (length 66.9 mm) from the Hagerman Lake Beds

Martin and Mengel: Pliocene Goose

79

Figure 3. Some elements of the holotype of A user thompsoni new species ( x 1 ). a, left carpometacarpus, in internal and external view, b, left
coracoid, in dorsal, ventral, and internal view, c, left humerus, in palmar and anconal view.

(Blancan) of Idaho. Although the crushed condition of the
present specimen prevents comparisons other than in size, A.
thompsoni was a much larger bird (femur approximately 83
mm). It was also much larger than Branta propinqua Shufeldt
(1892) from the late Pleistocene of Oregon (humerus 106.2 ver-
sus 167.0 mm) and B. esmeralda Burt (1929) from the Miocene

of Nevada (carpometacarpus 77.6 versus 97.8 mm). The latter
species also seems to have had a relatively shorter carpometa-
carpus. The shape of the process for digit II is similar in A.
thompsoni and B. esmeralda. Eremochen Brodkorb and Het-
erochen Short (Brodkorb 1961:174; Short 1970) are distinctive
extinct genera that do not appear to be closely similar to Bran-

80

Martin and Mengel: Pliocene Goose

Figure 4. Some elements of the holotype of Anser thornpsoni (XI). From left to right: anterior, internal, and posterior view of left tibiotarsus and
left tarsometatarsus, posterior view.

ta or Anser. Presbychen abavus Wetmore (1930) from the Mio-
cene of California and B. dickeyi L. Miller (1924) from the
Pleistocene of California and Oregon are giant species much
larger than Anser thornpsoni. Anser e quit um Bate (1916), from
the late Pleistocene of Malta, if all its elements belonged to the
type, was a weirdly proportioned bird perhaps incapable of
flight. Brant a howardae L. Miller (1930) from the late Miocene
is based solely on the distal end of a carpometacarpus and few
comparisons are possible. The distal metacarpal symphysis,
however, is shorter than that of A. thornpsoni. Anser azer-
baidzhanicus Serebrovsky (1940), from the Pleistocene of Azer-
baidzhan, USSR, apparently had a much larger cranium and
more bulging frontals (see Howard 1964:269).

Comparisons with Recent Taxa

Most if not all living geese have been recorded from the
Pleistocene and one or two from earlier time (Brodkorb
1964:234-237). Recent workers are increasingly reluctant,
however, to credit the existence of living species before the
Pleistocene (Brodkorb 1966; Feduccia 1975). Although Anser
thornpsoni shares osteological characters with various living
geese (see Systematics section), several are diagnostic, individ-
ually or in combination. Features of the carpometacarpus
alone separate it from Recent forms.

Because fossil remains of birds usually consist of a few ele-
ments at most, these often fragmentary, comparisons are gen-
erally restricted to a few characters and often to relative esti-
mates of general size. The relative completeness of Anser

thornpsoni, however, permits unusually extensive comparison,
at least with Recent geese (Tables 1 and 2).

SOME STATISTICAL CONSIDERATIONS: A full-scale
statistical study of proportions in modern geese, while much
to be desired, is far beyond the scope of this paper. Although
we restricted ourselves to readily available samples of the liv-
ing geese presently numerous in continental North America,
we think we have gained some insights into the probable range
of dimensions in Anser thornpsoni and its general proportions
compared with some familiar living forms. A few explanations
are necessary.

1. A modest, but significant, sexual difference in size re-
quired that males and females of both A. caerulescens and A.
rossii be separated in statistical analysis of direct measure-
ments (Table 1). Sexes of Branta canadensis andd. albifrons
were not so separated because the sample of the former in-
cluded several of the larger subspecies, while that of the latter
consisted of unsexed individuals. Hence both would be ex-
pected to show exaggerated variances.

2. We combined sexes when considering relative proportions
(ratios) because there was no evidence of significant sexual
dimorphism in any of these species. This improved the sample
sizes (Table 2).

3. Ratios, individually determined for each specimen and
summed, were treated statistically in the same way as (more
or less) normally distributed linear measurements. Although
numerous precedents exist in similar cases (e.g., Engels 1940
and citations therein), caution is required (Simpson, Roe, and
Lewontin 1960:163-165; Sokal and Rohlf 1969:17-18). In the

Martin and Mengel: Pliocene Goose

81

Table 1. Measurements (in mm) of various geese with means ± standard errors, standard deviations, and N (Anser caerulescens, A. rossii)\ or
mean, N, and range ( Branta canadensis, A. albifrons).*

Character

Anser
thompsoni
(new species)

Anser

caerulescens

Anser

rossii

Branta
canadensis
(large races)

Anser

albifrons

Anser

anser

Anser

fabalis

Skull + bill

120.3

66

116.3 ±
2.6 (15)

0.7

66

86.2 ±
1.8 (7)

0.7

120.5 (10)
116.2-128.9

105.6 (5)
102.2-110 .2

—

—

$9

111.8 ±
3.6 (8)

1.3

9 9

83.7 ±

3.7 (8)

1.4

Bill

56.4

66

61.8 ±
5.2 (15)

1.3

6 6

44.4 ±
0.9 (7)

0.3

63.2 (7)
56.9-69.0

56.9 (2)
55.2, 58.5

—

—

99

58.0 ±
1.0 (8)

0.4

99

42.7 ±
2.6 (8)

1.0

Humerus

167.0

6 6

148.3 ±
4.9 (17)

1.2

6 6

128.0 ±
2.2 (6)

0.9

181.8 (11)
168.6-194.9

147.9 (5)
143.8-152.7

117.6 (1)

161.4 (1)

99

144.0 ±
3.9 (12)

1.1

9 9

124.4 ±
3.9 (8)

1.5

Ulna

166 ± 1
(estimate)

66

148.3 ±
4.9 (17)

1.2

6 6

127.3 ±

3.3 (7)

1.3

171.8 (11)
157.3-182.6

143.0 (5)
138.3-146.2

—

—

99

143.5 ±
4.4 (12)

1.3

9 9

124.0 ±
4.1 (7)

1.7

Carpometarcarpus

97.8

6 6

83.1 ±
3.5 (16)

0.9

6 6

73.4 ±

2.4 (7)

0.9

101.1 (10)
92.9-104.8

83.1 (5)
78.9-84.7

98.5 (1)

91.3 (1)

99

80.0 ±
2.5 (12)

0.7

99

70.4 ±
1.9 (8)

0.7

Femur

83 ± 1
(estimate)

66

73.0 ±
2.8 (17)

0.7

62.5 ±
1.2 (7)

0.5

84.1 (8)
77.8-91.9

71.1 (2)
68.4, 73.9

—

—

99

71.1 ±
1.2 (12)

0.3

9 9

61.0 ±
2.1 (8)

0.8

Tibiotarsus

133.9

66

132.5 ±
5.3 (17)

1.3

66

114.2 ±
2.3 (7)

0.9

148.6 (11)
140.7-159.8

121.7 (5)
118.7-124.2

149.6 (1)

133.4 (1)

99

129.9 ±
3.8 (12)

1.1

9 9

113.2 ±
3.0 (8)

1.1

T arsometatarsus

90.3

6 6

84.0 ±
4.7 (17)

1.1

6 6

71.6 ±
1.2 (7)

0.5

90.9 (11)
83.6-98.8

72.4 (5)
70.6-74.5

91.9 (1)

72.9 (1)

99

80.9 ±
2.6 (12)

0.7

99

68.8 ±
3.9 (8)

1.5

* The abridged statistical treatment of B canadensis and A. albifrons is explained in text.

present case, although the distributions of our ratios somewhat
resemble normal ones (their coefficients of variation average
somewhat smaller than those of the direct measurements), we
regard their confidence limits with some suspicion and have
avoided firm conjectures except where differences and t values
were considerable.

GENERAL SIZE: Relations among the Recent taxa and
their elements are observable in Table 1. Standard deviations
in populations of equal variability are directly proportional to
size (i.e. , the mean). Thus, in comparing the fossil with Recent
geese, one may infer the standard deviation (Fisher 1952) of
a single specimen from that of a near relative (e.g., Anser
caerulescens, assuming similar variability) and plot the theo-
retical ranges of its population within any chosen limits (here
±2cr or 95 percent of a theoretically normal population). This
may be done considering the single specimen either as an av-
erage, a very small, or a very large example (Simpson, Roe,
and Lewontin 1960:207-208). The results, for the absolute size
of the fossil humerus compared with that of the Snow Goose,
are shown in Figure 5a.

Clearly, if the holotvpe of Anser thompsoni is a relatively
very large example, the measurements of its population would
extensively overlap those of male A. caerulescens and other
geese of comparable size.

In short, while there is no question that Anser thompsoni
(humerus length 167 mm) averaged very much to considerably
larger than .4. caerulescens (mean length of humerus, 6 6, 144
mm), even given the seemingly large difference of 23 mm (16
percent above the Snow Goose mean), very extensive overlap
of individuals is not ruled out. This assumes particular im-
portance whenever size is the only meaningful character in a
comparison, and is especially worth noting among geese,
which are usually polytypic and often highly so (individual
Canada Geese from various populations range from 1.3 to more
than 10 kg in weight).

PROPORTIONS: Of special interest here are ratios involv-
ing the limbs and their components (Table 2), characters re-
lated to locomotion. Anser thompsoni lacks any conventional
index of “absolute” body size (e.g., trunk length, Engels
1941:63), so we have expressed the lengths of appendicular

82

Martin and Mengel: Pliocene Goose

Table 2. Proportions of various geese (as percent) with means ± standard errors, standard deviations, and N.

Character
(Ratios of lengths)

Anser
thompsoni
(new species)

Anser

caerulescens

Anser

rossii

Branta
canadensis
(large races)

Anser

albifrons

Anser

anser

Anser

fabalis

Skull/bill

213.2

190.3 ± 2.0
8.5 (23)

196.8 ± 1.6
5.7 (11)

192.0 ± 2.2
5.8 (7)

186.9 (2)
186.5, 187.3

—

—

Ulna/humerus

99.4

(estimate)

100.0 ± 0.5
2.8 (29)

99.0 ± 0.7
2.6 (11)

94.5 ± 0.4
1.2 (11)

96.7 ± 0.7
1.6 (5)

—

—

Carpometacarpus/humerus

58.6

55.9 ± 0.2
1.0 (28)

56.7 ± 0.2
0.5 (12)

56.0 ± 0.3
0.8 (10)

56.2 ± 0.6
1.2 (5)

55.5 (1)

56.6 (1)

Femur/humerus

49.7

(estimate)

49.4 ± 0.5
2.5 (29)

48.8 ± 0.2
0.7 (12)

46.4 ± 0.4
1.2 (8)

48.2 (2)

—

—

Tibiotarsus/humerus

80.2

89.9 ± 0.4
1.9 (29)

90.0 ± 0.5
1.8 (12)

81.6 ± 0.5
1.7 (11)

82.3 ± 0.8
1.8 (5)

84.2 (1)

82.7 (1)

Tarsometatarsus/humerus

54.1

56.5 ± 0.3
1.5 (29)

55.2 ± 0.4
1.3 (12)

50.0 ± 0.5
1.7 (11)

48.9 ± 0.5
1.1 (5)

51.7 (1)

49.5 (1)

“Index of locomotion”*

140.2

(estimate)

130.7 ± 0.5
3.0 (28)

131.7 ± 0.7
2.2 (11)

140.4 ± 1.5
3.8 (7)

141.0 (2)
141.0, 141.0
145**, 146**

142**

Humerus/trunk

—

77.6 (5)

80.6 (2)

76.6 (2)

79.4 (2)

—

—

* Humerus + Ulna + Carpometacarpus/Femur + Tibiotarsus + Tarsometatarsus.
** From Verheyen, 1955b.

elements as percentages of the length of the humerus. This
element has been found to be a fair indicator of general body
size in comparatively homogeneous groups of anseriforms
(Humphrey and Clark 1964:189).

In assessing the validity of the humerus as an indicator of
absolute size, we measured the trunk lengths of a few of the
Recent geese in our samples and obtained humerus/trunk
length ratios (Table 2). The differences found between species
in these small samples are uniformly insignificant (p < 0.20 to
0.40). Trunk length is difficult to measure even in articulated
specimens and prohibitively difficult in disarticulated ones.

For present purposes, therefore, we accept the humerus as
at least a moderately good standard of overall size in these
geese. Even if it were not, the ratios presented here provide
direct intramembral ratios for wing elements and standard-
ized, indirect ones for hind limb elements and intermembral
comparisons. As such, all of them are directly comparable
among the species considered.

Three sets of relationships are of special interest.

1. The length of the hind limb and its elements. Anser cae-

rulescens and A. rossii, virtually sibling species, are distinctly
long-legged geese. This is clearly perceptible in the field when
the birds are standing. The “index of locomotion” (i.e. , wing
length/leg length; see Verheyen 1955b: 10—12) ford, caerules-
cens compared with Branta canadensis in the present samples
shows this very clearly (p < 0.001). Anser thompsoni differs
from A. caerulescens in this respect, being short-legged like
the Canada Goose (0.001 < p < 0.01), a difference also re-
vealed by comparison of the tibiotarsus/humerus ratios (Fig.
5b).

The relative length of the leg in these geese results princi-
pally from different combinations of tibiotarsal and tarso-
metatarsal lengths (Table 3), since their femora seem to be
comparatively uniform in relative size. In long-legged Anser
caerulescens (and A. rossii, which, being relatively similar in
most respects is omitted from further comparisons), the tibio-
tarsus and tarsometatarsus are both long. In short-legged
Branta canadensis and A. albifrons (and probably in A. anser
and A. fabalis), both elements are short. The fossil also has a
short leg, but this results from a very short tibiotarsus offsetting

Table 3. Differences and their significances for each comparison of the relative length of the tibiotarsus (lower left) and tarsometatarsus (upper
right) among Anser thompsoni and several Recent geese.*

Anser

thompsoni

Anser

caerulescens

Branta

canadensis

Anser

albifrons

Anser thompsoni

—

longer

shorter

shorter

n.s.

0.02 < p < 0.05

0.01 < p < 0.02

Anser caerulescens

shorter

—

shorter

shorter

p < 0.001

p < 0.001

p < 0.001

Branta canadensis

shorter

longer

—

shorter

n.s.

p < 0.001

n.s.

Anser albifrons

shorter

longer

shorter

—

n.s.

p < 0.001

n.s.

* Read down (e.g. , the tibiotarsus of A. thompsoni is relatively shorter than that of A. caerulescens, etc.). The letters n.s. stand for not significant.

Martin and Mengel: Pliocene Goose

83

Figure 5. Some statistics of A user thompsoni and Recent taxa. A, length of humerus, compared with ,4. caerulescens caernlescens males. B,
length of tibiotarsus/length of humerus, compared with A. c. caerulescens. C, length of ulna (estimatedl/length of humerus, compared with Branta
canadensis. For Recent taxa: white boxes = mean ± 2 standard errors; black rectangles = mean ± 2 standard deviations; vertical lines = means.
For A. thompsoni new species, M, , M2, and M3 = theoretical means with the holotype regarded, respectively, as average, 2a below, and 2a above
the mean (a inferred from A. caerulescens)', horizontal line = M, ± 4a. The horizontal scale is logarithmic. For details see text.

the lesser effect of a comparatively long tarsometatarsus . In

Table 3 and the text, probabilities are those resulting from
Student’s t test of the difference between means (for its variant
formula for comparing single specimens with samples, see
Simpson, Roe, and Lewontin 1960:182-183; or Sokal and
Rohlf 1969:224-225).

It is difficult to assess these differences in a functional con-
text with present knowledge. Our initial tendency to regard
the long legs of the Snow Goose as an adaptation for grazing
per se was dampened by indications (Palmer 1976:150, 233)
that the Snow Goose is more addicted than the Canada Goose
(and probably more than the White-fronted Goose) to aquatic
feeding. We now hazard the guess that short legs in geese are
related to grazing on comparatively bare ground or in low
cover on land or in shallow water, where a long neck, as in
Branta canadensis, usefully increases the area covered. Longer

legs appear better suited to marshy substrates and higher cov-
er. A user thompsoni seems to belong in the first category.

2. Relative lengths of forelimb elements. Anser caerulescens
and A. rossii have long ulnae in relation to their humeri (Ver-
heyen 1955b: 10-12). In our samples this relative difference in
length is highly significant when either is compared with Bran-
ta canadensis (p < 0.001). Anser albifrons has an intermediate
ulna/humerus index, but its ulna, relatively, is still signifi-
cantly shorter than that of the Snow Goose (0.01 < p < 0.02).
With its long ulna, A. thompsoni resembles A. caerulescens,
in comparison (Fig. 5c) with/?, canadensis (0.001 < p < 0.01).
The length of the ulna of A. thompsoni is approximate, but
even if it were 2 mm shorter than estimated, or 164 mm, it
would still be significantly longer than that of B. canadensis
(p < 0.02).

The carpometacarpus of Anser thompsoni is also relatively

84

Martin and Mengel: Pliocene Goose

longer than those of Recent geese considered, which are rather
uniform in this regard. Compared with A. caerulescens, this
difference in length is significant (0.01 < p < 0.02).

Proportions in wing elements are related to the mechanics
of flight in ways that are often obscure. Further complications
are introduced by feather geometry, which is little studied in
present-day forms and unknown in nearly all fossils (see Engels
1941 on hawks and Fisher 1946 on New World vultures).

All anseriforms, however, are comparatively short-winged
and heavily wing-loaded, conditions resulting in rapid, flap-
ping flight. Their wings are therefore in the same general cat-
egory (Boker 1927) as those of falcons and swallows
(“Schwirrflug”). With this wing-form, speed and maneuver-
ability are evidently increased by lengthening of the outer ele-
ments, particularly those of the hand, and the primary feath-
ers. In geese, this would mean a more duck-like wing.
(Mengel, after many hours of field observation, thinks that the
Snow Goose is more agile and maneuverable aloft than the
Canada Goose.) Further, the proximal extension and relative
length of the extensor attachment of the process of metacarpal
I in Anser thompsoni, in comparison with all Recent geese, are
features associated with maneuverability and quick response.
These features are not as well adapted to sustained heavy work
loads (Fisher 1946:559), however, as the more distally situated
and elongate processes of Recent geese and, particularly,
swans.

From these considerations we conjecture that A. thompsoni
was probably a rapid and, for its size, a comparatively ma-
neuverable flier. Its skeleton, as noted above, also appears to
have been less pneumatic than those of the modern geese we
have studied. We hestitate to assess the significance of this,
particularly since Prange et al. (1979) have recently posed
searching questions concerning the whole relationship of flight
and pneumaticitv. The very small furcula of A. thompsoni is
also of interest here because a very appreciable portion of M.
pectoralis major, the principal adductor of the wing, inserts
on this element. Conceivably this late Pliocene goose did not
perform annual migrations as long as those of Recent geese.
If this is true, the features discussed above may somehow be
related to long distance migratory flight.

3. Cranial proportions. The type oi Anser thompsoni is very
short-billed. Although the ratio of bill length to total head
length is rather variable in Recent geese, the difference be-
tween this ratio in the fossil and its mean in the Recent Snow
Goose is very significant (0.01 < p < 0.02). This, together
with inspection of the values for other Recent geese supports
the supposition that A. thompsoni was indeed a short-billed
goose. The only comparably short-billed Recent Anser is the
little A. canagicus, or Emperor Goose, a tidal flat grazer and
grubber of the Bering Sea region. We are not prepared to
consider the possible functional significance of these facts.

SUMMATION

While our total picture of Anser thompsoni is blurred by
imperfect preservation, it is still the most complete of any
known fossil goose. This bird seems to have been a big,
dumpy, short-legged, rather long-winged, short-billed Anser,
larger than a Snow Goose but resembling that species in flight
silhouette and maneuverability. In size certainly, and in gen-
eral appearance probably, it resembled the Old World Greylag
(A. anser). It may not have been highly migratory.

The proportional configurations of Recent geese are rather
variable and doubtless represent subtle adaptations to behav-
ioral modes that could easily be rather plastic. Anser thomp-
soni displays a unique and perhaps rather specialized combi-
nation, particularly in its short tibiotarsal-long tarsometatarsal
combination.

The carpometacarpus is also intriguing, specifically the
shape of the process of metacarpal I, which distinguishes Anser
thompsoni from at least the Recent species of both Anser and
Branta, and which we suspect may be the primitive condition
for geese. Whether or not improved knowledge of fossil geese
supports this position, we suspect that further information on
this and other characters will require some systematic realign-
ments consistent with the distribution of shared-derived char-
acters when, and if, these characters can be identified.

ACKNOWLEDGMENTS

We thank Mike Voorhies and C.B. Schultz for permission to
study the fossil birds in the University of Nebraska State Mu-
seum, and Orville Bonner for skillful preparation of the ho-
lotvpe of Anser thompsoni. Jeannie Robertson and Dawn
Adams provided the drawings, except Figure 5 for which we
thank Caryn Lowther. John Chorn assisted in various ways.
Jessica Harrison and J.D. Stewart assisted in checking refer-
ences.

We also thank John Bucher and Tracy Lynn Mengel for
assistance with measurements of Recent geese, the former on
a visit to the National Museum of Natural History, Smithson-
ian Institution; Richard Zusi for a loan from the same museum;
Norman A. Slade for statistical consultation; and Marion A.
Jenkinson for a helpful reading of the manuscript.

LITERATURE CITED

Bate, D.M.A. 1916. On a small collection of vertebrate re-
mains from the Har Dalam Cavern, Malta; with note on
a new species of the genus Cygnus. Proc. Zool. Soc. Lon-
don 1916:421-430.

Boellstorff, J. 1976. The succession of late Cenozoic vol-
canic ashes in the Great Plains; a progress report. Guide-
book to the 24th Annual Meeting, Midwestern Friends of
the Pleistocene. Guidebook Series 1. Stratigraphy and
Faunal Sequence — Meade County, Kansas.

Boker, H. 1927. Die biologische Anatomie der Flugarten der
Vogel und ihre Phvlogenie. Jour. f. Orn. 75:304-371.
Brodkorb, P. 1961. Birds from the Pliocene of Juntura,
Oregon. Quart. Jour. Florida Acad. Sci. 24:169-184.

. 1964. Catalogue of fossil birds, Part 2 (Anseriformes

through Galliformes). Bull. Florida State Mus. Biol. Sci.
8:195-335.

. 1966. Did living species of birds arise in the Pliocene?

XIV Internat. Orn. Congress, Abstracts:43-44.

Burt, W.H. 1929. A new goose (Branta) from the lower
Pliocene of Nevada. Univ. California Publ., Bull. Dept.
Geol. Sci. 18:221-224.

Engels, W.L. 1940. Structural adaptations in thrashers
(Mimidae: Genus Toxostoma) with comments on interspe-
cific relationships. Univ. California Publ. Zool. 42:341-
400.

. 1941. Wing skeleton and flight of hawks. Auk 58:61-

69.

Martin and Mengel: Pliocene Goose

85

Feduccia, A. 1975. Professor Hibbard’s fossil birds. Univ.
Michigan Mus. Paleontology, Misc. Pap. No. 1 2(3):67—
70.

Fisher, H I. 1946. Adaptations and comparative anatomy
of the locomotor apparatus of New World vultures. Amer.
Mid. Nat. 35:545-727.

. 1952. The validity of the fossil crane Grus nannodes.

Condor 54:205-206.

Howard, H. 1929. The avifauna of Emeryville Shellmound.

Univ. California Publ. Zool. 32:301-394.

. 1964. Fossil Anseriformes. Pp. 233-326 in The Wa-
terfowl of the World by J Delacour. Vol. 4. Country Life
Ltd. London.

Humphrey, P.S., and G.A. Clark, Jr. 1964. The anatomy
of waterfowl. Pp. 167-232 in The Waterfowl of The
World by J. Delacour. Vol. 4. Country Life Ltd. London.
Mercer, J.H. 1978. Age of earliest mid-latitude glaciation.
Nature 274:926.

Miller, L. 1924. Branta dickeyi from the McKittrick Pleis-
tocene. Condor 26:178-180.

. 1930. A fossil goose from the Ricardo Pliocene. Con-
dor 32:208-209.

Palmer, R.S. 1976. Handbook of North American Birds.

Vol. 2. Yale Univ. Press. New Haven and London.
Prange, HD, J.F. Anderson, and H. Rahn. 1979. Scal-
ing of skeletal mass to body mass in birds and mammals.
Amer. Nat. 113:103-121.

Schultz, C.B., and L.D. Martin. 1977. Biostratigraphy
of the Neogene-Quaternary boundary in North America.
Ann. Mus. Geol. Bologna (ser. 29) 41:285-295.

Schultz, C.B., and T.M. Stout. 1948. Pleistocene mam-
mals and terraces in the Great Plains. Bull. Geol. Soc.
Amer. 59:553-588.

Serebrovsky, P.V. 1940. New species of birds from Bina-
gadv beds. Comptes Rendus (Doklady) Acad. Sci.
U.R.S.S. 27(7):766— 768.

Short, L.L. 1970. A new anseriform genus and species from
the Nebraska Pliocene. Auk 87:537-543.

Shufeldt, R.W. 1892. A study of the fossil avifauna of the
Equus beds of the Oregon desert. Jour. Acad. Natl. Sci.
Philadelphia 9:389-425.

Simpson, G.G., A. Roe, and R.C. Lewontin. 1960. Quan-
titative Zoology. Rev. ed. Harcourt, Brace and Co. New
York. 440 pp.

Sokal, R.R., and F.J Rohlf. 1969. Biometry. W.H. Free-
man and Co. San Francisco.

Verheyen, R. 1955a. La systematique des Anseriformes ba-
see sur l’osteologie comparee. Bull. Inst, royal des Sci.
nat. de Belgique 31 (nr. 35): 1—18.

. 1955b. La systematique des Anseriformes basee sur

l’osteologie comparee (suite). Bull. Inst, royal des Sci. nat.
de Belgique 31 (nr. 36): 1—1 6.

Wetmore, A. 1930. Fossil bird remains from the Temblor
Formation near Bakersfield, California. Proc. California
Acad. Sci. 4th Ser. 19:85-93.

. 1933. Pliocene bird remains from Idaho. Smiths.

Misc. Colls. 87(20): 1-12.

Woolfenden, G.E. 1961. Postcranial osteology of the wa-
terfowl. Bull. Florida State Mus., Biol. Sci. 6:1-129.

A NEW FOSSIL HERON (AVES: ARDEIDAE) FROM THE OMO
BASIN OF ETHIOPIA, WITH REMARKS ON THE
POSITION OF SOME OTHER SPECIES
ASSIGNED TO THE ARDEIDAE

By Pierce Brodkorb1

ABSTRACT: Ardea howardae new species, from the Pliocene/Pleistocene of Shungura, Ethiopia, is
the largest known fossil heron. Changes are made in the systematic position of some other fossils assigned
to the Ardeidae. Goliathia andrewsi Lambrecht of the Paleogene of Egypt is transferred to Balaenicipitidae
as the earliest member of that family. Ardea lignitum Giebel from the Miocene of Germany is transferred
to the family Strigidae, order Strigiformes, as Bubo lignitum (Giebel), new combination. Ardea brunhuberi
Ammon from the Miocene of Germany is transferred to the family Phalacrocoracidae, order Pelecani-
formes, as Phalacrocorax brunhuberi (Ammon), new combination; Phalacrocorax praecarbo from the same
locality is synonymized with it. Ardeacites molassicus Haushalter from the Miocene of Germany is placed
with the Aves Incertae Sedis.

RESUMEN: Ardea howardae , especie nueva del plioceno/pleistoceno de Etiopia, es la mas grande de
todas las garzas fosiles conocidas hasta la fecha. Ademas necesitan cambios en la posicion sistematica de
unos miembros o miembros presuntos de la familia Ardeidae. Goliathia andrewsi Lambrecht del eoceno/
oligoceno egipciano se refiera a la familia Balaenicipitidae. Ardea lignitum Giebel del mioceno de Alemania
esta traspasada a los buhos del orden Strigiformes con denominacion de Bubo lignitum (Giebel), combi-
nacion nueva. Ardea brunhuberi Ammon tambien del mioceno aleman no es garza, es corbejon (pato
coche) del orden Pelecaniformes; denominole con la combinacion nueva Phalacrocorax brunhuberi (Am-
mon). En conformidad de la ley de prioridad Phalacrocorax praecarbo de la misma formacion debe ser
sinonima del precedente. Ardeacites molassicus Haushalter, tambien del mioceno aleman, esta colocado
con las Aves Incertae Sedis.

The Omo River rises in the Ethiopian highlands and flows
southward to discharge into Lake Rudolf (elevation 370 m)
along the border between Ethiopia and Kenya. The Lake Ru-
dolf trough and its extension up the lower Omo Valley are part
of the Kenya Rift. Repeated volcanic episodes began in the
early Miocene (Butzer 1971), providing sediments suitable for
radiometric dating.

The littoral, alluvial, and deltaic beds in the Omo Basin
have been extensively studied by the international Omo Re-
search Expedition because of the presence of early hominid
remains (Brown, Heinzelin, and Howell 1970). The Omo
Group includes four formations. The eroded uppermost basalt
of the Mursi Formation has K/Ar dates of 4.05-4.25 million
years and is unconformably overlain by the Nkalabong For-
mation (3.90-3.99 million years). The Shungura Formation
includes many dates between 3.75 and 1.84 million years; its
upper members are thus contemporaneous with the classic Bed
I of Olduvai Gorge. The Kibish Formation has dates of Middle

1 Professor of Zoology, University of Florida, Gainesville, Florida
32611.

(130,000 B.P.) and Upper (30,000 B.P.) Pleistocene and Ho-
locene (9,500-3,250 B.P ).

Although great progress has been made in our knowledge
of the evolution of man in Africa, it remains the least known
continent in regard to its present-day unique and rich avifau-
na. Nevertheless, many thousands of avian fossils from East
Africa are now at hand and include members of nearly all the
nonpasserine families that occur in Africa today, as well as a
rich representation of the Passeriformes. This paper is the first
in a projected series on the Pliocene/Pleistocene birds from
African hominid localities. Reports on other taxa are planned
to appear at frequent intervals.

SYSTEMATICS

Order Ardeiformes (Wagler)

Family Ardeidae Vigors

The new heron is referable to the family Ardeidae by the
following characters: (1) upper portion of coracoid very wide;
(2) acrocoracoid (caput) slightly elevated; (3) coracohumeral
groove deep; (4) brachial tuberosity with very slight posterior

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 330:87-92.

88

Brodkorb: Ethiopian Fossil Heron

Table 1. Measurements (mm) of the coracoid of several species of Ardea.

Species

Length

Width
of Head

Depth
of Head

Width through
Glenoid Facet

Width through
Scapular
Facet and
Procoracoid

Maximum
Width of
Glenoid
Facet

A. goliath (3)

86.3-95.5

14.7-15.3

8. 3-8. 8

14.9-17.3

13.3-14.1

10.7-11.3

A. howardae (type)

(80)1

12.8

6.4

12.3

10.3

9.4

A. sumatrana (2)

73.6-73.8

11.9-12.4

5. 4-5. 6

11.4-11.9

9.5-10.3

7. 2-7. 8

A. cinerea (9)

56.4-63.4

10.3-11.9

5. 5-5. 9

10.3-11.9

8. 9-9.6

7. 2-7. 8

A. melanocephala (3)

59.7-60.5

10.5-11.2

5. 2-5. 9

10.5-11.0

8. 2-8. 7

6. 0-6. 9

A. pacifica (3)

47.3-48.1

10.3-10.4

4. 8-5. 2

9. 2-9. 5

8.3 (1)

6. 1-6.2

A. purpurea (2)

54.7-55.4

9.0-9. 2

4. 1-4.6

8. 7-9. 8

6. 9-7. 7

5.4-6. 7

' Estimate; length as preserved 40.4.

overhang; (5) glenoid facet almost flat; (6) scapular facet a
deep, round cup; (7) procoracoid well developed, curved, im-
perforate, and unnotched; (8) triosseal canal almost flat.

Genus Ardea Linnaeus

As I have been unable to find generic characters in the upper
part of the coracoid in the Ardeidae, the species is referred to
Ardea Linnaeus on the basis of its very large size, which ex-
ceeds that of all known herons except the living Ardea goliath
Cretzschmar of Africa and India (see Table 1).

Ardea howardae new species

Figure 1

HOLOTYPE: Upper (cephalic) half of left coracoid, Uni-
versity of California Department of Anthropology, No. F504-
27.

TYPE LOCALITY: Shungura, Omo Basin, southwestern
Ethiopia. Shungura Formation, Member G, unit 27; age about
1.94 million years.

DIAGNOSIS: Differs from A. goliath as follows: (1) size
about 20 percent less; (2) head less swollen but relatively wider,
extending laterally almost to level of scapular facet (narrower
in A. goliath, occupying only the medial two-thirds of the
width of the bone); (3) scapular facet with lower margin hor-
izontal, then swinging upward to glenoid facet, where it forms
a prominent notch (in A. goliath the anterior margin of the
scapular facet forms a smooth arc that continues to the glenoid
facet, where it cuts only a slight notch); (4) procoracoid much
less developed; (5) triosseal canal shallower.

Differs from Recent Ardea sumatrana Raffles of southeast
Asia to Australia as follows: (1) size larger; (2) head more swol-
len, and relatively and absolutely wider; (3) notch separating
scapular and glenoid facets much larger; (4) triosseal canal
shallower.

Differs from Recent Ardea cinerea Linnaeus of Eurasia and
Africa as follows: (1) size larger; (2) head more swollen; (3)
coracohumeral groove shallower; (4) glenoid facet nearly flat
(more concave in A. cinerea); (5) brachial tuberosity reduced,
merging gradually with shaft (forms a swollen lip that over-
hangs posterior face of shaft in A. cinerea); (6) procoracoid less
developed and less recurved.

Differs from Recent Ardea melanocephala Vigors and Chil-
dren of Africa as follows: (1) size larger; (2) head of coracoid
more swollen although of about the same relative width; (3)
scapular facet much wider with its lateral margin curved (al-

most vertical in A. melanocephala); (4) notch separating gle-
noid and scapular facets much larger.

Differs from Recent Ardea pacifica Latham of Australia as
follows: (1) size very much greater; (2) head more swollen but
of similar lateral extent; (3) inner corner of head merging grad-
ually with shaft (inner corner of head forms a swollen lip over-
hanging shaft in A. pacifica); (4) scapular facet much wider
and rounder (laterally compressed in A. pacifica, with lateral
and medial edges nearly vertical); (5) notch separating glenoid
and scapular facets much more developed; (6) procoracoid less
developed and slightly curved (well developed and straighter
in A. pacifica).

Differs from Recent Ardea purpurea Linnaeus of southern
Eurasia and Africa as follows: (1) size very much greater; (2)
head more swollen with its inner corner gradually merging
with shaft; (3) notch between glenoid and scapular facets larg-
er; (4) procoracoid less developed and less recurved.

Ardea howardae is very much larger than the known Plio-
cene herons. These are Ardea polkensis Brodkorb (1955) and
Nycticorax fidens Brodkorb (1963) from the early Pliocene of
Florida, Nyctanassa kobdoenus Kurochkin (1976) from the
early Pliocene of Mongolia, and Botaurus hibbardi Moseley
and Feduccia (1975) from the late Pliocene of Kansas.

Ardea rupeliensis Van Beneden (1873) from the Oligocene
of Belgium barely escapes being a nomen nudum and has been
relegated to the Aves Incertae Sedis (Brodkorb 1978).

ETYMOLOGY: On the occasion of the 52nd anniversary
of her association with the Natural History Museum of Los
Angeles County, I dedicate this species to my friend Dr. Hilde-
garde Howard, in recognition of her many contributions to our
knowledge of the fossil birds of the Pliocene/Pleistocene.

COMMENTS ON SOME OTHER FOSSILS
ASSIGNED TO THE ARDEIDAE

My conclusions on the proper systematic position of some
other fossils described in the family Ardeidae are presented
below.

Goliathia andrewsi Lambrecht

Lambrecht (1930:30, Fig. 7) described Goliathia andrewsi
as a new genus and species of heron based on a very large
ulna from an unknown locality in the Upper Eocene/Lower
Oligocene Faiyum series of Egypt. Earlier in the same paper
he erected a new genus and species of large stork, Palaeoephip-

Brodkorb: Ethiopian Fossil Heron

89

Figure 1. Ardea howardae new species. From left to right, anterior,
preserved, 40.04 mm.

piorhynchus dietrichi, based on a skull and mandible from
“Kasr el Querun” ( = Kafr el Qeren?) in the Faiyum. He cor-
rectly thought the ulna too small to go with the skull of the
stork. He further stated that the olecranon, internal cotyla,
and cubital tubercles were all weakly developed as in herons,
rather than prominent as in Ciconiidae. Andrews (1907) had
previously remarked that the ulna was somewhat smaller and
notably stouter than that of the Recent Ardea goliath, but the
muscle scars were similar.

It is not possible to judge the depth of the internal cotyla
from Lambrecht’s drawing, but it clearly shows the reduction
of the olecranon and cubital tubercles. These characters are
not found in species of the family Ciconiidae, but they are
shared by and are even more pronounced in the Recent Ba-
laeniceps rex Gould (Ardeae: Balaenicipitidae) than they are
in Ardeidae.

The stoutness of the bone mentioned by Andrews is appar-
ent in the drawing, and this too more closely resembles that
of Balaeniceps than it does the slender ulna of Ardeidae. In
respect to the muscle scars I see no trenchant differences be-
tween Balaeniceps and Ardeidae.

lateral, posterior, medial, and below, cranial views of coracoid. Length as

In view of the above considerations I believe that Goliathia
andrewsi should be placed in the family Balaenicipitidae. The
only previous fossil record for the family is the tentative re-
ferral of the distal end of a tibiotarsus from the Miocene of
Tunisia (Rich 1974).

Ardea lignitum Giebel

This species was based on the distal half of a left femur
from the Brown Coal of Rippersroda in Thuringia, Germany
(Giebel 1860:152, PI. 1, Fig. 2, erroneously called Fig. 3 in his
text). He attributed the site to the Pliocene, but the formation
is now placed in the Sarmatian, i.e. , Upper Miocene (Thenius
1959).

Lambrecht (1933) was unable to locate any of Giebel’s types
so we must rely on published sources. Giebel wrote of the type
(in my translation), “This so strikingly resembles the femur of
the Gray Heron, Ardea cinerea, that comparison with other
genera seems completely superfluous. Still it is not identical.”
He then described several differences and illustrated anterior
and posterior aspects of the type.

90

Brodkorb: Ethiopian Fossil Heron

Table 2. Measurements (mm) of the femur of several species of Ardea
and large owls.

Species

Length

Least
Width
of Shaft

Distal

Width

Ardea goliath (2c? 9 )

126.8-135.0

10.2-10.4

23.1-23.9

A. herodias (17c? 9 )

100.0-114.2

6. 7-7. 9

17.1-18.8

A. cinerea (9c ? 9 )

79.1-92.8

6.4-7. 1

14.4-16.0

A . purpurea (2 c? 9 )

84.3-93.4

5.5-6. 1

12.5-14.0

''Ardea” lignitum

(l)1

(100)

8

18

Bubo bubo ( 1 9 )

100.2

9.6

21.2

B. bubo (19)2

99.9

9.0

20.7

B. b. davidi ( 2 —5 )3

106.0

9.5

22.6

B. binagadensis

(l)4

98.8

9.2

—

B. lacteus (1 9 )

88.8

8.6

20.1

B. virginianus

(lie? 9)

78.9-85.5

7. 1-8.4

16.2-18.0

B. africanus (9c ? 9)

65.0-69.6

5. 3-6.0

12.4-13.8

Nyctea scandiaca

(4c? 9)

82.6-90.9

7. 6-8. 6

16.8-20.1

N. s. gallica (2-3)5

92.4

8.5

19.3

Strix nebulosa

(2c? (?)

83.9-84.4

5.7-6. 1

15.5-15.9

S. uralensis (3)6

—

6.3

14.9

S. brea (4)7

75.6-76.6

—

—

S. varia (13c? 9 )

69.6-77 .2

5. 6-6.2

13.1-14.7

5. intermedia (l)8

—

5.0

12.5

S. aluco ( 1 <? )

59.3

4. 1

10.6

S. aluco (7)9

—

4.6

11.2

1 Holotype, Miocene, Germany, length estimated (Giebel 1847).

2 Mean of Recent specimens (Mourer-Chauvire 1975:165).

3 Mean of paratypes, Mindel stage, France (Mourer-Chauvire
1975:165).

4 Holotype, Upper Pleistocene, Azerbaijan (Burchak-Abramovich
1965:453).

5 Mean of paratypes, Mindel stage France (Mourer-Chauvire 1975:
162).

6 Mean of 3 Recent specimens (Mourer-Chauvire 1975:172).

7 Paratypes, Upper Pleistocene, California (Howard 1933:68).

8 Referred specimen, Mindel stage, France (Mourer-Chauvire
1975:172); the holotype is a coracoid, Middle Pleistocene, Czechoslo-
vakia (Janossy 1972:53).

9 Mean of Recent specimens (Mourer-Chauvire 1975:172).

For some unexplained reason the femora and some other
bones of herons and owls are rather similar. Pertinent differ-
ences in the distal half of the femora of owls (both Strigidae
and Tytonidae) are enumerated below.

In anterior view, (1) the external side of the external condyle
of owls flares strongly laterad from the edge of the shaft (more
so in Bubo and Nyctea than in Strix and Tyto), whereas in
Ardea the condyle is more compressed and lies more nearly in

line with the shaft; (2) in owls the distal notch of the external
condyle is quite hidden, but it is visible in Ardea; (3) the lateral
edge of the shaft is concave above the condyle (most so in
Bubo), contrasting with a marked prominence in the same area
in Ardea\ (4) the intercondylar sulcus is deep and wide (espe-
cially in Bubo and Nyctea), narrow and shallow in Ardea-, (5)
the distal end of the intermuscular line shows some individual
variation, but in owls it usually terminates far above the in-
ternal condyle (especially in Bubo), whereas in Ardea it con-
tinues to the proximal rim of the internal condyle.

In posterior view of the femora of owls (6) the internal con-
dyle is relatively narrow, especially in Bubo (wide and long in
Ardea)-, (7) the shelf of the external condyle swings gradually
toward the shaft in both strigid and tytonid owls, but in Ardea
the shelf bulges laterad before swinging to the shaft; (8) the
two intermuscular lines show some individual variation in
owls, but they usually merge about half way up the shaft; they
remain separate in Ardea.

In medial and lateral views, (9) the femur is more strongly
curved in owls than in Ardea.

Giebel’s plate clearly shows characters 1-8 enumerated
above, and the ninth is mentioned in his text. They were used
by him to differentiate his fossil from the living Ardea cinerea
Linnaeus, but they really prove he had an owl. In both quan-
titative (see Table 2) and qualitative characters the Brown
Coal species is closest to the living Eagle Owl, Bubo bubo
(Linnaeus). It must be removed from Ardeidae and placed in
Strigidae as Bubo lignitum (Giebel), new combination.

Giebel (1847) also described eight supposedly extinct birds
from the late Pleistocene in the then widely held belief that all
remains from the “Diluvium” represented species that failed
to survive the Noachian Flood. All appear to be synonyms of
living species.

Ardea brunhuberi Ammon

This species was based on the proximal half of a left car-
pometacarpus from the Upper Miocene Brown Coal Forma-
tion in the clay works of Mayer and Reinhard between Deck-
betten and Priifening in Wiirttemburg, Germany. The detailed
description and photograph (Ammon 1911:33, Fig. S) are re-
peated in a later and more accessible paper (Ammon 1918:30,
Fig. 4).

Although the author compared it with the Recent Purple
Heron, Ardea purpurea Linnaeus, a glance at the illustration
suffices to show that this is a cormorant, not a heron. The
alular metacarpal (metacarpal I) is thrust proximad, with its
tip obliquely truncate as in Phalacrocorax, in contrast with
the condition in Ardea, in which it is thrust less proximad and
has a pointed tip. The facet for the alular digit (digit I) is very

Table 3. Measurements (mm) of the carpometacarpus of three species of Phalacrocorax.

Measurement

P brunhuberi

P . carbo
(n = 2)

P. miocaenus

Length

75 (estimate)

73.8-74.3

43.2

Height through alular metacarpal

15

14.6-14.9

9.6

Length of inner rim of trochlea

10

10.4-11.5

—

Height of alular metacarpal

5

4. 0-5.0

—

Width of proximal articulation

7

6. 9-7. 2

5.0

Length of pisiform process

3

3.3-3. 5

—

Brodkorb: Ethiopian Fossil Heron

91

Table 4. Measurements (mm) of the coracoid of four species of Phalacrocorax.

Measurement

P praecarbo

P carbo (2)

P. miocaenus

P. littoralis

Length

75-80 (estimate)

75.0-78.3

52

54

Length of furcular facet

9

11.0-11.5

—

—

Width of furcular facet

6

7.6-8. 1

—

—

Width of glenoid facet

10

10.4-10.8

—

—

Width below glenoid facet

—

10.7-11.4

6.7

8.2

Head to tip of procoracoid

20

21.1-22.5

-

—

large and deeply excavated, but in Ardea it is small and only
slightly excavated. Furthermore, the trochlea is large (small
in Ardea), and the external rim extends much farther proximad
than in Ardea. It is therefore necessary to refer to this species
as Phalacrocorax brunhuberi (Ammon), new emendation.

In his later paper Ammon (1918:28, Fig. 3) erected Phala-
crocorax praecarbo, based on the upper half of a left coracoid
from the same locality. Tables 3 and 4 show that the types of
both names under discussion agree in size with the correspond-
ing elements of European specimens of Recent Phalacrocorax
carbo (Linnaeus). As it is unproven and highly unlikely that
two species of cormorants of the same size class occurred at
this site, I place Phalacrocorax praecarbo in synonymy with
Phalacrocorax brunhuberi .

Phalacrocorax intermedins (Milne-Edwards 1867) from the
Orleanais Sand, known only from an incomplete humerus, is
slightly smaller than P. carbo. It is also somewhat older than
P. brunhuberi, so I hesitate to synonvmize the latter with
Milne- Edward’s species.

ArcLeacites molassicus Haushalter

This species was based on a right humerus from the railroad
cut through Miocene/Pliocene sands, between Augsburg and
Landau in the former district of Algau, southwestern Bavaria
(Haushalter 1855:11, PI 2, Fig. 1). The type was formerly in
the Munich Museum, but according to Lambrecht (1933) it
has been lost. Haushalter compared it with the living Euro-
pean Bittern, Botaurus stellaris (Linnaeus), but his description
is practically useless. The type is shown in palmar view, with
most of the deltoid crest and the condyles badly damaged. The
published measurements are length 140 mm, mid-width 7 mm.
My measurements from the plate are length 140, mid-width
of shaft 8, and least width of shaft 7 mm.

If the drawing is accurate, this is not a heron. The head of
the humerus is too long and pointed. The deltoid crest is too
long, the bicipital crest much too narrow, and the distal end
is too narrow. It obviously represents a water bird of some
sort, but I am unable to get it in an order, much less a family.
Therefore I assign it to the Aves Incertae Sedis.

ACKNOWLEDGMENTS

I am deeply indebted to Mary D. Leakey, who entrusted me
with the study of the immense collection of avian fossils from
Olduvai Gorge, and who also provided financial aid. For the
opportunity to study the birds from the Ethiopian sites I am
also indebted to J. Desmond Clark, F. Clark Howell, D C.
Johanson, and their associates. Some of the comparative ma-
terial used in the present study was made available through
the kindness of Richard K. Brooke, M.P. Stuart Irwin, Storrs
L. Olson, and Oscar T. Owre.

LITERATURE CITED

Ammon, L. von. 1911. Bayerische Braunkohlen und ihre
Verwertung. Munich. 82 pp.

. 1918. Tertiare Vogelreste von Regensburg und die

jungmiocene Vogelwelt. Abhandl. naturwiss. Vereines zu
Regensburg 12:1-70.

Andrews, C.W. 1907. Notes on some vertebrate remains
collected in the Fayfim, Egypt, in 1906. Geol. Mag., ser.
5 4:97-100.

Brodkorb, P. 1955. The avifauna of the Bone Valley For-
mation. Florida Geol. Surv. Rept. Invest. No. 14:1-57.

. 1963. Fossil birds from the Alachua Clay of Florida.

Florida Geol. Surv. Special Publ. No. 2, Paper No. 4:1-
17.

. 1978. Catalogue of fossil birds, part 5 (Passeri-
formes). Bull. Florida State Mus., Biol. Sci. 23(3): 139—
228.

Brown, F.H., J de Heinzelin, and F.C. Howell.
1970. Pliocene/Pleistocene formations in the lower Omo
Basin, southern Ethiopia. Quaternaria 13:247-268.

Burch ak-Abramovich, N.I. 1965. Noviya vid iskopaemogo
filina iz Binagadov. Ornitologiya 7:452-454.

Butzer, K.W. 1971. The Lower Omo Basin: Geology, fauna
and hominids of Plio-Pleistocene formations. Naturwis-
senschaften 58:7-16.

Giebel, C.G. 1847. Fauna der Vorwelt mit steter Berfick-
sichtigung der lebenden Thiere, 1, part 2, xi + 2 17 + 1
pp. Leipzig.

. 1860. Zur Fauna der Braunkohlenformation von

Rippersroda in Thfiringen. Zeitschr. f. Gesammt. Natur-
wissensch. 16:147-153.

Haushalter, C.L. 1855. Merkwiirdige fossile Thierfiber-
reste aus des Algauer Molasse. Inaugural-Abhandlung.
Munich. 16 pp.

Janossy, D. 1972. Die Mittelpleistozane Vogelfauna der
Stranska Skala. Studia Musei Moraviae, Anthropos.
20:35-64.

Kltrochkin, E.N. 1976. Novie dannie o ptitsakh Pliotsena
Zapadnoi Mongolii. Paleont. i Biostrat. Mongolii, Trudy
3:51-67.

Lambrecht, K. 1930. Studien fiber fossile Riesenvogel. Geo-
logica Hungarica, ser. pal. 7:1-37.

. 1933. Handbuch der Palaontologie. Berlin, xx +

1024 pp.

Milne-Edwards, A. 1867-1871. Recherches anatomiques
et paleontologiques pour servir a l’histoire des oiseaux de
la France. Paris. Vol. 1. 475 pp.

Moseley, C., and A. Feduccia. 1975. Upper Pliocene her-
ons and ibises from North America. Univ. Michigan Mus.
Paleont., Papers on Paleontology No. 12:71-74.

92

Brodkorb: Ethiopian Fossil Heron

Mourer-Chauvire, C. 1975. Les oiseaux du Pleistocene
moyen et superieur de France. Docum. Lab. Geol. Fac.
Sci. Lyon, no. 64, fasc. 1:1-262.

Rich, P. 1974. Significance of the late Tertiary avifaunas
from Africa (with emphasis on a mid- to late Miocene
avifauna from southern Tunisia). Annals Geol. Surv.
Egypt 4:167-210.

Thenius, E. 1959. Tertiar. Wirbeltierfaunen. Handbuch der
stratigraphischen Geologie, 3, part 2: v + 328. Stuttgart.
Van Beneden, P.J. 1973. Paleontologie des vertebres.
Oiseaux. Patria Belgica 1:353-373. Brussels.

THE AUSTRALIAN BROMORNITHIDAE: A GROUP OF
EXTINCT LARGE RATITES

By Pat Vickers Rich1

ABSTRACT: The Dromornithidae was a family of large ground birds, presumably ratite, that has

a Miocene to Pleistocene (minimum age, 26,000 years B.P.) record in Australia. Five genera are now
known, including Dromornis , Genyornis, Ilbandornis, Bullockornis, and Barawertornis . The known
forms range in size and agility from slightly larger than, and about as gracile as, the living emu, to truly
gigantic, ponderous forms such as Dromornis stirtoni that rivaled or exceeded the size and proportions
of the Malagassy elephant bird (Aepyornis maximus). Dromornithids and emus (Casuariidae) appear in
the record simultaneously in the Miocene, the former ranging over most of Australia and reaching their
greatest known diversity in the Miocene. Detailed osteological analyses of this group indicate that it is
monophyletic, and more closely related to the Casuariidae, including emus, than to any other avian
family.

The family Dromornithidae has been known for some time
in Australia (Mitchell 1839), but because specimens were few
and poorly preserved, little attention was given this group until
recently. This paper presents a brief history of work on the
dromornithids, familial and generic diagnoses, and hypotheses
for phylogenetic relationships within the Dromornithidae as
well as between the dromornithids and other avian families.

The first known reference to dromornithids lies in the oral
traditions and art of the Aborigines. Tindale (1951) and Hall
et al. (1951) noted traditions among the Tjapwurong tribe in
western Victoria concerning “mihirung paringmal,” or giant
emus that supposedly lived “long ago when the volcanic hills
[of the Western District of Victoria] were in a state of erup-
tion.” Some lava flows in this area are as young as 8000 years
B.P. (Gill 1972). Temporal overlap of the dromornithids with
Aborigines at about 26,000 years B.P. has recently been con-
firmed (Lancefield Swamp, Victoria; Gillespie et al. 1978), and
younger records would not be surprising.

The first convincing evidence of a group of ratite birds dis-
tinct from the Casuariidae in Australia was the spectacular
discovery in 1892 of partial skeletons in Pleistocene sediments
of Lake Callabonna, South Australia. Newton (1893) reported
this find, and subsequently Stirling (1896) and Stirling
and Zietz(1896, 1900, 1905) described the Callabonna material
as a new genus and species, Genyornis newtoni. Prior to this
work all evidence of the extinct dromornithids, including the
namesake of the family from Peak Downs in Queensland, Dro-
mornis australis (Owen 1872, 1874), consisted of isolated and
often fragmentary specimens. All of the specimens came from
localities restricted to the eastern half of Australia.

Information on the Dromornithidae remained sparse until

1 Earth Sciences Department, Monash University, Clayton, Victoria
3168; National Museum of Victoria, 285-321 Russell Street, Mel-
bourne, Victoria 3000, Australia.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 330:93-103.

the middle of the twentieth century. Several expeditions into
central and northern Australia in the mid-1950’s and thereafter
led to the discovery of three new genera: Barawertornis, Il-
bandornis and Bullockornis (Rich 1979). Additional specimens
of the previously known genera Dromornis and Genyornis
were also recorded.

The dromornithids can now be delimited on the basis of the
osteological material presently available, and meaningful com-
parisons with other ratite groups can be detailed. The track-
ways of probable dromornithids from southeastern Australia
(Rich and Green 1974; Rich and Gill 1976) and a possible
dromornithid egg from Pleistocene dune deposits in Western
Australia offer negligible diagnostic data, however.

Abbreviations used are: AM, Australian Museum, Sydney;
CPC, Commonwealth Palaeontological Collections, Bureau of
Mineral Resources, Canberra; SIAM, American Museum of
Natural History, Department of Vertebrate Paleontology Field
Number, New York; UCMP, University of California, Mu-
seum of Paleontology, Berkeley.

DIAGNOSTIC FEATURES OF THE
DROMORNITHIDAE

The dromornithids comprise a group of medium-sized to
truly gigantic ground birds that were endemic to Australia
from at least the Miocene until the late Pleistocene, perhaps
as late as the Holocene. The smallest form, Barawertornis
tedfordi, was about the size of the living ostrich or slightly
smaller. The largest, Dromornis stirtoni, equaled or possibly
exceeded the weight of any bird previously known (Fig. 1),
including Aepyornis maximus, but was surpassed in height by
the moas of New Zealand. Limb proportions vary from the
massive-limbed Dromornis stirtoni, to the moderately slender,
elongate-limbed Genyornis, to the extremely gracile-limbed
Ilbandornis lawsoni. Barawertornis is restricted to the early

94

Rich: Australian Dromornithidae

Rich: Australian Dromornithidae

95

Miocene, Bullockornis to the late Miocene, Ilbandornis to the
late Miocene or early Pliocene, Dromornis to the late Miocene
and Pliocene, while Genyornis is restricted to the Pleistocene.
The stratigraphic distribution of the dromornithids is sum-
marized in Figure 2, and Figure 3 shows their geographic
distribution.

The dromornithids were ratites in the classic sense, in that
they lacked a true keel on the sternum. No palatal region of the
skull is known, however, and allocation of this group to the
palaeognathous birds is based entirely upon the similarities of
their postcranial skeletons to those of emus and cassowaries.
Vertebrae are also relatively rare elements, and in no specimen
is a complete vertebral series known. The dromornithids can
be characterized from their known elements by the following
combination of characters: (1) sternum more elongate than
broad with costal area occupying 60 percent of the lateral mar-
gin; (2) sternal notches lacking; (3) articulation of sternum with
scapulocoracoid restricted to far lateral margins of anterior
border of sternum; (4) glenoid facet of scapulocoracoid and
entire wing greatly reduced, decidedly more than in ostriches
and rheas; (5) humerus lacking well defined articular surfaces;

(6) radius and ulna fused at several points along their shafts;

(7) carpometacarpus lacking intermetacarpal space and with
only one phalangeal articulation; (8) synsacrum of moderate
width, not broad as in moas and aepyornithids; (9) synsacrum
with articulation of hindlimb about midway between anterior
and posterior ends, in contrast to condition in struthionids,
where this articulation lies forward of midpoint, and in apter-
ygids, where it lies posterior to midpoint; (10) pubes, ischia,
and ilia subequal in posterior extension; (11) pubes not fused
as in struthionids; (12) pubes fuse with ischium and in turn
with ilium in adult specimens to produce elongate ilioischiatic
fenestra and short ischiopubic fenestra, a condition found else-
where within the ratites only in aepyornithids; (13) femur with
trochanter projecting about same distance proximad of shaft
as head; (14) femur with external condyle moderately exceed-
ing internal condyle in distal extension, but not as dispropor-
tionately as in rheas and ostriches; (15) femur lacking massive
muscle scars in popliteal area, thus differing from the femora
of moas and aepyornithids; (16) femur with internal condyle
with distalmost extension occurring anterior to condylar mid-
point, approaching elliptical shape with major axis forming
acute angle with posterior margin of shaft (semicircular in
shape in emus, moas, elephant birds; distally flattened in
rheas, ostriches); (17) femur with condyles of equal depth; (18)
tibiotarsus not decidedly mediolaterallv compressed near prox-
imal end; (19) tibiotarsus with inner cnemial crest extending
far proximad to proximal articular surface; (20) tibiotarsus

with supratendinal bridge present, differing from those of all
other ratites except moas, and tendinal canal centrally located;
(21) tarsometatarsus with hypotarsal region broad and trian-
gular in shape in proximal view, with two shallow hypotarsal
canals located near medial and lateral boundaries of hypotar-
sus (this arrangement differs from the rectangular structure
dissected by a single, deep canal found in moas and kiwis, the
laterally offset hypotarsus of ostriches and rheas, and the low,
rectangular hypotarsus of aepyornithids); (22) tarsometatarsus
with single, prominent ridge extending most of length of pos-
terior surface (this differs markedly from the short, double
ridges in moas and kiwis, the short, narrow ridge in rheas and
ostriches, and the absence of a ridge in aepyornithids); (23)
tarsometatarsus without articulation for metatarsal I, indicat-
ing the absence of the first digit; (24) tarsometatarsus with
three trochleae present, although internal trochlea often quite
reduced; (25) pes with phalangeal count of 3-4-4, not the more
characteristic ratite count of 3-4-5, and a tendency in at least
the geologically younger forms (Ilbandornis and Genyornis ) to
develop blunted, hooflike unguals, rather than claws with a
triangular or rounded cross section.

RELATIONSHIPS OF THE
DROMORNITHIDAE TO OTHER
AVIAN FAMILIES

In a recent paper (Rich 1979) I used three different methods
of analysis in an attempt to determine the relationship of the
Dromornithidae to the remaining ratites. Although I distinctly
favor Method I, other methods that have frequently been used
in phylogenetic analysis were explored to determine if all meth-
ods resulted in similar conclusions. In the first method a phe-
netic analysis based on the postcranial skeleton was performed
initially to determine the family most closely related to the
ratites, and their probable nearest relatives, the tinamous (Tin-
amidae) (Bock 1963). Then — using the definition that primi-
tiveness was defined by common occurrence of a character in
both the ratites and their nearest sibling group, the tinamous —
an analysis to determine whether a character was primitive or
derived was performed (see Rich 1979 for the data input for
all analyses, and detailed tables summarizing the results).

The second method involved surveying all of the ratites and
defining as primitive the most commonly occurring state for
each character. Those characters that occurred rarely within
the ratites were considered derived or advanced. Theoretical-
ly, this method would help to determine phyletic branching,
where a derived character or suite of characters is held by
three or less families. Derived characters that appeared early

Figure 1. A selection of diagnostic bones of Dromornithidae from mid- to late Cenozoic deposits of Australia. Femora of: (a) Barawertornis
tedfordi, CPC 7341 (Type), Riversleigh, Queensland, Miocene (distal width 87 mm); (B-C) Bullockornis planet, CPC 13844 (Type) and CPC 13845
respectively, Bullock Creek, Northern Territory, Miocene (distal widths 160 mm and >152 mm); (D) Dromornis stirtoni, CPC 13851 (Type),
Alcoota, Northern Territory, Late Miocene or early Pliocene (distal width 202 mm); (E) Ilbandornis woodburnei, CPC 13850 (Type) Alcoota,
Northern Territory, late Miocene or early Pliocene (distal width 112 mm); (F) Dromornis australis, AM F 10950, Peak Downs, Queensland,
probably Pliocene (distal width 120 mm); (G) Genyornis newtoni, SIAM 61, proximal right humerus, Lake Callabonna, South Australia, Pleistocene
(depth from external to internal tuberosity 25 mm); (H-K (Ilbandornis sp., UCMP 67038, characteristic ungual phalanx of pes, Alcoota, Northern
Territory, late Miocene or early Pliocene (total length 28 mm); (L-M) Dromornis stirtoni, UCMP 113049, sternum, Alcoota, Northern Territory,
late Miocene or early Pliocene (maximum width across sternocoracoidal processes approx. 225 mm); (N) Dromornis stirtoni, UCMP 113050,
scapulocoracoid, Alcoota, Northern Territory, late Miocene or early Pliocene (total length >239 mm); (O) Ilbandornis lawsoni, UCMP 70118,
proximal view of left tibiotarsus, Alcoota, Northern Territory, late Miocene or early Pliocene (maximum depth about 88 mm); and (P) Ilbandornis
sp., UCMP 70649, distal end, right tibiotarsus, Alcoota, Northern Territory, late Miocene or early Pliocene (distal width 76 mm).

96

Rich: Australian Dromornithidae

Figure 2. Stratigraphic distribution of the Dromornithidae in Australian Cenozoic deposits. Black squares represent occurrences of most dro-
n979)thld m3teriaL B’ Bamwertornis’ Bu- Bullockornis ; Dr, Dromornis; D, Dromornithidae ; G, Genyornis; I, Ilbandornis. Modified from Rich

Rich: Australian Dromornithidae

97

Figure 3. Geographic distribution of the Dromornithidae in Australia. A, Pleistocene; ▲ , Pliocene; ■ , Miocene localities. Same abbreviations for
generic names as in Fig. 2. (1) Diamantina River, D; (2) Warburton River, D; (3) Lancefield, D; (4) Cooper Creek, D; (5) Lake Callabonna, G;
(6) Brother’s Island, Pt. Lincoln, G; (7) Cuddie Springs, G; (8) Riversleigh, B; (9) Bullock Creek, Bu; (10) Canadian Lead (Gulgong, Mudgee), D;
(11) Endurance Pit, South Mt. Cameron, D; (12) Wellington Caves, D; (13) Lake Ngapakaldi, D; (14) Snake Dam Locality, D; (IS) Baldina Creek,
G; (16) Normanville (Salt Creek), G; (17) Alcoota, Dr and I; (18) Big Cave (Naracoorte), D; (19) Penola, D; (20) Mt. Gambier, D; (21) Peake
Downs, Dr; (22) Lake Palankarinna, D; (23) Scott River, ? D; (24) Mammoth Cave, D; (25) Thornbindah, D.

in the history of ratites, however, and thus possibly were pos-
sessed by a large number of the members of this group, would
be misinterpreted by this method.

The third method involved an initial phenetic analysis of 56
characters (chosen because they could be used to diagnose the
Dromornithidae; Rich 1979). The 56 characters studied were
summed for each ratite group regardless of their polarity
(primitive or advanced ( = derived)), in essence a strictly nu-
merical taxonomic approach. Only the sternum, synsacrum,
and hindlimb elements were considered because of the lack of
information about other elements of various fossil groups, and
because of the near or total loss of the forelimb in the Dinor-
nithidae-Emeidae lineage as well as its marked reduction in
other ratites.

The primary purpose of using the three different methods
was the determination of the avian group most closely related
to the Dromornithidae. Thus, relationships between the re-

maining ratite groups are only briefly mentioned in the follow-
ing discussion.

The analysis using Method I indicates that the Dromornith-
idae share decidedly more derived characters (19 of the 56
studied) with the Casuariidae (including emus) than with any
other ratite family, but that the two families share few prim-
itive characters (7) (see Table 1). The Casuariidae and the
Dromornithidae show a large number of derived characters
within the ratites, with 33 and 38 (respectively) of the 56 char-
acters derived rather than primitive. The Struthionidae and
Rheidae have nearly the same number of derived characters
(37 and 33, respectively), but the Aepyornithidae, Apterygi-
dae, Dinornithidae-Emeidae have fewer derived characters
(30, 23, and 22, respectively). Among the 19 derived characters
shared between the Casuariidae and Dromornithidae are four
unique to these two groups. These are: (1) synsacrum with
ilium, ischium, and pubis all protruding about the same dis-

98

Rich: Australian Dromornithidae

Table 1 . Number of characters of the sternum, synsacrum, and hind-
limb shared by the ratites and their sibling group, the Tinamidae.
Method I approach

a. Shared derived characters (determined by lack of occurrence in the
Tinamidae).

J3

'3

Rheidae

Casuariidae

Dromornithidae

Aepvornithidae

Apterygidae

Dinornithidae-

Emeidae

4

7

9

12

IS —

8 7

10 13

18 20

17

19

7

9

13

17 —

20 29

tance posteriad; (2) femur with internal condyle triangular, or
elliptical closely approaching triangularity, with apex forming
distalmost projection of condyle; (3) tibiotarsus with inner cne-
mial crest extending far proximad of proximal articular sur-
face; and (4) tarsometatarsus with hvpotarsus and intercotylar
prominence extending about equal distances proximad to prox-
imal articular surfaces. The remaining 15 derived characters

Figure 4. Phylogenetic hypotheses expressing possible interrelation-
ships of the ratites.

idae (16 and 8, respectively) and Rheidae (18 and 11, respec-
tively), which are quite closely related to one another (sharing
26 derived characters). Without further, more expanded anal-
ysis, however, it is difficult to determine how many of the
characters shared by the Struthionidae-Rheidae and the Casu-
ariidae are the result of convergent evolution. Three phyletic
hypotheses are suggested (Fig. 4) based on the Method I ap-
proach.

The Method II approach, i.e. , using commonality of a char-

shared between Casuariidae and Dromornithidae are likewise
shared with at least one and often more ratite groups.

For each group of ratites, the following number of the 56
characters mentioned above for Method III were present in
the derived condition using the Method I approach: Struthion-
idae 37; Rheidae 33; Casuariidae 33; Dromornithidae 38; Ae-
pyornithidae 30; Apterygidae 23; Dinornithidae-Emeidae 22.

Thus, the analysis of Method I indicates that the Dromor-
nithidae are most closely related to the Casuariidae. A single,
common ancestral stock could have given rise to each of these
two groups, and two separate colonizations within Australia-
New Guinea are not necessary to account for their presence.
The Australian ratites appear decidedly distinct from the New
Zealand moas and kiwis (which form a close-knit osteological
group) and are apparently the most primitive of all the ratite
groups. The Casuariidae, but not the Dromornithidae, in turn
share a large number of derived characters with the Struthion-

Table 2. Number of derived characters (determined by commonality
of occurrence within the ratites) of the sternum, synsacrum, and hind-
limb shared by several ratite groups. Method II approach

1)

3

C

o

3

c n

<v

<5

’o3

-C

PC

<v

cd

-o

3

3

c /)

3

u

<v

(3

"5b

>.

<v

a

<

O Qj

■5 E

Q fcd

Struthionidae

Rheidae

9

—

—

—

-

—

—

Casuariidae

2

2

—

—

—

—

—

Dromornithidae

1

2

7

—

—

—

—

Aepvornithidae

4

1

3

6

—

—

—

Apterygidae

Dinornithidae-

1

0

2

1

1

—

—

Emeidae

3

2

1

7

4

8

—

Rich: Australian Dromornithidae

99

Table 3. Number of unweighted characters of the 56 studied by Rich
(1979) that are shared by the various ratite groups. Method III ap-
proach.

Struthionidae

Rheidae

Casuariidae

Dromornithidae

Aepvornithidae

Apterygidae

Dinornithidae-

Emeidae

Struthionidae

—

—

—

—

—

—

—

Rheidae

39

—

—

—

—

—

—

Casuariidae

24

33

—

—

—

—

—

Dromornithidae

12

19

26

—

—

—

—

Aepvornithidae

20

21

26

19

—

—

—

Apterygidae

Dinornithidae-

18

30

33

18

27

Emeidae

22

27

28

23

31

38

—

acter within the ratites to indicate primitiveness and rarity to
indicate a derived or advanced condition, reinforced many of
the conclusions reached in the Method I analysis, although not
all were in agreement. The Dromornithidae shared the greatest
number of derived characters (7) with both the Casuariidae
and the Dinornithidae-Emeidae, but only one with the Apter-
ygidae (see Table 2). The suggested similarity between moas
and dromornithids can perhaps be accounted for by size and
functional similarities, a result of convergent evolution. Using
Method II, casuariids clearly share far more characters con-
sidered to be derived with dromornithids than with any other
ratite group, and the relationship between the moas and kiwis
of New Zealand is strongly supported.

Method III, the numerical taxonomic approach, also rein-
forced several of the relationships suggested by the two pre-
vious methods, but obscured or contradicted others (Table 3).
The Dromornithidae still shared more characters (26) with the
Casuariidae, but the Casuariidae shared more characters with
the Rheidae (33) and the Apterygidae (33) than with any other
ratite group. Method III suggests that the Struthionidae and
Rheidae are more similar to one another than to any other
ratite group (sharing 39 characters), as are the moas and kiwis
(sharing 38 characters). The Aepvornithidae share the most
characters with the Dinornithidae.

GENERIC RELATIONSHIPS WITHIN
THE DROMORNITHIDAE

As illustrated above, three different types of phylogenetic
analysis suggest that the Dromornithidae share the greatest
number of similarities with the Casuariidae (Fig. 4). Because of
this, I also evaluated the characters of the latter group when I
considered intra-dromornithid relationships.

In my analyses, a character was considered to be primitive
within the family Dromornithidae if it was shared by both the
dromornithids and the casuariids, or in some cases if it fre-
quently occurred in a wide range of avian families. Addition-
ally, a second criterion used in some analyses was to assume
a character to be primitive within the dromornithids if it was
the more common state in this group. All of these methods of
determining the polarity of a character, i.e. , whether primitive
or advanced, have been discussed in numerous publications

(Kluge and Farris 1969; Kluge 1971; Schaeffer et al. 1972;
Hecht 1976). Time was not a factor that influenced determi-
nation of whether a character was primitive or advanced. In
fact, one of the oldest occurring dromornithids, Bullockornis ,
appears to be one of the most specialized in the family. Once
the primitive or advanced nature of each character or char-
acter complex was estimated, a phylogeny was derived based
upon the minimum number of alterations to a primitive mor-
photype needed to produce the five genera and eight species
known for the family. Analysis was restricted to the femur and
tarsometatarsus because these are the only elements repre-
sented for all of the dromornithid genera, and most species.
In Figure 5,21 characters of these two elements are tabulated
for each of the dromornithid genera as well as Dromaius, a
casuariid genus. The phylogeny implied in Figures 5 and 6
resulted from analyses using the following two approaches.

First, an analysis of the dromornithid complex using com-
monality of occurrence within the Dromornithidae as an in-
dicator of primitiveness was performed. The results are sum-
marized in the following paragraphs (see Figs. 7 and 8 for
illustrations of the femur, tibiotarsus, and tarsometatarsus).

Barawertornis, the oldest known dromornithid from Mio-
cene sediments in the Northern Territory, is the most primitive
and the smallest known member of the family, having 14 of
the 21 characters summarized in Figure 5 in the primitive
state. This genus is presumably specialized over the postulated
primitive morphotype for the Dromornithidae in that on the
femur, the internal condyle is broader than the external (char.
22), 1 the fibular condyle protrudes only moderately laterad
(char. 15), the popliteal area is deep (char. 14); and, on the
tarsometatarsus, trochleae II and IV are subequal in distal
extension (char. 9).

The remaining dromornithids, including Dromornis, Ilban-
dornis, Genyornis , and Bullockornis, are derived with respect
to Barawertornis because on the femur, the neck is not decid-
edly constricted (char. 3), the internal condyle has a decidedly
elliptical shape (char. 18); and on the tarsometatarsus, trochlea
II is moderately to highly reduced (char. 11).

Within the above lineage it is evident that Dromornis, II-
bandornis, and Genyornis are closely related and form a nat-
ural grouping separate from Bullockornis . Shared derived
characters for this former lineage include the femur with the
medial surface of the internal condyle decidedly ridged (char.
19), and the fibular condyle extending as far or nearly as far
posteriad as the internal condyle (char. 26); and the tarsome-
tatarsus with trochlea III moderately to decidedly broader than
trochlea IV (char. 16).

Although decidedly more derived than Barawertornis, the
Dromornis-Ilbandornis-Genyornis lineage, even if its most
specialized member ( Genyornis ) is considered, does not possess
as many derived characters within the Dromornithidae as does
Bullockornis (12 out of 21, Genyornis ; 14 out of 21, Bullock-
ornis).

Bullockornis, with two species, is characterized by having
the following derived characters: femur with shaft deep (char.
10), long axis of the external condyle nearly in line with the

1 Character numbers refer to those cited by Rich (1979) in a detailed
analysis of the Dromornithidae. Of the 200 characters analyzed, only
a few were actually useful in the final phylogenetic analyses.

100

Rich: Australian Dromornithidae

:

I

J a

: -c

03

-C 03

U I

s ~

■a _c

a ^

a £

•5 Q
§Q

T3 C

0) <**>

a "

o3

S £■

"O (L»

5 c

£ c •

2 S,

Q |

JJ 43

-5 D.

co a;
<v w

3

o <D

Sg

>

o -O

•rj <d

<d o « £

bi> u -n

13 ^ L 5

I _. - .2 *

1 c^j -— <d hh

1 ^ r ^ ^

. - £ o 03

: jd ^ cu i>

; >. "c a> —

i T3 .3 > o

1 C .3-0

: o B +£ £

‘ u ^ 2 "

v <* £

S' g" <u

I *

0,3 s

<D

2 -G

u <=>

o P

o3 w

c a

j- co
dJ d>

a <d

; -c -r
I c £

03 „

O U H ^

''“" 03

- a ^

S ^ ?

£ c ^ «
o 5

3 u

a;

N

C —
<D o3
bC c
£ c3

>-» <jj

-C CD

44 ,5

E 7=

Si

got

Jn CO «
<y <D "3

* a -S

(U 03 ^
^ bo °
° g J=

■£ — •S
3 o 'g

* o ST

-C

^ a

-C <D
u 'O

-a s

.5 ^

<n 2

CNJ 3

<n Ji, aj
' D 3

•rj O

C o3

- d> jD

1 1 ’

<d H

‘.go,

-O Ch -
<d ^

13 2

3 0J

2 -P 03

— 13 £

03 >-

> £ 73

^ a i:

«2 £ £

D O o3

*-. o3
bo g
<d 03
T3 -C

rsj 3

T3

C

o3

Sic

-3

2

13

‘ Id

S

o

CD

Ctf i

H

-a

I S

CD C+H

o

o3

P 03 ^

• -C — v

vo c/i cj to

3 C C a3
§ £ «
Uh C U* 03

Rich: Australian Dromornithidae

101

Casuariidae

Barawertornia

Dromornia Ilbandornia

Bullockornis

internal condyle

jT-ahallowing of popliteal area (fem.
-deepening of neck (fem.)

-increase in depth of fibular condyle (fem.)
-condyles line up with main axis of shaft (fem.)
-narrowing of internal condyle (fem.)

-deepening of shaft and condyles (tint.)

-widening of intertrochlear space III-IV (tmt.)
-increased prominence of subhypotarsal ridge (tmt.)
-increase in length of trochlea II relative to
IV (tmt.)

-increase in depth of trochlea III (tmt.)

-increase in width of trochlea IV with respect to
III (tmt.)

-increase in narrowing of trochlea III posteriorly
( tmt. )

f^antero posterior compression of condyles and shaft (fem.)

-splaying of condyles (not parallel in distal view) (fem.)

-development of elliptical-shaped internal condyle (fem.)

-equalising of depth of external and internal condyles (fem.)

-reduction in grooving of posterior shaft surface (tmt.)

-deepening of trochleae (tmt.)

-splaying of margins of trochlea III (tmt.)

-increase in length of trochlea II relative to IV (tmt.)

-decrease in prominence of subhypotarsal ridge (tmt.)

Figure 6. Phylogenetic hypothesis expressing possible interrelationships of the genera of dromornithids. Major changes occurring from most
primitive {Barawertornis) to most advanced genera are enumerated.

C

Figure 7. Comparison of femora of Dromornithidae and Casuari-
idae. A, proximal view; B, distal view; C, medial views.

long axis of the shaft (char. 16), and the internal and external
condyles nearly subequal in depth (char. 23); tarsometatarsus
with intertrochlear space between trochleae III and IV broad
(char. 2), the subhypotarsal ridge prominent (char. S), trochlea
III extremely deep (char. 15), and the medial and lateral mar-
gins of trochlea III markedly convergent posteriorly (char. 17).

The second analytical approach used the common occur-
rence of a character state within the nearest sibling group, the
Casuariidae, to indicate the primitive state of a character and
produced the results described below.

Barawertornis has the greatest number of characters in a
primitive state (11), even though two could not be evaluated,
and Bullockornis the least (5). This reinforces the interpreta-
tion of the previous analysis. The following trends for the
dromornithids are suggested by this analysis. In the femur
there is a broadening of the neck region, and overall antero-
posterior compression of shaft and distal end ( Bullockornis
may well be primitive in having a deep shaft, contrary to the
suggestion of the previous analysis, or the deep condition is
secondarily derived), a shallowing of popliteal area, an in-
crease in lateral protrusion of external condyle, a decrease in
twisting of condyles away from long axis of shaft, the devel-
opment of an elliptical internal condyle (viewed medially), a
broadening and narrowing of the internal condyle relative to
width of external condyle, an increase in depth of external
condyle so it becomes subequal in depth to internal condyle,
and, in distal view, the long axes of condyles become parallel.

In the tarsometatarsus there is a decrease in the relative
separation of trochleae III and IV, a deepening of shaft, a
decrease in prominence of subhypotarsal ridge, an increase in
distal extension of trochlea II, a reduction in size of trochlea
II and thus in size of second digit, an increase in depth of
trochlea III, and an increase in width of trochlea IV with

102

Rich: Australian Dromornithidae

DROMORNITHID

EMU-CASSOWARY

Figure 8, Comparison of tibiotarsi (A-C), and tarsometatarsi (D) of
Dromornithidae and Casuariidae. A, D, proximal views; B, anterior
views; C, distal views.

respect to trochlea III, and thus an equalization in the size of
the two digits.

Other trends that occur at some time within the dromorni-
thid lineage (but just when and in which genera is difficult to
determine because of the lack of complete enough specimens
of some genera) after it diverged from the cassowary and emu
lineage include: (1) deepening and strengthening of the lower
jaw, (2) deepening of the atlas vertebra, (3) flattening of the
sternum and reduction in area of scapulocoracoid contact, as
well as slight overall reduction in size of scapulocoracoid (the
scapula and coracoid become more in line with one another),

(4) equalization of overall dimensions of radius and ulna, and
lengthening of both relative to humerus, (5) decrease in depth
of synsacrum dorsal to acetabulum and an increase in depth
of pubic bar, (6) increase in mediolateral compression of cne-
mial crests and a decrease in depth of internal cnemial crest
of tibiotarsus, (7) development of a supratendinal bridge on
tibiotarsus (or retention from ancestral stock that gave rise to
both Casuariidae and Dromornithidae), (8) increase in depth
of internal condyle on tibiotarsus, and (9) development of hoof-
like ungual phalanges on digits and loss of one phalanx in digit
IV, resulting in a phalangeal formula for the pes of 3-4-4.

SUMMARY

The family Dromornithidae is composed of five genera and
eight species of extinct ground birds restricted to the middle
and late Cenozoic of Australia (Figs. 2 and 3). Although some
forms were only slightly larger than their closest relatives, the
emus and cassowaries, one form may have exceeded the weight
of the largest known bird, Aepyornis maximus of Madagascar.

The history of the dromornithids extends back only into the
Miocene, but the presence of four genera in the Miocene in-
dicates that the group had its origins much earlier. The fossil
record for most Cenozoic vertebrates in Australia is poorly
known before the Miocene because of a lack of known older
localities, a result of the geology of the Australian continent
and the inaccessibility of certain areas that have paleontolog-
ical promise.

The use of three different methods of phylogenetic analysis
consistently suggested that the Dromornithidae were most
closely related to the only other Australian ratite group, the
Casuariidae. The dromornithids are quite distinct from the
moas (Dinornithidae-Emeidae) of New Zealand and probably
originated from a common stock with the Casuariidae, and
possibly the Rheidae and Struthionidae. The ancestral form
may have been advanced over that which gave rise to the
moas, kiwis, and elephant birds. It is conceivable that a single
ancestral stock on the Australian landmass could have given
rise to both the casuariids and dromornithids, and two inva-
sions are not an absolute requirement to account for the Aus-
tralian ratite diversity. If such an event took place during the
early to mid-Cretaceous, the ancestral form could conceivably
have flown or walked between Australia and the remaining
southern continents when tenuous connections still existed be-
tween these land masses.

ACKNOWLEDGMENTS

Thanks are due R.A. Stirton (deceased), R.H. Tedford,
M.C. McKenna, W. Bock, and T.H. Rich for assistance, dis-
cussion, and moral support throughout this study; to M.L.
Vickers, U. Gowronski, C. Armstrong, and R. Sheehan for
persistence in typing; and to R.E. Lemley for his support with
the most recent large collections of dromornithid material.

LITERATURE CITED

Bock, W.J. 1963. The cranial evidence for ratite affinities.

Proc. XIII Int. Ornith. Congress 1:39-54.

Gill, E.D. 1972. Eruption date of Tower Hill volcano,
western Victoria, Australia. Victorian Nat. 89(7): 188-192.
Gillespie, R., D.R. Horton, P. Ladd, P.G. Macumber,

Rich: Australian Dromornithidae

103

T.H. Rich, R. Thorne, and R.V.S. Wright. 1978.
Lancefield swamp and the extinction of the Australian
megafauna. Science 200(4345): 1044-1048.

Hall, F.J., R.G. McGowan, and G.F. Guleksen. 1951.
Aboriginal rock carvings: a locality near Pimba, S. A. Rec.
So. Aust. Mus. 9:375-380.

Hecht, M.K. 1976. Phylogenetic inference and methodology
as applied to the vertebrate record. Evol. Biol. 9:335—
363.

Kluge, A. G. 1971. Concepts and principles of morphologic
and functional studies. Pp. 3-41 in Chordate structure
and function (Waterman, A.J., Ed.). Macmillan Co.,
New York.

Kluge, A. G., and Farris, J.S. 1969. Quantitative phyletics
and the evolution of anurans. System. Zool. 18:1-32.

Mitchell, T.L. 1839. Three expeditions into the interior
of eastern Australia, 2. T. and W. Moore, London. 415 pp.

Newton, A. 1893. Palaeontological discovery in Australia.
Nature 47( 122 6):606 .

Owen, R. 1872. Notice of his nineteenth memoir on the ex-
tinct birds of the genus Dinornis. Proc. Zool. Soc. London
1872:682-683.

. 1874. On Dinornis (Part XIX): containing a descrip-
tion of a femur indicative of a new genus of large wingless
birds ( Dromornis australis) from a post-Tertiary deposit
in Queensland, Australia. Trans. Zool. Soc. London
8:381-384.

Rich, P.V. 1979. The Dromornithidae, a family of large,

extinct ground birds endemic to Australia: Systematic and
phylogenetic considerations. Canberra Bur. Min. Res.,
Bull. 184:1-196.

Rich, P.V., and E.D. Gill. 1976. Possible dromornithid
footprints from Pleistocene dune sands of southern Vic-
toria, Australia. Emu 76(4):22 1—223.

Rich, P.V., and R.H. Green. 1974. Bird footprints at South
Mt. Cameron, Tasmania. Emu 74:245-248.

Schaeffer, B., M.K. Hecht, and N. Eldredge. 1972.
Phylogeny and paleontology. Evol. Biol. 6:31-46.

Stirling, E.C. 1896. The newly discovered extinct gigantic
bird of South Australia. Ibis 7( 1 1 ): 5 93 .

Stirling, E.C. , AND A. H.C. Zietz. 1896. Preliminary notes
on Genyornis newtoni; a new genus and species of fossil
struthious bird found at Lake Callabonna, South Austra-
lia. Trans. Roy. Soc. South Australia 20:171-190.

— . 1900. Fossil remains of Lake Callabonna. I. Geny-

ornis newtoni. A new genus and species of fossil stru-
thious bird. Mem. Roy. Soc. So. Aust. 1(2):4 1—80.

. 1905. Fossil remains of Lake Callabonna. Part III.

Description of the vertebrae of Genyornis newtoni. Mem.
Roy. Soc. So. Aust. 1(3):8 1—1 10.

. 1913. Fossil remains of Lake Callabonna. Part IV.

Description of some further remains of Genyornis new-
toni. Mem. Roy. Soc. So. Aust. 1(4): 1 1 1-126.

Tindale, N.B. 1951. Comments on supposed representa-
tions of giant bird tracks at Pimba. Rec. So. Aust. Mus.
9:381-382.

THE PRESENT STATE OF KNOWLEDGE OF THE
CENOZOIC BIRDS OF ARGENTINA1

By Eduardo P. Tonni2

ABSTRACT: An annotated list by age and a summary by order of the Cenozoic birds of Argentina

is presented. A re-evaluation of fossil taxa over which there has not been general agreement as to systematic
position allows the following conclusions: (1) Neculus rotlii Ameghino 1905 is a synonym of Palaeosphe-
niscus gracilis Ameghino 1899; (2) Palaeoapterodytes ictus Ameghino 1905 is a synonym of Palaeosphe-
niscus bergi Moreno and Mercerat 1891; (3) Eudromia and perhaps Nothura are present in Monteher-
mosan sediments of Buenos Aires Province; (4) Phalacrocorax pampeanus Moreno and Mercerat 1891,
from late Pleistocene deposits of Buenos Aires Province, is assignable to the living species P brasilianus',
(5) Dryornis pampeanus Moreno and Mercerat 1891 is a vulturid closely related to Vultur.

Preliminary analysis of new material currently under study reveals (1) an anseriform that shows affinities
with the Tachyerini from late Oligocene-early Miocene marine deposits of Patagonia; (2) that the earliest
known specimens of phorusrhacoids appear in Casamayoran (early Eocene) deposits of Patagonia.

RESUMEN: Se presenta una lista anotada por edad y un resumen por orden de las aves cenozoicas

de Argentina. La reevaluacion de los taxa fosiles sobre los cuales no hay un consenso respecto a su posicion
taxonomica permite llegar a las siguientes conclusiones: (1) Neculus rotlii Ameghino 1905 es sinonimo de
Palaeospheniscus gracilis Ameghino 1899; (2) Palaeoapterodytes ictus Ameghino 1905 es sinonimo de
P alaeospheniscus bergi Moreno and Mercerat 1891; (3) Eudromia y quizas Nothura estan presentes en
sedimentos Montehermosenses de la Provincia de Buenos Aires; (4) Phalacrocorax pampeanus Moreno
and Mercerat 1891, de depositos del Pleistoceno tardio de la Provincia de Buenos Aires es asignado a la
especie viviente P. brasilianus] (5) Dryornis pampeanus Moreno and Mercerat 1891 es un vulturido de
cercana relacion a Vultur.

Un analisis preliminar de material nuevo actualmente en estudio revela (1) un anseriforme con afinidad
a los Tachyerini de depositos marinos de Oligoceno superior-Mioceno inferior de Patagonia; y (2) que los
primeros especimenes conocidos de phorusrhacoides aparecen en los depositos casamayorenses (Eoceno
tardio) de Patagonia.

The record of Cenozoic birds of Argentina is the most com-
plete for any region of South America. Most of the described
forms come from Patagonia, in southern Argentina, thus lim-
iting the record geographically. This has resulted in a restricted
knowledge of the origin and evolution of certain groups, es-
pecially those presently endemic to the Neotropical Region. In
addition, the record for the rest of South America is limited
primarily to the Pleistocene (for a brief review of fossil birds
of South America see Campbell 1979).

At the end of the 1800’s and the beginning of the 1900’s
Florentino Ameghino described most of the known taxa of
Cenozoic Argentinian birds, but his work requires revision as
a result of new contributions. A partial revision has been ac-
complished for his extensive work on the fossil mammals of
Argentina, but this is not the case with the birds. Only two
avian groups, the phorusrhacoids and penguins, have been

1 Translated by Lidia Lustig.

2 Division Paleontologia Vertebrados. Facultad de Ciencias Naturales
y Museo. 1900 — La Plata, Argentina.

extensively revised. Any revisions have been hindered in part
by the scattering of the original materials in different Argen-
tinian and foreign institutions, a fact that has brought about
great confusion.

It must be re-emphasized here, as I have before (Tonni
1973), that there are no uniform taxonomic criteria to be used
with both fossil and Recent birds. In general, the criteria of
paleontologists differ from those of neontologists. When one
considers some of the taxonomic problems regarding Recent
birds, e.g., polytypic species, sibling species, sexual dimor-
phism, etc., it is evident that the study of fossil birds, the
remains of which are generally very fragmentary, can be dif-
ficult and complicated. This points again to the need to revise
the works of the pioneer authors using the more rigorous taxo-
nomic standards of today.

In spite of these difficulties, and others common to the study
of all fossil vertebrates, the paleontological record of Argen-
tinian birds contributes valuable data on some groups, espe-
cially if the information is taken as a whole and not partially.

A list of Cenozoic Argentinian birds, with their geographic

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 3 30:105-114.

106

Tonni: Cenozoic Birds of Argentina

Figure 1. Cenozoic epochs and the corresponding South American
Land Mammal Ages. Modified from Marshall et al. 1977. (1) After
Berggren and van Couvering (1974). (2) After Savage (1975).

and stratigraphic data, follows. The scheme of Land Mammal
Ages follows that of Pascual et al. (1965) and Pascual and
Odreman Rivas (1973); see Figure 1. The Tertiary stratigraph-
ic chart has recently changed significantly, but it is not the
purpose of the present study to discuss these changes and the
reader is referred to the pertinent literature (Berggren and Van
Couvering 1974; Marshall et al. 1977; Marshall et al. 1979).
I have re-studied most of the material of doubtful taxonomic
position. Preliminary data on material under study are included.

Those orders with a fossil record sufficient to warrant ad-
ditional comments are discussed at the end of this paper. Such
a discussion based on the data presently available is quite
limited, and will remain so until the study of fossil birds from
the Neotropical Region becomes the focus of attention of a
greater number of investigators.

SYSTEMATIC LIST OF CENOZOIC
BIRDS FROM ARGENTINA

Casamayoran

Order Phoenicopteriformes,

Family Presbyornithidae:

Presbyornis antiquus (Howard 1955)

LOCALITY AND HORIZON: Cafiaddn Hondo, near Paso
Niemann, Cluibut Province. Casamayor Fm.

Presbyornis pervetus Wetmore 1926

LOCALITY AND HORIZON: Cafiadbn Hondo, near Paso
Niemann, Chubut Province. Casamayor Fm.

REMARKS: This specimen was described as Telmabates
howardae by Cracraft (1970). See Feduccia (1976) for further
information.

Order Opisthocomiformes,

Family Onychopterygidae:
Onychopteryx simpsoni Cracraft 1971

LOCALITY AND HORIZON: Cafiaddn Hondo, near Paso
Niemann, Chubut Province. Casamayor Fm.

REMARKS: Brodkorb (1978:2 15) places this species in Aves
Incertae Sedis.

Divisaderan

Order Ralliformes, Family Cunampaiidae:
Cunampaia simplex Rusconi 1946

LOCALITY AND HORIZON: 500 m west of Mina Atala,
Departamento de Las Heras, Mendoza Province. Divisadero
Largo Fm.

REMARKS: The systematic position of Cunampaia simplex
is uncertain. The general concensus is to place it in its own
family within the Ralliformes, but its placement within the
order varies according to the author. Patterson and Kraglie-
vich (1960:12) state that “su posicion dentro de la Clase es por
ahora incierta” — “its position within the Class is for now un-
certain.” Wetmore (1960) considered the position of the family
within the Ralliformes to be uncertain, a treatment followed
by Cracraft (1968). Brodkorb (1967) includes the family Cu-
nampaiidae in the suborder Cariamae.

Tonni: Cenozoic Birds of Argentina

107

Deseadan

Order Ardeiformes, Family Ciconiidae:
Ciconiopsis antarctica Ameghino 1899
LOCALITY AND HORIZON: Santa Cruz Province. De-
seado Fm. (“Guaranitic Fm.” of Ameghino).

Order Pelecaniformes,

Family Cladornithidae:

Cladornis pachypus Ameghino 1895
LOCALITY AND HORIZON: Rio Deseado, Santa Cruz
Province. Deseado Fm.

Order Anseriformes,

Family incertae sedis :

Teleornis impressus Ameghino 1899
LOCALITY AND HORIZON: Santa Cruz Province. De-
seado Fm.

Loxornis clivus Ameghino 1895
LOCALITY AND HORIZON: Rio Deseado, Santa Cruz
Province; Rio Chico, west of Puerto Visser, Chubut Province.
Deseado Fm.

Order Accipitriformes,

Family Accipitridae?:

Cruschedula revola Ameghino 1899
LOCALITY AND HORIZON: Golfo de San Jorge, Santa
Cruz Province. Deseado Fm.

REMARKS: This species is based on the proximal end of
a right scapula, as noted by Brodkorb (1964), and not on a
tarsometatarsus as stated by Ameghino. The material is not
very diagnostic, but its placement in the Accipitridae is prob-
able.

Climacarthrus incompletus Ameghino 1899
LOCALITY AND HORIZON: Santa Cruz Province. De-
seado Fm.

REMARKS. This species is based on the distal end of a
right tarsometatarsus that lacks the external trochlea, most of
the middle trochlea, and the posterior portion of the internal
trochlea. This species can be tentatively assigned to the Ac-
cipitridae on the basis of characters of the internal trochlea;
i.e., it extends clistad more than the middle trochlea and has
a marked concavity in the proximal part of the internal mar-
gin.

Order Ralliformes

The family sequence followed here corresponds to that of
Cracraft (1968, 1969).

Family Aramidae:

Aminornis excavatus Ameghino 1899
LOCALITY AND HORIZON: Rio Deseado, Santa Cruz
Province. Deseado Fm.

Loncornis erectus Ameghino 1899
LOCALITY AND HORIZON: Rfo Deseado, Santa Cruz
Province. Deseado Fm.

Family Phorusrhacidae:

Andrewsornis abbotti Patterson 1941

LOCALITY AND HORIZON: Cabeza Blanca, Chubut
Province. Deseado Fm.

Family Brontornithidae:

Physornis fortis Ameghino 1895

LOCALITY AND HORIZON: Rio Deseado, Santa Cruz
Province. Deseado Fm.

Family Psilopteridae:

Smiliornis penetrans Ameghino 1899

LOCALITY AND HORIZON: Santa Cruz Province. De-
seado Fm.

Pseudolarus guaraniticus Ameghino 1899

LOCALITY AND HORIZON: Santa Cruz Province. De-
seado Fm.

Family Cariamidae:

Riacama caliginea Ameghino 1899

LOCALITY AND HORIZON: Santa Cruz Province. De-
seado Fm.

“Patagoniano”

The age of the marine sediments generally referred to as
“Patagoniano” has been a matter of great controversy. Ac-
cording to Camacho (1974), the “Patagoniano” is divisible into
the following units: the San Julian Formation, strata with
Monophoraster and Venericor, and the Monte Leon Forma-
tion. The type locality of the San Julian Fm. is in the Gran
Bajo of San Julian, Santa Cruz Province. The latter two units
outcrop from the Golfo de San Jorge and Santa Cruz and
Chubut Provinces, northwards to the proximity of Trelew and
the lower course of the Rio Chubut. These three units were
referred by Camacho to the early Eocene, late middle Eocene,
and late Oligocene, respectively.

Bertels (in press) considers the San Julian Formation to be
late Eocene to early Oligocene in age. Cione and Exposito (in
press), on the basis of their study of an icthyofauna from a
unit referred to the Monophoraster and Venericor strata of the
Golfo de San Jorge, state that that unit is at least late Oligocene
in age.

Marshall et al. (1977), on the basis of isotopic dating of the
sediments of the Santa Cruz and Colhuehuapi Formations that
respectively overlie and underlie the Monte Leon Fm., assigns
the Monte Leon Fm. to the late Oligocene.

Recently, Riggi (1979) recognized a single formational unit
(Patagonia Fm.) in the “Patagoniano” sediments. He divides
the Patagonia Fm. into two members, the San Julian and the
Monte Leon, both of Oligocene age.

Unfortunately, the penguin remains described by Ameghino
and Moreno and Mercerat lack precise geographic and strati-
graphic data. There are precise data on only the material stud-
ied by Simpson (1946, 1972) from the lower course of the Rio
Chubut, between Gaiman and Trelew. Field work carried out
by the author in this area (Tonni in press) has verified that the
associated icthyofauna indicates the deposits are at least late
Oligocene to early Miocene in age.

There is no conclusive evidence demonstrating that the pen-

108

Tonni: Cenozoic Birds of Argentina

guins from the more southerly outcrops (Golfo de San Jorge,
San Julian) are of a different age. The following nine species
are all from the “Patagoniano.”

Order Sphenisciformes,

Family Spheniscidae:
Palaeospheniscus gracilis Ameghino 1899

LOCALITY: Golfo de San Jorge, Trelew, Gaiman; Chubut
Province.

REMARKS: Neculiis rothi Ameghino 1905 is included in
this species. The holotype is strongly weathered and thus the
trochleae appear to be disproportionately small. Other sup-
posedly diagnostic characters, e.g., tarsometatarsus very flat-
tened anteroposteriorlv with the anterior surfaces of the troch-
leae lying in the same plane, the intertrochlear notches deep,
and a large metatarsal facet, correspond to those of the holo-
type of P. medianus (=P. gracilis). The transverse diameter
at mid-shaft of the tarsometatarsus of “ Neculus rothi" is 12
mm, while that of the holotype of P. medianus (=P. gracilis)
is 13.5 mm. Consequently, taking into consideration individ-
ual size variation, N. rothi lies within the size range of P
gracilis, of which it is a synonym.

Palaeospheniscus patagonicus
Moreno and Mercerat 1891
LOCALITY: Golfo de San Jorge, Trelew, Gaiman; Chubut
Province.

Palaeospheniscus wimani (Ameghino 1905)
LOCALITY: San Julian, Santa Cruz Province.
Palaeospheniscus bergi
Moreno and Mercerat 1891
LOCALITY: Golfo de San Jorge, Trelew; Chubut Province.
REMARKS: Palaeoapterodytes ictus Ameghino 1905,
which was based on the proximal portion of a strongly weath-
ered right humerus, is included in this species. Simpson (1946)
stated/3, ictus was indeterminate, and that Palaeoapterodytes
( =Apterodytes ) was probably a synonym of Palaeospheniscus
or some other genus from the same beds. Simpson questioned
placing Palaeoapterodytes in synonymy with Palaeospheniscus
because “[. Palaeoapterodytes ] probably did not have a bipartite
tricipital fossa, which is always present in Palaeospheniscus"
(Simpson 1972:29). A re-evaluation of the holotype verifies that
the tricipital fossa is undoubtedly bipartite and that the spec-
imen shares the size range and other morphological character-
istics of the humeri referred to Palaeospheniscus bergi.

Cliubutodyptes biloculata Simpson 1970
LOCALITY: Cerro Castillo, Chubut Province.
Paraptenodytes antarcticus
(Moreno and Mercerat 1891)

LOCALITY: The mouth of the Rio Santa Cruz, Santa Cruz
Province.

Paraptenodytes brodkorbi Simpson 1972

LOCALITY: San Julian, Santa Cruz Province.

Paraptenodytes robustus (Ameghino 1895)
LOCALITY: La Cueva, San Julian, Santa Cruz Province;
Gaiman, Chubut Province.

Arthrodytes grandis Ameghino 1901

LOCALITY: San Julian, Santa Cruz Province.

Order Procellariiformes,

Family Procellariidae?:

Argyrodyptes microtarsus Ameghino 1905

LOCALITY AND HORIZON: Rio Seco, Santa Cruz Prov-
ince. Basal “Patagoniano.”

REMARKS: This species was based upon a supposedly as-
sociated distal half of a tibiotarsus and a distal half of a femur.
As pointed out by Simpson (1946, 1972), it is not from a species
of the Sphenisciformes. Its placement in the Procellariidae,
following Brodkorb (1963), is quite probable.

Order Anseriformes, Family Anatidae:
Anatidae new genus new species

LOCALITY: Gaiman, Chubut Province.

REMARKS: This new genus new species, with affinities
with the Tachyerini, is currently being described (Tonni in

press).

Santacrucian

Order Rheiformes, Family Rheidae:
Opisthodactylus patagonicus Ameghino 1895
LOCALITY AND HORIZON: Santa Cruz Province. Santa
Cruz Fm.

REMARKS: Patterson and Kraglievich (1960) correctly con-
cluded that Opisthodactylus, type of the family Opisthodac-
tylidae Ameghino 1895, is not a phorusrhacoid, but a rhea.
Brodkorb (1963) included the Opisthodactylidae in the Rhei-
formes.

Order Pelecaniformes, Family Pelecanidae;
Liptornis hesternus Ameghino 1895
LOCALITY AND HORIZON: Santa Cruz Province. Santa
Cruz Fm.

Order Ardeiformes, Family Plataleidae:
Protibis cnemialis Ameghino 1891
LOCALITY AND HORIZON: Monte Observacion, Santa
Cruz Province. Santa Cruz Fm.

Order Anseriformes, Family Anatidae:
Eoneornis australis Ameghino 1895
LOCALITY AND HORIZON: Monte Observacion, Santa
Cruz Province. Santa Cruz Fm.

Eutelornis patagonicus Ameghino 1895
LOCALITY AND HORIZON: Monte Observacion, Santa
Cruz Province. Santa Cruz Fm.

Order Accipitriformes,

Family Accipitridae:

Thegornis musculosus Ameghino 1895
LOCALITY AND HORIZON: Tagua Quemada, Santa
Cruz Province. Santa Cruz Fm.

Tonni: Cenozoic Birds of Argentina

109

Thegornis debilis Ameghino 1895

LOCALITY AND HORIZON: Corriguen-Kaik, Santa
Cruz Province. Santa Cruz Fm.

REMARKS: Brodkorb (1964) includes these two species in
the subfamily Circinae. If this is correct, they represent the
oldest record for the subfamily.

Family Falconidae:

Badiostes patagonicus Ameghino 1895

LOCALITY AND HORIZON: “Patagonia.” Santa Cruz
Fm.

Order Galliformes, Family Cracidae:

Anisolornis excavatus Ameghino 1891

LOCALITY AND HORIZON: “Southern Patagonia.” San-
ta Cruz Fm.

Order Ralliformes, Family Phorusrhacidae:
Phorusrhacos longissimus Ameghino 1887

LOCALITY AND HORIZON: La Cueva, Tagua Quema-
da, Monte Observacion, Rfo Shehuen; Santa Cruz Province.
Santa Cruz Fm.

REMARKS: See Brodkorb (1967) for the synonymy of the
species included in this family and all other large extinct South
American ralliforms.

Family Palaeociconiidae:

Palaeociconia cristata Moreno and Mercerat 1891

LOCALITY AND HORIZON: Monte Le6n, Tagua Que-
mada, Monte Observacion, La Cueva; Santa Cruz Province.
Santa Cruz Fm.

Family Psilopteridae:

Pseudolarus eocaenus Ameghino 1891

LOCALITY AND HORIZON: “Patagonia.” Santa Cruz
Fm.

Psilopterus australis
Moreno and Mercerat 1891

LOCALITY AND HORIZON: Killik-Aike, Monte Lehn,
Monte Observacion, Take Harvey, La Cueva, Corriguen-
Kaik, Tagua Quemada, Karaiken; Santa Cruz Province. Santa
Cruz Fm.

Psilopterus communis
Moreno and Mercerat 1891

LOCALITY AND HORIZON: Monte Observacion, Lago
Pueyrredon, La Cueva, Rio Shehuen; Santa Cruz Province.
Santa Cruz Fm.

Psilopterus minutus (Ameghino 1891)

LOCALITY AND HORIZON: Monte Observacion, Santa
Cruz Province. Santa Cruz Fm.

Lophiornis obliquus Ameghino 1891

LOCALITY AND HORIZON: Monte Observacion, Santa
Cruz Province. Santa Cruz Fm.

Friasian

Order Accipitriformes,

Family Accipitridae:

Accipitridae genus and species
indeterminate

LOCALITY AND HORIZON: Ing. Jacobacci, Neuquen
Province. Unnamed formation.

REMARKS: This and the following falconid species are the
first reported fossil birds from mammal-bearing Friasian sed-
iments. The material is very fragmentary and identification
beyond the family level cannot be attempted.

Family Falconidae;

Falconidae genus and species
indeterminate

LOCALITY AND HORIZON: Ing. Jacobacci, Neuquen
Province. Unnamed formation.

Chasicoan

Order Ralliformes, Family Psilopteridae:
Psilopterus new species

LOCALITY AND HORIZON: Arroyo Chasico, Buenos
Aires Province. Vivero member of the Arroyo Chasico Fm.

REMARKS: This species is currently being described (Ton-
ni in press).

Huayquerian

Order Accipitriformes,

Family Teratornithidae:
Teratornithidae new genus new species

LOCALITY AND HORIZON: Salinas Grandes de Hidal-
go, Departamento Atreuco, La Pampa Province. Epecuen Fm.

REMARKS: This specimen is described by Campbell and
Tonni, this vol.

Order Ralliformes, Family Phorusrhacidae:
Andalgalornis ferox
Patterson and Kraglievich 1960

LOCALITY AND HORIZON: Chiquimil, Catamarca
Province. Andalgala Fm.

Onactornis depressus Cabrera 1939

LOCALITY AND HORIZON: Lago Epecuen, Buenos
Aires Province. Epecuen Fm.

Family Psilopteridae:

Procariama simplex Rovereto 1914

LOCALITY AND HORIZON: Catamarca Province.
“Araucanense stage.”

Hermosiornis incertus Rovereto 1914

LOCALITY AND HORIZON: Catamarca Province.
“Araucanense stage.”

Montehermosan

Order Rheiformes, Family Rheidae:
Heterorhea dabbenei Rovereto 1914

LOCALITY AND HORIZON: 17 km SW of Pehuen-Co,
Buenos Aires Province. Monte Hermoso Fm.

110

Tonni: Cenozoic Birds of Argentina

Order Tinamiformes, Family Tinamidae:
Tinamisornis parvulus Rovereto 1914

LOCALITY AND HORIZON: 17 km SW of Pehuen-Co,
Rio Quequen Salado, Buenos Aires Province. Monte Hermoso
Fm. , “Irenense.”

REMARKS: For a discussion of this and the following
species of tinamous see Tonni (1977a).

Eudromia intermedia (Rovereto 1914)

LOCALITY AND HORIZON: 17 km SW of Pehuen-Co,
Buenos Aires Province. Monte Hermoso Fm.

Eudromia sp.

LOCALITY AND HORIZON: Rfo Quequen Salado, Bue-
nos Aires Province. “Irenense.”

Order Ralliformes, Family Psilopteridae:
Prophororhacus rapax
(J. Kraglievich 1946)

LOCALITY AND HORIZON: Atlantic coast near Arroyo
Loberia, Partido de Grab Pueyrredon, Buenos Aires Province.
Chapadmalal Fm.

Family Cariamidae:

Chunga incerta Tonni 1974

LOCALITY AND HORIZON: 17 km SW of Pehuen-Co,
Buenos Aires Province. Monte Hermoso Fm.

Order Accipitriformes, Family Vulturidae:
Dryornis pampeanus
Moreno and Mercerat 1891

LOCALITY AND HORIZON: 17 km SW of Pehuen-Co,
Buenos Aires Province. Monte Hermoso Fm.

REMARKS: A revision of the type material resulted in the
placement of this species, originally presumed to be a phorus-
rhacoid, in the Vulturidae, as stated by Brodkorb (1967). It is
closely related to Vultur.

“Mesopotamian”

Order Ralliformes, Family Phorusrhacidae:
Andalgalornis steulleti
(L. Kraglievich 1931)

LOCALITY AND HORIZON: Cliffs of the Rio Parana,
Entre Rios Province. “Mesopotamian” (Basal part of the
Ituzaingo Fm. of Acenolaza (1976)).

REMARKS: For comments on this and the following two
species see Patterson and Kraglievich (1960).

Andalgalornis deautieri
(L. Kraglievich 1931)

LOCALITY AND HORIZON: Cliffs of the Rfo Parana,
Entre Rios Province. “Mesopotamian” (Basal part of the Itu-
zaingo Fm. of Acenolaza (1976)).

Onactornis? pozzi (L. Kraglievich 1931)

LOCALITY AND HORIZON: Cliffs of the Rio Parana,
Entre Rios Province. “Mesopotamian” (Basal part of the Itu-
zaingo Fm. of Acenolaza (1976)).

Ensenadan

Order Tinamiformes?, Family Tinamidae?:
Querandiornis romani Rusconi 1958

LOCALITY AND HORIZON: “Toscas of the Rfo de la
Plata,” in the proximity of the Estacion Anchorena, Buenos
Aires Province. Ensenada Fm.

REMARKS: The type material could not be located for re-
study, but some of the characters cited by Rusconi (1958) in
his description of the species, e.g., foramen magnum pyriform
and skull globular, make its taxonomic position questionable.
He compared the material, a skull, with two living species,
Eudromia elegans and Fulica leucoptera, finding similarities
and differences that are not only non-diagnostic, but so general
that they could be applied to the description of the skull of
almost any living bird. In his description Rusconi himself ex-
pressed doubts as to whether he should place the fossil in the
Tinamiformes, or in the Ralliformes; he felt “inclined” to place
it in the Tinamidae.

Order Rheiformes, Family Rheidae:

Rhea anchorenensis
(C. Ameghino and Rusconi 1932)

LOCALITY AND HORIZON: “Reefs of the Rfo de la Plata
opposite the Estacion Anchorena,” Buenos Aires Province.
Ensenada Fm.

REMARKS: Brodkorb (1963) established Rhea americana
anchorenense C. Ameghino and Rusconi as a separate species,
R. anchorenensis. A re-evaluation of the holotype verifies this
assignment.

Order Anseriformes, Family Anatidae:

Anas leucophrys Vieillot 1816

LOCALITY AND HORIZON: Partido de Grab Alvarado,
Buenos Aires Province. Miramar Fm.

REMARKS: See Tonni (1969) for a report on the occurrence
of this species.

Order Psittaciformes, Family Psittacidae:
Cyanoliseus ensenadensis (Cattoi 1957)

LOCALITY AND HORIZON: Olivos, Buenos Aires Prov-
ince. Ensenada Fm.

REMARKS: For comments on this species see Tonni (1972).

Order Passeriformes, Family Furnariidae:
Cinclodes major Tonni 1977

LOCALITY AND HORIZON: Mar del Plata, Buenos Aires
Province. Miramar Fm.

Family Emberizidae:

Zonotrichia robusta Tonni 1970

LOCALITY AND HORIZON: Miramar, Buenos Aires
Province. Miramar Fm.

Sicalis sp.

LOCALITY AND HORIZON: Mar del Plata, Buenos Aires
Province. Miramar Fm.

REMARKS: Material referable to this genus was studied by
Tonni (1973). The size of this specimen must have been similar
to that of the living species Sicalis olivascens orS. auriventris .

Tonni: Cenozoic Birds of Argentina

111

Because of the poor representation of the material it was not
given a specific assignation. It is, to the present, the oldest
record for the genus.

Lujanian

Order Tinamiformes, Family Tinamidae:
Nothura paludosa Mercerat 1897

LOCALITY AND HORIZON: Arrecifes, Buenos Aires
Province. Buenos Aires Fm.

REMARKS: This species, like others described by Mercer-
at, was not adequately described or figured. As in other in-
stances, a re-evaluation of the type was not possible because
it could not be located.

Order Rheiformes, Family Rheidae:

Pterocnemia fossilis Ameghino 1882

LOCALITY AND HORIZON: Olivera, Buenos Aires
Province. Buenos Aires Fm.

REMARKS: Rhea pampeana Moreno and Mercerat 1891 is
a synonym of Pterocnemia fossilis Ameghino.

Rhea azarae (Moreno and Mercerat 1891)

LOCALITY AND HORIZON: Monte Hermoso, Buenos
Aires Province. “Pampeano.”

REMARKS: The incomplete femur on which Moreno and
Mercerat founded Protorhea azarae is assignable to Rhea, and
not to a camel, “ Auchenia lujanensis," as assumed by Ame-
ghino (1891). Its placement in a species separate from Rhea
americana is only tentative because of the fragmentary nature
of the material.

Order Anseriformes, Family Anatidae:
Neoclien debilis (Ameghino 1891)

LOCALITY AND HORIZON: La Plata, Buenos Aires
Province. “Belgranense stage.”

Order Ralliformes, Family Rallidae?:

Euryonotus brachypterus Mercerat 1897

LOCALITY AND HORIZON: Arrecifes, Buenos Aires
Province. Buenos Aires Fm.

Euryonotos argentimis Mercerat 1897

LOCALITY AND HORIZON: Arrecifes, Buenos Aires
Province. Buenos Aires Fm.

Order Charadriiformes, Family Laridae?:

Pseudosterna degener Mercerat 1897

LOCALITY AND HORIZON: Lujan, Buenos Aires Prov-
ince. Lujan Fm.

Pseudosterna pampeana Mercerat 1897

LOCALITY AND HORIZON: Arrecifes, Buenos Aires
Province, Buenos Aires Fm.

Order Accipitriformes, Family Falconidae:
Lagopterus minutus
Moreno and Mercerat 1891

LOCALITY AND HORIZON: Lujan, Buenos Aires Prov-
ince. Buenos Aires Fm.

REMARKS: The holotvpe of this species, which should
have been in the collections of the Museo de la Plata, could
not be found. The figure given by the authors is unclear, but
the specimen could probably be assigned to Polyborus.

Order Pelecaniformes,

Family Phalacrocoracidae:
Phalacrocorax brasilianus (Gmelin)

LOCALITY AND HORIZON: Lujan, Buenos Aires Prov-
ince.

REMARKS: Moreno and Mercerat (1891) named Phalacro-
corax pampeanus on the basis of an incomplete humerus,
which was actually a specimen of P. brasilianus (Gmelin).

Recent

The sediments named “lacustrine post-pampean” by
Ameghino (1898), or “platense stage” by Ameghino (1898) and
Doering (1882), were assigned, even in recent works, to the
Pleistocene, following the scheme used by Ameghino through-
out his works. But these sediments, at least those from typical
localities from which bird remains were collected, e.g., Cana-
da de Rocha, Lujan, and Mercedes, are actually post-Pleis-
tocene in age. The sediments of the “platense stage” were de-
posited in enclosed basins and flood plains as lentic and lotic
deposits in Buenos Aires Province, and no extinct megafauna
has ever been recovered from them. All the species recovered
as fossils belong to the Recent indigenous fauna of the area.
With reference to the birds, even Ameghino (1898) pointed out
that their remains are identical to those of Recent species. The
following species are all from these deposits.

Sarcoramphus fossilis Moreno and Mercerat 1891 and Ca-
thartes fossilis Moreno and Mercerat 1891 were correctly as-
signed to Vidtur gryphus Linnaeus and Cathartes aura Lin-
naeus by Ameghino (1891). The assignment of Sarcoramphus
fossilis to S. papa by Brodkorb (1963:257) is incorrect.

Foetopterus ambiguus Moreno and Mercerat 1891 was as-
signed to Cliloephaga picta Gmelin by Tonni (1970).

Rhea fossilis Moreno and Mercerat 1891 is assignable to R.
americana, as stated by Ameghino ( 1891) and verified by Brod-
korb (1963:201). Rhea subpampeana Moreno and Mercerat
1891 is also assignable to R. americana.

DISCUSSION

SPHENISCIFORMES. The penguins from the “Patagoni-
ano” comprise nine species distributed in four genera. Of the
nine species, seven are found in deposits of similar age (late
Oligocene-early Miocene) and in the same locality (lower
course of the Rio Chubut between Gaiman and Trelew). Ac-
cording to Simspon (1972), this is the greatest known diversity
of penguins for a restricted area and geologic age, including
the present.

The numerous new collections obtained by the Museo de La
Plata in recent years have partially verified Simpson’s (1972)
hypothesis that the larger-sized species like Arthrodytes gran-
dis and Paraptenodytes brodkorbi are found in more southerly
deposits (San Julian), whereas those of smaller size come from
more northerly deposits (Gaiman, Trelew). An exception to
this is a record of Paraptenodytes robustus from Gaiman. Pa-
laeospheniscus gracilis appears to be represented exclusively
in those sediments of the “Patagoniano” immediately overlying

1 12

Tonni: Cenozoic Birds of Argentina

the contact with the Colhue-Huapi Fm. in the lower course of
the Rio Chubut, in association with Lamna cattica totuserrata
(Chondrichthyes: Isuridae) which is very abundant in the same
levels, but which totally disappears 20 to 40 m above the
above-mentioned contact.

RHEIFORMES. Opistodactylus patagonicus Ameghino
1895, from the early Miocene (Santacruzian) of Patagonia, is
the first record of the order Rheiformes. Patterson and Krag-
lievich (1960) correctly assigned this species to the Rheidae.
One record of a Recent genus is represented in the middle
Pleistocene (Ensenadan) of Buenos Aires Province, while fos-
sils of Recent species have been found in late Pleistocene (Lu-
janian) deposits of the same area. All known fossil rheas have
been found within the present area of distribution of the fam-
ily.

TINAMIFORMES. The first representatives of this order
are found in Pliocene (Montehermosan) sediments of Buenos
Aires Province. Two living genera, Eudromia and Nothura,
are reported from these deposits, and perhaps also the Recent
species Nothura maculosa (see Tonni 1977a).

PROCELLARIIFORMES. Argyrodyptes microtarsus
Ameghino 1905, from marine sediments of the “Patagoniano,”
was described by Ameghino as a species of penguin. Simpson
(1946) doubted that the species could be assigned to the Sphe-
niscidae, and demonstrated its probable relationships with the
Procellariiformes. Brodkorb (1963) correctly concluded that A.
microtarsus belonged to the Procellariidae, a fact corroborated
by Tonni (in press).

The Procellariidae are well represented in the early and
middle Miocene of North America and Europe, but Argyro-
dyptes microtarsus is the only fossil record of the family for
South America.

PELECANIFORMES. Cladornis pachypus Ameghino
1895, from lower Oligocene (Deseadan) sediments of Santa
Cruz Province, is the first record for the order in South Amer-
ica. Wetmore (1960) placed this species in a new suborder,
Cladornithes, that would also include Cyphornis magnus Cope
1894 and Palaeochenoides mioceanus Shufeldt 1916, both
from the early Miocene of North America.

Liptornis hesternus Ameghino 1895, of the early Miocene
(Santacruzian) of Santa Cruz Province, represents, together
with Pelecanus gracilis Milne-Edwards 1863 from the early
Miocene of Europe, the oldest record for the Pelecanidae. This
family is now comprised of only one genus, Pelecanus, that is
widely distributed in almost all tropical and warm temperate
areas of the world. Over a dozen extinct species are known,
all but Liptornis hesternus referred to Pelecanus.

The only records of the Phalacrocoracidae in South America
are the reports of the Recent Phalacrocorax brasilianus in late
Pleistocene deposits of Argentina (Buenos Aires Province) and
Brazil, andP. olivaceus andP. bougainvillii in late Pleistocene
deposits of Peru (Campbell 1979).

PHOENICOPTERIFORMES. The order Phoenicopteri-
formes is represented by two species of the extinct family Pres-
byornithidae, Presbyornis antiquus (Howard 1955) and P. perv-
etus Wetmore 1926 from the early Eocene (Casamayoran) of
Chubut Province. P. pervetus has also been reported for de-
posits of a similar age in North America.

ARDEIFORMES. Ciconiopsis antartica Ameghino (1899,
from lower Oligocene (Deseadan) deposits of Santa Cruz Prov-
ince, is the oldest record of the Ciconiidae for the Americas,

and for the world as well since this family is recorded for the
first time in the Old World (Europe) inthe late Eocene-early
Oligocene.

Protibis cnemialis Ameghino 1891, from the early Miocene
(Santacruzian) of Patagonia, is also the oldest record of the
Plataleidae for the Americas.

ANSERIFORMES. The position of Teleornis impressus
Ameghino 1899 and Loxornis clivus Ameghino 1895 within
this order is uncertain. Species undoubtedly belonging to the
Anatidae are recorded from the early Miocene (Santacruzian)
with the appearance of Eoneornis australis Ameghino 1895
and Eutelornis patagonicus Ameghino 1895, and earlier in the
marine sediments of the “Patagoniano” (late Oligocene-early
Miocene) where an extinct species with affinities with the
Tachyerini has been found (Tonni in press). Neospecies have
been recorded from middle Pleistocene deposits onward in
Buenos Aires Province (Tonni 1969).

ACCIPITRIFORMES. Cruschedula revola Ameghino 1899
and Climacarthrus incompletus Ameghino 1899 are both based
on non-diagnostic material and their position within the order
is uncertain, but their assignment to the Accipitridae seems
reasonable. Undoubted members of the Accipitridae are re-
ported for the Santacruzian, with the appearance of Thegornis
musculosus Ameghino 1895 and T. debilis Ameghino 1895.
Badiostes patagonicus Ameghino 1895 is also from Santacru-
zian deposits.

The first records of these birds of prey coincides with those
of the caviomorph rodents, during the early Oligocene (Desea-
dan). In the Miocene (Santacruzian and Friasian), such rodents
and other small rodent-like mammals, e.g., the Interatheriidae
and Hegetotheriidae, acquire a great species diversity and pop-
ulation density, coinciding with a notable increase in the num-
ber of accipitriforms represented during the Friasian (in ad-
dition to those listed above, two or three species have been
collected but have yet to be studied).

GALLIFORMES. Anisolornis excavatus Ameghino 1891,
from the early Miocene (Santacruzian) of Patagonia, is the
oldest record for the Cracidae in South America. The family
is at present restricted to the warm forested areas of the Neo-
tropical Region. Only a few fossils, all attributed to neospecies,
are known from the late Pleistocene (Brazil, Peru), so their
evolutionary history is almost totally unknown.

OPISTHOCOMIFORMES. This order is represented in the
early Eocene (Casamayoran) of Patagonia by Onychopteryx
simpsoni Cracraft 1971. The ecological requirements of this
species seem to have been similar to those of the living rep-
resentatives of the order, as suggested by the presence of other
faunistic indicators, e.g., Leptodactylidae, Alligatoridae, and
Boidae. During the early Eocene a fauna evolved in Patagonia
under conditions of relatively high temperature and humidity,
along with a shrubby or grassy steppe vegetation (indicated by
the presence of grazing mammals) and isolated forests.

RALLIFORMES. This is the best represented avian order,
both in terms of diversity and number of specimens, of the
Cenozoic of Argentina. At the present, numerous species dis-
tributed in six families live in South America. In Tertiary sed-
iments dating from the early Eocene of Chubut Province (ma-
terials from Canadon Hondo are still being studied) are found
representatives of five other, extinct, families. These five fam-
ilies, generally called “Phorusrhacoids,” differentiated into
several different adaptive types, such as cursorial predators

Tonni: Cenozoic Birds of Argentina

113

(Phorusrhacidae), graviportal scavengers (Brontornithidae),
and cursorial predators capable of limited flight (Psilopteri-
dae).

In the Deseadan sediments of Patagonia three of the extinct
families are represented: Phorusrhacidae, Brontornithidae,
and Psilopteridae. A species of the Cariamidae, Riacama calig-
inea Ameghino 1899, is recorded for the first time. This family
is restricted to South America, being represented at the present
time by two monotypic genera, Cariama Brisson and Chunga
Hartlaub. A new species of Chunga is currently being de-
scribed from Montehermosan sediments of Buenos Aires Prov-
ince.

Cunampaia simplex Rusconi 1946 is recorded from the late
Eocene-early Oligocene of Mendoza Province. It represents a
monotypic family of questionable taxonomic position.

During the Santacruzian the large phorusrhacoids reach
their greatest diversity. At this time there existed representa-
tives of four families: the three mentioned above for the De-
seadan, and the Palaeociconiidae. The decline of this group,
both in numbers and diversity, began before the end of this
period. During the Huayquerian, only the Phorusrhacidae and
Psilopteridae are present, represented by a total of four species.
In Montehermosan deposits there are also records of the last
two mentioned families, but the number of specimens found
in well-known localities is low. There are no records of these
large ralliforms in Uquian deposits.

The oldest record for the family Aramidae is Aminornis ex-
cavatus Ameghino 1899, from the early Oligocene of Santa
Cruz Province. Loncornis erectus Ameghino 1899, from the
same locality and horizon, is placed with doubt in that family
by Brodkorb (1967).

Euryonotus brachypterus Mercerat 1897 and E. argentinus
Mercerat 1897 were placed in the Rallidae by Mercerat. The
type material could not be located and the descriptions are too
brief to enable verification of that assignment. These fossils
came from late Pleistocene deposits of Buenos Aires Province.

CHARADRIIFQRMES. The Laridae are represented in the
late Pleistocene of Buenos Aires Province by two presumed
extinct species: Pseudosterna degener Mercerat 1897 and P.
pampeana Mercerat 1897. But, as with the case with the rallids
described by Mercerat, their taxonomic position is very doubt-
ful.

PSITTACIFORMES. One extinct species of the Psittacidae,
Cyanoliseus ensenadensis (Cattoi 1957), is known from Argen-
tina (middle Pleistocene of Buenos Aires Province). The living
species C. patagonus was present in the late Pleistocene of
Argentina (Buenos Aires Province).

PASSERIFORMES. The Furnariidae are represented by an
extinct species of Cinclodes, C. major Tonni 1977, from the
Ensenadan of Buenos Aires Province. C. major shows a degree
of specialization similar to that of C. patagonicus (Gmelin).

The only other family of this order unquestionably repre-
sented in Cenozoic sediments of Argentina is the Emberizidae,
of which one extinct species, Zonotrichia robusta Tonni 1969,
and one undetermined species of Sicalis are known. Both are
from middle Pleistocene deposits of Buenos Aires Province.

LITERATURE CITED

Acenolaza, F.G. 1976. Consideraciones bioestratigraficas
sobre el terciario marine de Parana y alrededores. Acta,
Geol. Lilloana 13(2):9 1—108 .

Ameghino, C., and C. Rusconi. 1932. Nueva subespecie
de avestruz fosil del plioceno de Buenos Aires. An. Soc.
Cient. Arg. 114:38-42.

Ameghino, F. 1882. Catalogo explicativo de las colecciones
de antropologia prehistorica y de paleontologia de Flo-
rentine Ameghino. Cat. Exp. Continental Sud Ameri-
cana, anexo A:35-42.

. 1887. Enumeracion sistematica de las especies de

mamiferos fosiles coleccionadas por Carlos Ameghino en
los terrenos eocenos de la Patagonia austral y depositados
en el Museo La Plata. Bol. del Museo La Plata 1:1-26.

. 1891. Enumeracion de las aves fosiles de la Republica

Argentina. Rev. Arg. Hist. Nat. 1:441-453.

. 1895. Sur les oiseaux fossiles de la Patagonia. Bol.

Inst. Geog. Arg. 15:501-602.

. 1898. Sinopsis geologica-paleontologica. Pp. 111-255

in Segundo Censo de la Republica Argentina; Cap. I:
Territorio, Parte III, Vol. I.

. 1899. Sinopsis geologica-paleontologica. Suplemento

(Adiciones y correcciones). 13 pp. La Plata.

. 1901. L’age des formations sedimentaires de Pata-

gonie. An. Soc. Cient. Arg. 51:20-39, 65-91.

. 1905. Enumeracion de los impennes fosiles de la Pat-
agonia e Isla Seymour. An. Mus. Nac. Buenos Aires 6:97—
167.

Berggren, W.A., AND J.A. van Couvering. 1974. The
Late Neogene. Biostratigraphy, geochronology and paleo-
climatology of the last 15 million years in marine and
continental sequences. Palaeogeog., Palaeoclim., Palaeo-
ecol. 16( 1—2): 1—2 16.

Bertels, A. Micropaleontology and paleogeography of the
Upper Cretaceous and Cenozoic of Argentina. 7th African
Micropaleont. Colloquium, (in press).

Brodkorb, P. 1963. Catalogue of fossil birds. Part 1 (Ar-
chaeopterygiformes through Ardeiformes). Bull. Florida
State Mus., Biol. Sci. 7(4): 179-293.

— . 1964. Catalogue of fossil birds. Part 2 (Anseriformes

through Galliformes). Bull. Florida State Mus., Biol. Sci.
8(3): 195-335.

. 1967. Catalogue of fossil birds. Part 3 (Ralliformes,

Ichthyornithiformes, Charadriiformes). Bull. Florida
State Mus., Biol. Sci. 1 1(3):99-220.

. 1978. Catalogue of Fossil Birds. Part 5 (Passeri-
formes). Bull. Florida State Mus., Biol. Sci. 23(3): 139—
228.

Camacho, H.H. 1974. Bioestratigrafia de la formaciones
marinas del Eoceno y Oligoceno de la Patagonia. An.
Acad. Nac. Cienc. Exac. Fis. y Nat. 26:39-57.

Campbell, K.E., Jr. 1979. The Non-Passerine Pleistocene
Avifauna of the Talara Tar Seeps, Northwestern Peru.
Royal Ontario Mus., Life Sci. Contrib. 118:1-203.

Campbell, K.E., Jr., and E. Tonni. 1980. A New Genus
of Teratorn from the Huayquerian of Argentina (Aves:
Teratornithidae). (this vol.)

Cabrera, A. 1939. Sobre vertebrados fosiles del plioceno de
Adolfo Alsina. Rev. Mus. La Plata (n.s.), Paleont. 2:15—
21.

Cattoi, N. 1957. Contribucion al estudio de la avifauna ex-
tinguida del pleistoceno de la Republica Argentina.
Ameghiniana 1(1-2): 17-24.

Cione, A.L., and E. Exposito. Chondrichthyes (Pisces) del

114

Tonni: Cenozoic Birds of Argentina

“Patagoniano” s. 1. de Astra, Golfo de San Jorge, provin-
cia de Chubut, Argentina. Act. II Cong. Arg. Paleont. y
Bioestrat. y I Cong. Latinoamer. de Paleont. (in press).

Cracraft, J. 1968. A review of the Bathornithidae (Aves,
Gruiformes) with remarks on the relationships of the sub-
order Cariamae. Amer. Mus. Nov. 2326:1-46.

. 1969. Systematics and evolution of the Gruiformes

(Class Aves) 1. The Eocene family Geranoididae and the
early history of the Gruiformes. Amer. Mus. Nov. 2388:1-
41.

. 1970. A new species of Telmabates (Phoenicopteri-

formes) from the lower Eocene of Patagonia. Condor
7 2(4):479— 480.

. 1971. A new family of Hoatzin-like birds (order Op-

isthocomiformes) from the Eocene of South America. Ibis
1 1 3(2 ): 2 2 9 — 2 33 .

Doering, A. 1882. Informe Oficial de la Comision Cientifica
agregada al Estado Mayor General de la Expedition al
Rio Negro. 3a. parte, Geologia:29S-S30.

Feduccia, A. 1976. Osteological evidence for shorebird af-
finities of the flamingos. Auk 93:587-601.

Howard, H. 1955. A new wading bird from the Eocene of
Patagonia. Amer. Mus. Nov. 1710:1-25.

Kraglievich, J. 1946. Noticia preliminar acerca de un nue-
vo y gigantesco estereornito de la fauna Chapadmalense.
Anal. Soc. Cient. Arg. 142:104-121.

Kraglievich, L. 1931. Restos de vertebrados vivientes y
extinguidos hallados por los senores E.R. Wagner y her-
mano en tumulos precolombianos de Santiago del Estero.
Phvsis 10:229-241.

Marshall, L.G., R. Pascual, G.H. Curtis, and RE.
Drake. 1977. South American geochronology: radio-
metric time scale for middle to late Tertiary mammal-
bearing horizons in Patagonia. Science 195:1325-1328.

Marshall, L.G., R.F. Butler, R.E. Drake, G.H. Curtis,
R.H. Curtis, and R.H. Tedford. 1979. Calibration
of the great American interchange. Science 204(4390): 2 72—
279.

Mercerat, A. 1897. Note sur les oiseaux fossiles de la Re-
publica Argentina. An. Soc. Cient. Arg. 43:222-240.

Moreno, F.P., and A. Mercerat. 1891. Catalogo de los
pajaros fosiles de la Republica Argentina conservados en
el Museo de La Plata. An. Mus. La Plata, Paleont. Arg.
1:7-71.

Pascual, R., and O.E. Odreman Rivas. 1973. Las uni-
dades estratigraficas del Terciario portadoras de mami-
feros. Su distribution y sus relaciones con los aconteci-
mientos diastroficos. Act. V° Cong. Geol. Arg. 3:293-338.

Pascual, R., E.J. Ortega, D. Gondar, and E. Tonni.
1965. Las Edades del Cenozoico mamalifero de la Ar-
gentina con especial atencion a aquellas del territorio bon-
aerense. An. Com. Inv. Cient. Prov. Buenos Aires 6: 1 65—
193.

Patterson, B. 1941. A new phororhacoid bird from the
Deseado Formation of Patagonia. Field Mus. Nat. Hist.,
Geol. Ser. 8:49-54.

Patterson, B., and J.L. Kraglievich. 1960. Sistematicas

y nomenclatura de las aves fororracoideas del Plioceno
Argentino. Publ. Mus. Munic. Cs. Nat. y Trad. Mar del
Plata 1(1): 1-51.

Riggi, J.C. 1979. Nuevo esquema estratigrafico de la For-
mation Patagonia. Rev. Asoc. Geol. Arg. 34(1): 1-11.

Rovereto, C. 1914. Los Estratos Araucanos y sus fosiles.
An. Mus. Nac. Hist. Nat. Buenos Aires 25:1-249.

Rusconi, C. 1946. Ave y reptil oligoceno de Mendoza. Bol.
Paleont. de Buenos Aires 21:1-3.

. 1958. Ave fosil del Plioceno de Buenos Aires. Rev.

Mus. Hist. Nat. 11:157-160.

Savage, D.E. 1975. Cenozoic. The Primate episode. Pp. 2-
27 in Approaches to Primate Paleobiology. Contributions
to Primatology. (F.S. Szalay, Ed.). Vol. 5. Karger, Basel
and New York.

Simpson, G.G. 1946. Fossil penguins. Bull. Amer. Mus.
Nat. Hist. 87(1): 1-99.

. 1970. Miocene penguins from Victoria, Australia,

and Chubut, Argentina. Mem. Nat. Mus. Victoria 3 1: 17-
24.

. 1972. Conspectus of patagonian fossil penguins. Am.

Mus. Nov. 2488:1-37.

Tonni, E.P. 1969. La presencia de Anas leucophrys (Aves,
Anseriformes) en sedimentos de Edad Ensenadense (Pleis-
toceno medio) de la provincia de Buenos Aires. Ameghi-
niana 6(4):309-313.

. 1970. Foetopterus ambiguus Mor. et Merc. 1891: su

asignacion a Chloephaga picta (Aves, Anseriformes).
Ameghiniana 7(3): 2 7 9—2 80.

. 1970a. Zonotrichia robusta n. sp. (Aves, Passeri-
formes) del Pleistoceno medio de la provincia de Buenos
Aires. Ameghiniana 7(2): 161-165.

. 1972. Cyanolisens ensenadensis (Cattoi) nov. comb.

(Aves, Psittaciformes) del Pleistoceno medio de la provin-
cia de Buenos Aires. Ameghiniana 9(2): 145-148.

. 1973. Las aves de Edad Ensenadense (Pleistoceno

medio) de la provincia de Buenos Aires. Tesis Fac. Cs.
Nat. y Museo de La Plata No. 213. 90 pp.

. 1974. Un nuevo cariamido (Aves, Gruiformes) del

Plioceno superior de la provincia de Buenos Aires.
Ameghiniana 1 1(4):366-372.

. 1977. Un furnarido (Aves, Passeriformes) del Pleis-
toceno medio de la provincia de Buenos Aires. Publ. Mus.
Munic. Cs. Nat. Mar del Plata 2(6): 141-147.

. 1977a. Notas sobre tinamidos fosiles argentinos. I:

el genero Tinatnisornis Rovereto, 1914. Ameghiniana
14( 1 — 4): 2 2 4 — 232 .

. Un nuevo anseriforme de sedimentos marinos terci-

arios de Chubut, Argentina. El Hornero, Rev. Soc. Ornit.
del Plata (in press).

. Un nuevo psilopterido (Aves, Gruiformes) del Pli-
oceno temprano de la provincia de Buenos Aires. Publ.
Mus. Munic. Cs. Nat. Mar del Plata (in press).

Wetmore, A. 1926. Fossil birds from the Green River de-
posits of eastern Utah. Ann. Carnegie Mus. 16:391-402.

. 1960. A classification for the birds of the world.

Smith. Misc. Coll. 139( 1 1 ): 1—3 7 .

A THICK-KNEE (AVES: BURHINIDAE) FROM THE PLEISTOCENE
OF NORTH AMERICA, AND ITS BEARING
ON ICE AGE CLIMATES

By Alan Feduccia1

ABSTRACT: A new species of thick-knee (Aves: Burhinidae) from the Pleistocene (Sangamon) of

North America (Decatur County, Kansas) is described; it is named Burhinus aquilonaris new species. The
genus Burhinus is a good indicator of dry, tropical savannah, and its presence in Kansas during the
Sangamon is evidence that similar habitat existed in the High Plains at that time.

The late Professor Claude W. Hibbard of the University of
Michigan amassed an exceptionally large collection of fossil
birds in his quest for small mammals. These avian fossils are
fundamental to our understanding of the late Pliocene avi-
fauna of North America and have been reviewed and discussed
elsewhere (Feduccia 1975).

Hibbard’s theories on the climates of the Pliocene and Pleis-
tocene of North America were less than well received over the
years and even now are not generally accepted; this was a
constant plague to him and resulted in many heated debates.
“Hibbie,” as he was affectionately known, was constantly
awaiting additional fossil evidence to bolster his radical views
of Plio-Pleistocene climates, but unfortunately there was little
among the collections of fossil birds to provide clues to paleo-
climate or paleoecologv. One such fossil had been sent to the
late Dr. Alexander Wetmore in 1943. I have examined the
considerable volume of correspondence between Hibbard and
Wetmore over this specimen. This began in 1943 and in it
Hibbard continually urged Wetmore to publish a description
of the fossil. Wetmore was extremely busy during those years,
including among his duties being Secretary of the Smithsonian
Institution. The fossil concerned was a thick-knee ( Burhinus )
from the Pleistocene of North America, and Hibbard had
known its generic identity since the 1940’s.

In a letter from Hibbard to Wetmore in 1962, Hibbard
writes: “Back in the early 40’s I collected a bird humerus. . . .
I sent the material to Professor Loye Miller who considered it
nearest to the Thick-knees of Tropical America. He returned
it due to the lack of comparative material and suggested that
I send it to you, which I did immediately. . . . Over the years
I have told my classes of the occurrence of a Thick-knee in the
fauna of Kansas from the very late Illinoian or early Sanga-
mon. Have I been wrong? ... Is the specimen good enough
for identification? If so, and a Thick-knee, could you possibly

1 Department of Zoology, University of North Carolina, Chapel Hill,
North Carolina 27514; and Research Associate, Department of Ver-
tebrate Zoology, National Museum of Natural History, Smithsonian
Institution, Washington, D C. 20560.

find time to report its occurrence. If it is such, I think it is
quite important . . . .” Modern species of Burhinus occur only
in dry, tropical savannahs, and Hibbard was most anxious to
see the description appear. Over the years “Hibbie” constantly
attempted to impress his students and colleagues with the im-
portance of a Pleistocene thick-knee from North America. In
an effort to complete some of Dr. Wetmore’s unfinished work,
the now famous thick-knee is here described. Wetmore left an
incomplete description of this fossil and the name used here
is that which he had chosen for it.

SYSTEMATICS

Order Charadriiformes
Family Burhinidae
Burhinus aquilonaris new species
Figure 1

HOLOTYPE: Left humerus (complete except for minor
breakage in middle of outer margin of deltoid crest), University
of Kansas Museum of Natural History, Vertebrate Paleontol-
ogy No. 6822. Collected in August 1943, by C.W. Hibbard.

LOCALITY AND HORIZON: From the late Pleistocene
(Sangamon) Sanborn Formation, bed No. 2, at level of high-
way culvert in bottom of draw on south side of Route U.S.
36, in NE 14, Sec. 3, T.3, R.27W., Decatur County, Kansas.
The holotype was mentioned specifically in an early report by
Hibbard et al. (1944), who, in listing a jaw fragment of a
prairie dog, Cynomys ludovicianus (Ord), “collected 6 feet
[ — 1.8 mj below the soil zone in bed 2 at locality no. 10,”
remarked that “the remains of a bird (KUMVP No. 6822)
. . . were found associated” with it.

MEASUREMENTS OF HOLOTYPE (in mm): Length
88.2; transverse width of shaft at center 5.2; transverse width
across distal end 13.2.

ASSOCIATED PARATYPES (probably from same indi-
vidual): A nearly complete left radius, the proximal and distal
portions of a left ulna, three rib fragments, and a fragmented
left ilium, same data and catalog number as holotype. These

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 330:115-118.

Feduccia: Pleistocene Burhinidae

116

Figure 1. Burhinus aquilonaris, new species, holotype left humerus, University of Kansas Museum of Natural History, Vertebrate Paleontology
No. 6822. From left to right, and bottom left to right: palmar, ventral, anconal, proximal, and distal views. All views X1U.

were found associated with the holotype humerus and almost
certainly belong to the same individual. The fossil radius mea-
sures 94.3 mm in total length, those of the continental races
of B. bistriatus (Wagler), 98.0 to 119.6 mm (x = 109.6, s.d. =
7.3, n = 13); ofB. b. dominicensis (Cory), 91.9-92.9 mm; and
of B. superciliaris (Tschudi), 87.7 and 94.2 mm.

ETYMOLOGY: Latin, aquilonaris, meaning northern or
northerly.

DIAGNOSIS: Intermediate in size between the continental
American races of Burhinus bistriatus and the smaller Bur-
hinus bistriatus dominicensis of Hispaniola; slightly larger
than Burhinus superciliaris. Fourteen specimens of humeri
representing the continental races of B. bistriatus range from
91.0 to 103.6 (x = 96.8, s.d. = 4.6) in length; three specimens

of B. b. dominicensis range from 82.2 to 84.1 mm (x = 83.5,
s.d. = 1.1); and two specimens of B. superciliaris measure
80.7 and 86.9 mm. Closest in characters to the living conti-
nental races of B. bistriatus, but differs in having (1) deltoid
crest slightly longer; (2) bicipital furrow smoother, without
definite depression at proximal end below head found in mod-
ern forms; (3) outline of head more angular, less smoothly
rounded; (4) internal tuberosity slightly heavier; (5) ectepicon-
dylar prominence with proximal external margin more angu-
lar; (6) entepicondyle more produced distally and more angu-
lar, with distal margin extending to level of distal margin of
trochlea; and (7) brachial depression more sharply marked,
with the impression of M. brachialis anticus more clearly de-
fined.

Feduccia: Pleistocene Burhinidae

117

REMARKS: The only record of an extinct species of Bur-
hinus is that of Burhinus nanus Brodkorb (1959), described
from Upper Pleistocene cave deposits on New Providence Is-
land, Bahamas. This species, based on a left coracoid, is dis-
tinguished from modern Burhinus bistriatus dominicensis of
Hispaniola by smaller size.

In addition, Burhinus superciliaris, a small South American
burhinid that is common in semi-arid regions of northwestern
South America (extending to northern Chile) is known as a
fossil from the late Pleistocene of southwestern Ecuador, and
from the late Pleistocene Talara Tar Seeps in northwestern
Peru (Campbell 1976, 1979). A larger American burhinid, Bur-
hinus bistriatus, that is known in arid regions from northern
South America northward into southeastern Mexico, is not
definitely recorded in the fossil record, but Howard (1971) has
reported burhinid fossils from the Pleistocene of Eddy County,
New Mexico, and from San Josecito Cave in northeastern
Mexico as approximating the living B. bistriatus in size.

HIBBARD’S INTERPRETATION OF
PLIOCENE AND PLEISTOCENE
CLIMATES

Hibbard was greatly impressed with the presence of large
land tortoises ( Geochelone ) in North America during the Plio-
cene and Pleistocene. In a now classic paper (Hibbard 1960),
he outlined a general climatic interpretation based on the as-
sumption that the overall physiology of large tortoises did not
differ significantly from that of the living species now restricted
to the Galapagos Islands, South America, Africa, Madagascar
and other islands of the Indian Ocean, Ceylon, India, and
southeastern Asia.

There is considerable dispute as to what constitutes the end
of the Upper Pliocene and the beginning of the Pleistocene, a
time that falls within a North American land mammal age,
the Blancan. The reason for this is that the first Pleistocene
fauna was also the last Pliocene fauna. In addition, there was
no boreal fauna (as at present) to be shifted southward with
the southward movement of the first continental glacier and
provide a paleontological indication of the onset of glaciation.
Nonetheless, Hibbard considered the very large, thick-shelled
land tortoise, Geochelone rexroadensis (Oelrich 1952), found
in association with a smaller tortoise, G. turgida (Oelrich
1957), in the Rexroad Formation, to indicate an equable cli-
mate in Kansas in the Blancan, a climate in which freezing
conditions did not exist. As he stated (Hibbard 1960:17), “It
is not known how far north the large tortoises ranged during
the Upper Pliocene, but it can be assumed that part of the
Upper Pliocene was as equable as the climate during part of
the first interglacial which allowed these large tortoises to live
as far north as Brown County, Nebraska (McGrew 1944:48-
49).” It seemed clear to Hibbard that there was no evidence
to indicate that, “Conditions essentially the same as those of
the present were reached by the Upper Pliocene . . . but the
outbreaks of polar air so characteristic of the present winters
appear to be a development of the late Pliocene” (MacGinitie
1958:70-71).

Faunas of the first glacial, or Nebraskan, are not well
known, and no fauna has been found that would have been
characteristic of the maximum extent of continental glacial
advance. However, immediately prior to the Nebraskan, the

giant land tortoises are known from the High Plains region as
far north as northern Nebraska (Sand Draw Locality — see
Zakrzewski 1975 for age), and parts of a large tortoise have
been found in deposits of the first interglacial (Aftonian) in
Meade County, Kansas (Deer Park local fauna). The pygal
scute of a tortoise from Nebraska represents a specimen of
approximately 2 m length (McGrew 1944). Hibbard (1960:19)
concluded that, “These large tortoises lived in a subtropical
climate . . . .’’In addition, the occurrence of a large, thick-
shelled Geochelone from the Aftonian Borchers local fauna of
Meade County, Kansas (above the Pearlette ash — see Za-
krzewski 1975), indicated to Hibbard (1960:21), “a subtropical
climate with winter temperatures not lower than 32°F [0°C].”
Even during the second glacial, the Kansan, remains of a
large, thick-shelled Geochelone (estimated length, 2 m) are
known from the Seymour Formation of northern Texas below
the Pearlette ash, in association with remains of alligator, glyp-
todonts, a giant armadillo, ground sloths, Stegomastodon,
mammoths, tapirs, horses, and a leptodactylid frog. This as-
sociation indicated to Hibbard that in Texas during the late
Kansan the winters must have been considerably warmer than
at present. By the Illinoian, or third glacial period, the High
Plains faunas take on an aspect of a more northern Recent
fauna than those of either the Nebraskan or Kansan. But even
in the third interglacial, or Sangamon, the large tortoises (Geo-
chelone) appear again in southwestern Kansas (Cragin Quarry
local fauna). These indicated to Hibbard that during the San-
gamon there existed in southwestern Kansas subtropical con-
ditions with winter temperatures seldom or never lower than
0°C. The last glacial, or Wisconsin, period is characterized by
the massive extinctions of the vertebrate fauna that charac-
terized the end of the Pleistocene of North America. Even the
giant tortoises that had been present continuously in Florida
throughout the Pleistocene became extinct. These extinctions,
which began approximately 12,000 years ago, are attributed
by Martin (1958 and numerous subsequent papers) to man
rather than climatic change, but Grayson (1977) has pointed
out that birds and mammals underwent comparable amounts
of generic extinction. The magnitude of avian extinctions (at
the same time as the mammalian extinctions) is incompatible
with Martin’s “Pleistocene Overkill Hypothesis.” This evi-
dence would, of course, lend support to Hibbard’s view that
the harsh climate and associated climatic zoning of Recent
times was a consequence of the Wisconsin glaciation. This
view probably also explains the massive Pleistocene extinctions
without the intervention of man.

BURHINUS IN THE PLEISTOCENE OF
NORTH AMERICA

Hibbard was extremely anxious for avian paleontologists to
describe the fossil birds that he had discovered, but as the
descriptions emerged there was little in the way of avian in-
dicators of climatic conditions in the Pleistocene. Collins (1964)
had identified some of the ibis bones as belonging to taxa now
confined to the Neotropics, but the bone fragments used in his
identifications are unidentifiable (Feduccia 1975). Burhinus
aquilonaris represents the first of Hibbard’s avian fossils to
clearly indicate something of the Pleistocene paleoclimate and
paleoecology of the High Plains. The various species of thick-
knees (Burhinidae) are widespread in tropical regions of the

118

Feduccia: Pleistocene Burhinidae

Old World. In the New World, the genus Burhinus is at pres-
ent confined to Hispaniola and arid portions of the mainland
from southern Mexico to Peru and Brazil. The genus is char-
acteristic of dry, scrubby, open country. Slud (1964:102) states
that in Costa Rica Burhinus bistriatus is “an indicator species
of the scrubby . . . cattle country, frequents openings in bushy
and thickety growth amid low woodland as well as open ranges
and coastal grassy ‘plains’ akin to salt marshes. A terrestrial
bird, occurring singly or in small groups, it is both diurnal and
nocturnal.” Slud’s description is typical for the American
species of Burhinus, and can be reasonably extended to char-
acterize the Pleistocene species as well. The species of Bur-
hinus are thus good indicators of tropical, dry savannah, and
we can assume this to have been the habitat in Decatur Coun-
ty, Kansas, during the Sangamon when Burhinus aquilonaris
lived there.

ACKNOWLEDGMENTS

I thank Storrs L. Olson who encouraged this study, placed
the specimen (on loan from the University of Kansas Museum
of Natural History) at my disposal, and offered helpful sug-
gestions on the manuscript. Specimens of modern Burhinus
were borrowed through the courtesy of W.E. Lanyon (Amer-
ican Museum of Natural History), J.P. O’Neill (Louisiana
State University Museum of Zoology), R. W. Storer (University
of Michigan Museum of Zoology), N.K. Johnson (Museum of
Vertebrate Zoology, Berkeley), and R.W. Schreiber (Natural
History Museum of Los Angeles County). The illustration for
Figure 1 was skillfully rendered by L.B. Isham. This is con-
tribution Number 1 of the Wetmore Papers, a project sup-
ported in part by Trust Funds from the Smithsonian Institu-
tion for completing unfinished work and study of undescribed
material left by the late Alexander Wetmore.

LITERATURE CITED

Brodkorb, P. 1959. Pleistocene birds from New Providence

Island, Bahamas. Bull. Fla. State Mus., Biol. Sci.

4(1 1):349 — 37 1.

Campbell, K.E., Jr. 1976. The late Pleistocene avifauna of

La Carolina, southwestern Ecuador. Smithsonian Contr.
Paleobiology No. 27:155-168.

. 1979. The Non-Passerine Pleistocene Avifauna of the

Talara Tar Seeps, Northwestern Peru. Royal Ontario
Mus., Life Sci. Contr. No. 118:1-203.

Collins, C.T. 1964. Fossil ibises from the Rexroad fauna
of the upper Pliocene of Kansas. Wilson Bull. 76:43-49.
Feduccia, A. 1975. Professor Hibbard’s fossil birds. Univ.
Michigan Mus. Paleontology, Misc. Pap. No. 1 2(3):67—
70.

Grayson, D.K. 1977. Pleistocene avifaunas and the overkill
hypothesis. Science 195:691-693.

Hibbard, C.W. 1960. An interpretation of Pliocene and
Pleistocene climates in North America. Mich. Acad. Sci.,
Arts, and Letters, 62nd Ann. Rep. Pp. 5-30.

Hibbard, C.W., J.C. Frye, and A.B. Leonard. 1944. Re-
connaissance of Pleistocene deposits in North-central
Kansas. State Geol. Surv. Kansas, Bull. 50(Pt. 1):28 pp.
Howard, H. 1971. Quaternary avian remains from Dark
Canyon Cave, New Mexico. Condor 73(2):237— 240.
MacGinitie, H.D. 1958. Climate since the late Cretaceous.
Pp. 61-79 in Zoogeography (Carl L. Hubbs, Ed.). Amer.
Assoc. Adv. Sci., Publ. 51.

Martin, P.S. 1958. Pleistocene ecology and biogeography of
North America. Pp. 61-79 in Zoogeography (Carl L.
Hubbs, Ed.). Amer. Assoc. Adv. Sci., Publ. 51.
McGrew, P.O. 1944. An early Pleistocene (Blancan) fauna
from Nebraska. Field Mus. Nat. Hist., Geol. Ser. 9 (2):33-
66.

Oelrich, T.M. 1952. A new Testudo from the Upper Plio-
cene of Kansas with additional notes on associated
Rexroad mammals. Trans. Kans. Acad. Sci. 55(3):300—
311.

Oelrich, T.M. 1957. The status of the Upper Pliocene tur-
tle, Testudo turgida Cope. J. Paleo. 31(1):228-241.

Slud, P. 1964. The birds of Costa Rica. Bull. Amer. Mus.
Nat. Hist. 128:1-430.

Zakrzewski, R.J. 1975. Pleistocene stratigraphy and pa-
leontology in western Kansas: the state of the art, 1974.
Univ. Michigan Mus. Paleontology, Misc. Pap. No.
12(3): 12 1—128.

A REVIEW OF THE RANCHOLABREAN AVIFAUNA OF THE
ITCHTUCKNEE RIVER, FLORIDA

By Kenneth E. Campbell, Jr.1

ABSTRACT: A collection of 1363 fossil bird bones representing 56 species is described, bringing the

known avifauna from the late Pleistocene (Rancholabrean) deposits of the Itchtucknee River, Florida, to
67 species represented by over 1750 specimens. A predominance of aquatic species indicates the presence
of extensive freshwater pond and marsh habitat. A mid-Wisconsinan age, or younger, is suggested for the
deposits. The recognition of a paleospecies of Milvago (Accipitriformes: Falconidae) in Florida marks the
first record of the genus for North America.

The Itchtucknee River, located in Columbia County, Flor-
ida, has long been known as a rich source of Pleistocene fossils,
including mammals (Simpson 1929, 1930, 1932; Kurten 1965;
Martin 1969), reptiles (Auffenberg 1963), and birds (Wetmore
1931; McCoy 1963).

In the years subsequent to McCoy’s report on the paleoavi-
fauna, over 1300 additional specimens of fossil birds have been
recovered from the Itchtucknee River by numerous collectors;
these form the basis for the present paper. Fossils from the
Itchtucknee River are generally found as stream float along
the bottom of the river, but they may also be collected as they
wash out of the deposits. The avian fossils reported here were
collected from a number of sites. Although the locality is given
as simply “Itchtucknee River,” some specimens have more spe-
cific locality data. The specimens reported below are part of
the collections of the Florida State Museum, University of
Florida.

As discussed by McCoy (1963), the Itchtucknee River is con-
sidered a Pleistocene, possibly mid-Wisconsinan, locality.
Webb (1974) lists it among the Rancholabrean sites of Florida
based on the mammalian taxa contained therein. The exact
age of most of the specimens reported here cannot be deter-
mined because the river is eroding deposits of probably dif-
ferent ages, and often redepositing fossils on younger sedi-
ments. Fossils of different ages are often mixed on the river
bottom. Also, because the river meanders in places, the pro-
cesses of erosion and redeposition continued over time have
resulted in deposits containing fossils of mixed ages; these are
now being exposed on the bottom of the river once again. I do
not believe any of the fossils are older than Rancholabrean,
but because of the mixing action of the river there may be a
few post-Rancholabrean fossils inadvertently included. Com-
ments by McCoy (1963) as to the type of preservation of the
fossil materials and the nature of the river and specific sites
are still pertinent and will not be repeated here.

1 Natural History Museum of Los Angeles County, 900 Exposition
Boulevard, Los Angeles, CA 90007.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 330:119-129.

The following faunal list includes only those specimens I
identified; the fossils identified by McCoy are listed in Table
1. Many of the specimens identified by McCoy (1963) have
been rechecked, either by myself or others, resulting in the
correction of some misidentifications (see Table 1). A check of
all McCoy’s identifications would be very difficult because he
reported on fossils from four different collections and did not
give catalog numbers of most of the specimens he listed.

FAUNAL LIST

Podilymbus podiceps (Linnaeus 1758)
Pied-billed Grebe

MATERIAL: Coracoids (19); humeri (55); ulnae (7); carpo-
metacarpi (2); femora (17); tibiotarsi (22); tarsometatarsi (31).
UF/FSM 15209-15213, 15215-15219, 15221-15222, 15224-
15324, 22305-22341.

REMARKS: Storer ( 1976) studied most of the material listed
above for his report on Pleistocene Pied-billed Grebes. I concur
with his conclusion that P. magnus should be regarded as a
synonym of P. podiceps.

Podilymbus wetmorei Storer 1976

MATERIAL: Two complete left femora, 1 complete left
tarsometatarsus. UF/FSM 15214, 15220, 15223.

REMARKS: Storer (1976) referred these specimens to P
wetmorei in his original description of that species on the
grounds that the femora are much heavier and the tarsometa-
tarsus is much thicker than those of P. podiceps. Although I
agree with Storer that the specimens referred to P. wetmorei
are significantly larger than corresponding bones of P. podi-
ceps, I think there is a good possibility that the former may
have been a resident race of P. podiceps in peninsular Florida
during the Pleistocene. The type locality forP. wetmorei is the
Dixie Lime Products quarry, Locality IA, 1.6 km south of
Reddick, Marion County, Florida. The lack of any good os-
teological characters other than size that separate P. podiceps
and P. wetmorei might be taken as evidence for this possibil-
ity.

120

Campbell: Itchtucknee Avifauna

Table 1. Avian species from the late Pleistocene deposits of the Itchtucknee River, Florida.

(From McCoy 1963)

Added

Total

Minimum

Minimum

Minimum

Number of

Number of

Number of

Number of

Number of

Number of

Species

Specimens

Individuals

Specimens

Individuals

Specimens

Individuals

Podilymbus podiceps

601

4

150

27

210

31*

fPodilymbus wetmorei

0

0

3

2

3

2

Podiceps auritus

1

1

0

0

1

1

Phalacrocorax auritus

21

3

74

9

95

12*

Anhinga anhinga

1

1

3

2

4

2

Ardea herodias

8

2

39

8

47

8

Casmerodius albus

O2

0

1

1

1

1

Butorides virescens

1

1

0

0

1

1

Egretta caerulea

l3

1

1

1

2

2*

Nycticorax nycticorax

2

1

4

1

6

2*

Botaurus lentiginosus

24

1

3

1

5

1

t Ciconia maltha

2

1

9

2

11

2

Mycteria americana

05

0

3

1

3

1

Eudocimus albus

1

1

0

0

1

1

Olor columbianus

1

1

9

4

10

4

Olor buccinator

15

5

47

5

62

5

Branta canadensis

25*

4

83

9

108

9

Anas platyrhynchos

197

3

144

24

163

24

Anas rubripes

6

2

10

4

16

6

Anas acuta

6

2

25

4

31

5

Anas discors

16

6

30

9

46

9

t Anas itchtucknee

1

1

0

0

1

1

Anas crecca

3

1

4

2

7

2

Anas strepera

4

1

1 1

4

15

4

Anas americana

5

4

3

1

8

4

Anas clypeata

5

2

32

8

37

9

Aix sponsa

9

3

27

7

36

8*

Aythya ajfinis

17

4

40

14

57

17*

Aythya collaris

25

10

115

22

140

32*

Aythya americana

2

1

2

1

4

2

Avthva valisineria

16

4

95

15

111

15

Bucephala clangula

0

0

3

2

3

2

Bucephala albeola

1

1

0

0

1

1

Lophodytes cucullatus

4

2

12

5

16

8*

Mergus merganser

1

1

0

0

1

1

Mergus serrator

0

0

1

1

1

1

Oxyura jamaicensis

0

0

5

1

5

1

Cathartes aura

0

0

2

1

2

1

Coragyps atratus

1

1

1

1

2

1

t Gymnogyps amplus

1

1

0

0

1

1

Accipitridae gen. et sp. indet.

l8

1

1

1

2

1

Buteo lineatus

0

0

3

1

3

1

Buteo jamaicensis

1

1

6

3

7

3

Haliaetus leucocephalus

0

0

3

2

3

2

t Milvago readei

0

0

2

1

2

1

Pandion haliaetus

10

3

15

5

25

6*

Colinus virginianus

0

0

4

1

4

1

Meleagris gallopavo

27

9

141

18

168

23*

Grus americana

3

1

16

3

19

3

Grus canadensis

0

0

3

1

3

1

Aramus guarana

0

0

4

1

4

1

Rallus elegans

2

1

4

1

6

2*

Rallus limicola

1

1

0

0

1

1

Porzana Carolina

1

1

1

1

2

1

Gallinula chloropus

15

3

63

7

78

7

t Gallinula brodkorbi

5

3

29

5

34

5

Porphyrula martinica

0

0

7

3

7

3

Fulica americana

389

11

55

16

93

27

jEctopistes migratorius

0

0

1

1

1

1

Zenaidura macroura

0

0

l

1

1

1

Tyto alba

l.o

1

1

1

2

1

Otus asio

0

0

2

1

2

1

Strix varia

0

0

1

1

1

1

Colaptes auratus

1

1

0

0

1

1

Campbell: Itchtucknee Avifauna

121

Table 1. Continued.

Species

(From McCoy 1963)

Added

Total

Number of
Specimens

Minimum
Number of
Individuals

Number of
Specimens

Minimum
Number of
Individuals

Number of
Specimens

Minimum
Number of
Individuals

Corvus brachyrhynchos

0

0

9

2

9

2

Agelaius phoeniceus

1

1

0

0

1

1

Quiscalus quiscula

2

1

0

0

2

1

t Extinct species.

* Approximate only. Total obtained by adding figures of McCoy with those reported on here, which will not necessarily give an accurate minimum
number.

1 Including 2 specimens listed as Podilymbus magnus Shufeldt. However, in his formal list McCoy stated there were 47 specimens of P. podiceps,
not 58 as listed in his table.

2 McCoy referred this specimen to Casmerodius albus.

3 McCoy listed this species in his table, but did not include it in his faunal list.

4 McCoy referred this specimen, a complete humerus, to Branta canadensis.

5 Including two specimens referred to Palaeophoyx Columbiana by McCoy.

6 I have combined here the two subspecies of Branta canadensis listed by McCoy.

7 Including specimens listed as Anas fulvigula by McCoy.

8 McCoy referred this specimen to Teratornis merriami

9 Including specimen assigned to Fulica minor Shufeldt, considered a synonym of F. americana (Olson 1977).

10 Including specimen referred to Palaeophoyx columbiana by McCoy.

Phalacrocorax auritus (Lesson 1831)
Double-crested Cormorant
MATERIAL: Maxilla; mandible; coracoids (11); scapulae
(2); humeri (11); ulnae (22); radius; carpometacarpi (5); femora

(4) ; tibiotarsi (8); tarsometatarsi (8). UF/FSM 15454— 15522,
15646, 20014-20015, 22451.

Anhinga anhinga (Linnaeus 1766)

Anhinga

MATERIAL: Humerus; ulnae (2). UF/FSM 22459-22551.

Ardea herodias Linnaeus 1758
Great Blue Heron

MATERIAL: Coracoids (4); scapulae (2); humeri (3); ulnae

(5) ; radii (2); carpometacarpi (4); femora (2); tibiotarsi (5); tar-
sometatarsi (12). UF/FSM 15535-15571, 20012, 22559.

Casmerodius albus (Linnaeus 1758)

Great Egret

MATERIAL: Ulna. UF/FSM 15572.

Egretta caerulea (Linnaeus 1758)

Little Blue Heron

MATERIAL: Tibiotarsus. UF/FSM 15534.

Nycticorax nycticorax (Linnaeus 1758)
Black-crowned Night-Heron
MATERIAL: Ulna; carpometacarpus; femur; tarsometatar-
sus. UF/FSM 15531-15533, 20013.

Botaurus lentiginosus (Rachett 1813)
American Bittern

MATERIAL: Humerus; tibiotarsus; tarsometatarsus. UF/
FSM 20019, 22536-22537.

REMARKS: McCoy (1963) described a new genus and

species of heron, Paleophoyx columbiana, from the Itchtuck-
nee River on the basis of two coracoids and an ulna. Olson
(1974) published a reappraisal of this heron, stating that he
could find no characters to separate the coracoids from those
of Botaurus lentiginosus, while the ulna represented the Barn
Owl, Tyto alba. He stated that the coracoids fell at or slightly
below (1 mm) the lower size range of Recent B. lentiginosus,
and suggested that#, lentiginosus may have been smaller dur-
ing the Pleistocene. The specimens listed above all fall well
within the size range of Recent B. lentiginosus, indicating that
B. lentiginosus was not significantly smaller during the Pleis-
tocene.

Ciconia maltha L. Miller 1910

MATERIAL: Coracoid; scapula; humeri (2); carpometacar-
pus; tibiotarsi (3); tarsometatarsus. UF/FSM 1963, 1552 5—
15530, 15591, 22555.

Mycteria americana Linnaeus 1758
American Wood-Ibis

MATERIAL: Ulna; tarsometatarsi (2). UF/FSM 15592-
15594.

REMARKS: Olson (pers. comm.) has suggested that be-
cause of their large size these specimens may represent Mycteria
wetmorei Howard.

Olor columbianus (Ord 1815)

Whistling Swan

MATERIAL: Scapula; humeri (2); ulna; femur; tarsometa-
tarsi (4). UF/FSM 15619-15620, 15659-15663, 15665, 15668.

Olor buccinator Richardson 1831
Trumpeter Swan

MATERIAL: Cranium; coracoids (4); scapula; humeri (7);
ulnae (8); radii (6); carpometacarpi (6); pelvis; femora (2); tib-

122

Campbell: Itchtucknee Avifauna

iotarsi (5); tarsometatarsi (5). UF/FSM 13988, 15613-15618,
15621-15658, 15666, 22449.

REMARKS: Although Olor columbianus has been reported
as a rare winter resident in Florida (AOU 1957), 0. buccinator
is known from Florida only on the basis of fossils, and only
from the Itchtucknee River (Wetmore 1931). This may suggest
that the range of 0. buccinator was shifted southward during
glacial periods of the Pleistocene, and retreated northward
during inter- and postglacial periods. It should be noted that
there are six times as many specimens of O. buccinator re-
ported here as there are of O. columbianus .

Branta canadensis Linnaeus 1758
Canada Goose

MATERIAL: Coracoids (16); scapulae (4); humeri (12); ul-
nae (16); radii (9); carpometacarpi (8); sternum; femora (5); tib-
iotarsi (7); tarsometatarsi (5). UF/FSM 14104, 15669-15705,
22492-22535, 22548.

Anas platyrhynclios Linnaeus 1758
Mallard

MATERIAL: Coracoids (26); scapulae (8); humeri (50); ul-
nae (25); radius; carpometacarpi (24); femora (2); tibiotarsi (15);
tarsometatarsi (2). UF/FSM 2393, 2978, 15774-15783, 19477—
19617, 22483.

REMARKS: Undoubtedly included in the material listed
above are specimens of Anas rubripes, a species so similar
osteologicallv to A. platyrhynclios that bones of the two are
extemely difficult, if not impossible, to separate. If any of the
material above is from A. rubripes, I consider specimens UF/
FSM 15774—15783 to be the best possibilities.

Anas acuta Linnaeus 1758
Pintail

MATERIAL: Coracoids (7); humeri (6); ulnae (4); carpo-
metacarpi (7); tibiotarsus. UF/FSM 15784—15808.

Anas discors Linnaeus 1766
Blue-winged Teal

MATERIAL: Coracoids (4); humeri (14); ulnae (5);
carpometacarpi (4); tibiotarsi (2); tarsometatarsus. UF/FSM
15881-15910.

Anas crecca Linnaeus 1758
Green-winged Teal

MATERIAL: Humeri (4). UF/FSM 15911-15914.

Anas strepera Linnaeus 1758
Gad wall

MATERIAL: Coracoids (6); ulnae (2); carpometacarpi (3).
UF/FSM 15724-15734.

Anas americana Gmelin 1789
Baldpate

MATERIAL: Coracoid; carpometacarpi (2). UF/FSM
15720-15721, 16831.

Anas clypeata Linnaeus 1758
Shoveler

MATERIAL: Coracoids (8); humeri (8); ulnae (6); carpo-
metacarpi (10). UF/FSM 15809-15840.

Aix sponsa (Linnaeus 1758)

Wood Duck

MATERIAL: Coracoids (4); scapulae (2); humeri (12); ulnae
(5); carpometacarpi (2); femur; tibiotarsus. UF/FSM 15 747—
15773.

Ay thy a afftnis (Eyton 1838)

Lesser Scaup

MATERIAL: Coracoids (7); humeri (24); ulnae (3); carpo-
metacarpi (5); tarsometatarsus. UF/FSM 15841-15880.

Aythya collaris (Donovan 1809)
Ring-necked Duck

MATERIAL: Coracoids (35); scapulae (5); humeri (52); ul-
nae (8); carpometacarpi (6); femur; tibiotarsi (5); tarsometatarsi
(3). UF/FSM 1749, 14092, 16007-16100, 14458-14476.

Aythya americana (Eyton 1838)

Redhead

MATERIAL: Coracoid; ulna. UF/FSM 15722-15723.

Aythya valisineria (Wilson 1814)
Canvasback

MATERIAL: Crania (3); coracoids (25); scapulae (4); humeri
(34); ulnae (10); carpometacarpi (12); femora (4); tibiotarsi (2);
tarsometatarsus. UF/FSM 2000, 14092, 15915-16006, 22484.

Bucephala clangula (Linnaeus 1758)
Common Goldeneye

MATERIAL: Coracoid; humeri (2). UF/FSM 15716-15718.

Lophodytes cucullatus (Linnaeus 1758)
Hooded Merganser

MATERIAL: Coracoid; humeri (8); ulnae (2); femur. UF/
FSM 15735-15746.

Mergus serrator Linnaeus 1758
Red-breasted Merganser
MATERIAL: Carpometacarpus. UF/FSM 15719.

Oxyura jamaicensis (Gmelin 1789)

Ruddy Duck

MATERIAL: Coracoid; humerus; ulna; tibiotarsus; tarso-
metatarsus. UF/FSM 15711-15715.

Cathartes aura (Linnaeus 1758)

Turkey Vulture

MATERIAL: Humerus; tibiotarsus. UF/FSM 15595, 20016.

Coragyps atratus (Bechstein 1793)

Black Vulture

MATERIAL: Tarsometatarsus. UF/FSM 20018.

Campbell: Itchtucknee Avifauna

123

Accipitridae genus and species
indeterminate

MATERIAL: Proximal end of left tibiotarsus (Collection of
Pierce Brodkorb No. PB 1874); distal end of left tarsometa-
tarsus (UF/FSM 22560).

REMARKS: These specimens are from a large eagle, but
they are too fragmentary to yield any diagnostic characters.
They do not represent a living species, or Amplibuteo wood-
wardi (L. Miller). The tibiotarsus was referred to Teratornis
merriami L. Miller by McCoy (1963). Olson (pers. comm.) has
reported a fragmentary distal end of a tarsometatarsus (USNM
209535) from the late Pleistocene deposits of the Aucilla River,
Florida, that is also from a large eagle, but it is different from
the tarsometatarsus listed above. So there were at least two
large, presently unknown, eagles living in Florida in the late
Pleistocene.

Buteo lineatus (Gmelin 1788)
Red-shouldered Hawk

MATERIAL: Humerus; carpometacarpus; tibiotarsus. UF/
FSM 22453-22454, 22491.

Buteo jamaicensis (Gmelin 1788)
Red-tailed Hawk

MATERIAL: Femur; tibiotarsus; tarsometatarsi (4). UF/
FSM 15608-15612, 22452.

Haliaetus leucocephalus (Linnaeus 1766)

Bald Eagle

MATERIAL: Carpometacarpi (3). UF/FSM 22446-22448.

Genus Milvago Spix 1824
Milvago readei (Brodkorb 1959)

MATERIAL: Distal end and shaft of right tibiotarsus; distal
end of left tarsometatarsus. UF/FSM 22561-22562. Figures
la-c, 2b-d, 3.

REMARKS: Milvago readei was described by Brodkorb
(1959) as a species of Falco on the basis of a fragmentary distal
end of a left tibiotarsus (Fig. lb, d). The type locality is Pit 2,
Arredondo (Arredondo II), Alachua County, Florida. The type
horizon is the Arredondo clay member of the Wicomico For-
mation, which Brodkorb considered to be Illinoian. In a study
of the fossil box turtles (Terrapene) of Florida, Auffenberg
(1958) also considered Arredondo II to be Illinoian, although
he considered the possibility that the nearby Arredondo IA
deposits might be referred to the Sangamon. Martin (1968)
considered the Arredondo IA deposits to be either Illinoian,
Sangamon, or W'isconsinan in age; the thrust of his paper,
however, requires that the deposits be of a glacial age. No
other fossils have been referred to this species since its descrip-
tion.

The holotype of Milvago readei was described as being sim-
ilar to Falco mexicanus, but differing by having “(1) external
ligamental prominence a small knob; (2) no shelf connecting
external ligamental prominence and groove for peroneus pro-
fundus; (3) intercondylar pit elongated in an obliquely trans-
verse direction toward anterior edge of external condyle; (4)
size about half. Distal width, 8.0; depth of external condyle,

6.5; anterior height of external condyle, 4.7; depth of internal
condyle, about 6.5 mm” (Brodkorb 1959:274— 275). I consider
characters (1) and (2) to be valid at the generic level, and that
character (3) may also serve to separate Falco from Milvago.

The tibiotarsus from the Itchtucknee River differs from the
holotype of M. readei in only minor details, such as having a
smaller opening in the intercondylar fossa and a more prom-
inent posteroproximal corner to the external condyle. Both of
these are variable characters. The holotype tibiotarsus is
slightly, but insignificantly, smaller than that from the Itch-
tucknee River. Measurements (in mm) of the latter are: Distal
width 8.8; depth of external condyle 7.2; anterior height of
external condyle 5.2; depth of internal condyle 7.3; least width
of shaft 4.9. (For additional measurements of Milvago spp.,
see Olson 1976a:358; Campbell 1979:96-99.)

The Itchtucknee River tibiotarsus (Fig. la, c) is character-
ized by having the (1) external condyle with very deep con-
cavity on external surface (moderately deep in M. chimango
Vieillot; deep in M. chimachima Vieillot; similar in M. brod-
korbi Campbell 1979); (2) internal condyle with moderate con-
cavity on posterointernal side (shallow in M. chimango ; deep
in M. chimachima ; similar in M. brodkorbi)', (3) anterior in-
tercondylar fossa with very deep pit on external side (shallow
to moderately deep pit in M. chimango and M. chimachima',
deep pit in M. brodkorbi ); (4) intercondylar groove shallow in
anterior view (deeper in M. chimango, M. chimachima, and
M. brodkorbi ); (5) external condyle deeply undercut by very
large tendinal fossa (very restricted fossa in M. chimango and
M. chimachima', similar in M. brodkorbi)', (6) supratendinal
bridge with proximal end positioned near external edge of shaft
(positioned farther from external edge of shaft in M. chimango,
M. chimachima, and M. brodkorbi)', (7) shaft sturdy (similar
in M. chimango and M. brodkorbi', more slender in M. chi-
machima). No tibiotarsus is known for M. alexandri Olson
1976.

The tarsometatarsus (Figs. 2b-d, 3) from the Itchtucknee
River assigned to M. readei is referable to the subfamily Po-
lyborinae on the basis of having the middle trochlea project
distad much more than the inner trochlea (in the subfamily
Falconinae the inner trochlea projects distad as much as or
more than the middle trochlea). With the exception of Spi-
ziapteryx (Olson 1976a), the species of the subfamilies Her-
petotherinae (Forest Falcons) and Polihieracinae (Pigmy Fal-
cons) have very distinctive tarsometatarsi that do not closely
resemble those of the caracaras or falcons.

The tarsometatarsus agrees with that of Milvago and differs
from those of all other genera of the subfamily Polyborinae
( Daptrius , Phalcoboenus, and Polyborus) by having (1) troch-
leae laterally compressed; (2) middle trochlea widening mark-
edly distad in both anterior and posterior view, and with ex-
ternal edge projecting more posteriad than internal edge; (3)
inner trochlea not rotated or rotated slightly to moderately
posteriad, with anterior surface flush with that of shaft; i.e.,
no point of trochlear surface extending anterior to shaft sur-
face; (4) inner trochlea with wing (posterior projection) roughly
perpendicular to long axis of shaft without turning distad at
tip; (5) inner trochlea with broad flattened area on inner half
of posterior side; delineated by distinct sharp edge leading from
anteroproximal edge of base to tip of wing and a ridge leading
from end of metatarsal facet to tip of wing.

124

Campbell: Itchtucknee Avifauna

Campbell: Itchtucknee Avifauna

125

The tarsometatarsus is characterized by having (1) inner
trochlea with wing moderately large (much smaller and pro-
jecting posteriad less inM. chimango andM. alexandri ; broad-
er, but projecting posteriad less in M. chimachima', similar,
but more massive and with only slight convexity on distal, or
internal, surface in M. brodkorbi) (Fig. 2a, e, f); (2) inner troch-
lea rotated moderately posteriad (not rotated in M. chimango,
M. chimachima, and M. alexandri ; rotated slightly in M. brod-
korbi); (3) internal intertrochlear notch narrow and deep (nar-
row and shallow in M. chimango', narrow and moderately deep
in M. chimachima', moderately wide and deep inM. alexandri',
wide and deep in M. brodkorbi)', (4) middle trochlea with
trochlear groove broad and shallow (similar, but not as broad
inM. chimango', narrow and deeper inM. chimachima', similar
in M. alexandri', similar, but slightly deeper in M. brodkorbi)',

(5) groove leading proximad from internal intertrochlear notch
and skirting medial edge of metatarsal facet quite well marked;
i.e., distinctly visible (slightly marked in M. chimango, M.
alexandri, and M. brodkorbi', not visible in M. chimachima)',

(6) shaft broad with very pronounced ridge leading proximad
from metatarsal facet (highly variable width, but only slight
ridge in M. chimango and M. chimachima', slightly narrower
with moderately pronounced ridge in M. alexandri', broader,
with similar ridge in M. brodkorbi)', (7) metatarsal facet large
(smaller in M. chimango, M . chimachima, and M. alexandri,
similar in M. brodkorbi). The tarsometatarsus is most similar
in size to those of M. chimachima, but as the external trochlea
is broken the only measurements that can be taken are as
follows (in mm): Distal width (est.) 8.8; width of middle troch-
lea 3.4; depth of middle trochlea 4.3.

No tarsometatarsus is known for M. readei, but I doubt
that there was more than one species in Florida at any given
time during the Pleistocene. I consider referral of the tarso-
metatarsus to M. readei appropriate.

DISCUSSION: Prior to 1976 there were only two known
species of Milvago, M. chimango and M. chimachima, both
living in and restricted to South America, Panama, and Costa
Rica. Olson (1976) described the first recognized paleospecies
of Milvago, M. alexandri, on the basis of a complete tarso-
metatarsus from late Pleistocene cave deposits in Haiti, His-
paniola. Recently, I described the second known paleospecies,
M. brodkorbi, on the basis of 160 specimens from the late
Pleistocene Talara Tar Seeps of northwestern Peru (Campbell
1979). To these four species we can now add another, M.
readei, from the late Pleistocene of Florida. This is the first
record of the genus from North America. Not only is the rapid
proliferation of fossil species in such a short time for a genus
that previously had no fossil record remarkable, it is even more
so considering that all three paleospecies have been found out-
side the present ranges of the living species. Also, the fact that
the paleospecies are all known from late Pleistocene deposits
is most interesting. Why should a genus, which appears to
have been so successful, suddenly, relatively speaking, have
three species in widely separated areas become extinct? Olson

(1976:359) also puzzled over the lack of a clear reason for the
extinction of M. alexandri on Hispaniola.

The proximity of M. readei and M. alexandri in time and
space poses the question of whether they are both valid species,
a question that becomes even stronger considering how few
specimens there are for each. Based upon an examination of
tibiotarsi (25) and tarsometatarsi (19) of M. brodkorbi, I con-
sider intraspecific variation in Milvago to be relatively small.
There are variable characters, but the species of Milvago do
not appear to be nearly as variable osteologically as species of
the related genus Poly bonis. Of the seven characters listed for
the tarsometatarsus, I consider the first two to be the most
important. They reflect rather significant structural differ-
ences, and they varied only minimally in the sample of tar-
sometatarsi of M. brodkorbi. Although there is the possibility
that larger samples would demonstrate greater variability, I
consider M. readei and M. alexandri to be valid paleospecies
on the basis of the currently available material.

Any speculation as to the evolution and development within
the genus Milvago is very tenuous, based as it would be on
such widely scattered data points; but a few comments can be
made. Olson (1976) considered M. alexandri from Hispaniola
more similar to M. chimachima than to M. chimango, while
I considered M. brodkorbi from Peru to be more similar to M.
chimango (Campbell 1979). I believe M. readei is more similar
to M. brodkorbi than to either of the living species, but more
similar to M. chimachima than M. chimango. There are so
few comparable characters between M. alexandri and M.
readei that it is impossible to suggest a similarity, or lack of
same.

The living species of Milvago are open country, savanna,
scrub forest, or forest edge inhabitants (Brown and Amadon
1968; Blake 1977). M. chimachima ranges from Panama and
Costa Rica southward east of the Andes to northern Argentina,
while M. chimango ranges southward from eastern Bolivia on
both sides of the Andes. The two species are sympatric over
a wide area of their ranges, but show no tendency to hybridize
(Vuilleumier 1970). While the two species are clearly related,
and some would even place them as members of a superspecies
(Brown and Amadon 1968), they are quite distinct morpho-
logically (Olson 1976).

For now it appears best to assume a South American origin
for the genus, and also, to assume that habitat changes re-
sulting from Pleistocene climatic events (Haffer 1974, 1978;
Simpson and Haffer 1978; Muller 1973; Vanzolini and Wil-
liams 1970; Campbell in press) were instrumental in the spe-
ciation within this genus. M. readei probably extended its
range into Florida via the Gulf Coast Savanna Corridor (Webb
1974) during one of the glacial periods of the Pleistocene. Flor-
ida was covered by much open savanna during dry phases of
the Pleistocene when the corridor was open (Webb 1974; Watts
1971), and would have been ideal habitat for a species ol Mil-
vago. At the end of the Pleistocene, Florida lost most of its
open country, retaining only the grasslands of the southern-

Figure 1. Stereophotographs of tibiotarsi of Milvago'. Milvago readei, referred right tibiotarsus, UF/FSM 22562, in anterior (a) and external (c)
view; Milvago readei, holotype left tibiotarsus, PB 1632, in anterior (b) and external (d) view; Milvago brodkorbi, right tibiotarsus, Royal Ontario
Museum ROM 17580, in anterior (e) and external (f) view. All X1.5.

126

Campbell: Itchtucknee Avifauna

Figure 2. Stereophotographs of left tarsometatarsi of Milvago-. Milvago brodkorbi, holotype, ROM 17447, in anterior (a), posterior (e), and internal
(f) view; Milvago readei, referred, UF/FSM 22S61, in anterior (b), posterior (c), and internal (d) view. All xi.S.

Campbell: Itchtucknee Avifauna

most part of the state. Many other factors may have been
responsible for the extinction of M. reader, but the climatic
and vegetational changes at the end of the Pleistocene that
resulted in the loss of the dry savanna habitat in Florida were
probably the most important. Whether there were similar hab-
itat changes on Hispaniola is not known at the present time.
There are wide areas of xeric vegetation present on Hispaniola
today (Wetmore and Swales 1931), although some of this is a
result of human habitat alteration. But factors other than cli-
mate may also have played a role in the extinction of M. al-
exandri.

Pandion haliaetus (Linnaeus 1758)

Osprey

MATERIAL: Coracoids (3); ulnae (2); tibiotarsi (4); tarso-
metatarsi (6). UF/FSM 2725, 15596-15607, 22547, 22548.

Colinus virginianus (Linnaeus 1758)
Bobwhite

MATERIAL: Humeri (2); tibiotarsi (2). UF/FSM 15523-
15524, 22455-22456.

Meleagris gallopavo Linnaeus 1758
Wild Turkey

MATERIAL: Coracoids (9); scapula; humeri (12); ulnae (11);
carpometacarpi (14); sternum; pelves (2); femora (10); tibiotarsi
(34); tarsometatarsi (47). UF/FSM 2069, 2332, 13988, 14104,
14126, 15328-15453, 15664, 15667, 16832, 22298-22304.

REMARKS: This series of fossils is discussed by Steadman
(this vol.)

Grus americana (Linnaeus 1758)
Whooping Crane

MATERIAL: Coracoids (2); scapulae (2); ulna; radii (2); car-
pometacarpus; tibiotarsi (4); tarsometatarsi (4). UF/FSM
15575-15589, 22552.

Grus canadensis (Linnaeus 1758)

Sandhill Crane

MATERIAL: Ulna; radius; tarsometatarsus. UF/FSM
15590-15591, 22554.

Aramus guarauna (Linnaeus 1766)
Limpkin

MATERIAL: Humerus; tibiotarsi (3). LTF/FSM 15325-
15327, 20017.

Rallus elegans Audubon 1834
King Rail

MATERIAL: Coracoid; humerus; tibiotarsus; tarsometatar-
sus. UF/FSM 22485-22488.

Porzana Carolina (Linnaeus 1758)

Sora

MATERIAL: Tibiotarsus. LTF/FSM 22558.

127

Figure 3. Referred left tarsometatarsus of Milvago readei, UF/FSM
22561, in anterior, internal, and posterior view. X2.

Gallinula chloropus (Linnaeus 1758)
Common Gallinule

MATERIAL: Coracoids (4); humeri (18); ulnae (6); carpo-
metacarpi (5); femora (9); tibiotarsi (8); tarsometatarsi (13). UF/
FSM 22371-22433.

Gallinula brodkorbi McCoy 1963

MATERIAL: Coracoids (2); humeri (3); ulnae (4); carpo-
metacarpus; femora (7); tibiotarsi (5); tarsometatarsi (7). UF/
FSM 22342-22370.

REMARKS: In a review of the Pleistocene rails of North
America, Olson (1974) considered G. brodkorbi a valid species
that was a large temporal representative of G . chloropus. Lat-
er, he synonvmized G. brodkorbi with G. chloropus (Olson
1977), considering the former a subspecies of the latter. Both
of us have examined the material from the Itchtucknee River
and we agree that two forms are present, and thatG. brodkorbi
is distinguished from G. chloropus primarily on the basis of
the larger size and greater heaviness, or stoutness, of its bones.
But as Olson (1977) pointed out, the specimens McCoy (1963)
referred to G. brodkorbi fall within the size range of Recent
G. chloropus, possibly requiring future nomenclatural changes
if no diagnostic characters other than size can be determined
for the two forms when a larger series is available. Because
of individual variation within gallinules I do not believe we
have a sufficient number of specimens to detail specific osteo-
logical differences between the two forms.

A problem with considering G. brodkorbi a temporal sub-
species of G. chloropus is thatG. chloropus is also recognizable
in the fauna and is represented by approximately twice as
many specimens as G. brodkorbi. If one accepts the premise
of Mayr (1963) that subspecies may form only if two popula-
tions of a species are isolated in space, than one must also
accept the restriction that temporal subspecies must be isolated
in time. Unfortunately, the specimens from the Itchtucknee
River were recovered from the bottom of the river over its
entire length, not in situ, so we do not know if the two forms
were contemporaneous. Until further work on delineating the
relationship between the two forms is possible, I prefer to
maintain G. brodkorbi at the specific rank.

128

Campbell: Itchtucknee Avifauna

Porphyrula martinica (Linnaeus 1766)
Purple Gallinule

MATERIAL: Humeri (4); femur; tibiotarsus; tarsometatar-
sus. UF/FSM 20040-20043, 22434-22436.

Fulica americana Gmelin 1789
American Coot

MATERIAL: Coracoids (3); humeri (12); ulna; carpometa-
carpus; femora (4); tibiotarsi (28); tarsometatarsi (6). UF/FSM
20020-20039, 22437-22445, 22457-22482.

Ectopistes migratorius (Linnaeus 1766)
Passenger Pigeon

MATERIAL: Tarsometatarsus. UF/FSM 22556.

Zenaidura macroura (Linnaeus 1758)
Mourning Dove

MATERIAL: Humerus. UF/FSM 22557.

Tyto alba (Scopoli 1769)

Barn Owl

MATERIAL: Carpometacarpus. UF/FSM 15573.

Otus asio (Linnaeus 1758)

Screech Owl

MATERIAL: Humeri (2). UF/FSM 22489-22490.

Strix varia Barton 1799
Barred Owl

MATERIAL: Femur. UF/FSM 15574.

Corvus brachyrhynchos Brehn 1822
Common Crow

MATERIAL: Coracoid; humeri (2); ulnae (2); tibiotarsus;
tarsometatarsi (3). UF/FSM 22538-22546.

DISCUSSION

McCoy (1963) identified 392 avian fossils representing 52
species (with corrections to his identifications, 47) from the
Itchtucknee River deposits. The 1362 specimens reported here
represent 56 species. The total number of species (see Table
1) reported from the Itchtucknee River now stands at 67.

In his paper, McCoy (1963) described three new species:
Palaeophoyx Columbiana, Anas itchtucknee, and Gallinula
brodkorbi. As mentioned earlier, Palaeophoyx Columbiana is
a synonym of Botaurus lentiginosus. Gallinula brodkorbi, al-
though considered a temporal subspecies of G. chloropus by
some (Olson 1977), is here considered a valid species. I have
examined the type of Anas itchtucknee, but on the basis of the
available comparative material, both fossil and Recent, I was
not able to prove either that it was or was not a valid paleo-
species. I did not find any additional specimens of anatids that
could be referred to the species, and am inclined to doubt its
validity. Late Pleistocene anatids are rather difficult to identify
because of the close relationship of the living species to which
we try to refer the fossils, and because of the large intraspecific
size and osteological variation that occurs in that group. Large

series of comparative material are required to screen out this
variation, and such series are not always available.

There are only seven (excluding the unknown accipitrid)
extinct species, or 10 percent of the total number, present in
the avifauna, and they represent only 3 percent of the speci-
mens and 4 percent of the individuals recorded. If Gallinula
brodkorbi is excluded, the six remaining extinct species rep-
resent only one percent of the specimens and three percent of
the individuals. Thus, the portion of the avifauna consisting
of extinct species is very small. When one considers that one
species, Ectopistes migratorius , was recently extirpated by
man, that two may be direct ancestors of living forms ( Gym-
nogyps amplus — Gymnogyps calif ornianus ; Gallinula brod-
korbi— Gallinula chloropus ), and that only one of the others (Ci-
conia maltha ) is reasonably well known, it becomes apparent
that we are dealing with essentially a modern avifauna. Three
of the seven extinct species (Ciconia maltha, Gymnogyps am-
plus, and Ectopistes migratorius) are also known from the de-
posits at Rancho La Brea, California, for which the oldest
available date is approximately 40,000 B.P. In view of all of
the above, McCoy’s (1963) suggestion of a mid-Wisconsinan
age for the Itchtucknee River deposits seems quite reasonable,
and if this date is in error I would suggest that it is too old.

From an analysis of the fossil material he studied, McCoy
(1963) concluded that extensive freshwater pond and marshy
conditions prevailed at the time of deposition of the fossils.
The additional material reported here completely supports his
conclusion. Of the 20 most abundant species, only two, Mele-
agris gallopavo and Pandion haliaetus, are not strictly pond
or marsh dwellers, although even the latter feeds primarily in
the water. Only 17 of the remaining 47 species known from
the avifauna can be considered non-aquatic, and they, plus
M. gallopavo, make up only 12 percent of the total number of
specimens, or 13 percent of the minimum number of individ-
uals. The 26 species, 39 percent of the total, of grebes and
anatids alone account for 62 percent of the specimens, or 58
percent of the minimum number of individuals. The predom-
inance of aquatic birds and the lack of any species in the
avifauna that could be considered marine strongly indicates
extensive freshwater pond and marsh conditions.

The significance of Milvago readei in the late Pleistocene of
Florida, in addition to what it suggests about evolution and
dispersal patterns within the genus, lies in the implication that
many avian species of South or Central American origin ac-
companied the large mammals of similar origin into Florida
via the Gulf Coast Savanna Corridor during glacial periods.
It would appear more likely that suites of savanna-adapted
birds rather than just one or two species moved into Florida
during the periods of lesser rainfall and more open country
than now, only to retreat or become extinct when the climate
returned to more humid conditions and the open country dis-
appeared during interglacial periods. We should, therefore,
expect to find additional South American forms in glacial-age
deposits of Florida. We must also bear in mind that Florida
could have acted as a major dispersal route for avian species
that populated the Antilles.

ACKNOWLEDGMENTS

I thank Pierce Brodkorb for the use of his comparative col-
lection of osteological specimens, without which this study

Campbell: Itchtucknee Avifauna

129

would not have been possible, and his advice and encourage-
ment thoughout the study. S. David Webb kindly made the
fossil collections of the Florida State Museum available for
study. Special thanks are extended to Jill Littlewood for her
excellent drawings of the tarsometatarsus.

LITERATURE CITED

American Ornithologists’ Union. 1957. Check-list of
North American Birds. Fifth edition. Lord Baltimore
Press, Inc., Baltimore. 691 pp.

Auffenberg, W. 1958. Fossil turtles of the genus Terrapene
in Florida. Bull. Florida State Mus., Biol. Sci. 3(2):53—
92.

. 1963. The fossil snakes of Florida. Tulane Stud.

Zool. 10:131-216.

Blake, E.R. 1977. Manual of Neotropical Birds. Vol. 1.

Univ. of Chicago Press, Chicago. 674 pp.

Brodkorb, P. 1959. The Pleistocene avifauna of Arredondo,
Florida. Bull. Florida State Mus., Biol. Sci. 4(9): 2 69— 2 9 1 .
Brown, L., and D. Amadon. 1968. Eagles, hawks and fal-
cons of the world. McGraw-Hill, New York. 2 vols. 945
PP

Campbell, K.E., Jr. 1979. The Non-passerine Pleistocene
Avifauna of the Talara Tar Seeps, Northwestern Peru.
Royal Ont. Mus., Life Sci. Contrib. 118:1-203.

. 1980. Late Pleistocene Events Along the Coastal

Plain of Northwestern South America, (in press).
Haffer, J. 1974. Avian speciation in tropical South Amer-
ica. Publ. Nuttall Ornith. Club 14:1-390.

. 1978. Distribution of Amazon Forest Birds. Bonner

Zool. Beitr. 29:38-78.

Kurten, B. 1965. The Pleistocene Felidae of Florida. Bull.

Florida State Mus., Biol. Sci. 9:215-273.

Martin, R.A. 1968. Late Pleistocene Distribution of Micro-
tus pennsylvanicus. J. Mamm. 49(2):265-271.

. 1969. Taxonomy of the giant Pleistocene beaver Cas-

toroides from Florida. J. Paleon. 43:1033-1041.

Mayr, E. 1963. Animal Species and Evolution. Harvard
Univ. Press, Cambridge. 797 pp.

McCoy, J.J. 1963. The fossil avifauna of Itchtucknee River,
Florida. Auk 80(3):335-351.

Muller, P. 1973. The dispersal centres of terrestrial verte-

brates in the neotropical realm. W. Junk, The Hague.
244 pp.

Olson, S.L. 1974. A reappraisal of the fossil heron Palaeo-
phoyx columbiana McCoy. Auk 91(1): 17 9 — 180.

. 1976. A new species of Milvago from Hispaniola,

with notes on other fossil caracaras from the West Indies
(Aves: Falconidae). Proc. Biol. Soc. Wash. 88(33):355-
366.

. 1976a. The affinities of the falconid genus Spiziap-

teryx. Auk 93(3):633-636.

. 1977. A synopsis of the fossil Rallidae. Pp. 339-373

in Rails of the World by S. Dillon Ripley. D.R. Godine,
Publ., Boston.

Simpson, B., and J. Haffer. 1978. Speciation Patterns in
the Amazonian Forest Biota. Ann. Rev. Ecol. Sys. 9:497-
518.

Simpson, G.G. 1929. The extinct land mammals of Florida.
Fla. Geol. Surv., 20th. Ann. Rep.:229-279.

. 1930. Additions to the Pleistocene of Florida. Amer.

Mus. Nov. 406:1-14.

. 1932. Miocene land mammals from Florida. Fla.

Geol. Surv. Bull. 10:11-41.

Steadman, D. 1980. A review of the Osteology and Paleon-
tology of Turkeys (Aves: Meleagridinae). (this vol.)

Storer, R.W. 1976. The Pleistocene Pied-billed Grebes
(Aves: Podicipedidae). Smiths. Contrib. Paleobiol. 27:147-
153.

Vanzolini, P.E., and E.E. Williams. 1970. South Amer-
ican anoles: The geographic differentiation and evolution
of the Anolis chrysolepis species group (Sauria, Iguani-
dae). Arq. Zool., Sao Paulo 19:1-298.

Vuilleumier, F. 1970. Generic relations and speciation pat-
terns in the caracaras (Aves: Falconidae). Breviora 355: 1-
29.

Watts, W.A. 1971. Postglacial and interglacial vegetation
history of southern Georgia and central Florida. Ecol
52:676-689.

Webb, S.D. 1974. Pleistocene mammals of Florida. Univer-
sity of Florida Press, Gainesville. 270 pp.

Wetmore, A. 1931. The avifauna of the Pleistocene in Flor-
ida. Smiths. Misc. Coll. 85:1-41.

Wetmore, A., and B.H. Swales. 1931. The Birds of Haiti
and the Dominican Republic. U.S. Nat. Mus., Bull.
155:1-483.

A REVIEW OF THE OSTEOLOGY AND PALEONTOLOGY OF
TURKEYS (AVES: MELEAGRIDINAE)

By David VV. Steadman1

ABSTRACT: A study of the comparative osteology of specimens of all known species of living and

extinct turkeys results in the recognition within the Meleagridinae of three genera: Rhegminornis Wetmore,
Proagriocharis Martin and Tate, and Meleagris Linnaeus. Meleagris contains all diagnostic specimens of
Blancan and younger ages, including both living species. The genera Agriocharis Chapman and Parapavo
L. Miller are synonymized with Meleagris, and the species Meleagris alta Marsh and Meleagris tridens
Wetmore are synonymized with Meleagris gallopavo. Evidence suggests that since at least Blancan times
species in the lineage leading to Meleagris gallopavo have continuously occupied the southern United
States. Meleagris gallopavo is absent from pre-RanchoIabrean deposits.

Turkeys are large gallinaceous birds comprising the subfam-
ily Meleagridinae of the family Phasianidae. All living and
extinct species of this subfamily are known only from the New
World. The fossil record of turkeys begins with Rhegminornis
calobates Wetmore (1943) from the Lower Miocene (Heming-
fordian) of Florida. This poorly known form has characters of
both the Phasianinae and the Meleagridinae. A much larger
unnamed species is represented by a single, rather undiagnos-
tic, element from the Upper Miocene deposits in Virginia. The
earliest certain turkey is Proagriocharis kimballensis Martin
and Tate (1970) from the Upper Pliocene (Hemphillian) of
Nebraska. Fossils of turkeys are relatively common in the
Pleistocene, and several distinct species are recognized. Table
1 lists the fossil localities from which the specimens came that
I examined in this study, while Figure 1 shows the geograph-
ical distributions of all species of extinct and living turkeys.
The nomenclature used in Table 1, although based on evidence
presented later in this paper, is given now to maintain consis-
tency throughout the paper.

The stimulus of this review of the osteology and paleontol-
ogy of turkeys was the discovery of large numbers of fossils
from two Pleistocene sites in Florida. Inglis IA, of earliest
Irvingtonian age; and Coleman IIA, late Irvingtonian in age.
These specimens differed significantly from the living Mele-
agris gallopavo Linnaeus, and appeared to resemble several
poorly known extinct forms previously unknown in eastern
North America. The Inglis IA site, with approximately 1240
specimens representing a minimum of 44 individuals, and the
Coleman IIA site, with approximately 320 elements repre-
senting a minimum of 17 individuals, provide the earliest ad-
equate samples for an assessment of individual variation in
fossil turkeys. These samples thus provide excellent standards
for comparison with fossil turkeys from other localities that
are typically more fragmentary and less numerous.

1 Paleoenvironmental Laboratory, Department of Geosciences, Uni-
versity of Arizona, Tucson, Arizona 85721.

Prior to the Inglis and Coleman finds, large samples of fossil
turkeys were known from only three species, all of Ranchola-
brean age. These species ar e M. crassipes L. Miller (1940), M.
californica (L. Miller 1909), and M. gallopavo. The relation-
ships between the later forms and the rarer, earlier species
were poorly known. This paper compares all extinct and living
species of turkeys, incorporating the large Irvingtonian sam-
ples in an effort to determine the relationships among the var-
ious meleagridines. A revised classification and possible phy-
logeny are proposed, based on these comparisons.

COMPARATIVE OSTEOLOGY

The study of fossil turkeys is complicated by their great
sexual dimorphism, males being much larger than females. In
the absence of qualitative osteological characters, a male of a
small species might easily be confused with a female of a larger
species. This problem is especially acute when a fossil site
yields too few specimens to assess intraspecific size variation.
This problem is largely alleviated, however, when enough fos-
sils are recovered and two distinct size classes can be distin-
guished. This is the case at Inglis IA, Coleman IIA, Ichetuck-
nee River, Rancho La Brea, and a few other sites. When
specimens from a particular site are not qualitatively separable
from two or more known species, the size of the fossils will
determine the identification. This identification must be con-
sidered tentative because of the great intraspecific size varia-
tion. The large size differences found in turkey bones often
make direct comparisons difficult, hence the extensive use of
the term “relatively” in the following comparisons.

In undertaking the studies of comparative osteology, my
original approach was to compare Meleagris gallopavo with
M. ( =Agriocharis ) ocellata and M. ( =Parapavo ) californica
and, by searching for generic characters, determine to which
genus the Inglis IA and Coleman IIA specimens belonged. It
soon became apparent, however, that diagnostic characters
are few and quite subtle. A broader approach was then taken

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 330:131-207.

132

Steadman: Turkey Osteology and Paleontology

in which each form was closely examined without reference to
its supposed generic status.

With the notable exception of the work of Howard (1927),
most characters of the various reputed genera and species of
fossil turkeys have been based on only one or two individual
specimens and do not hold when larger series of fossils or
comparative material are examined. My findings often contra-
dict previously published descriptions, the discrepancies being
due largely, I believe, to the use of inadequate series of modern
skeletons by earlier workers. The comparisons below are based
on (1) the complete skeletons of 16 individuals of Meleagris
gallopavo (8d, 8 9) and 7 of M. ocellata (3d, 4 9); (2) two
partial tarsometatarsi of Rhegminornis calobates\ (3) the ho-
lotype coracoid and three partial tarsometatarsi of Proagri-
ocharis kimballensis ; (4) all of the material of M. progenes
reported by Brodkorb (1964b); (5) the paratype tarsometatar-
sus of M. leopoldi', (6) a cast of the holotype humerus of M.
anza\ (7) all of the Inglis IA and Coleman IIA specimens; (8)
one to nine specimens of each element of M. californica from
Rancho La Brea; and (9) one to eight specimens of each known
element of M. crassipes from San Josecito Cave. These state-
ments in turn form the basis for comparing specimens from all
other sites.

The comparative osteological data presented below are
roughly quantified in Tables 2 and 3. These tables will be
discussed more fully in the Systematics section. Figures 2
through 7, based on a specimen of M. gallopavo osceola, ex-
plain the various measurements of Tables 4 through 22. In
Tables 4 through 22, an asterisk (*) follows a measurement
taken from a slightly damaged specimen. The actual value of
the measurements would range from 1 to 5 percent greater
than the reported value if the specimens were not damaged.
Measurements followed by two asterisks (**) were taken from
more severly damaged specimens in which the actual value
would range from 5 to 10 percent greater than the reported
value. Means that include one or more of these measurements
are also followed by an asterisk. Measurements followed by
“A” are careful approximations of the actual value.

Cranium

Howard 1927: Pis. 1, 2.

1. The interfrontal depression (Howard 1927: PI. 1, Fig. 4e)
is usually shallower in M. ocellata than in Inglis and Coleman
specimens, M. californica, or M. gallopavo. This contradicts
Howard (1927), who reported a greater depression in M. cal-
ifornica than in M. gallopavo, and a variable depth in M.
ocellata.

2. The supraorbital processes of the frontals in ventral view
are even with or lateral to the postorbital process in M. ocel-
lata', in other forms they do not extend to the postorbital pro-
cess. Inglis specimens are damaged in this area. This contra-
dicts Howard (1927), who reported a similar amount of flaring
of the supraorbital process in M. californica as in M. ocellata,
with a lesser amount in M. gallopavo.

3. The amount of tapering as the frontals approach parietals
in dorsal view (Howard 1927: PI. 1, Figs. 1-4) is usually great-
er in M. ocellata than in Coleman fossils or M. gallopavo', in
Inglis fossils and M. californica it may resemble any of the
above forms. Howard (1927) also noted more tapering in M.
ocellata than in M. gallopavo, with M. californica being in-
termediate.

4. The frontals in lateral view are usually more protrudent
above the parietals (Howard 1927: PI. 2, Fig. Id), resulting in
a larger depression along the fronto-parietal suture, in M. cal-
ifornica and M. gallopavo than in Coleman specimens or M.
ocellata. Inglis specimens may resemble any of the above. This
contradicts Howard (1927), who noted the lack of a fronto-
parietal depression in M. gallopavo and its presence in M.
californica and M. ocellata, while reporting a higher relative
parietal height (i.e., a lower relative frontal height) in M. gal-
lopavo than in M. ocellata, both of which were said to be
higher than in M. californica.

5. The foramen magnum is circular to oblong to triangular
in M. ocellata, circular to oblong in M. gallopavo, and circular
in Coleman and Inglis fossils and M. californica.

Howard (1927) found the cranium of M. californica to be
more similar to that of M. ocellata than to that of M . gallopavo,
while I find thatM. californica, M. gallopavo, andM. ocellata
are not consistently separable from each other by any cranial
characters. Inglis and Coleman specimens closely resemble
these three species, which are certainly not generically distinct
on the basis of their crania.

Premaxilla

Howard 1927: PI. 1.

1. The portion of the premaxilla anterior to the nares is
usually narrower relative to its length in M. gallopavo than in
M. ocellata or M. californica. M. progenes and Inglis speci-
mens are damaged in this area. Howard (1928) and Brodkorb
(1964b) found no individuals in which this character does not
hold.

2. The premaxilla is relatively deeper in M. gallopavo than
in other forms, as noted by Howard (1928) and Brodkorb
(1964b).

3. Two ventral foramina immediately anterior to the choa-
nae are present in M. progenes but are absent in other turkeys
(Brodkorb 1964b). These are not to be confused with the pair
(rarely more) of foramina located nearer to the anterior end
than to the choanae, which may occur in any of the forms.

M. progenes is distinct from all other forms in character 3.
Inglis fossils, M. ocellata and M. californica, are more similar
to each other than to M. gallopavo in other characters of the
premaxilla. Howard (1928) noted the greater resemblance of
M. californica to M. ocellata than to M. gallopavo. However,
none of the differences seen in the premaxilla of turkeys appear
to be of generic value, particularly in view of the apparent
evolutionary plasticity of bill morphology in certain birds.

Mandible

1. The postarticular process is more expanded dorsoven-
trally, usually longer, and usually more laterally compressed
in M. ocellata than in M. gallopavo. All fossil specimens ex-
amined are damaged in this area.

2. The rami are usually relatively larger in M. ocellata than
in M. gallopavo. All fossil specimens examined are damaged
in this area.

3. The mandibular foramen is usually relatively larger in
M. ocellata than in M. gallopavo. All fossil specimens exam-
ined are damaged in this area.

4. The dentary at the mandibular symphysis is slightly nar-
rower and usually slightly longer in M. gallopavo than in M.
ocellata. The other forms are similar to M. ocellata in width,

Steadman: Turkey Osteology and Paleontology

133

but are shorter at the symphysis than either M. ocellata or M.
gallopavo. Brodkorb (1964b) reported a narrower and longer
dentary in M. gallopavo than in M. progenes or M. ocellata,
but noted no variation in this character.

The mandible of M. gallopavo is usually distinct from that
of M. ocellata. Fossil forms resemble M. ocellata more than
M. gallopavo, but this is based only on character 4.

Sternum

Howard 1927: Pis. 3, 4.

1. The manubrium is directed more dorsad in M. gallopavo
than in other forms.

2. The manubrium in dorsal view is roughly triangular in
shape in all turkeys, the sides of the triangle being more in-
terrupted by a projection of the dorsal lip of the coracoidal
sulcus in some Coleman specimens and in all M. ocellata than
in other forms. This finding is consistent with that of Howard
(1927, 1963), who found the shape of the manubrium in M.
californica to be closer to that in M. gallopavo than in M.
ocellata. However, I see no consistent difference in the width
of any portion of the triangular area, which was reported by
Howard (1963) to be wider in M. gallopavo and M. ocellata
than in the Vallecito Creek specimen or M. californica.

3. The dorsal manubrial spine in lateral view is deeper in
some M. californica and in all M. gallopavo than in Inglis or
Coleman specimens orM. ocellata, in agreement with Howard
(1963), who also found this projection to be shallower in the
Vallecito Creek specimen than in M. gallopavo.

4. The ventral manubrial spine in lateral view is more
rounded in M. ocellata than in M. gallopavo, in which it is
pointed or squared. All fossils examined are damaged in this
area.

5. The coracoidal sulcus is relatively shallower in M. ocel-
lata than in M. californica or M. gallopavo, largely because of
the lesser development of the ventral lip. Inglis specimens are
usually similar to M. ocellata, while Coleman specimens are
usually similar to M. californica or M. gallopavo.

6. The dorsal surface is usually more pneumatic in Inglis
and Coleman specimens than in M. californica, M. gallopavo,
or M. ocellata.

The sternum in M. californica resembles that of M. gallo-
pavo more than that of M. ocellata, while Inglis and Coleman
specimens resemble that of M. ocellata or M. californica more
than that of M. gallopavo.

Furcula

Fig. 8; Howard 1927: PI. 5.

1. The rami in dorsal view are usually straighter in M.
gallopavo than in M. ocellata. All fossils examined are dam-
aged in this area. Howard (1927:5) found the rami to be
“straight-sided” in M. californica and M. gallopavo and
“curved outward” in M. ocellata, but she noted no variation
in this character.

2. The dorsal surfaces of the rami in lateral view are slightly
concave in M. ocellata ', in M. gallopavo they may be straight,
slightly concave, slightly convex, or sigmoid (slightly convex
on the sternal end and slightly concave on the scapular end).
All fossils examined are damaged in this area.

3. The rami are stoutest in Inglis fossils and narrowest in

M. gallopavo, with M. californica and M. ocellata being in-
termediate.

4. The rami near the symphysis are much more strongly
flattened dorsoventrally, especially on the internal side, in M.
gallopavo than in Inglis specimens or M. californica, with M .
ocellata being intermediate.

5. The intermuscular lines convergent on the dorsal surface
of the symphysis are much more distinct in Inglis specimens
than in the other forms.

6. The intermuscular lines convergent on the ventral surface
of the symphysis are usually more distinct in Inglis specimens
and M. californica than in M. gallopavo or M. ocellata. This
agrees with the findings of Howard (1927).

7. The hvpocleidium is usually relatively shorter in M. ocel-
lata than in M . californica or M . gallopavo. Inglis specimens
are damaged in area.

8. The hvpocleidium is much more expanded dorsoven-
trally in Inglis specimens than in M. gallopavo or M. ocellata,
with M. californica being intermediate.

Howard (1927) regarded M. californica as generically dis-
tinct from M. gallopavo and M. ocellata on the basis of furcular
characters, most of which are now seen to be only average
differences. In characters 3, 5, and 8, the Inglis specimens
differ from turkeys and resemble all species of Phasianinae
examined.

Coracoid

Howard 1927: Pis. 6, 7.

1. The sternal facet is deeper internally and more tapered
externally in most Coleman specimens, most M. californica,
and in M. ocellata than in Proagriocharis kimballensis , Inglis
specimens, or M. gallopavo. It is deep internally, but only
slightly tapered externally, in M. crassipes. The specimen of
M. progenes is damaged in this area.

2. The pneumatic foramen is extremely variable in relative
size in all forms. Martin and Tate (1970:215) used the small
pneumatic “fossa” (foramen?) of P. kimballensis to separate it
from other turkeys, but all other forms may have an equally
small pneumatic foramen.

3. The inner dorsal intermuscular line is curved farther
away from the inner edge of the shaft in P. kimballensis than
in other turkeys, as stated by Martin and Tate (1970). Howard
(1927) correctly reported a similar amount of curvature in M.
gallopavo and M. californica, but regarded this as being great-
er than in M. ocellata. Brodkorb (1964b) found this line to be
curved farther away in M. gallopavo and M. californica than
in M. ocellata, with M. progenes being intermediate. How-
ever, the individuals of each of these four species that I ex-
amined had a similar amount of curvature.

4. The inner dorsal intermuscular line is longer, extending
to the sternal facet, inM. crassipes than in other forms, among
which this line usually extends farther sternally in Inglis spec-
imens. The specimens of P. kimballensis and M. progenes are
damaged in this area.

5. The outer dorsal intermuscular line curves away from the
outer edge of the shaft more in M. progenes, M. crassipes, and
someM. ocellata than in other forms, contradicting Brodkorb
(1964b), and Martin and Tate (1970), who reported more cur-
vature in M. progenes and in Proagriocharis, respectively,
than in other species.

6. The procoracoid may be of similar shape and extent in

134

Steadman: Turkey Osteology and Paleontology

all forms, contradicting Martin and Tate (1970), who found
a more blunt procoracoid inP. kimballensis and M. californica
than in M. ocellata or M. gallopavo.

7. The scapular facet is more rounded in P. kimballensis,
some Inglis fossils, someM. californica, and som eM. crassipes
than in other turkeys, generally agreeing with Howard (1927)
and Martin and Tate (1970).

8. The triosseal canal is deeper, producing a thinner neck,
in P kimballensis than in other turkeys, as stated by Martin
and Tate (1970).

9. The humeral end in dorsal view may be inflected from
the axis of the shaft similarly in all forms, contradicting Brod-
korb (1964b), who reported more inflection in M. progenes
than in M. gallopavo or M. ocellata.

10. The furcular facet is less protrudent from the neck (join-
ing it more smoothly) in a few Coleman fossils, all M. califor-
nica, most M. gallopavo, and a few M. ocellata than in other
forms. This contradicts Brodkorb (1964b), who found no over-
lap between M. gallopavo and M. ocellata, and Martin and
Tate (1970:2 14-2 15), who stated that P. kimballensis “resem-
bles Agriocharis [ ocellata ] and differs from Parapavo
[californicus] and Meleagris [gallopavo] in that the head is
raised above the inner surface of the neck.”

11. The furcular facet is notched in P. kimballensis and in
Inglis and Coleman specimens, unnotched inM. progenes, and
variable in all other forms. Howard (1927:6) overlooked this
variability in reporting a “slight notch” in M. californica,
which she considered to be most similar to M. gallopavo, and
a “shallow notch farther removed from the interior border” in
M. ocellata. Martin and Tate (1970:214) also found no vari-
ability and reported an “indistinct mid-ventral notch” in P.
kimballensis and M. californica and the absence of the same
in M. gallopavo and M. ocellata.

12. The furcular facet in medial view is shallower relative
to the depth of the shaft in M. crassipes than in other forms.
However, no consistent difference in its overall shape was
noted, thus contradicting Howard (1927), who found it to be
oval in outline in M. gallopavo and M. californica, and more
rounded in M. ocellata.

13. The attachment of Lig. humero-coracoideum anterius
inferior (“coracohumeral ligament”) may be similarly shaped
in all turkeys, contradicting Martin and Tate (1970), who
found it to be elongate in P. kimballensis, but triangular in
other turkeys.

14. The attachment of Lig. humero-coracoideum anterius
inferior has a distinct external border in all turkeys as stated
by Howard (1927). Martin and Tate (1970) found it to be in-
distinct in P. kimballensis, but breakage obscures this char-
acter.

15. The external border of the attachment of Lig. humero-
coracoideum anterius inferior reveals no consistent differences
in any turkeys, contradicting Howard (1927), who found it to
be more abrupt in M. gallopavo than in M. ocellata, with M.
californica being intermediate.

16. The head is quite variable in shape in all forms, among
which no consistent differences were found, thus disagreeing
with Martin and Tate (1970), who reported the head to be oval
in P. kimballensis and M. californica and not oval in M. gal-
lopavo and M. ocellata.

17. The head in internal view is connected to the neck for
a lesser distance in P. kimballensis than in M. progenes, M.

californica, M. gallopavo, or M. ocellata. Meleagris crassipes
is intermediate. This contradicts Brodkorb (1964b), who found
the head to have a less extensive connection to the neck, i.e.,
to be more “free” (Brodkorb 1964b) from the neck, in M. pro-
genes and M. ocellata than in M. gallopavo. Inglis and Cole-
man fossils may be similar to any of the above forms, thus
contradicting Martin and Tate (1970), who found the head
connected to the neck less in P. kimballensis than in any other
turkey.

Martin and Tate (1970) regarded the coracoid of P. kim-
ballensis to be most similar to that of M. ocellata, with some
resemblances to M. californica. They did not compare it qual-
itatively with M. progenes. The coracoid of P. kimballensis is
most similar to that of the Inglis species and least similar to
that ofM. progenes and M. californica. The perceived lack of
similarity between P. kimballensis and M. progenes may be at
least partly due to the presence of only one fairly complete
coracoid for each.

Brodkorb (1964b) found the coracoid of M. progenes to re-
semble that of M. ocellata more than that of M. gallopavo,
while I find that M. progenes does not resemble either living
species more than the other.

Coracoids of M. gallopavo, M. ocellata, and the Coleman
species are very similar to each other. As with other elements,
Inglis and Coleman specimens are closer to each other than to
any other forms.

Howard (1927) considered the coracoid of M. californica
generically separable from that of M. gallopavo and M. ocel-
lata, with a closer resemblance to M. gallopavo. However, M.
californica does not consistently differ from M. gallopavo or
M. ocellata in any of the 15 characters listed.

Scapula

Fig. 9; Howard 1927: Pis. 7, 8.

1 . The acromion is usually less pointed in Coleman speci-
mens and M. gallopavo than in Inglis specimens, M. ocellata,
or M. californica. The specimens of M. progenes are damaged
in this area.

2. The furcular articulation in proximal view is usually
deeper relative to the depth of the glenoid facet in M. califor-
nica and M. ocellata than in M. progenes, Inglis and Coleman
specimens, or M. gallopavo, thus agreeing with Howard
(1927).

3. The base of the shaft in dorsal view lacks a pneumatic
foramen in M. progenes and in Inglis fossils, while Coleman
fossils, M. californica, M. gallopavo, and M. ocellata have a
pneumatic foramen of variable size. Wetmore (1944:98) re-
ported a non-pneumatic scapula from Rexroad as “Meleagrid-
idae sp.?”. Brodkorb (1964b) tentatively referred two similar
scapulae from Rexroad to M. progenes. Both authors noted
that the Rexroad specimens differed from other known species
of turkeys in this character.

The structure of the scapula is very similar in all forms
that have a pneumatic foramen. In agreement with Howard
(1927), I tentatively regard M. californica as being more sim-
ilar to M. ocellata than to M. gallopavo in its scapular struc-
ture. Evolutionary implications of a foraminate versus non-
foraminate scapula are discussed below in the section on
evolution.

Steadman: Turkey Osteology and Paleontology

135

Humerus

Howard 1927: PI. 2.

1. The head is similar in relative size in all forms, contra-
dicting Howard (1927), who reported a more pronounced head
in M. californica than in either living species.

2. The external tuberosity in males is usually less prominent
in Inglis specimens and M. crassipes than in other forms. No
differences were seen among females.

3. The head protrudes more anconally above the attachment
of the deltoid muscle in M. gallopavo than in most Inglis or
Coleman specimens, most M. californica, most M. crassipes,
or most M. ocellata, partially contradicting Howard (1927),
who reported greater protrusion in M. gallopavo than in M.
ocellata, with M. californica being intermediate. The Vallecito
Creek specimen is damaged in this area. Cracraft (1968) noted
the variability of this character.

4. The capital groove is usually shallower and narrower in
M. ocellata than in M . gallopavo. Other forms are similar to
one or the other of the above species. The Vallecito Creek
specimen is damaged in this area.

5. The medial rim of the pneumatic foramen has less prox-
imal extension and is less clearly defined in some Coleman
specimens and in all M. ocellata than in other forms.

6. The deltoid crest in lateral view usually has a more point-
ed apex in Inglis specimens and M. ocellata than in Coleman
specimens, M. californica, M. gallopavo or M. crassipes. The
Vallecito Creek specimen is damaged in this area.

7. The scar of M. latissimus dorsi is deeply depressed, es-
pecially its distal portion, in the Vallecito Creek specimen,
usually in M. gallopavo, rarely in Inglis and Coleman speci-
mens, M. californica and M. ocellata, and never in M. cras-
sipes. It is never quite as deep in M. ocellata as in the Vallecito
Creek specimen. This contradicts Howard (1927:8), who found
it to be deep in M. gallopavo and shallow in M. ocellata, with
M. californica resembling either species. She added, “A certain
variability may be noted in the Meleagris [gallopavo] speci-
mens and might be found in Agriocharis [ocellata] as well,
were more material available.” Howard (1963) found this mus-
cle scar to be deep in the Vallecito Creek turkey, but did not
consider it in determining the relationship of the fossil.

8. The shaft of the Vallecito Creek specimen in palmar or
anconal view is more decurved than in other forms. However,
this is probably an artifact of preparation, the specimen being
broken into two pieces and glued together immediately distal
to the scar of M. latissimus dorsi.

9. The brachial depression is deeper in the Vallecito Creek
specimen than in other forms, possibly because of crushing,
as stated by Howard (1963). It is nearly as deep in some Inglis
and Coleman specimens and some M. crassipes as in the Val-
lecito Creek specimen.

10. The attachment of M. pronator brevis faces more palmad
and less mediad in the Vallecito Creek specimen, most Inglis
and Coleman specimens, and most M . ocellata than in M.
californica, mostM. crassipes, or most M. gallopavo. Howard
(1963) noted no variation in reporting it to face more palmad
in the Vallecito Creek specimen and M. ocellata than in M.
gallopavo.

11. The prominence proximal to the attachment of the an-
terior articular ligament is usually sharper and more protru-
dent in Coleman and Inglis specimens than in other forms.

12. The attachment of the anterior articular ligament faces
more laterad in the Vallecito Creek specimen (Howard 1963)
and some Inglis specimens than in other forms.

13. The external condyle usually has a more pointed prox-
imal end, especially in males, in Coleman specimens and M.
gallopavo than in Inglis specimens, M. crassipes, or M. ocel-
lata', and M. californica may be similar to any of the above
forms. Although it is rounded in the Vallecito Creek specimen,
it resembles certain individuals of all other forms, thus con-
tradicting Howard (1963), who stated it to be more rounded
in the Vallecito Creek specimen and M. ocellata than in M.
gallopavo or M. californica.

14. The external condyle has the greatest medial curvature
in those specimens with the most pointed external condyles
(see character 13) and vice versa. This varies exactly as char-
acter 13.

The status of the Vallecito Creek specimen in characters 7,
8, and 9 is probably a result of its poor preservation, and these
characters are, therefore, not included in Table 2. See page
143 for a further discussion of the Vallecito Creek humerus.

With the exception of the Vallecito Creek specimen, the least
similarity in the humerus is seen between M. gallopavo and
specimens from Inglis, and between M. gallopavo and M.
ocellata. The most similarity is between M. californica and M.
crassipes. Only in character 5, in which M. ocellata differs
from all forms except some Coleman specimens, is an absolute
distinction seen between any two forms represented by more
than one specimen.

Ulna

Howard 1927: Pis. 1, 8.

1. The shaft in lateral view in males is more curved in M.
crassipes and M. ocellata than in M. californica or M. gallo-
pavo. Males from Inglis and Coleman may resemble any of
the above forms. No difference is seen between females of any
forms. Curvature of the shaft generally increases as ulnar size
decreases.

2. Howard (1927) reported a more marked distal overhang
and a narrower distal border in the external cotyla in M. gal-
lopavo than in M. ocellata or M. californica, but this is not
apparent to me.

I detected no other valid characters in the ulna, which ap-
pears to be the least diagnostic major limb bone in turkeys.

Radius

Howard 1927: PI. 7.

1. The lateral ridge in the distal portion of the shaft is usu-
ally less distinct in M. ocellata than in other forms. Howard
(1927) noted several differences in the shape and position of
this ridge in M. californica, M. ocellata, and M. gallopavo,
but these differences are not apparent to me.

2. The distal ligamental prominence is usually less protru-
dent mediad in M. ocellata than in other forms.

3. The distal tendinal groove is usually deeper, especially in
its distal portion, and of greater proximal extent, in M. ocellata
than in other forms. The condition in M. ocellata is similar to
that stated by Howard (1927). However, she found this groove
to be only a faint notch in M. californica, while being broad
and shallow in M. gallopavo, characters which I find to be
inconsistent.

136

Steadman: Turkey Osteology and Paleontology

Howard (1927) found the radius of M. calif ornica to resem-
ble that of M. ocellata more than that of M. gallopavo, but my
data show that M. californica resembles M. gallopavo more
than M. ocellata in this element. In fact, the radii from Inglis
and Coleman and those of M. californica, M. crassipes, and
M. gallopavo are alike in all three characters, and three av-
erage differences from all other known forms are seen in the
radii of M. ocellata.

Ulnare

1. The ulnar base is wider relative to its length in Inglis
fossils and M. ocellata than in M. progenes or M. gallopavo,
contradicting Brodkorb (1964b), who found it relatively wider
in M. progenes and M. ocellata than in M. gallopavo.

2. The ulnar base is not consistently different in height,
contradicting Brodkorb (1964b), who found it high in M. pro-
genes and M. ocellata and low in M. gallopavo.

Carpometacarpus

Howard 1927: PI. 9.

1. The inner trochlea is usually less notched in the proximal
edge by Lig. internum ossi carpi ulnaris et metacarpi in M.
progenes and M. ocellata than in other forms, contradicting
Brodkorb (1964b), who reported a deeper notch in M. progenes
and M. gallopavo than in M. ocellata.

2. The interior carpal fossa does not consistently differ in
depth in any form, but the posterior carpal fossa is usually
shallower in M. ocellata than in other forms, contradicting
Brodkorb (1964b), who found both carpal fossae to be shal-
lower in M. progenes and M. ocellata than in M. gallopavo.

3. The intermetacarpal tubercle, which may be fused or
unfused to metacarpal III in all forms, as noted by Howard
(192 7), does not consistently differ in position. Brodkorb
(1964b) found it positioned more distally in M. gallopavo than
in M. progenes, which is true for all but one individual of M.
gallopavo which I examined.

4. The facet for digit III does not consistently differ in rel-
ative length, contradicting Brodkorb (1964b), who reported a
longer facet in M. progenes than in living turkeys.

The carpometacarpus of M. progenes is very similar to those
of Inglis and Coleman specimens. Brodkorb (1964b) found M.
progenes to be more similar toM. ocellata than toM. gallopavo
in most characters of the carpometacarpus. However, my
qualitative data do not reveal a closer resemblance of M. pro-
genes to either of the living species. M. ocellata has the most
distinctive carpometacarpus. Howard (1927) found no basis
for relationships in this element.

Pelvis and Synsacrum

1. The median dorsal ridge is usually relatively wider in M.
ocellata than in M. gallopavo. All fossils examined are dam-
aged in this area.

2. The anterior portion of the ilium usually flares more lat-
erad in M. gallopavo than in M. ocellata. All fossils examined
are damaged in this area.

3. The antitrochanter usually has a more decurved ventral
border in Inglis and Coleman specimens and M. gallopavo
than in M. californica or M. ocellata.

4. The area between the antitrochanter and the obturator
foramen is usually more pneumatic in Inglis specimens than
in other forms.

5. The ischium is usually more inflected caudally, appearing
larger in caudal view, in M. gallopavo than in M. ocellata. All
fossils examined are damaged in this area.

No absolute distinction is ever seen between M. gallopavo
and M. ocellata, all characters being based upon average dif-
ferences. Little can be said of fossil forms because of their poor
condition.

Femur

Howard 1927: PI. 9.

1. The lesser trochanter is notched by the groove for Lig.
capsularis femoris to a variable depth in all forms, contra-
dicting Howard (1927), who found it less distinctly grooved in
M . ocellata than in M. californica or M. gallopavo. The single
specimen of M. progenes is similar to individuals of all other
forms, contradicting Brodkorb (1964b), who reported a more
strongly notched lesser trochanter in M. californica and M.
gallopavo than in M. progenes or M. ocellata.

2. The groove for Lig. capsularis femoris on the lateral sur-
face of the shaft does not consistently differ in depth in any
form, contradicting Brodkorb (1964b), who stated it to be
deeper in M. californica and M. gallopavo than in M . progenes
or M. ocellata.

3. The greater trochanter in medial view usually has a great-
er proximal extension in M. gallopavo than in M. ocellata.
Inglis and Coleman fossils and M. californica may resemble
any of the above forms. The specimens of M. progenes and M.
crassipes are damaged in this area.

4. The shaft in medial view is usually less curved in M.
ocellata than in other forms. The specimen of M. progenes is
damaged in this area.

5. The posterior intermuscular lines may be fused or unfused
in the middle portion of the shaft in all forms except possibly
M. progenes (in the single known specimen of this species these
lines are fused). Howard (1927) noted less convergence of these
lines in M. gallopavo than in M. ocellata or M. californica,
but Brodkorb ( 1 964b: 2 27) found these lines to be usually
“fused along middle third of their length” in M. ocellata and
M. californica, and “usually unfused, although the character
is variable” in M. gallopavo.

6. The intercondylar groove is usually relatively deeper in
M. crassipes than in other forms, all of which may resemble
each other in depth. This contradicts Howard (1927), who
found the intercondylar groove in M. californica to be broader
but similar in depth to that in M. gallopavo, and deeper and
narrower in M. ocellata than in either M. californica or M.
gallopavo.

7. No other distinguishing characters are seen in the distal
end. Howard (1927) noted a greater depth (relative to width)
of the inner condyle in M. californica than in M. gallopavo or
M. ocellata, and the latter two species were said to have an
oblique angle at the junction of the posterior surface of the
shaft and inner condyle, as opposed to nearly a right angle in
M. californica. These characters are not apparent to me.

The three characters (1,2, and 5) used by Brodkorb (1964b)
in stating that M. progenes resembled M. ocellata more than
M. gallopavo are found to vary in most species. Yet the femur

Steadman: Turkey Osteology and Paleontology

137

of M. progenes still resembles that of M. ocellata or M. cali-
fornica slightly more than that of M. gallopavo. Inglis and
Coleman femora show perfect resemblance to each other, and
both are also quite similar to that of M. gallopavo, M. ocellata,
and M. calif arnica. The femur of M. ocellata is least similar
to that of M. gallopavo or M. calif ornica.

Tibiotarsus

Howard 1927: Pis. 10, 11.

1. The inner cnemial crest is usually less protrudent laterally
in M. ocellata than in Coleman specimens or M. gallopavo,
and Inglis specimens and M. californica may resemble any of
the above forms. Meleagris progenes and M. crassipes are
damaged in this area. This finding contradicts that of Howard
(1927:18), who reported this crest as “Most abruptly thrust
toward outside in Agriocharis [ocellata]; Parapavo [calif amicus]
similar to Meleagris [gallopavo] and Pavo." She said that her
statement on M. ocellata agreed with that of L. Miller (1916a),
who did not include M. gallopavo in his comparison. However,
L. Miller did not say whether he was referring to the inner or
to the outer cnemial crest. (One must read L. Miller (1916a)
with caution because he used “ Agriocharis ” and “ Meleagris
ocellatus" interchangeably. When he spoke of “ Meleagris ," he
seemed to be referring either to both M. ocellata and M. gal-
lopavo, or only to M. ocellata.)

2. The internal ligamental prominence is larger in the only
undamaged specimen of M. progenes than in most Inglis and
Coleman specimens and all examples of other forms.

3. No other diagnostic characters are seen in the distal end.

L. Miller (1916a) reported the condyles to be closer together
and the tunnel under the supratendinal bridge to be larger in

M. ocellata than in M. californica, but these differences are
not apparent to me. Based on characters of the tibiotarsus, the
affinities of M. californica to M. ocellata or M. gallopavo were
unclear to Howard (1927). I agree with Howard and stress the
similarity of the tibiotarsi of all turkeys. This statement is not
without precedent, as A. Miller and Bowman (1956) and Brod-
korb (1964b) have noted the lack of useful taxonomic char-
acters in the distal end of the tibiotarsus in turkeys.

Fibula

The lateral surface of the fibula in all turkeys is convex to
some extent, but it is more convex in M. californica than in
most Inglis specimens, all Coleman specimens, most M. gallo-
pavo, and all M. ocellata, partially contradicting Howard
(1927), who found it to be convex in M. californica, straight
in M. gallopavo, and slightly convex in M. ocellata. Howard
also noted variability in the fibula of M. ocellata, and regarded
the fibula as being of little taxonomic value.

Tarsometatarsus

Figs. 10-14; Howard 1927: Pis. 12, 13.

The most diagnostic bone in turkeys is the tarsometatarsus,
and many characters of this bone have been used to define
various taxa of turkeys. Most important has been the bony
spur core on the plantar side of the shaft, which in life is
surrounded by a scutellum. Males have a spur, but it is nor-
mally absent in females. However, Williams and Austin (1969)
estimated that more than 1 percent of the females of M. gal-

lopavo osceola in their Florida study area had a spur on at
least one leg. Pattee and Beasom (1977) also found at least one
tarsal spur on 2 of 228 females of M. gallopavo intermedia in
Texas.

The normal presence of rudimentary spurs in females of M.
ocellata has often been cited as a significant difference between
that species and M. gallopavo, and was used in the original
generic diagnosis of Agriocharis (Chapman 1896). This rudi-
ment is only a single dark, hardened, and slightly elevated
scutellum that is located where the bony spur would be in a
male. Its presence is not normally reflected in any way on the
tarsometatarsal bone itself. Such a rudimentary spur is also
present in females of M. gallopavo and is occasionally as de-
veloped as in M. ocellata.

The scutellum and inner bony spur core of males become
increasingly pointed with age. Spur development in M. gal-
lopavo osceola may begin at 6 months of age and is not com-
pleted by the end of the first year (Lovett Williams pers.
comm.). Eaton and Moore (1965-66:39) found the spur core
of M. gallopavo silvestris to be “only a knob” at age 9-10
months, “more prominent . . . well formed” at age 20-22
months, and “slightly more massive (than at age 20-22
months]” at age 29-33 months. I have noted that males of
wild M. ocellata have a small (length, 8 to 15 mm), slightly
pointed spur by age 8-9 months and develop long, sharply
pointed spurs by age 2-3 years. However, the size and point-
edness of the spur core seem to show no relationship to the
total size of the bone. Eaton and Moore (1965-66:36), in ref-
erence to wild M. gallopavo silvestris, state, “The hind ap-
pendage [ = leg] , though reaching almost full length at 5-6
months continued to become more massive and developed the
bony core for the spur between the tenth and twentieth
months.” Both the shortest and the longest of ten tarsometa-
tarsi of males of M. gallopavo osceola that I measured had
short, blunted spur cores. L. Miller (1916a) reported the weak-
est spur in the longest, and the strongest spur in the shortest,
of 25 specimens of M. californica he examined.

I also examined the tarsometatarsi of the following species
of Galliformes from the collection of Pierce Brodkorb, in ad-
dition to those of turkeys, to determine the relationships of
Rhegminornis calobates : Odontophorinae — Dendrortyx ma-
croura, Callipepla squamata, Colinus virginianus; Phasiani-
nae — Alectoris graeca, Francolinus sephaena, Pternistes
swainsonii, Coturnix delegorguei, Caloperdix oculea, Gallus
gallus, Crossoptilon auritum, Catreus wallichii, Syrmaticus
ellioti, Chrysolophus pictus, Pavo muticus; and Numididae —
Acryllium vulturinum.

1. The small calcaneal ridge between the inner and outer
calcaneal ridges is not consistently different in size in any form
and may be absent in M. gallopavo. The specimens of Rheg-
minornis calobates and M. progenes are damaged in this area.
My findings contradict those of L. Miller ( 1 916a: 91), who not-
ed the well marked presence of this medial calcaneal ridge in
M. californica, but found it to be “almost entirely wanting
even in old specimens of Meleagris gallopavo . . . but faintly
indicated inM. ocellatus .” Although Wetmore (1924) also used
this character, he noted its variability in M. gallopavo and
commented on its decreased taxonomic significance. Howard
(1927) found this ridge to be well developed inM. ocellata and
M. californica, but 18 of her specimens of M. gallopavo lacked
the ridge, and its development and position in the remaining

138

Steadman: Turkey Osteology and Paleontology

13 was variable. A. Miller and Bowman (1956) correctly stated
that this ridge is not useful in separating Meleagris from Para-
pavo.

2. The large outer calcaneal ridge usually has less plantar
protrusion in M. ocellata than in other forms, partially con-
tradicting L. Miller (1916a), who reported it to be more prom-
inent in M. californica than in M. gallopavo or M. ocellata.
The specimens of R. calobates and M. progenes are damaged
in this area.

3. The larger outer calcaneal ridge usually is relatively long-
er in M. crassipes than in M. leopoldi, Inglis and Coleman
fossils, M. gallopavo or M. ocellata. It is either intermediate
in length or similar to the shorter specimens of the forms above
in P. kimballensis and M. californica. The specimens of R.
calobates and M. progenes are damaged in this area. This
contradicts L. Miller (1916a), who found it to be shorter in M.
gallopavo and M. ocellata than in M. californica.

4. The internal cotyla protrudes more abruptly mediad in
M. gallopavo than in R. calobates, P. kimballensis, M. leo-
poldi, several Inglis and Coleman specimens, several M. cal-
ifornica, several M. crassipes, mostM. ocellata, and all species
of Numidinae, Phasianinae, and Odontophorinae examined.
The specimens of M. progenes are damaged in this area. Wet-
more (1924:9) may have been referring to this same character
when he attributed a “proportionally broader and heavier”
head of the tarsometatarsus to M. gallopavo as compared to
M. californica.

5. The lateral prominence of the external cotyla is usually
slightly less developed in M. crassipes than in other turkeys.
The specimens of R calobates and M. progenes are damaged
in this area.

6. Two narrow ridges are present above the tubercle for
tibialis anticus in R. calobates, P. kimballensis, and all non-
meleagridine phasianids examined, but are absent in all known
post-Hemphillian turkeys. The specimens of M. progenes are
damaged in this area.

7. The tubercle for tibialis anticus is narrower but more
prominent in R. calobates than in other forms, as stated by
Olson and Farrand (1974). It is also narrower and more prom-
inent in R. calobates than in all other phasianids examined
except Francolinus sephaena and Gallus gallus, in which it is
more prominent but equally narrow. The specimens of M.
progenes are damaged in this area.

8. The acrotarsial groove on the shaft is usually shallower
in M. californica and M . crassipes than in M. leopoldi, Inglis
and Coleman specimens, or M. gallopavo. It is similar in R.
calobates to shallower individuals of the last three forms listed
above, but is shallower than in M . leopoldi. It is usually deeper
in P. kimballensis and M. ocellata than in R. calobates, M.
leopoldi, M. californica, or M. crassipes. Its depth is also quite
variable within other phasianid subfamilies studied. The spec-
imens of M. progenes are damaged in this area. This partially
contradicts Olson and Farrand (1974), who reported a shal-
lower groove in R. calobates than in M. gallopavo or M. ocel-
lata.

9. The thin ossified intertendinal septum that extends from
the hypotarsus nearly to the facet for the hallux is absent in
R. calobates, but is present in all other turkeys and many other
phasianids, as stated by Olson and Farrand (1974). They noted
the probable lack of taxonomic significance in this character.
The specimens of M. progenes are damaged in this area. This
septum, absent in young (but full-sized) specimens of fossil and

living turkeys, was found by Eaton and Moore (1965-66) to
develop at age 5-6 months in females and 9-10 months in
males of M. gallopavo silvestris.

10. The spur core is strongly curved upward in P. kimbal-
lensis, Inglis specimens, and M. ocellata, slightly to strongly
curved upward in Coleman specimens, and straight to slightly
upcurved in M. californica, M. gallopavo, and M. crassipes
(see Fig. 4). The specimens of M. progenes and M. leopoldi
are damaged in this area. No spur core is present on the known
specimens of R. calobates.

11. The surface of the spur core is of similar roughness in
all forms, contradicting Howard (1927:24), who described the
spur core of M. gallopavo as “roughly built” and that of M.
californica andM. ocellata as “well formed.” Developing spur
cores have a very rough surface, which becomes smoother with
age. Full-length, pointed spur cores are of the same approxi-
mate smoothness in all species. Thus the distinction noted by
Howard was probably due to age differences in her specimens.

12. The facet for hallux does not consistently differ in po-
sition or shape in any turkeys, while in other phasianids it may
resemble that of turkeys or be more rounded and deeper. This
contradicts Brodkorb (1964b), who reported it as low in M.
progenes, rather low in M. ocellata, and elevated in M. gal-
lopavo.

13. The inner intertrochlear foramen is rarely absent only
in Inglis and Coleman specimens, M. californica, M. gallo-
pavo, and M. ocellata. It is always present in varying degrees
of development in other turkey specimens. Note that the five
variable forms are the only turkeys for which a large sample
is available. Howard (1927) noted variation in the develop-
ment of this foramen in M. californica, M. gallopavo, and M.
ocellata', A. Miller and Bowman (1956) also noted the incon-
sistency of this character. Olson and Farrand (1974) noted its
presence in R. calobates as a meleagridine character to distin-
guish it from the Phasianidae (sensu stricto). However, this
foramen may be present in specimens of Colinus virginianus,
Alectoris graeca, Pternistes swainsonii, Coturnix delegorguei,
Caloperdix oculea, Gallus gallus, Crossoptilon auritum, Chry-
solophus pictus, and Pavo muticus.

14. The inner intertrochlear foramen is located lower in one
specimen off?, calobates (PB 8447) than in all other phasianids
examined. Olson and Farrand (1974) stated that it is in the
same location in the holotype of R. calobates (MCZ 2331) as
in M. gallopavo.

15. The lateral distal foramen is usually relatively lower in
Inglis and Coleman specimens than in other turkeys, including
R. calobates. Brodkorb (1964b) reported it to be low in M.
progenes and M. ocellata, slightly higher in M. leopoldi, and
high in M. gallopavo and M. californica, but these differences
are not apparent to me.

16. The lateral distal foramen in plantar view is usually
relatively larger in M. crassipes and usually smaller in M.
ocellata than in other turkeys, including R. calobates. Olson
and Farrand (1974) also reported it to be smaller inM. ocellata
than in M. gallopavo, but reported it to be relatively smaller
in R. calobates than in M. gallopavo. This is not true in spec-
imens that I have examined. They also reported it to be more
elongate in meleagridines than in phasianids ( sensu stricto).
However, it may be as relatively large and elongate in Gallus
gallus, Crossoptilon auritum, or Chrysolophus pictus as in tur-
keys.

17. The inner trochlea is usually relatively wider in M. eras-

Steadman: Turkey Osteology and Paleontology

139

sipes than in other turkeys. The specimens of R. calobates are
damaged in this area, and that of P. kimballensis was not
available for study. Brodkorb (1964b) found it to be narrower
in M. progenes and M. gallopavo than in M. leopoldi or M.
ocellata, but I found no consistent difference in this character
in these four species.

18. The inner trochlea in plantar view is situated more me-
dially in some M. ocellata and all non-meleagridine phasianids
examined than in other turkeys, including R. calobates. The
specimen of P. kimballensis was not available for study.

19. The inner trochlea is less elevated in R. calobates , M.
leopoldi, and two specimens of M. crassipes than in all other
turkeys or all other phasianids examined except Alectoris grae-
ca. The specimen of P. kimballensis was not available for
study, but, in Fig. IE of Martin and Tate (1970), it appears
to be more elevated than in the above forms. Olson and Far-
rand (1974) reported it to be more elevated in turkeys than in
other phasianids (sensu stricto), but this is not apparent to me.

20. The inner trochlea is usually situated more on the plan-
tar surface in females of M. ocellata than in other turkeys,
including R. calobates. The specimen of P. kimballensis was
not available for study. This character is variable among non-
meleagridine phasianids examined.

21. The middle trochlea has a similar amount of acrostarsial
rotation in certain specimens of each species of turkey, includ-
ing R. calobates, and may resemble certain species of Numi-
didae, Phasianinae, and Odontophorinae. The specimen of/5.
kimballensis was not available for study. This contradicts Ol-
son and Farrand (1974), who found the middle trochlea to be
rotated more acrotarsiad inM. gallopavo and/?, calobates than
in other Galliformes.

22. The ridge in the proximal portion of the medial plantar
border of the middle trochlea is usually less conspicuous in M.
californica and M. crassipes than in other turkeys, including
R. calobates. Wetmore (1924) also distinguished M. gallopavo
from M. californica on this basis, but noted no variation in
this character.

23. The outer trochlea does not vary significantly in its ex-
tent of plantar rotation in any turkey. The specimens of R
calobates and P. kimballensis were not available for study.
This contradicts Wetmore (1924), who observed that both lat-
eral trochleae were rotated more plantad in M. ocellata than
in M. californica or M. gallopavo.

24. The outer trochlea is usually less elevated in M. crassipes
than in other turkeys. It is more elevated in R. calobates than
in other turkeys, as stated by Olson and Farrand (1974), and
is also more elevated in R. calobates than in all other phasi-
anids examined. The specimens of P. kimballensis are dam-
aged in this area.

25. The intertrochlear notches are usually relatively narrow-
er in M. ocellata than in R. calobates (only inner notch pres-
ent), M. progenes, M . leopoldi, Inglis specimens, or M . cras-
sipes. Other turkeys may resemble any of the above forms,
but are least similar to M. ocellata. The specimens of P. kim-
ballensis are damaged in this area. The intertrochlear notches
of all non-meleagridine phasianids examined are similar to
those of certain turkeys. This contradicts Brodkorb (1964b)
who reported them to be wider in M. progenes than in other
turkeys, and also Olson and Farrand (1974), who noted less
divergent trochleae in/?, calobates andM. ocellata than inM.
gallopavo.

Table 3 presents a quantitative compilation of the 18 char-

acters of the tarsometatarsus in which a significant difference
is seen between at least two forms listed above. Rhegminornis
calobates shows the lowest overall similarity to other forms,
and possesses two typically phasianine characters (characters
6, 7). It is more similar to that of Proagriocharis kimballensis
or M. leopoldi than to all other forms, to which it is consis-
tently the most dissimilar species. The tarsometatarsus of P.
kimballensis is more similar to that of later forms than is the
tarsometatarsus of R. calobates, although retaining phasianine
character 6. It is least similar to those of M. crassipes, and its
modest degree of resemblance to all other forms is greatest
with M. progenes and M. leopoldi. Meleagris progenes is very
similar to M. leopoldi, Inglis and Coleman specimens, M. gal-
lopavo, and M. californica. Meleagris leopoldi is very similar
only to M. progenes, but it is quite different from only R.
calobates and M. crassipes. Inglis specimens are very similar
to those of M. progenes and the Coleman specimens, and
the latter are also very similar to M. progenes and M. gallo-
pavo. M. gallopavo has a high degree of resemblance to M.
progenes, Coleman specimens, and M. californica, with the
last also closely resembling M. progenes.

High levels of similarity of the magnitude noted between
the tarsometatarsi of M. progenes, M. leopoldi, Inglis and
Coleman specimens, M. gallopavo, andM. californica are not
noted in M. ocellata or M. crassipes. These two species are
very unlike each other and display only a modest degree of
similarity to all forms but/?, calobates, with the exception that
M. crassipes is fairly similar to M. californica and quite dis-
similar to P. kimballensis.

Ratios of all of the measurements of every element of the
various taxa were computed and compared. These data, avail-
able from the author on request, prove to be of limited taxo-
nomic value with the exception of the data on the tarsometa-
tarsi of males. To my surprise, the greatest difference among
all of the taxa in these proportions is often between Recent
specimens of M. gallopavo silvestris andM. g. osceola. Thus,
considering this high variability within a living species, I re-
gard the use of intra-elemental proportions in the definition of
extinct genera and species as being generally of very limited
value, except as discussed below.

The angle of the spur core (Table 20: K) and the relative
height of the spur core (Table 22: F/A, G/A) have traditionally
been very important in the definition of various taxa of tur-
keys. It may be seen from my data that the angle of the spur
core is smallest in P. kimballensis and M. crassipes, although
both species are represented by only one individual. Next ap-
pears the intermediate group of M. progenes, M. leopoldi, In-
glis and Coleman specimens, and M. californica. A gradation
is seen in the various large samples from Florida, with a stead-
ily increasing angle through time from Inglis to Coleman to
Ichetucknee to modern M. g. osceola. Note the large amount
of overlap between adjacent samples through this progression.
Also note the high degree of variability in this angle within
any given large sample. These data suggest the use of great
caution when making taxonomic conclusions based upon only
minor differences.

L. Miller (1940) reported angles of 39 and 79 degrees, re-
spectively, for M. crassipes andM. californica, while A. Miller
and Bowman (1956) reported angles of 53.0 to 58.5 degrees for
M. leopoldi, 62.0 degrees in M. californica, and 63.5 to 71.5
degrees in M. gallopavo. Brodkorb (1964b) reported angles of
60 to 80 degrees in M. gallopavo and M. californica, while

140

Steadman: Turkey Osteology and Paleontology

Martin and Tate (1970) reported angles of less than 60 degrees
in P. kimballensis (39 degrees) and “ Agriocharis and angles
greater than 60 degrees in M. californica or M. gallopavo.
Brodkorb (1964b) appears to have placed too much weight on
this angle when he said, “Both A. leopoldi (A.H. Miller and
Bowman 1956) andd. crassipes (L. Miller 1940) have the spur
core low on the shaft and at an angle of less than 60 degrees,
characters that require their removal from the genus Mele-
agris, in which they were described.” However, both M. gal-
lopavo and M. californica may commonly have a spur core
angle of under 60 degrees (Table 20). Also, M. crassipes has
a greater average relative height of the spur core than all other
species (Table 22: F/A, G/A), with M. californica and M. gal-
lopavo intermediate between it and the forms with lower spur
cores (M. leopoldi, Inglis and Coleman specimens). Thus, the
height of the spur core, which is very different in M. crassipes
than in M. leopoldi, cannot be used as a generic character to
separate both M. crassipes and M. leopoldi from M. gallopavo.
Proagriocharis kimballensis and M progenes have spur core
heights that resemble those of many other species, and large
samples of these species are needed before anything definitive
can be stated. Note the high variability of this character in
Coleman specimens, M. gallopavo, and M. californica, dem-
onstrating the variation that may occur in large samples.

L. Miller (1916a) first used spur core height as a taxonomic
character, reporting a height of 40 to 41 percent in M. cali-
fornica. Howard (1927) reported a height of 39.2 to 46.9 per-
cent in M. gallopavo, 41.3 to 49.7 percent in M. californica,
and 40.0 percent in M. ocellata, her figures corresponding to
F/A of Table 22; the figures of more recent authors pertain to
G/A. Noting no variation, L. Miller (1940) reported heights of
45 percent in M. crassipes, 43 percent in M. gallopavo, 42
percent in M. californica, and 35 percent in M. ocellata. The
height (39.8 percent) in M. leopoldi reported by A.H. Miller
and Bowman (1956) is a composite based on measurements
from two different specimens; the 36.3 percent reported herein
is based solely upon the paratype tarsometatarsus. Other val-
ues of A. Miller and Bowman (1956) are 41.7 to 46.0 percent
in M. californica, 41.2 to 43.4 percent in M. gallopavo, and
36.0 percent in M. ocellata. The relative spur core height in

M. progenes, described as “situated low” in the Rexroad spec-
imen by Brodkorb ( 1 964b: 2 25), can be computed only from
the tentatively referred specimen from Benson, Arizona. The
absolute height of the spur core (Table 20: F, G) of the Rexroad
specimen of M. progenes is about at the minimum range of
Inglis specimens. Martin and Tate (1970) listed a height of 36
percent in M. ocellata and 42 percent in P. kimballensis, say-
ing that the latter just overlaps the lower range of M. gallopavo
and M. californica. However, the figure of 42 percent is er-
roneous, for assuming that their figures for total length (98
mm) and spur core height (40 mm) are correct, the relative
spur core height would be 40.8 percent. The relative stoutness
of the spur core (Table 22: H/J) is greatest in M. crassipes,
and averages smallest in M. ocellata.

DISCUSSION OF FOSSIL TURKEYS BY
LOCALITY

Miocene (Hemingfordian)

THOMAS FARM, Gilchrist County, Florida, Rhegminornis
calobates. Tarsometatarsus (MCZ 2331, PB 8447-8449) — Ta-

ble 21, Fig. 14; Figs. 1-2 of Olson and Farrand (1974), in
which PB 8448 is mislabelled PB 1776. This species, originally
described by Wetmore (1943) in the Jacanidae (Charadri-
iformes), was removed from the Jacanidae and placed in the
Meleagridinae by Olson and Farrand (1974). While agreeing
that it is a member of the Phasianidae (sensu lato), I feel that
the known specimens of/?, calobates are insufficient to place it
unequivocally in the Meleagridinae, although such a place-
ment may very well be correct. The four meleagridine char-
acters of R. calobates used by Olson and Farrand to exclude
it from the Phasianidae ( sensu stricto) may be found in certain
species of Phasianidae (see characters 13, 16, 19, 21). Using
living species to define the characters of meleagridine versus
non-meleagridine phasianids, R. calobates has characters of
both groups, suggesting that a redefinition of these subfamilies
would be necessary if one were going to include R. calobates
in one of them. Such a redefinition would be of limited value
at present because of our fragmentary knowledge of R. calo-
bates. Also, it would tend to mask the intermediate nature of
the specimens, and, as more specimens of R. calobates and
other Tertiary phasianids become available, such redefinitions
would become increasingly difficult and meaningless. I have
not attempted to place R. calobates in a modern subfamily
because of the apparently mosaic nature of its tarsometatarsus
and its lack of other known elements.

An indication of the degree of similarity between R. calo-
bates and various turkeys can be seen in Table 3. It is most
similar to Proagriocharis kimballensis and Meleagris leopoldi,
although every form except P . kimballensis is less similar to
R. calobates than any other form. This, as well as overall size,
suggests that R. calobates could be an early form of turkey,
possibly near the ancestry of P. kimballensis . However, this
hypothesis is very tentative at present as it is based on but one
element. It must be kept in mind that the amounts of similarity
seen in any of the relatively poorly known forms in Tables 2
and 3 may be an artifact of small sample sizes.

Early Late Miocene (Clarendonian?)

WESTMORELAND STATE PARK, Westmoreland County,
Virginia. Claremont Member of Eastover Formation (Black-
welder and Ward, unpubl.;_/ide Lauck Ward). Meleagridinae,
cf. Meleagris . Tibiotarsus (USNM 237260) — Table 19. This
specimen is a left tibiotarsus with only a very small section of
the shaft missing. It is slightly eroded and possibly from a
juvenile and could not be distinguished qualitatively from tib-
iotarsi of M. gallopavo or M. ocellata. However, as the tib-
iotarsus of turkeys is nearly lacking in diagnostic characters
(see page 137), positive assignment to Meleagris is not jus-
tified. Most importantly, however, this specimen documents
the occurrence of a fairly large turkey as early as the late
Miocene. Lacking comparable elements, nothing can be said
of the affinities of this specimen to Rhegminornis calobates or
Proagriocharis kimballensis , both of which are much smaller.

I carefully compared this specimen with all other Tertiary
specimens of turkeys, as well as with skeletons of several gen-
era of the Cracidae, as this family of Galliformes is known to
occur in Tertiary fossil localities in North America (Brodkorb
1964a). All specimens reported herein have phasianid char-
acters and can be distinguished from specimens of the Craci-
dae.

Steadman: Turkey Osteology and Paleontology

141

Late Miocene or Early Pliocene
(Hemphillian)

UNSM COLL. LOC. FT-40, Frontier County, Nebraska.
Proagriocharis kimballensis. Coracoid (UNSM 20033) — Table
4. Tarsometatarsus (UNSM 20035-20037) — Tables 20-22,
Fig. 14; Fig. 1 (reduced by the printer beyond the authors’
wishes) of Martin and Tate (1970). Described as a new genus
and species by Martin and Tate (1970), I tentatively regard P.
kimballensis as distinct from Meleagris, cautiously noting that
this conclusion is based upon only two elements of the skele-
ton. Fortunately, both spurred and unspurred tarsometatarsi
exist of P. kimballensis, thus indicating the relative size of the
supposed males and females. Proagriocharis kimballensis is
smaller than any other turkey except R. calobates.

Proagriocharis kimballensis is most similar to R. calobates,
M. leopoldi, and the Inglis fossils (Table 2). However, only
the tarsometatarsi of M. leopoldi and R. calobates could be
compared with P. kimballensis. In Table 3, P. kimballensis
is seen to be most similar to M. progenes and M . leopoldi,
although several other taxa are nearly as similar. In reporting
a greater overall similarity of Proagriocharis to Agriocharis
than to Meleagris ( gallopavo ) or Parapavo ( californicus ), Mar-
tin and Tate (1970) followed Brodkorb (1964a, b) in regarding
Agriocharis as including the species progenes, leopoldi, anza,
crassipes, and ocellata, of which progenes and crassipes are
seen herein to be qualitatively the least similar of all turkeys
to Proagriocharis (Table 2). Meleagris progenes and M. cras-
sipes are, however, the most similar of the above to P. kim-
ballensis in size. The H/A ratio of P. kimballensis is greater
than that of all others (Table 22), but approaches that of M.
crassipes.

In their generic diagnosis of the tarsometatarsus, Martin and
Tate (1970) found Proagriocharis to differ from Meleagris (gal-
lopavo) and Parapavo (californicus), but to resemble Agri-
ocharis (progenes, leopoldi, crassipes, and ocellata), in having
an angle of the spur core of less than 60 degrees. An increased
sample size (Table 20: K) indicates that the spur core angle is
not a generic character (see discussion on pages 139-40).

According to Martin and Tate (1970:215), “The spur core
(cast) is more proximally placed (42 percent of the total length)
than it is in Agriocharis ocellata (36 percent of the total length),
and just overlaps the lower range of Parapavo [californicus]
and Meleagris [gallopavo] in this respect.” The relative height
of the spur core of P. kimballensis, if based on their reported
measurements (p. 217), would be 40.8 percent, as mentioned
above, not 42 percent as reported. However, as Martin and
Tate state, these measurements are from a cast and are there-
fore probably not extremely accurate. The figure of 40.8 per-
cent is within the range found for numerous other forms (Table
22: G/A), and it is no basis for separating the species from
other species at the generic level.

Martin and Tate ( 1970:2 17) distinguish M. progenes and M.
leopoldi from P. kimballensis “because of the difference in the
placement of the spur core in this species.” I agree with them
in the difference of the angle of spur core (Table 20: K), but
find that the difference they report in the relative height of the
spur core requires some clarification. It has been established
above that the true height of the spur core of the single spec-
imen of P. kimballensis is 40.8 percent, not 42 percent. They
report the spur core of M. leopoldi (1970:217) to be “placed

slightly lower (39.8 percent of the total length) than it is in P.
kimballensis (see Miller and Bowman 1956:44).” They do not
mention the fact that Miller and Bowman’s ratio was based
on the total length of the paratype and the height of the spur
core of the holotype. Had Martin and Tate examined the para-
type tarsometatarsus of M. leopoldi, a greater difference be-
tween it and P. kimballensis would have been seen (Table 22:
G/A). Martin and Tate (1970:217) continue that the spur core
of M. progenes is “slightly more distally placed than in A.
leopoldi .” This is true (Table 20: G), but it must be stressed
that the relative height of the spur core in M. progenes from
Rexroad, Kansas, is unknown, although it is 42.8 percent in
a specimen of Meleagris cf. M. progenes from Benson, Ari-
zona.

In summary, Proagriocharis is retained herein as a distinct
genus, mainly because of its low similarity to most other forms
as seen in Tables 2 and 3, although many characters used by
Martin and Tate (1970) to distinguish it from other turkeys are
shown to be invalid. As stated in the discussion of Rhegmi-
nornis above, Proagriocharis is probably not on or near the
lineage leading to the larger Pleistocene and Recent forms of
Meleagris. However, many more specimens of varying ages
are needed before the relationships of Proagriocharis to earlier
or later forms can be clearly stated.

BUCKHORN, Grant County, New Mexico. Meleagridinae,
cf. Meleagris. Tibiotarsus (F:AM 10434) — Table 18. This
specimen is from a form much larger than indicated for the
known species of Rhegminornis and Proagriocharis. It is not
qualitatively distinguishable from Meleagris, and it also re-
sembles several species of Meleagris in size.

CLIFTON COUNTRY CLUB, Graham County, Arizona.
Meleagridinae, genus and species indeterminate. Coracoid
(F:AM 10421) — Table 4. This specimen, a humeral end of a
coracoid, is similar in size to that of Proagriocharis kimballen-
sis, if it represents a male. If it represents a female, it is similar
in size to those of M. californica, M. ocellata, and M. cras-
sipes. It resembles that of P . kimballensis in characters 7 and
10, but resembles the coracoids of all turkeys in characters 9,
13, 14 and 15. It differs from that of P. kimballensis in char-
acter 11, and is intermediate between that of P. kimballensis
and other forms in character 17. Although agreeing with M.
crassipes in all of its characters, this specimen cannot justifi-
ably be referred to a known species.

BONE VALLEY (PALMETTO MINE), Polk County, Flor-
ida. Meleagridinae, cf. Meleagris. Distal end of tibiotarsus (UF
21033) — Table 19. No characters apparent other than size.
This is the earliest definite record of a turkey in Florida, pend-
ing further elucidation of the affinities of Rhegminornis. This
tibiotarsus and those from Westmoreland Park, Virginia, and
Buckhorn, New Mexico, suggest a rather widespread occur-
rence of fairly large turkeys in the late Miocene and early
Pliocene.

Pliocene (Hemphillian or Blancan)

UNIVERSITY DRIVE, Orange County, California. Mele-
agris sp. Femur (LACM 64001) — Table 16. No characters ap-
parent other than size. The primary significance of this spec-
imen is that it documents the earliest known occurrence of a
turkey in southern California.

142

Steadman: Turkey Osteology and Paleontology

Late Pliocene (Blancan)

HAILE XVA, Alachua County, Florida. Meleagris sp. Tibio-
tarsus (UF 17545) — Table 18. No characters apparent other
than size. Reported as M. gallopavo by Campbell (1976).

BENSON, Cochise County, Arizona. Meleagris cf. M. pro-
genes. Tarsometatarsus (AMNH 6330, USNM 10551) — Tables
20, 22, Fig. 10; Fig. 5 of Wetmore (1924). USNM 10551 was
referred to the genus Agriocharis by Wetmore (1924), while
Brodkorb (1964b) referred this specimen to A. progenes be-
cause it was “similar in size and position of the spur core”
( 1 964b: 2 25). I have also examined a nearly complete tarso-
metatarsus from this site in which the spur core is broken.
These specimens agree closely in size only with M. progenes
from Rexroad, Kansas, and with M. crassipes. I refer them
only tentatively to M. progenes because they are qualitatively
separable only from Rhegminornis calobates (characters 14,
19) and M. crassipes (characters 5, 8, 16).

CITA CANYON, Randall County, Texas. Meleagris leopoldi.
Tibiotarsus (PPHM 3174). Tarsometatarsus (PPHM 3169) —
Tables 20, 22, Fig. 10; Fig. 1 of A. Miller and Bowman (1956).
Reported as Parapavo californicus by L. Miller and Johnston
(1937) and as cf. Meleagris by Johnston and Savage (1955),
these specimens were described as a new species, Meleagris
leopoldi, by A. Miller and Bowman (1956). Brodkorb (1964b)
transferred M. leopoldi to Agriocharis on the basis of char-
acters that are not considered herein to be of generic value, as
detailed on pages 139-40 of the Comparative Osteology sec-
tion.

Unfortunately, the tarsometatarsus is the only diagnostic
element of M. leopoldi known from Cita Canyon. Meleagris
leopoldi resembles rather closely all forms except Rhegminor-
nis, M. ocellata, andM. crassipes, being closest to M. progenes
(Table 3). The tarsometatarsus of M. leopoldi is consistently
larger than that of M progenes, although some degree of over-
lap would probably be seen if larger samples of each form
were available. Potentially significant differences between M.
leopoldi and M. progenes (specimens from both Rexroad and
Benson considered) are seen in the B/E, E/A, and G/A ratios
(Table 22). Meleagris leopoldi resembles specimens from Inglis
in every measurement (Table 20) except A and E, the first of
which can be extremely variable (note, for example, the range
of all Recent specimens ofM. gallopavo in Table 20). Meleagris
leopoldi differs from M. californica in measurements E, F,
and G (Table 20) and in the F/A and G/A ratios. Meleagris
leopoldi is consistently smaller than M. gallopavo and differs
in its G/A ratio. Meleagris leopoldi resembles M. ocellata in
all measurements except D, and all ratios of these two forms
are similar. Meleagris leopoldi is larger and of different pro-
portions than M. crassipes.

It is seen from the comparisons above that the Cita Canyon
specimen closely resembles both qualitatively and quantita-
tively only the Inglis specimens, although its high degree of
qualitative similarity with M. progenes also suggests close af-
finities to that form. Considering the amount of variation seen
in M. gallopavo in Table 20, it is quite likely that, if large
samples of turkeys were available from Benson, Cita Canyon,
and Rexroad, a case could be made for their conspecificity.

REXROAD, Meade County, Kansas. Meleagris progenes. All
specimens have UMMP numbers. Premaxilla (31052). Man-

dible (47783). Sternum (31039). Coracoid (20940) — Table 4.
Scapula (45930, 45965) — Table 7, Figure 9. Ulnare (48109).
Carpometacarpus (20941, 48188) — Tables 14, 15. Femur
(45912) — Table 16. Tibiotarsus (45970, 48191) — Table 19.
Tarsometatarsus (31034, 48189) — Tables 20, 21, Figures 10,
14; PI. I of Brodkorb (1964b). This species was described as
Agriocharis progenes by Brodkorb (1964b) because of its sup-
posed greater similarity to M. (Agriocharis) ocellata than to
M. gallopavo. However, M. progenes shows no greater affinity
to one of the living species than to another (Table 2). Never-
theless it is more similar to them than to either Rhegminornis
or Proagriocliaris, which supports its inclusion in the genus
Meleagris. The apparent similarity between M. progenes and
M. leopoldi seen in Table 3 is based only on the tarsometa-
tarsus and thus may not be an accurate reflection of affinities
(compare Table 2 to Table 3). M. progenes is of a size that
may be distinct from all other forms except M. crassipes, to
which it does not show enough qualitative resemblance to sug-
gest a close relationship.

Early Pleistocene (Irvingtonian)

GILLILAND, Knox County, Texas. Meleagridinae, genus
and species indeterminate. Coracoid (PB unnumbered) — Table
5. Femur (UMMP 39387) — Table 17; PI. I of Brodkorb
(1964b). No characters other than size are apparent in the
badly crushed coracoid from the Bruce Burnett Ranch. The
femur from Rattlesnake Point, reported as Agriocharis sp. in
Hibbard (1960), was referred to M. (. Agriocharis ) anza by
Brodkorb ( 1 964b: 2 28) on the basis of “agreement in geologic
horizon, general size, and marked expansion of the shaft of
the femur (shaft of humerus expanded in type of A. anza)." I
have not examined this specimen, but do not consider it safely
referable to M. anza for the following reasons: (1) chronological
agreement is not a sound basis for biological relationship; (2)
agreement in size is not guaranteed because of the uncertainty
of the sex of the specimens involved; (3) a mediolateral expan-
sion of the shaft of the femur is not known to be related to
such an expansion in the humerus (in the case of the Vallecito
Creek specimen, the expansion is due to crushing); (4) as noted
by Brodkorb (1964b), no femur is known from Vallecito Creek,
making direct comparison of specimens impossible. As neither
the coracoid nor the femur from Gilliland are large enough to
be safely referable to Meleagris, no generic allocation is made.

INGLIS IA, Citrus County, Florida. Meleagris cf. M. leopoldi
or M. anza. All specimens have UF numbers. Numbers in
parentheses are numbers of specimens examined. Individual
catalogue numbers available on request. Cranium (9). Pre-
maxilla (1). Quadrate (8). Mandible (7). Basihyal (1). Axis (5).
Cervical Vertebra (128). Sixth Thoracic Vertebra (12). Fused
Thoracic Vertebrae (24). Synsacrum (41). Caudal Vertebra (6).
Pygostyle (3). Vertebral Rib (32). Sternal Rib (14). Sternum
(26). Furcula (5) — Figure 8. Coracoid (67) — Tables 4, 5. Scap-
ula (34) — Tables 8, 9, Fig. 9. Humerus (94) — Tables 12, 13.
Ulna (73) — Tables 10, 11. Radius (46) — Tables 12, 13. Ulnare

(10) . Radiale (8). Carpometacarpus (49) — Tables 14, 15. Pollex

(11) . Manus Digit II, Phalanx I (24). Manus Digit II, Phalanx
II (10). Manus Digit III (4). Pelvis (31). Femur (81) — Tables
16, 17. Patella (1). Tibiotarsus (93) — Tables 18, 19. Fibula
(28). Tarsometatarsus (108) — Tables 20-22, Figs. 10, 12, 14.
Hallux (6). Pes Digit I, Phalanx I (6). Pes Digit II, Phalanx I

Steadman: Turkey Osteology and Paleontology

143

(13). Pes Digit II, Phalanx II (13). Pes Digit III, Phalanx I (24).
Pes Digit III, Phalanx II (13). Pes Digit III, Phalanx III (11).
Pes Digit IV, Phalanx I (13). Pes Digit IV, Phalanges II, III,
IV (28). Ungual Phalanx (16). Misc. Pedal Phalanx (3). Known
from about 1240 specimens, this is the earliest turkey for which
any appreciation of individual variation can be attained. Pre-
viously (Steadman 1975), and with much hesitation, I referred
the Inglis fossils to M. anza because the humeri could not be
distinguished from the poorly preserved holotype of M. anza.
However, as already noted, the tarsometatarsus of M. leopoldi
also bears much resemblance to those from Inglis, and the two
populations may represent a single species. Again I am faced
with the problem of referring a large series of specimens to a
species known essentially from one element. Meleagris leopoldi
is based on a tarsometatarsus (unknown in M. anza), a very
diagnostic and commonly preserved element of turkeys. As the
tarsometatarsus is spurred, we can be reasonably sure that the
known specimens of M. leopoldi represent males, thereby also
providing an idea of the size of the females. That a strong
sexual dimorphism in size was developed even before the time
of M. leopoldi is shown by the carpometacarpus and tarso-
metatarsus of M. progenes. Although the holotype humerus of
M. anza equals those of females from Inglis in size, the pos-
sibility exists that it actually represents a male of a very small
species. Referral of the Inglis specimens to either M. leopoldi
or M. anza is somewhat tentative, each species being based
on specimens of only one skeletal element. However, I consider
this action to be better than describing the turkey from Inglis
as a new species, thereby adding another name of dubious
validity to the already excessively long list of specific names
of turkeys. I believe that future discoveries of turkeys of ages
similar to those of Cita Canyon, Inglis, and Vallecito Creek
may show that the turkeys from these sites represent one lin-
eage that slowly increased in size through time. In this case,
the delimitation of species would be arbitrary, and M. leopoldi
and M. anza might even be regarded as conspecific. That a
single species of turkey could have such a wide geographical
range is not at all surprising when one considers the present
range of M. gallopavo and the diverse types of habitat that it
occupies. If new fossils prove that the turkeys from Inglis are
distinct from those of Cita Canyon or Vallecito Creek, appro-
priate nomenclatural steps can be taken at that time.

Closest affinities of the turkey from Inglis appear to be with
fossils from Coleman; and it is least similar to Rhegminornis
calobates, Meleagris gallopavo, and M. ocellata (Table 2). As
shown in Table 3, the greatest resemblance is with M. progenes
and fossils from Coleman, with only R. calobates being very
dissimilar. The Inglis turkey is larger in size than R. calobates,
P. kimballensis , M. progenes, M. ocellata, and M. crassipes',
equal to or slightly larger than M. leopoldi or M. californica;
slightly smaller than the population from Coleman; smaller
than M. gallopavo.

Distinctive ratios of the Inglis fossils in Table 22 include the
following: G/A is less than in all other forms except M. leo-
poldi', J/A is greater than in all forms except Coleman speci-
mens and M. ocellata -, K/A is less than in all forms except M.
crassipes ; P/A is less than in all forms except M. g. osceola.

Were it not for the series of intermediate fossils from Cole-
man, the degree of difference between the Inglis fossils and
M. gallopavo apparent from Table 2 would suggest a possible
sudden replacement of the early leopoldi-anza- like form by the

larger gallopavo at some point in the Pleistocene. It now ap-
pears that the Florida peninsula, if not much of the southern
portion of North America, has been occupied since at least
Blancan times by a series of populations of turkeys that in-
creased slightly in size through time.

The turkey from Inglis possesses certain features, such as
the stout furcula and the non-pneumatic scapula, that are thus
far unknown in any later forms. The evolutionary implications
of these structures will be discussed in the section on evolution.

VALLECITO CREEK, San Diego County, California. Mele-
agris anza. All are LACM 3753. Sternum. Humerus — Table
9; PI. Ill in Howard (1963). Ulna. Synsacrum. This material
was described as a new species, Agriocliaris anza, by Howard
(1963), with a humerus as the holotype. Unfortunately, speci-
mens other than the humerus are damaged beyond usefulness.
The holotype humerus is crushed more severely on the anconal
side than on the palmar side. I consider its broad, flat shaft
(Table 9: C, D) and deep brachial depression (character 9) to
be due to this crushing, a possibility recognized by Howard
(1963). I also regard the curvature of the shaft (character 8) as
an artificial condition. Therefore, characters 8 and 9 are not
considered in compiling Table 2, in which the humerus from
Vallecito Creek is seen to be slightly closer to Inglis specimens
and those of M. gallopavo than to those of other forms. Based
upon only seven characters of just one element, these figures
may be of little value. The holotype of M. anza differs from
other forms in the following qualitative characters: Coleman
specimens — character 12; M. gallopavo — character 12; M. cal-
ifornica— characters 10, 12; M. ocellata — characters 5, 7, 12;
M. crassipes — characters 7, 12. It cannot be distinguished
from humeri from Inglis by any of these characters. In refer-
ring this humerus to Agriocliaris instead of Meleagris, How-
ard’s (1963) characters included the shape of the external con-
dyle (character 13), now known to be variable and therefore
of little value, and the orientation of the attachment of M.
pronator brevis (character 10), which varies in some forms,
but appears to be consistent in others. This single specimen
cannot provide any indication of individual variation of its
population.

Except for the width and depth of the shaft, which are
altered by crushing, this specimen resembles in size the females
of several different forms. If the material from Vallecito Creek
had not already been the basis for the description of a new
taxon, I would have called the material “Meleagridinae, cf.
Meleagris.” As this material cannot be clearly distinguished
quantitatively from several species of Meleagris, its tentative
referral to Meleagris would seem justified. However, as the
available specimens from Vallecito Creek represent the type
material of Meleagris anza, instead of regarding this name as
a nomen dubium, I retain it with the hope that more diagnostic
specimens will be recovered at this site in the future. An ef-
fective suite of characters simply cannot be based on poorly
preserved material of only one or several elements, especially
when dealing with such intraspecifically variable organisms as
turkeys. Diagnoses based on single specimens of variable or-
ganisms cannot account for individual variation and will al-
most surely be altered by the discovery of new specimens. The
old idea that avian fossils are very rare has promoted descrip-
tions of “species” based on material that is often at best indic-
ative only of a family or genus, thereby forcing future workers

144

Steadman: Turkey Osteology and Paleontology

either to tentatively refer similar new fossil material to the
poorly defined taxon, or else not recognize the old taxon. Only
through the discovery of new, more diagnostic fossils at Val-
lecito Creek can the true affinities of M. anza be learned.

PORT KENNEDY CAVE, Montgomery County, Pennsyl-
vania. Meleagridinae cf. Meleagris. Carpometacarpus (ANSP
165) — Table 14. Without naming the element(s) involved,
Wheatley reported “a turkey” (187 la:237) and “ Meleagris

?” (187 1 b:385 ) from this site. Mercer (1899:280) included

“ Meleagris altus, leg bone with spur” in his faunal list of Port
Kennedy Cave, noting that the single specimen was excavated
by Wheatley in 1871. The location of this specimen is unknown
to me. Wetmore (1931b) listed this record as M. superba ( =
alta) as did Brodkorb (1964a), under the name M. alta. The
only fossil of a turkey from this site that I have examined is
a previously unreported complete carpometacarpus, which
differs from those of Inglis, Coleman, and M. gallopavo in
having metacarpal I less protrudent anteriorly, although sev-
eral specimens from Inglis closely approach its condition. It is
larger than females of any species, and, if it represents a male,
it is the size of those of Inglis, Coleman, or M. californica,
being slightly smaller than in M. gallopavo.

The exact affinities of this specimen cannot be determined
from the available data. Although resembling Meleagris in
size, it differs from this genus in its small metacarpal I. Generic
characters of the carpometacarpus of turkeys are at present
undetermined, as this element is unknown in Rhegminornis or
Proagriocharis.

HAILE XVI A, Alachua County, Florida. Meleagris sp. Scap-
ula (UF 22083, 22084) — Table 7. Humerus (UF 22081,
22082) — Table 8. Femur (UF 22086). Tibiotarsus (UF 22085).
Pedal Phalanx (UF 22087). The scapulae have a pneumatic
foramen, thus resembling post-Inglis forms. They agree in size
with the Coleman fossils more than with the Inglis form or M.
g. osceola. The distal end of a humerus shows no characters
other than size, in which it resembles Inglis specimens more
than the Coleman fossils or M. gallopavo.

These fossils are referred to Meleagris, but not to any
species, on the basis of the above characters. These specimens
suggest a form of turkey between those of Inglis and Coleman
if one assumes that the same lineage is involved. A refinement
of the age of this site is expected once the mammalian fauna
is studied further. This will help to pinpoint the time of the
development of the pneumatic foramen in the scapula.

WILLISTON, Levy County, Florida. Meleagris sp. Femur
(PB 2321). No characters apparent other than size. Reported
asM. gallopavo by Holman (1959:5) and Brodkorb (1964a:335).

COLEMAN IIA, Sumter County, Florida. Meleagris sp. (in-
termediate between leopoldi-anza and gallopavo). All speci-
mens have UF numbers. Numbers in parentheses are numbers
of specimens examined of each element. Individual catalogue
numbers available on request. Cranium (3). Premaxilla (1).
Axis (2). Cervical Vertebra (53). Sixth Thoracic Vertebra (4).
Fused Thoracic Vertebra (1). Synsacrum (7). Sternal Rib (2).
Sternum (12). Coracoid (18) — Tables 4, 5. Scapula (6) — Tables
6, 7, Fig. 9. Humerus (37) — Tables 8, 9. Ulna (23) — Tables
10, 11. Radius (9) — Tables 12, 13. Radiale (1). Carpometa-
carpus (12) — Tables 14, 15. Manus Digit II, Phalanx I (2).
Manus Digit II, Phalanx II (2). Pelvis (7). Femur (27) — Tables

16, 17. Tibiotarsus (33) — Tables 18, 19. Tarsometatarsus
(35) — Tables 20-22, Figs. 10, 12, 14. Pes Digit II, Phalanx I
(4). Pes Digit III, Phalanx I (8). Pes Digit III, Phalanx II (2).
Pes Digit III, Phalanx III (1). Pes Digit IV, Phalanx I (1).
These 320 fossils provide a link between the older M. leopoldi-
awsa-like forms and the younger M. gallopavo and M. califor-
nica. I noted previously (Steadman 1975) an intermediateness
between the Inglis fossils and later forms and referred the
Coleman fossils to M . anza, thereby attempting to stress their
similarity to the Inglis fossils. Although showing a slightly
greater overall resemblance to the Inglis specimens than to
those of M. gallopavo or M. californica, the Coleman fossils
are not herein regarded as conspecific with those from Inglis
since a distinct change, the attainment of a pneumatic foramen
in the scapula, occurred in the interval between deposition of
the Inglis and Coleman faunas. In the absence of the Coleman
specimens, the relatively low amount of similarity of the spec-
imens from Inglis to M. gallopavo (Table 2) would suggest a
more distant relationship than is proposed herein. The Cole-
man turkey shows less similarity toM. ocellata or M. crassipes
than to M. progenes, M. gallopavo, or M. californica (Table
2). Its lowest similarity is with Rhegminornis calobates and
Proagriocharis kimballensis .

The Coleman specimens are definitely larger in overall size
than those of Rhegminornis calobates, Proagriocharis kimbal-
lensis, M. progenes, or M. crassipes\ almost always larger than
those of M. ocellata\ usually larger than Inglis specimens or
M. californica, overlapping more with the former than with
the latter; and usually smaller than those of M. gallopavo,
often with overlap.

Distinctive ratios of the Coleman specimens in Table 22 are
as follows: B/E is greater than in Inglis specimens; J/A is great-
er than in all others except Inglis specimens and M. ocellata.
A general trend in these ratios is the relative stoutness of the
Coleman specimens as compared to M. g. osceola. In every
case, however, specimens from Rancholabrean sites in Florida
either more strongly resemble Coleman specimens than those
of M. g. osceola, or are intermediate, providing further evi-
dence that the Coleman form was directly ancestral to M.
gallopavo.

SANTA FE RIVER IIA, Gilchrist County, Florida. Meleagris
cf. M. gallopavo. Ulna (UF 14930) — Table 10. Femur (UF
14928) — Table 16. Tibiotarsus (UF 22078) — Table 18. Tarso-
metatarsus (UF 22076, 22077) — Tables 20, 22. These specimens
resemble specimens of late Rancholabrean and Recent M. gallo-
pavo more closely than those of Inglis or Coleman in size and
H/C ratio (Table 22). However, the tarsometatarsus differs
from that of M. gallopavo in character 4.

Late Pleistocene (Rancholabrean)

RANCHO LA BREA, Los Angeles County, California. Mele-
agris californica. All specimens from LACM. Numbers in
parentheses are numbers of specimens of each element exam-
ined. Individual catalogue numbers available on request.
Cranium (2). Premaxilla (3). Mandible (2). Fused Thoracic
Vertebrae (1). Synsacral Vertebra (1). Pvgostyle (1). Ster-
num(4). Furcula(4). Coracoid (56) — Tables4, 5. Scapula(51) —
Tables 6, 7, Fig. 9. Humerus (65) — Tables 8, 9. Ulna (64) —
Tables 10, 11. Radius (74) — Tables 12, 13. Carpometacarpus
(64) — Tables 14, 15. Pelvis (4). Femur (105) — Tables 16,

Steadman: Turkey Osteology and Paleontology

145

17. Tibiotarsus (74) — Tables 18, 19. Fibula (4). Tarso-
metatarsus (82) — Tables 20-22, Figs. 12, 14; Plate 25

of L. Miller (1909); Figs. 44, 45 of L. Miller (1940). All
major elements illustrated in Pis. 1-13 of Floward (1927).

L. Miller (1909) described this species as Pavo calif ornicus, a
phasianine. L. Miller (1916a) later erected the genus Parapavo
for these specimens, which he regarded (L. Miller 1916a,
1916b, 1925) as being intermediate between Pavo and Mele-
agris ocellata, and more closely related to these two species
than to M. gallopavo. Interestingly, L. Miller (1912:78) in-
cluded “ Meleagris ?” alongside Pavo californicus in a list of
birds from Rancho La Brea, with no mention of the elements
involved. Wetmore (1924), referring only to a tarsometatarsus,
regarded Parapavo not as a peacock but as a turkey interme-
diate between Agriocharis ( ocellata ) and Meleagris {gallopavo ),
although closer to the latter. After a very thorough study of
Parapavo, Howard (1927) agreed with Wetmore (1924), ex-
cept for considering Parapavo more closely related to Agri-
ocharis than to Meleagris. Although Howard (1927) considered
Parapavo to be generically separable from Meleagris gallopavo
and M. ocellata, only in the fibula did she find M. californica
to be consistently generically distinct from both M. gallopavo
and M. ocellata, although noting (Howard 1927:24) “Owing
to the variability in the fibula of Pavo and Agriocharis, little
import can be attached to this element.” Howard (1928), with
a study of the premaxilla, gave support to her previous con-
clusion that Parapavo is different at the generic level from
other turkeys. Sushkin (1928), comparing the coracoid, car-
pometacarpus, and tarsometatarsus of M. californica with
those of M. ocellata and Pavo, concluded that M. californica
is not related to Pavo but is sufficiently similar to M. ocellata
that it could be congeneric. Wetmore (1931b), L. Miller (1942)
and Brodkorb (1964a, b) all recognized Parapavo, while A.
Miller and Bowman (1956) found no basis for a generic dis-
tinction in the tarsometatarsi of Meleagris and Parapavo.

Analysis of the various supposed generic characters of the
turkey from Rancho La Brea is provided in the “Comparative
Osteology” section, where M. californica is seen to lack char-
acters that warrant generic separation from M. gallopavo or

M. ocellata. Meleagris californica is not an unusual form at
all, but simply another species of Meleagris with rather close
affinities to other species of the genus (Tables 2 and 3), which
it also resembles in size. Thus, I consider the genus Parapavo
L. Miller to be a synonym of Meleagris Linnaeus.

IMPERIAL HIGHWAY (LACM 1052), Orange County, Cal-
ifornia. Meleagris sp. Radius (LACM 1052/2009) — Table 12.
No characters apparent other than size. Referred to Parapavo
californicus by Howard (1936) on the basis of characters herein
regarded as inconsistent (see page 135). Also reported as P.
californicus by L. Miller (1942), Brodkorb (1964a), and W.
Miller (1971). L. Miller (1942) and Brodkorb (1964a) refer to
this site as La Habra, California.

CARPINTERIA, Santa Barbara County, California. Mele-
agris californica. All specimens are LACM. Numbers in pa-
rentheses are numbers of specimens examined of each element.
Individual catalogue numbers available on request. Coracoid
(17) — Tables 4, 5. Scapula (10) — Tables 6, 7. Humerus (17) —
Tables 8, 9. Ulna (22) — Tables 10, 11. Radius (16) — Tables
12, 13. Carpometacarpus (17) — Tables 14, 15. Femur (16) —
Tables 16, 17. Tibiotarsus (21) — Tables 18, 19. Tarsometa-

tarsus (16) — Tables 20-22. There are no significant qualitative
or quantitative differences between these specimens and those
ofM. californica from Rancho La Brea. Reported as Parapavo
californicus by L. Miller (1927, 1942), Wetmore (1931b), and
Brodkorb (1964a).

WORKMAN AND ALHAMBRA STREETS (LACM 1023),
Los Angeles County, California. Meleagridinae, cf. Meleagris.
Tibiotarsus (LACM 982) — Table 19. No characters apparent
other than size. Howard (1936:250) found that this specimen
“may be assignable to Parapavo, but unfortunately they do
not possess any diagnostic generic characters by which to make
definite identification.” Reported as Parapavo californicus by

L. Miller (1942) and Brodkorb (1964a), and as cf. Parapavo
by W. Miller (1971:55), who called this site “Workman and
Alameda Sts.”

LA MIRADA, Los Angeles County, California. Meleagris cf.

M. californica. Radius (LACM 2009) — Table 12. No charac-
ters apparent other than size. W. Miller (1971) referred a par-
tial coracoid (not examined by me) and the radius from this
site to Parapavo californicus .

POTTER CREEK CAVE, Shasta County, California. Mele-
agris sp. Coracoid (UCMP 1055/8368) — Table 5. Humerus
(UCMP 1055/114545) — Table 9. Qualitatively separable from
all species except M. gallopavo and M. californica, each of
these specimens may also resemble either of the above two
species in size. Reported as Meleagris sp. by L. Miller (1911);
as “referable either to Parapavo or to Meleagris ” by L. Miller
(1925:67); and as Parapavo californicus by Brodkorb (1964a).
Mention is made here of the report of ‘ Meleagris sp.” from
Hawver Cave, Eldorado County, California, by L. Miller
(1912:75). I have not located the specimen involved, nor were
any turkeys reported from Hawver Cave by L. Miller (1911)
or Brodkorb (1964a).

AMERICAN FALLS, Power County, Idaho. Meleagris gal-
lopavo. Tarsometatarsus (ISUM 1736) — Tables 20, 22. Re-
ported also as M. gallopavo by Hopkins et al. (1969), this
specimen agrees with tarsometatarsi of M. gallopavo in size,
proportions, and in all characters except no. 4.

PAPAGO SPRINGS CAVE, Santa Cruz County, Arizona.
Meleagris sp. Humerus (AMNH 8683, 8687) — Table 9. Femur
(AMNH 8684, 8685) — Table 17. These specimens are indistin-
guishable from females of M. gallopavo or males of M. cras-
sipes. North Papago Cave, which is an extension of Papago
Springs Cave, contained a tarsometatarsus of M. crassipes
(Rea this vol.).

ARIZPE, Sonora, Mexico. Meleagris cf. M. gallopavo. Hu-
merus (AMNH 6823) — Table 8. No character apparent other
than size. Reported as M. gallopavo by Cracraft (1968). Also
regarded as Meleagris cf. M. gallopavo by Rea (this vol.).

BLTRNET CAVE, Eddy County, New Mexico. Meleagris cf.
M. gallopavo. Humerus (ANSP 14161) — Table 8. Carpometa-
carpus (ANSP 13495) — Table 15. Femur (ANSP 14134) — Ta-
ble 17. Tibiotarsus (ANSP 14133) — Table 19. No characters
apparent other than size. Reported as M. gallopavo by Schultz
and Howard (1935). Although the humerus is too large forM.
crassipes, a tarsometatarsus referable to M. crassipes has also
been recovered from Burnet Cave (Rea this vol.). The strati-
graphic positions of the various fossils of turkeys from Burnet
Cave are unknown.

146

Steadman: Turkey Osteology and Paleontology

HOWELL’S RIDGE CAVE (Zeller-Howard pit), Grant
County, New Mexico. (Includes both Pleistocene and Holo-
cene strata.) Meleagris sp. Scapula (LACM 33889) — Table 7.
Ulna (LACM 33891, 33892) — Table 11. M. crassipes. Cora-
coid (LACM 33890). The foramen of the scapula is very small
and located nearer to the glenoid facet than in all other forms.
Howard (1962:242) recorded these four specimens, saying that
the scapula “resembles Meleagris more closely than Parapavo
and is tentatively assigned to M. gallopavo." Brodkorb (1964a)
listed M. gallopavo from this site. Neither the size nor quali-
tative characters of the ulnae or the scapula are diagnostic
enough to refer these specimens to any species. The distinctive
scapula could possibly represent M. crassipes, the scapula of
which is not yet known. Amadeo M. Rea and I have jointly
studied the coracoid and agree that it is of M. crassipes. Van
Devender and Worthington (1977) present an analysis of the
stratigraphy and chronology of Howell’s Ridge Cave, which
also contains, but not in the Zeller-Howard pit, a tarsometar-
sus of M. crassipes (Rea this vol.).

SAN JOSECITO CAVE, Nuevo Leon, Mexico. Meleagris
crassipes. All specimens are LACM or UCMP. Numbers in
parentheses are numbers of specimens examined of each ele-
ment. Individual catalogue numbers available on request.
Coracoid (4) — Tables 4, 5. Humerus (8) — Tables 8, 9. Ulna
(10) — Tables 10, 11. Radius (2) — Tables 12, 13. Carpometa-
carpus(lO) — Tables 14, 15. Lemur (10) — Tables 16, 17. Tibio-
tarsus (6) — Tables 18, 19. Tarsometatarsus (16) — Tables 20-
22, Pigs. 10, 14; Ligs. 44, 45 of L. Miller (1940). Described as
Meleagris crassipes by L. Miller (1940), a name also used by
L. Miller (1943) and A. Miller and Bowman (1956). Brodkorb
(1964b), solely on the basis of characters of the tarsometatarsus
that are largely refuted in the comparative osteology section
of this paper, placed M. crassipes in Agriocharis. This generic
reallocation was followed by Brodkorb (1964a) and Martin and
Tate (1970). However, the “Great-footed Turkey” of San Jo-
secito Cave (L. Miller 1943) is herein reassigned to Meleagris,
though admittedly it deviates somewhat from other members
of the genus, both in its small size and in its qualitative char-
acters. It resembles the various forms of Meleagris more than
Rhegminornis or Proagriocharis (Tables 2 and 3). If generic
distinction of M. crassipes were warranted, the name Agri-
ocharis is inapplicable because the type is M. ocellata, which
is definitely referable to Meleagris. This is further supported
by the relatively large qualitative difference between M. cras-
sipes and M. ocellata (Tables 2 and 3).

The carpometacarpus and tarsometatarsus, elements which
are clearly much larger in males than females in other species
of turkeys, exhibit only a gradational change in size in the
specimens of M. crassipes that I have examined (Tables 14,
15, 20, 21 herein; L. Miller 1943:157-158). Other elements in
the fairly large series of turkey fossils from San Josecito Cave
are separable, upon close inspection, into two size classes that
presumably represent males and females. The tarsometatarsus
of M. crassipes is quite distinctive, being very stout relative
to its length (Table 22: B/A, C/A, D/A, L/A, M/A, N/A, P/A).
Also, the stout spur core is located higher on the shaft (Table
22: F/A, G/A) than in other species. If differences approaching
the magnitude of those of the tarsometatarsus were seen in
other elements ofM. crassipes, a case could be made for erect-
ing a new genus for this small turkey, which appears to have
survived well into the Holocene (Rea this vol.).

INGLESIDE, San Patricio County, Texas. Meleagris gallo-
pavo. Coracoid (TMM 30967-1741) — Table 4. Tibiotarsus
(TMM 30967-1139, 30967-1063B, 30967-1564)— Tables 18,

19. Tarsometatarsus (TMM 30967-1169, 30967-1467) — Tables

20, 22. Reported as M. gallopavo by Feduccia (1973) and Rea
(this vol.), these fossils differ in certain measurements from
Irvingtonian fossils, but agree in every way with Recent spec-
imens of M . gallopavo.

SHEFFIELD GRAVEL PITS, Texas. Meleagris sp. Ulna
(F:AM 11249) — Table 10. No characters apparent other than
size. The age of this deposit is uncertain, but it is probably
late Pleistocene.

NORTH LIBERTY, St. Joseph County, Indiana. Meleagris
sp. Humerus (USNM 17004) — Table 8. No character apparent
other than size (immature). Reported as M. gallopavo by Wet-
more (1945) and Brodkorb (1964a).

CARLISLE CAVE, Cumberland County, Pennsylvania. Me-
leagris cf. M. gallopavo. Humerus (USNM 345074, 345727) —
Table 21. No characters apparent other than size. Reported
as “the wild turkey” by Leidy (1889:2), and as M. gallopavo
by Wetmore (1931b) and Brodkorb (1964a).

FRANKSTOWN CAVE, Blair County, Pennsylvania. Mele-
agris cf. M. gallopavo. All are CM 11053. Plate XVII of Pe-
terson (1926). Humerus — Table 8. Pelvis. Femur — Table 16.
Tibiotarsus — Table 18. Tarsometatarsus. No characters ap-
parent other than size. Reported as “a large species undoubt-
edly belonging to the genus Meleagris" by Holland (1908:232,
1912:751), as M. superba by Peterson (1926) and Wetmore
(1931b), and asM. alia by Brodkorb (1964a). After comparison
with the description of M. superba (Cope 1871) and the illus-
trations of M. superba in Shufeldt (1915), Peterson (1926:254)
said, “The only discrepancy appears to be the slightly shorter
tarsometatarsus of the specimen from Frankstown Cave.

. . . Whether or not M. superba Cope should be accepted as
a valid species cannot be determined from the material avail-
able.” I regard both M. superba and M. alta an synonyms of
M. gallopavo (see below). The scapula reported from Franks-
town Cave by Peterson (1926), seen to be foraminate in his
Plate XVII, could not be located.

MANALAPAN, Monmouth County, New Jersey. Meleagris
gallopavo (includes the synonyms M. alta, M. celer, and M.
superba). Coracoid (AMNH 1443) — Table 4. Scapula (PU
22356) — Table 6. Humerus (?YPM 533-536, not examined by
the author) — Table 8. Ulna (PU 22356) — Table 10. Radius
(AMNH 2341) — Table 12. Femur (AMNH 1443, PU 22356) —
Table 16. Tibiotarsus (AMNH 1443, PU 22356; also 1-3 spec-
imens not examined by the author — PYPM 533-536) — Tables
18, 19. Tarsometatarsus (ANSP 12137; AMNH 2341; also 2
specimens not examined by the author — PYPM 533-536) —
Tables 20-22. Otherwise harmless turkey bones from this site
form the type series of three invalid species, namely Meleagris
altus Marsh (1870b), M. superba Cope (1871), and M. celer
Marsh (1872), all of which I regard as synonyms of M. gallo-
pavo. Based on Marsh’s descriptions, Shufeldt (1897) correctly
synonymized M. alta and M. celer with M . gallopavo, but did
not discuss M. superba. Shufeldt (1913), again criticizing
Marsh for describing a new species on such fragmentary ma-
terial, regarded M. alta to be a synonym of M. superba. Shu-
feldt (1913), after examination of the type tarsometatarsus of

Steadman: Turkey Osteology and Paleontology

147

M. celer, said that the specimen was not of any form of turkey,
but he suggested that it was similar to a heron. Ironically,
Shufeldt himself later (1917) described the tibiotarsus of a tur-
key from Vero, Florida, as a new species of heron (“ Ardea
sellardsi'y. Later, Shufeldt (1915) again considered M. alta to
be a synonym of M. superba, which was (p. 66), “a very tall
turkey indeed” [Shufeldt’s italics]. Peterson (1926) questioned
the validity of M. superba. Wetmore (1931b) recognized M.
celer and M. superba, but did not mention M. alta. A. Miller
and Bowman (1956) noted that M. celer was similar in size to
M. gallopavo and may be synonymous. Howard (1963) sus-
pected thatM. celer was the female of M. superba, but Brod-
korb (1964a) correctly regarded M. superba and M. celer as
subjective synonyms of M. alta Marsh, a treatment which
must be followed regardless of the affinities of M. alta. Martin
and Tate (1970) followed Brodkorb (1964a) in recognizing M.
alta .

A cast of a “type” tarsometatarsus of M. alta Marsh (ANSP
12137; for history of this specimen, see Olson and Gillette 1978)
reveals nothing to distinguish it from M. gallopavo. Marsh
(1872:260) reported the absence of a bridge over the inner canal
of the hypotarsus in M. alta, saying “The bridge, if it existed
in an ossified condition, was less massive than in the wild
turkey; otherwise some portions of it would have been pre-
served in the present specimens.” However, the absence of the
bridge is simply due to breakage of the plantar portion of the
hypotarsus, thus exposing the inner canal.

Marsh (1872) also described the tarsometatarsus of M. alta
as being slender and elongated. This cannot be determined
from the available cast, but the length reported by Marsh is
within the range of both Pleistocene and Recent specimens of
M. gallopavo. The supposedly lower placement of the spur
core of M. alta reported by Marsh is within the range of M.
gallopavo (Table 22: G/A). The large pneumatic foramen in
the coracoid ofM. superba reported by Cope (1871) is probably
due to individual variation (see character 2 of the coracoid.)

My measurements of the bones of M. alta never exceed the
length of one or more modern specimens of M. gallopavo, and,
in only one case (Table 4) are the measurements of Cope (1871),
Marsh (1872), or Shufeldt (1915) greater than those of modern
M. gallopavo. As their measurements often tend to be greater
than my measurements of apparently the same bones, and are
also often only estimations, this slight difference is of little or
no significance.

Turkey fossils from Manalapan in the Princeton University
Museum of Natural History, which are probably some of the
same specimens used by Cope and Marsh, are not separable
from M. gallopavo. These 12 specimens were in a tray that
contained the type femur of Grus proavus Marsh from Mana-
lapan, and an old pencilled slip that reads “Dr. Thompson
probably.” Dr. C.C. Thompson collected the specimens from
which M. alta, M. superba, and M. celer were described. I
have examined two lots of turkey fossils in the collection of
the American Museum of Natural History with preservation
identical to that of the Princeton specimens (dark brown, brit-
tle, poorly mineralized). One lot (AMNH 2341), labelled “Bird
bones.- — Undetermined. Probably of as many individuals.
Pleistocene. Freehold? New Jersey,” contained two tarsometa-
tarsi and a radius of a turkey; the other (AMNH 1443) is
labelled uMeleagris superbus Cope. Femur, tibia, coracoid.
Post-Pliocene. Freehold, Monmouth Co., N.J. Coll. Dr. C.C.
Thompson.” These specimens resemble M. gallopavo both

qualitatively and quantitatively as do those from Princeton.
M. alta Marsh, M. superba Cope, andM. celer Marsh are thus
placed in synonymy of M. gallopavo Linnaeus because of their
lack of distinguishing features.

Brodkorb (1964a) reported two other localities for M. alta,
both of which are discussed more fully above. A carpometa-
carpus from Port Kennedy Cave, Pennsylvania, is herein re-
ported as Meleagridinae, cf. Meleagris, and the fossils from
Frankstown Cave, Pennsylvania, are regarded herein as Mele-
agris cf. M. gallopavo.

The following 14 Rancholabrean localities in Florida have
mammalian faunas that have made possible their correlation,
as summarized by Webb (1974:13). In no case have any mele-
agridine fossils contradicted Webb’s correlation, thus further
supporting his proposed sequence of local faunas. The names
of sites that are usually followed by IA, IIA, etc. are not so
designated in certain cases because such information did not
accompany the specimens and could not be derived from the
associated data.

BRADENTON, Manatee County. Meleagris sp. Tarsometa-
tarsus (USNM 12074) — Table 20. No characters apparent oth-
er than size. Reported as M. gallopavo by Wetmore (1931a, b)
and Brodkorb (1964a).

ROCK SPRING, Orange County. Meleagris cf. M. gallo-
pavo. Coracoid (PB 7917) — Table 5. Humerus (PB 7918) —
Table 9. Radius (PB 7919-7921) — Tables 12, 13. Carpome-
tacarpus (PB 1481, 7922) — Table 14. Tibiotarsus (PB 7923) —
Table 18. These fossils agree closely with Recent specimens of
M. gallopavo, but differ from Inglis and Coleman specimens
only in character 1 1 of the coracoid. Reported as M. gallopavo
by Woolfenden (1959) and Brodkorb (1964a).

WITHLACOOCHEE RIVER, Citrus County. Meleagris sp.
Tarsometatarsus (PB 8444) — Tables 20, 22. No characters ap-
parent other than size.

HAILE VIIA, Alachua County. Meleagris sp. Carpometacar-
pus (PB 8497) — Table 15. Femur (PB 1263) — Table 17. This
carpometacarpus differs from that of all other species in having
a larger pisiform process. Otherwise it is similar to specimens
from Coleman and of Recent M. gallopavo. Reported as M.
gallopavo by Brodkorb (1964a).

REDDICK IB, Marion County. Meleagris cf. M. gallopavo.
Coracoid (PB 705, 9012) — Table 5. Scapula (PB 1028> — Table
7. Humerus (PB 1222) — Table 8. Ulna (PB 1264) — Table 10.
Radius (PB 846) — Table 12. Carpometacarpus (PB 1756,
9013-9015, UF 2437) — ' Tables 14, 15. Femur (PB 1094, 9017,
9018, UF 2437) — Tables 16, 17. Tibiotarsus (PB 291, 699,
1123) — Table 18. Tarsometatarsus (PB 292, 494, 685) — Tables
20, 22. This series of specimens falls within the range of quan-
titative and qualitative variation of Recent M. gallopavo with
two exceptions: the humerus resembles specimens from Inglis
in character 3; the tarsometatarsus resembles specimens of M.
crassipes in character 3. These differences are outweighed by
the similarity in size of these fossils, especially elements of the
leg, to Pleistocene and Recent specimens of M. gallopavo. Re-
ported asM. gallopavo by Brodkorb (1957, 1964a) and Hamon
(1964). Although listed herein as “Reddick IB,” some of these
fossils are labelled simply “Reddick” or “Reddick I” and may
be from another of the several localities within the Reddick
limestone mine.

148

Steadman: Turkey Osteology and Paleontology

MELBOURNE, Brevard County. Meleagris cf. M. gallopavo.
Scapula (USNM 17035, 17036) — Table 7. Humerus (USNM
12123) — Table 8. Ulna (USNM 12113). Carpometacarpus
(USNM 17028, 17038, 17039) — Tables 14, 15. Tibiotarsus
(USNM 121 14, 17034). Tarsometatarsus (USNM 12108,
17037, 17040) — Table 20. No characters apparent other than
size. Reported as M. gallopavo by Wetmore (1931a, b) and
Brodkorb (1964a).

ARREDONDO, Alachua County. Meleagris cf. M. gallopavo.
Humerus (PB 1630) — Table 8. No characters apparent other
than size. Reported as M. gallopavo by Brodkorb (1959,
1964a).

SABERTOOTH CAVE, Citrus County. Meleagris cf. M. gal-
lopavo. Tarsometatarsus (USNM 12188) — Tables 20, 22. No
characters apparent other than size. Reported as M. gallopavo
by Wetmore (1931a, b) and Brodkorb (1964a).

AUCILLA RIVER IA, Jefferson County. Meleagris gallopavo.
Humerus (USNM 209602) — Table 8. Radius (USNM 209709) —
Table 12. Tibiotarsus (USNM 209607, 209705) — Tables 18,
19. Tarsometatarsus (USNM 209605, 209609) — Tables 20, 22.
Referral of these specimens to M. gallopavo is based on size
and proportions of the tarsometatarsus, in which they differ
from those of Inglis or Coleman.

ICHETUCKNEE RIVER, Columbia County. Meleagris
gallopavo. All specimens have either UF or PB numbers.
Numbers in parentheses are numbers of specimens examined.
Individual catalogue numbers available on request. Sternum
(1). Coracoid (11) — Tables 4, 5. Scapula (1) — Table 7. Hu-
merus (1 1) — Tables 8, 9. Ulna (9) — Tables 10, 11. Radius (1) —
Table 12. Carpometacarpus (14) — Tables 14, 15. Femur (8) —
Tables 16, 17. Tibiotarsus (37) — Tables 18, 19. Tarsometa-
tarsus (44) — Tables 20-22. This large series of fossils generally
agrees very closely with specimens of Recent M . gallopavo and
differs in many ways from the fossils from Inglis or Coleman.
They usually resemble M. g. silvestris in size and proportions
more than M. g. osceola. Reported as M. gallopavo by Wet-
more (1931a, b), McCoy (1963), Brodkorb (1964a), and Camp-
bell (this vol ).

KENDRICK IA, Marion County. Meleagris sp. Tibiotarsus
(PB 1410) — Table 19. Tarsometatarsus (PB 1409) — Table 20.
These fragmentary specimens resemble M. gallopavo except in
character 4 of the tarsometatarsus. Reported as M. gallopavo
by Brodkorb (1964a).

VERO, Indian River County. Meleagris cf. M. gallopavo. Car-
pometacarpus (PB 8467) — Table 15. No characters apparent
other than size. Shufeldt (1917) unknowingly was the first to
report a turkey from Vero, describing the worn distal half of
a meleagridine tibiotarsus as Ardea sellardsi, a supposed new
species of heron. Wetmore (1931a) recognized Shufeldt’s inter-
ordinal error and synonymized Ardea sellardsi with Meleagris
gallopavo, and also reported several other specimens from this
site as M. gallopavo. Also listed as M. gallopavo by Wetmore
(1931b) and Brodkorb (1964a).

SEMINOLE FIELD, Pinellas County. Meleagris gallopavo
(includes the synonym M. tridens). All are USNM 244387 ex-
cept a humerus (USNM 12207), a tibiotarsus (USNM 12214),
and a tarsometatarsus (USNM 12052). Numbers in parenthe-
ses are numbers of specimens examined. Coracoid (3) — Tables

4, 5. Scapula (1) — Table 7. Humerus (3) — Tables 8, 9. Ulna
(7) — Tables 10, 11. Carpometacarpus (10) — Tables 14, 15.
Femur (8) — Tables 16, 17. Tibiotarsus (7) — Tables 18, 19.
Tarsometatarsus (28) — Tables 20-22, Fig. 13; Fig. 13 of Wet-
more (1931a). This series of fossils, reported as M. gallopavo
(all specimens except USNM 12052) and M. tridens Wetmore
(USNM 12052 only) by Wetmore (1931a, b) and Brodkorb
(1964a), agrees qualitatively and quantitatively with specimens
of Recent M. gallopavo. Brodkorb (1964a) noted that M. tri-
dens may merely be a specimen of M. gallopavo with an ab-
normal development of three tarsal spurs. Williams (1967) doc-
umented the occurrence of double spurs in living M. g. osceola
from Florida. There is also a tarsometatarsus from Inglis (UF
20680) with two spurs. Figure 13 illustrates the holotype ofM.
tridens next to a specimen of M. g. osceola with three spurs,
and a specimen of M. ocellata with two spurs. I agree with
Wetmore (1931a) in noting the lack of differences other than
the aberrant spurs between M. tridens and M. gallopavo.
Meleagris tridens is therefore a synonym of M. gallopavo.

Mention may be made here of a record of M. gallopavo at
“Pleistocene cavern deposits at Ocala, Florida” by Shufeldt
(1918:358; for further references to this site, see Ray 1957).
These specimens could not be located at the United States
National Museum, where Shufeldt said they would probably
be deposited. Pending re-examination of these specimens, this
record should not be considered as a valid occurrence of M.
gallopavo.

The following are 13 Pleistocene localities in Florida whose
faunas are either very limited or unstudied. All are regarded
herein as Rancholabrean (M. Frazier and S.D. Webb pers.
comm.), an age that is not refuted by the turkey specimens
from these sites. Because their ages are not as refined as the
other Floridian sites discussed above, they are simply listed in
alphabetical order.

BOWMAN IA, Putnam County. Meleagris cf. M. gallopavo.
Tarsometatarsus (PB 8606) — Tables 20, 22. No characters ap-
parent other than size.

DAVIS QUARRY, Citrus County. Meleagris gallopavo. Cor-
acoid (UF 22702, 22703) — Table 4. Femur (UF 22074) — Table
16. These specimens agree with those of M. gallopavo in all
ways, and differ from specimens from Inglis and Coleman in
character 11 of the coracoid.

ECONFINA RIVER, Taylor County. Meleagris sp. Tibiotar-
sus (USNM 243754) — Table 19. No characters apparent other
than size.

FLORIDA LIME COMPANY, Marion County. Meleagris sp.
Coracoid (PB 8439) — Table 5. Radius (PB 8440) — Table 12.
Carpometacarpus (PB 8441, 8442) — Table 14. The coracoid
resembles those from Coleman, not those of Recent M. gallo-
pavo, in character 1, while the opposite is true in character 10.
No other characters are apparent. Reported as M. gallopavo
by Brodkorb (1964a).

HAILE IIA. Alachua County. Meleagris cf. M. gallopavo.
Ulna (PB 1577) — Table 10. Tarsometatarsus (PB 1575). No
characters apparent other than size. Reported as M. gallopavo
by Brodkorb (1964a).

HOG CREEK, Sarasota County. Meleagris sp. Femur
(USNM 12096) — Table 17. Tibiotarsus (USNM 12098) — Table

Steadman: Turkey Osteology and Paleontology

149

19. No characters apparent other than size. Reported as M.
gallopavo by Wetmore (1931a, b) and Brodkorb (1964a).

MEFFORD CAVE I, Marion County. Meleagris cf. M. gal-
lopavo. All are UF 2119. Premaxilla. Mandible. Coracoid —
Table 4. Scapula — Table 7. Ulna — Table 10. Radius — Table
12. Carpometacarpus — Table 14. Synsacrum. Pelvis. These
elements agree with those of Recent M . gallopavo in all re-
spects except that the premaxilla is wider.

OAKHURST QUARRY, Marion County. Meleagris sp. Tib-
iotarsus (PB 8495) — Table 18. No characters apparent other
than size.

ST. MARK’S RIVER, Leon and Wakulla Counties. Meleagris
sp. Femur (USNM 209922) — Table 16. Tibiotarsus (USNM
209921). No characters apparent other than size.

SANTA FE RIVER IA, Gilchrist County. Meleagris cf. M.
gallopavo. Femur (UF 22080) — Table 17. Tibiotarsus (UF
10664) — Table 18. Tarsometatarsus (UF 10667) — Tables 20,
22. Although these specimens lack qualitative distinctions,
their size and proportions, especially of the tarsometatarsus,
are more similar to those of Recent M. gallopavo than to those
of Inglis or Coleman fossils. This site contains a mixture of
Blancan and Rancholabrean fossils (Webb 1974), but the tur-
key fossils are regarded herein as Rancholabrean because of
their similarity to M. gallopavo. These are probably the same
specimens as those upon which Brodkorb (1963) reported M.
gallopavo from Santa Fe I.

SANTA FE RIVER IVA, Gilchrist County. Meleagris sp.
Ulna (UF 16806) — Table 10. Femur (UF 22079). Tibiotarsus
(UF 16806, 22079) — Tables 18, 19. No characters apparent
other than size.

STEINHATCHIE RIVER, Taylor and Dixie Counties.
Meleagris cf. M. gallopavo. Radius (USNM 243755) — Table
12. No characters apparent other than size.

WEKIVA RUN III, Levy County. Meleagris sp. Tibiotarsus
(UF 14214) — Table 18. No characters apparent other than
size.

Holocene

WACISSA RIVER, Jefferson County, Florida. Meleagris gal-
lopavo. Tarsometatarsus (USNM 239842) — Tables 20, 22.
This specimen differs from those of Inglis and Coleman and
agrees with Recent M. gallopavo in size and proportions.

NICHOL’S HAMMOCK, Dade County, Florida. Meleagris
gallopavo. All are UF 22075. Numbers in parentheses are
numbers of specimens examined. Coracoid (5) — Tables 4, 5.
Scapula (1) — Table 9. Ulna (5) — Table 11. Radius (1) — Table
12. Carpometacarpus (3) — Tables 14, 15. Femur (2) — Table
17. Tibiotarsus (8) — Tables 18, 19. Tarsometatarsus (3) — Ta-
ble 21. This series agrees in every way with specimens of Re-
cent M. gallopavo, being more similar in size to those of M.
g. osceola than to those of late Pleistocene M. gallopavo, such
as from Ichetucknee River. Reported as M. gallopavo by
Hirschfeld (1968).

GOOD’S SHELLPIT, Volusia County, Florida. Meleagris
gallopavo. Coracoid (PB 1709, 2163) — Tables 4, 5. Humerus
(PB 1646, 1723, 1734, 1776)— Tables 8, 9. Ulna (PB 1757,
1828) — Tables 10, 11. Carpometacarpus (PB 1617, 1725) —

Tables 14, 15. Femur (PB 1698, 1710, 1735, 1777, 1810) —
Tables 16, 17. Tibiotarsus (PB 1758, 1778) — Tables 18, 19.
These specimens agree in every way with those of Recent M.
gallopavo. Reported as M. gallopavo by Brodkorb in Neill et
al. (1956) and by Brodkorb (1964a).

SILVER GLEN SPRINGS, Lake County, Florida. Meleagris
sp. Femur (PB 8498, 8499) — Table 16. No characters apparent
other than size. Reported asM. gallopavo by Brodkorb in Neill
et al. (1956) and Brodkorb (1964a).

BUFFALO SITE, Putman County, West Virginia. Meleagris
gallopavo. All specimens are SBU. Numbers in parentheses
are numbers of specimens examined. Individual catalogue
numbers available on request. Coracoid (234) — Tables 4, 5.
Scapula (30) — Tables 6, 7. Humerus (277) — Tables 8, 9. Ulna
(116) — Tables 10, 11. Radius (69) — Tables 12, 13. Carpometa-
carpus (308) — Tables 14, 15. Femur (63) — Tables 16, 17. Ti-
biotarsus (113) — Tables 18, 19. Tarsometatarsus (131) — Ta-
bles 20-22. All major elements are illustrated in Figs. 1-10 of
Kooliath (1975). This large series of bones, although repre-
senting birds eaten by 1 7 th century Amerindians, are regarded
by Kooliath (1975) and me as representing a wild population
of M. gallopavo. They thus provide an unexcelled sample of
M. g. silvestris from the period prior to extensive contact with
Europeans. These bones were measured and described by
Kooliath (1975), who compared them with modern M. g. sil-
vestris from New York, finding the modern birds to be 2 to
3 percent smaller in all linear measurements (generally only
total length considered) than those of AD 1650. My data in-
clude more types of measurements on a larger number of bones
and yield the same slight difference in size. Kooliath suggested
that selection for increased wildness because of increased hunt-
ing pressure since European contact may be the main factor
leading to the apparent reduction in size.

HARTMAN’S CAVE, Monroe County, Pennsylvania. Mele-
agris gallopavo. Coracoid (ANSP 761) — Table 4. Humerus
(ANSP 758-760, 771) — Tables 8, 9. Carpometacarpus (ANSP
753) — Table 14. Tibiotarsus (ANSP 753) — Table 19. Tarso-
metatarsus (ANSP 753) — Tables 20-22. Originally reported as
M. gallopavo by Leidy (1889), who noted that this deposit,
apparently collected without any stratigraphic control, con-
tained extinct genera of mammals (Mylohyus, Castoroides) as
well as advanced Amerindian artifacts, including ceramics.
Although Leidy (1889) gave no provenience for the turkey
bones contained therein, this site is listed as a Pleistocene rec-
ord for M. gallopavo by both Wetmore (1931b) and Brodkorb
(1964a). Upon examination of the specimens involved, I dis-
covered that some of the turkey bones not only bear butcher
marks made from steel knives, but also have obviously been
shot by a shotgun, lead pellets from which are still contained
within several of the bones. Thus Hartman’s Cave can no
longer be regarded as a Pleistocene locality for M. gallopavo.
The lack of mineralization of these bones further supports their
recency of deposition, as does the fact that they are inseparable
from modern skeletons of M. gallopavo. The alternative hy-
pothesis of Pleistocene firearms is rejected.

Holocene Mayan Archaeological Sites

DZIBILCHALTUN, Yucatan, Mexico. Meleagris ocellata.
All specimens have UFZA numbers. Coracoid (31 M-101, 603

150

Steadman: Turkey Osteology and Paleontology

M-825, 726 M-626)— Tables 4, 5. Humerus (728 M-567, 172
M-101)— ' Table 8. Ulna (M-110, 339 M-176) — Table 10. Car-
pometacarpus (M-110, 341 M-176, 164 M-101, 724 M-558) —
Table 14. Femur (517 M-179) — Table 17. Tibiotarsus (340
M-176, 759 M-176, 174 M-101, 335 M-1336)— Tables 18, 19.
Tarsometatarsus (775 M-300, 170 M-101, 725 M-645, 401
M-108D, 783 M-2027) — Tables 20-22. Reported asM. ocellata
by Wing and Steadman (in press), this series agrees with mod-
ern skeletons of M. ocellata.

MAYAPAN, Yucatan, Mexico. Meleagris ocellata. All are
MCZ 2536-2539, 2543, 2445. Numbers in parentheses are
numbers of specimens examined. Coracoid (41) — Tables 4, 5.
Scapula (39) — Tables 6, 7. Humerus (31) — Tables 8, 9. Car-
pometacarpus (27) — Tables 14, 15. Tarsometatarsus (20) — Ta-
bles 20-22. This very large series of bones, reported as M.
ocellata by Pollock and Ray (1957) and as Agriocharis ocellata
by Brodkorb (1964a), includes many specimens that are larger
and stouter than the corresponding elements in comparative
skeletons of wild M. ocellata. This is strong evidence for ar-
tificial fattening of these birds, such as by feeding them corn,
in combination with a sedentary existence. This in turn sug-
gests that these birds were kept in some state of confinement
by the people of Mayapan. To what extent M. ocellata was
tamed at Mayapan is presently impossible to ascertain. Mele-
agris ocellata is famous for its wildness, and numerous refer-
ences attest to the impossibility of taming this beautiful bird.
However, an example of its potential to attain at least some
degree of tameness has been observed by the author at Tikal
National Park, Peten, Guatemala, where because of protection
from hunting for about 20 years, the two local flocks of M.
ocellata living near the ruins are so tame as to come up to
homes and be fed corn and rice by the residents. These birds
may easily be approached to within about 8 meters, while
birds from flocks in adjacent areas, which are subjected to
hunting, are extremely wary and fly or run away at first sight
of a person (Steadman et al. 1979). The presence also at Maya-
pan of apparently wild M. ocellata suggests that wild birds
were eaten along with those supposedly reared. The sample
of wild turkeys hunted by the Maya was probably biased to-
ward those birds that fed heavily in the corn fields and were
thus in good flesh.

CANCUN ISLAND, Quintana Roo, Mexico. Meleagris cf. M.
ocellata. Tarsometatarsus (UFZA Q-509) — Table 20. No char-
acters apparent other than size (immature).

TULUM, Quintana Roo, Mexico. Meleagris cf. M. ocellata.
Ulna (MCZ 25 13) — Table 10. Carpometacarpus (MCZ 2531) —
Table 15. These specimens lack distinctive qualitative char-
acters, and are very tentatively referred to M. ocellata on the
basis of their very small size.

BARTON RAMIE SITE, Belize. Meleagris cf. M. ocellata.
Tarsometatarsus (PB 8492) — Table 21. No characters appar-
ent other than size. Reported as Agriocharis ocellata by Brod-
korb (1964a).

MACANCHE, Peten, Guatemala. Meleagris cf. M. ocellata.
Both are UFZA unnumbered. Carpometacarpus — Table 14.
Tibiotarsus — Table 18. No characters apparent other than
size.

SYSTEMATICS

The nomenclatural status of the turkeys from each of the
sites listed above and in Table 1 is based on data in the Com-
parative Osteology section (roughly quantified in Tables 2 and
3) and also on the measurements presented in Tables 4-22. As
previously noted, turkeys are very similar osteologically, and
most quantitative differences are only average ones. With the
exception of the problematical Rhegminornis calobates, from
which only one element is known, each taxon of turkey av-
erages more than partial agreement (i.e., a similarity index
value greater than 50) with all other known forms. Tables 2
and 3 are based only on those characters in which a difference
was seen between at least two forms. The degree of similarity
would be much higher if previously published characters that
do not hold were included. It must be understood that biases
exist in Tables 2 and 3 because of differences in elements and
characters being compared between any two taxa. These tables
present only an approximate, but useful, estimate of the degree
of similarity between the various forms of turkeys.

Recognition of the genera Rhegminornis and Proagriocharis
is based on their low overall resemblance to other forms (Ta-
bles 2-4, 20-22). A more detailed discussion of their affinities
is given in the accounts of individual sites above. Meleagris
is the only other genus recognized in this study. It includes all
known diagnostic turkey fossils from Blancan through Recent
times, as well as the two living species. The two other genera
recognized by Brodkorb (1964a) and most other workers, Agri-
ocharis and Parapavo, are herein regarded as synonyms of
Meleagris, for reasons outlined as follows.

Agriocharis was originally diagnosed on the external mor-
phology of M. ocellata. Chapman ( 1896:288) described the new
genus as follows: “The differences in the form and distribution
of the warty excrescences of the head and neck, and in the
character of the erectile appendages of the forehead, the more
highly graduate tail and the more rounded rectrices, the ab-
sence of a beard in the male and the presence of rudimentary
spurs in the female are all characters which entitle ocellata to
generic distinction. ...” Rudimentary spurs in females, dis-
cussed earlier in the Comparative Osteology section, are not
always present in females of M. ocellata, and they may be
present in females of M. gallopavo. Meleagris ocellata is char-
acterized by a lower average amount of similarity to other
turkeys than any other post-Hemphillian form (Tables 2 and
3), being approached in this respect only by M. crassipes.
However, M. ocellata resembles post-Hemphillian forms more
than it does Rhegminornis calobates or Proagriocharis kim-
ballensis, and the rather low level of agreement between M.
ocellata and M. gallopavo is skewed downward because com-
plete skeletons of these living species permitted me to find
characters that were imperceptible in fossil forms. In Table 2,
M. ocellata and M. gallopavo are compared in 85 different
characters, ten more than are used between any other forms.
All of these ten characters have a similarity value of 0 to 50.
Regardless, the amount of dissimilarity between the various
Pleistocene and Recent forms and M. ocellata is not enough
to justify generic separation. Although both M. ocellata and
M. crassipes appear to be somewhat unique within the genus,
I feel that the similarities between these species, which in the
past have been largely overlooked in a search for differences,

Steadman: Turkey Osteology and Paleontology

151

may best be emphasized by their inclusion in a single genus.
Ridgway (Ridgway and Friedmann 1946:458) recognized Agri-
ocharis but said, “Agriocharis is, in fact, so closely related to
Meleagris that I am somewhat doubtful as to the expediency
of recognizing it as a genus.” Paynter (1955) found the char-
acters of M. ocellata to be of no more than specific value after
several years of field and museum work with the birds of the
Yucatan peninsula. Nearly two months of observation of M.
ocellata at Tikal, Guatemala, has revealed many similarities
between the life histories of M. ocellata and M. gallopavo
(Steadman et al. 1979). Thus, non-osteological data exist that
support the inclusion of Agriocharis in Meleagris.

Meleagris californica is more similar to M. gallopavo than
to M. ocellata (Tables 2 and 3), and is consistently intermediate
between the two living species in size (Tables 4-22). L. Miller
(1916a) said that because M. californica was intermediate be-
tween M. ocellata and Pavo cristatus, species that are in dif-
ferent subfamilies, it was necessary to erect the genus Para-
pavo for that species. Howard (1927) correctly regarded M.
californica as distinctly meleagridine, not phasianine. Howard
(1927) considered M. ( Parapavo ) californica to be generically
separable from M . gallopavo and M . ocellata, although resem-
bling M. ocellata more than M. gallopavo, and stated
(1927:27): “Grouping the characters for each element together,
we find that the following elements [of M. californica ] possess
characters of each of the modern genera: sternum, coracoid,
humerus, radius, femur, tibiotarsus, tarsometatarsus; another
element (furcula) is in all characters nearer Meleagris
[gallopavo 1; others in all characters approach Agriocharis
[ocellata]: skull, scapula, ulna. The remaining elements either
possess characters distinct from both genera (fibula), or have
structure features common to all three genera (carpometacar-
pus, pelvis, pygostyle).” Thus, M. californica was supposedly
distinct from both M. gallopavo and M. ocellata only in the
fibula. However, the fibula of these species is not as definitely
different as stated by Howard (see page 137). The lack of
diagnostic features in M. californica, especially as compared
to M. gallopavo, strongly argues for its inclusion in Meleagris.
A. Miller and Bowman (1956) found no generic differences
between the tarsometatarsi of Meleagris (gallopavo) and Par-
apavo (californica). I concur with this and add that I cannot
find generic differences between M. californica and M. gallo-
pavo in any skeletal element.

The general recognition of the genus Parapavo for the past
60 years can perhaps be attributed largely to two factors. First,
it was originally described as a peacock in the genus Pavo (L.
Miller 1909); when its true subfamilial affinities became ap-
parent (L. Miller 1916a), to synonymize Pavo with a living
genus of turkey probably seemed to be a bit drastic, if indeed
it was even considered. Second, throughout the subsequent
systematic history of M. californica (Howard 1927, 1928;
Sushkin 1928), the two living species of turkeys were placed
in separate genera. Thus comparisons among M. californica,
M. gallopavo, and M. ocellata were carried out with a bias
toward thinking in terms of differences on the generic level
(Parapavo vs. Meleagris vs. Agriocharis). I have shown above
that the majority of those differences either do not hold at all,
or are only average differences that must therefore be consid-
ered on a specific, not a generic, level.

Revised Classification

Order Galliformes (Temminck 1820)

Family Phasianidae Vigors 1825
Subfamily Meleagridinae (Gray 1840)

Rhegminornithidae Wetmore 1943 (23 June), Proc. New En-
gland Zool. Club, vol. 22, p. 60 (type Rhegminornis Wet-
more). — Rhegminornithinae Brodkorb 1967 (12 June), Bull.
Florida State Mus., Vol. 11, no. 3, p. 201 (new rank).

Genus Rhegminornis Wetmore 1943
Rhegminornis Wetmore 1943 (type Rhegminornis calobates
Wetmore)

Rhegminornis calobates Wetmore 1943
Rhegminornis calobates Wetmore 1943 (23 June), Proc. New
England Zool. Club, vol. 22, p. 61, pi. 9, figs. 1-5 (type
from Thomas Farm, distal end of right tarsometatarsus, Mus.
Comp. Zool. no. 2331). — Rhegminornis calobates, Olson
and Farrand 1974 (June), Wilson Bull., vol. 86, no. 2, p.
114 (reassignment from Charadriiformes to Meleagridinae).

Early Miocene (Hemingfordian): Thomas Farm local fau-
na. Florida: Gilchrist County: Thomas Farm, 8 miles N of
Bell.

Genus Proagriocharis Martin and Tate 1970

Proagrio charts Martin and Tate 1970 (type Proagriocharis
kimballensis Martin and Tate)

Proagriocharis kimballensis
Martin and Tate 1970

Proagriocharis kimballensis Martin and Tate 1970 (5 June),
Wilson Bull., vol. 82, no. 2, p. 215, fig. 1 (type from S of
Lime Creek, left coracoid, Univ. Nebraska State Mus. no.
20033).

Late Miocene or early Pliocene (Hemphillian): lower part of
Kimball Formation, Univ. Nebraska Coll. Loc. Ft-40. Nebras-
ka: Frontier County: S of Lime Creek.

Genus Meleagris Linnaeus 1758

Meleagris Linnaeus 1758 (type Meleagris gallopavo Linnaeus)
Agriocharis Chapman 1896 (type Meleagris ocellata Cuvier)
Eumeleagris Coues 1903 (type Meleagris ocellata Cuvier)
Meleagrops (Marsh ms.) Shufeldt 1913 (type Meleagris celer
Marsh)

Parapavo L. Miller 1916a (type Pavo californicus L. Miller)

Meleagris progenes (Brodkorb 1964)

Meleagris gallopavo ; Meleagrididae, sp.P, Wetmore 1944.
Univ. Kansas Sci. Bull., vol. 30, pt. 1, no. 9, p. 98 (Rexroad
ranch, misidentification).

Agriocharis progenes Brodkorb 1964b (4 Nov.), Quart. Jour.
Florida Acad. Sci., vol. 27, no. 3, p. 223, pi. 1, figs. 1-3
(type from Rexroad ranch, distal part of right tarsometatar-
sus, Univ. Michigan Mus. Paleo. no. 31034).

Late Pliocene (Blancan): Rexroad Formation, Rexroad
local fauna. Kansas: Meade County: Rexroad ranch,
locality 3.

152

Steadman: Turkey Osteology and Paleontology

Meleagris leopoldi
A. Miller and Bowman 1956

Parapavo calif ornicus , L. Miller and Johnston 1937, Condor,
vol. 39, no. 5, p. 229 (Cita Canyon, misidentification).

cf. Meleagris, Johnston and Savage 1955, Univ. Califor-
nia Publ., Geol. Sci., vol. 31, no. 2, p. 39.

Meleagris leopoldi A. Miller and Bowman 1956 (5 March),
Wilson Bull., vol. 68, no. 1, p. 42, figs, la-lc (type from
Newton Harrell-Edd Reynolds ranch, distal end of right
tarsometatarsus, Panhandle Plains Hist. Mus. no. 753) —
Agriocharis leopoldi, Brodkorb 1964a (26 June), Bull. Flor-
ida State Mus., vol. 8, no. 3, p. 324. — Brodkorb 1964b (4
Nov.), Quart. Jour. Florida Acad. Sci., vol. 27, no. 3, p.
225.

Late Pliocene (Blancan): Cita Canyon beds. Texas: Ran-
dall County: Cita Canyon local fauna, UCMP locality
V-3721, at Newton Harrell-Edd Reynolds ranch, 2.2 km
S and 21 km E of Canyon.

Meleagris anza (Howard 1963)

Agriocharis anza Howard 1963 (30 Dec.), Los Angeles County
Mus., Contr. in Sci., no. 73, p. 19, pi. 3, fig. A (type from
Arroyo Tapiado, right humerus, Los Angeles County Mus.
no. 3753).

Early Pleistocene (Irvingtonian): upper Palm Spring For-
mation, upper 4000 ft. of Vallecito-Fish Creek section, Nat.
Hist. Mus. Los Angeles County Loc. no. 1358. California:
San Diego County: Anza-Borrego State Park: Arroyo Ta-
piado.

Meleagris gallopavo Linnaeus 1758

Meleagris gallopavo Linnaeus 1758, Systema Naturae, ed. 10,
vol. 1, p. 156 (type from Mexico).

Meleagris altus Marsh 1870a (March), Proc. Acad. Nat. Sci.
Philadelphia, p. 11 (nomen nudum). — Marsh 1870b (July),
Amer. Naturalist, vol. 4, no. 5, p. 317 (type from Mana-
lapan, New Jersey, portions of 3 skeletons). — Marsh 1872
(October), Amer. Jour. Sci., ser. 3, vol. 4, no. 22, p. 260
(descr. humerus, coracoid, femur, tibia, tarsometatarsus). —
Mercer 1899, Jour. Acad. Nat. Sci. Philadelphia, ser. 2,
vol. 11, pt. 2, p. 280 (referred tarsometatarsus from Port
Kennedy Cave, Pennsylvania).

Meleagris superbus Cope 1871, Trans. Amer. Philos. Soc.,
n.s., vol. 14, pt. 1, p. 238 (types from Manalapan, New
Jersey, 2 tibiae, 2 femora, 1 coracoid). — Peterson 1926, Ann.
Carnegie Mus., vol. 16, no. 2, p. 254, pi. 17, figs. 1-10
(referred 1 scapula, 1 humerus, 1 pelvis, 2 femora, 2 tibio-
tarsi, 2 tarsometatarsi from Frankstown Cave, Pennsylva-
nia).— Meleagris superba, Shufeldt 1915, Trans. Connecti-
cut Acad. Arts Sci., vol. 19, p. 66, pi. 10, figs. 71 — 73; pi.
11, figs. 74-77 (Marsh’s types from Manalapan, 3 humeri,
1 radius, 1 ulna, 1 coracoid, 1 scapula, 2 femora, 2 tibiotarsi,
1 tarsometatarsus, Yale Peabody Mus. nos. 533-536); M.
altus considered a synonym.

Meleagris celer Marsh 1872 (October), Amer. Jour. Sci., ser.
3, vol. 4, no. 22, p. 261 (types from Manalapan, New Jersey,
tibiotarsus and tarsometatarsus). Meleagrops celer (Marsh
ms.), Meleagris celer. — Shufeldt 1913, Auk, vol. 30, no. 1,
p. 29, pi. 3, figs. 3-5 (Marsh’s type tarsometatarsus from
Manalapan, Yale Peabody Mus.).

Ardea sellardsi Shufeldt 1917, Florida Geol. Surv., Ninth An-
nual Rept., p. 38, pi. 2, fig. 15 (type tibiotarsus from Vero
Beach, Florida). — Meleagris gallopavo, Wetmore 1931a,
Smithsonian Misc. Coll., vol. 85, no. 2, p. 33 (Shufeldt’s
type from Vero Beach, formerly Florida Geol. Surv. no.
7551, now in U.S. Nat. Mus.).

Meleagris Widens Wetmore 193 la (13 Apr.), Smithsonian Misc.
Coll., vol. 85, no. 2, p. 33, fig. 13, pi. 6 (type from Seminole
Field, Florida, tarsometatarsus, U.S. Nat. Mus. no. 12052).

Late Pleistocene (Rancholabrean) through Holocene: east-
ern, central, and southwestern United States, and parts of
eastern and western Mexico.

Meleagris calif ornica (L. Miller 1909)

Pavo californicus L. Miller 1909 (14 Aug.), Univ. Calif. Publ.,
Bull Dept. Geol., vol. 5, no. 19, p. 285, pi. 25 (type from
Rancho La Brea, right tarsometatarsus, Univ. Calif. Mus.
Paleo. no. 11300). — Parapavo californicus, L. Miller 1916a
(10 March), Univ. Calif. Publ., Bull. Dept. Geol., vol. 9,
no. 7, p. 96.

Meleagris richmondi Shufeldt 1915 (Feb.), Trans. Connecticut
Acad. Arts Sci., vol. 19, p. 67, pi. 2, fig. 19 (type from
Mission San Jose, California, fragmentary sternum, Yale
Peabody Mus. no. 905). — Parapavo californicus, Brodkorb
1964a (26 June), Bull. Florida State Mus., vol. 8, no. 3, p.
326.

Late Pleistocene (Rancholabrean): asphalt pits, Rancho
La Brea local fauna. California: Los Angeles County, Ran-
cho La Brea (L. Miller 1909).

Meleagris ocellata Cuvier 1820

Meleagris ocellata Cuvier 1820, Mem. Mus. Hist. Nat., vol.
5, no. 1, p. 4, pi. 1 (type from Gulf of Honduras). — Agri-
ocharis ocellata, Chapman 1896, Bull. Amer. Mus. Nat.
Hist., vol. 8, p. 287. — Eumeleagris ocellata, Coues 1903,
Key to North Amer. Birds, vol. 2, ed. 5, p. 727.

Holocene: Belize, northern Guatemala, eastern Chiapas,
and eastern Tabasco through Campeche, Yucatan, and
Quintana Roo, Mexico.

Meleagris crassipes L. Miller 1940
Meleagris crassipes L. Miller 1940 (15 May), Condor, vol. 42,
no. 3, p. 154, figs. 44-45 (type from San Josecito Cavern,
tarsometatarsus, Calif. Inst. Techn. no. 2708, now in Nat.
Hist. Mus. Los Angeles County). — Agriocharis crassipes,
Brodkorb 1964b (4 Nov.), Quart. Jour. Florida Acad. Sci.,
vol. 27, no. 3, p. 225.

Late Pleistocene (Rancholabrean): cave deposit, San Jo-
secito Cave local fauna. Mexico: Nuevo Leon: San Josecito
Cave, near Aramberri.

EVOLUTION

Rhegminornis calobates is morphologically unique among
the species included in this study (Tables 3, 21). Its relation-
ships to younger species are poorly understood, although there
is a suggestion of relatively close affinities to Proagriocharis
kimballensis (Table 3). The somewhat younger (Clarendoni-
an?) tibiotarsus from Westmoreland Park, Virginia, is much
too large for R. calobates and is also larger than P. kimballen-
sis. Typically for a tibiotarsus, this specimen suffers a lack of

Steadman: Turkey Osteology and Paleontology

153

diagnostic characters that prevents further elucidation of its
place in meleagridine phylogeny.

The relationships of Proagriocharis kimballensis to any of
the younger turkeys are also quite uncertain. If a trend of
increasing size with decreasing age occurred in late Hemphil-
lian and early Blancan times, as it did from the late Blancan
through the Rancholabrean in the M. gallopavo lineage, P.
kimballensis could qualify as a possible ancestor of M. pro-
genes on the basis of its size and age. But the low similarity
between these two species (Table 2), although probably due
at least in part to the sample sizes involved, does not give
strong support for such a lineage. Sadly, other Hemphillian
turkey fossils, all larger than P. kimballensis, are represented
only by tibiotarsi from Bone Valley, Florida, and Buckhorn,
New Mexico. These previously unreported fossils, both of
which are tentatively referred to Meleagris, and the Upper
Miocene (Clarendonian?) tibiotarsus from Virginia, suggest
that Proagriocharis and perhaps Rhegminornis represent a sib-
ling group of the larger main line of turkeys.

Meleagris progenes is most similar to M. leopoldi and fossils
from Inglis and Coleman (Table 2), suggesting that only one
lineage is represented by these four populations, with M. leo-
poldi intermediate, both morphologically and temporally, be-
tween M. progenes and the Inglis-Coleman forms.

The turkeys from Inglis and Coleman are very closely related
and certainly represent only one lineage. The qualitative char-
acters of the Vallecito Creek specimen (. M . ansa) are within
the range of variation of Inglis specimens. All humeral char-
acters that supposedly distinguish the Vallecito Creek turkey
from Inglis turkeys are either definitely or quite possibly due
to the crushing of the Vallecito Creek fossil, hence the tentative
referral of the Inglis specimens to this species. Extending the
range of a species of fossil turkey from California to Florida
is not unreasonable in light of the present distribution and
varied choice of habitat of M. gallopavo. The paleoecological
evidence (Downs and White 1968; Hibbard 1970; Hibbard and
Dalquest 1966; Howard 1963; Klein 1971; Martin 1974) sug-
gests fairly similar habitats for these Irvingtonian sites, further
increasing the likelihood of a single, wide-ranging species or
an osteologically similar superspecies of turkey existing during
the early and middle Irvingtonian. In this regard, however,
I must say that turkeys are very limited in their utility as
paleoecological indicators. Aside from needing trees in which
to roost, it is quite difficult to generalize about the habitat of
M. gallopavo, discussions of which are found in Hewitt (1967),
Leopold (1948), and Schorger (1966); see also Leopold (1948),
and Steadman et al. (1979), for habitat requirements of M.
ocellata. The use of M. ( uAgriocharis ”) ansa by Hibbard and
Dalquest (1966) to suggest environmental conditions in the ear-
ly Pleistocene of north-central Texas as being perhaps similar
to those of the modern Yucatan peninsula is particularly in-
appropriate, not only because of the above mentioned vague-
ness in the definition of turkey habitat, but also because the
femur from Gilliland, Texas, which Brodkorb (1964b) referred
to M. ( llAgriocharis ”) ansa, is not even safely identifiable to
genus (see discussion of Gilliland in Systematics section).

No turkey fossils older than Rancholabrean are even ten-
tatively referable to M. gallopavo, although the Colemen spec-
imens, when compared to earlier forms, are definitely ap-
proaching the M. gallopavo grade and are clearly intermediate
between specimens from Inglis and M. gallopavo. There is

little doubt that M. gallopavo evolved from the Coleman-tvpe
turkey. How much of the present range of M. gallopavo was
occupied by this form in late Irvingtonian times is not known.
Perhaps the M. gallopavo grade was attained initially in only
a small portion of this present range. This contradicts the fol-
lowing statement of Hibbard and Dalquest (1966:14), “It ap-
pears that these southern members [M . progenes, M. leopoldi,
and M. anza', at the time thought to be more closely related to
M. ocellata than to M. gallopavo] were slowly displaced south-
ward by the progressively cooler climates produced by each
successive glaciation. The more northern turkey, Meleagris
gallopavo, was able to extend its range southward with the
development of the strong continentality of the climate during
Wisconsin time.” This statement implies that the earlier forms
(“southern members”) were dead-end taxa that existed contem-
poraneously with M. gallopavo. It is a basic thesis of my stud-
ies, however, that these Blancan and Irvingtonian turkeys are
not drastically different from the living M. gallopavo and, in
fact, probably represent its direct ancestors.

Holman (1964) noted that the pneumatic foramen on the
dorsal base of the shaft of the scapula distinguishes living
Meleagridinae from other gallinaceous birds. As noted in the
Comparative Osteology section, Meleagris progenes of
Rexroad (Blancan) and Meleagris cf. M. leopoldi or M. anza
from Inglis IA (earliest Irvingtonian) lack this foramen, which
is present in the late Irvingtonian species of Meleagris from
Haile XVIA and Coleman IIA, as well asM. californica (Ran-
cholabrean), and the living M. gallopavo and M. ocellata. All
species of Phasianinae that I have examined lack this foramen
except Pavo cristatus and P. muticns. Increased pneumaticity
of the scapula may be related to increased size, as Pavo and
M . gallopavo are the largest phasianines and meleagridines,
respectively, and Meleagris sp. of Coleman is larger than M.
progenes or the Inglis specimens. The foramen in Pavo is be-
tween the glenoid facet and the furcular articulation, but, in
meleagridines, the foramen lies more posteriorly on the base
of the shaft, away from articulating surfaces. This difference
in position, combined with documentation from fossils from
Rexroad, Inglis, Haile XVIA, and Coleman on the develop-
ment of this foramen during Irvingtonian times, and the lack
of a foramen in all Phasianinae but Pavo, suggests that a non-
foraminate scapula is the primitive condition in both the Pha-
sianinae and Meleagridinae, and the foraminate scapula is the
derived state. The presence of a pneumatic foramen in M.
gallopavo, M. ocellata, and M. californica thus strongly argues
for their common ancestry in the middle Irvingtonian. Both
M. ocellata and M. californica are smaller than Inglis speci-
mens. If they branched off from the Inglis-Coleman lineage
before reaching the foraminate condition, the foramen may
not have been expected to evolve in these species. Therefore,
M. gallopavo, M. ocellata and M. californica are probably
derived from isolated populations of a Coleman-like turkey,
after the development of the foraminate scapula. This is fur-
ther supported by these three forms each being more similar
to specimens from Coleman than to those from Inglis (Table
2). The ancestors of M. californica and M. ocellata in the late
Irvingtonian were not necessarily identical to the Coleman tur-
key, because as much variation probably occurred in this
widespread species or superspecies as occurs in M. gallopavo
today (compare tarsometatarsal measurements and ratios of
M. g. silvestris and M. g. osceola in Tables 20 and 22).

154

Steadman: Turkey Osteology and Paleontology

Meleagris ocellata, which has no fossil record in the Pleis-
tocene, may have evolved from a population of turkeys similar
to those of Coleman that became isolated in the Yucatan region
by a high stand of the sea during an interglacial period in
middle to late Irvingtonian times. No turkeys live today in the
very wet coastal lowlands of western Tabasco to central Ve-
racruz, Mexico. Thus it appears that this wet tropical forest
forms a barrier between present populations of M. ocellata in
the Yucatan region and M. gallopavo intermedia in coastal
northern Veracruz and Tamaulipas. This barrier probably did
not exist continuously in the Pleistocene, as documented by
Martin (1974), who found strong western and neotropical af-
finities in the Coleman mammals, many of which probably
entered Florida via a Gulf Coast savanna corridor. This im-
plies somewhat more xeric conditions than at present, increas-
ing the likelihood that turkeys were living in even the wettest
portions of the Gulf Coast. Thus, it is likely that the Coleman-
type turkey occurred around the entire Gulf Coast area when
the “savanna corridor” existed. The distance involved would
not be great, as a drop in sea level of 100 meters during a
glacial period would have decreased the land distance from
Tampa, Florida, to Merida, Yucatan, from about 3600 to
about 2400 kilometers (Webb 1974).

Meleagris californica undoubtedly evolved from populations
that became isolated in California, where turkeys are known
to occur as early as the Hemphillian or Blancan (University
Drive site). The high degree of similarity between M. califor-
nica and M. gallopavo suggests either that these two species
were subject to fairly similar selective forces after populations
of their common ancestors became isolated, or that the ances-
tors of M . californica became isolated in California only after
reaching the M. gallopavo grade. Meleagris gallopavo is not
known west of central Arizona today. The quite arid condi-
tions in western Arizona and southeastern California that pre-
vail today could easily have provided a barrier to gene flow
between the turkeys of southwestern California and south-
eastern Arizona.

The relationships of M. crassipes to other turkeys is difficult
to assess at this point. It is the smallest of the various species
of Meleagris and is perhaps best characterized by its rather
distinctive tarsometatarsus. It resembles M. californica more
than other congeners (Tables 2 and 3), but it also has a fair
degree of qualitative and quantitative similarity to M. pro-
genes, a form known only from Kansas and Arizona. A more
complete discussion of M. crassipes is found in Amadeo Rea’s
paper in this volume.

The idea of a phasianine origin of the Meleagridinae is sup-
ported by the phasianine nature of the furcula (characters 3,

5, 9), scapula (character 3), and tarsometatarsus (characters 4,

6, 7) in the first meleagridines in which these elements are
known, as well as general osteological similarity. Therefore,
I propose that turkeys originated from a phasianine stock that
either (1) became isolated in the New World and evolved in
situ or (2) invaded the New World after reaching the mele-
agridine grade in the Old World. The presence of the above
mentioned phasianine characters in various turkeys favors the
first hypothesis, as does the lack of recognized meleagridine
fossils in the Old World. If one considers turkeys as having
come to North America via a land corridor over the Bering
Strait, then they would have dispersed through North America
from west to east, rather than vice versa. Details of this dis-

persal cannot be presently defined. However, the apparent
turkey from the late Miocene of Virginia and Rhegminornis
from the early Miocene of Florida suggest the potential for
Miocene and perhaps earlier turkey-like fossils from central
and western North America. A plausible alternative to the
Bering Strait dispersal route is that of an early Tertiary route
from western Europe to eastern North America. Available fos-
sil evidence favors neither hypothesis.

ACKNOWLEDGMENTS

It is with great pleasure that I dedicate this paper to Pierce
Brodkorb for his supervision during early phases of this study,
and to Hildegarde Howard in recognition of her innumerable
contributions in avian paleontology, especially her monograph
in 1927 of Meleagris californica, which still remains the most
useful single reference on the osteology of turkeys.

Kenneth E. Campbell, David W. Johnston, and S. David
Webb critically reviewed my Masters Thesis at the University
of Florida, upon which much of this paper is based. Storrs L.
Olson and Amadeo M. Rea read and commented on the manu-
script. Discussion and correspondence with Michael K. Fra-
zier, David D. Gillette, Jean-Georges Klein, H. Gregory
McDonald, Thomas R. Van Devender, S. David Webb, and
Elizabeth S. Wing have helped to answer many paleontolog-
ical questions, and discourse with Stephen W. Eaton and Lov-
ett E. Williams, Jr., has enriched by knowledge of living tur-
keys. Working with Amadeo M. Rea, whose complementary
study appears in this volume, has been most productive. Ken-
neth E. Campbell, Alison Habel, Charles Krantz, and Victor
E. Krantz helped with the photographs, and Thomas Ritchie
aided with the illustrations.

I also thank John Anderson, Edward Steadman, and The-
resa Steadman for aid in calculations, and Deborah Gaines for
her courageous typing efforts. Partial financial support was
provided by teaching assistantships from the Department of
Zoology of the University of Florida, and the Josselyn Van
Tyne Memorial Fund of the American Ornithologists’ Union
through James King. For working space and other resources
during preparation of this manuscript, I thank the Division of
Birds and the Division of Vertebrate Paleontology of the Na-
tional Museum of Natural History, Smithsonian Institution,
and the Laboratory of Paleoenvironmental Studies of the De-
partment of Geosciences, University of Arizona. This is con-
tribution number 842 of the Department of Geosciences, Uni-
versity of Arizona.

I am especially grateful to my parents, Norman and Theresa
Steadman, and to my revered friend and colleague, Storrs L.
Olson, whose constant encouragement have made this study
possible.

Listed below are abbreviations of collections from which
specimens were studied, followed by the names of persons who
kindly made them available. All of these people have my sin-
cere gratitude.

AMNH American Museum of Natural History (Charlotte
Holton, Malcolm McKenna).

ANSP Academy of Natural Sciences of Philadelphia (David
Gillette, Storrs Olson).

CM Carnegie Museum of Natural History (John Guilday,
Mary Dawson, Wendy Pollock, Mary Clench, Ken-
neth Parkes, Miriam Stern).

Steadman: Turkey Osteology and Paleontology

155

F:AM Frick Collection, American Museum of Natural His-
tory (Malcolm McKenna).

GW Glen Woolfenden.

ISUM Idaho State University Museum (H. Gregory Mc-
Donald, John White).

LACM Natural History Museum of Los Angeles County
(Robert McKenzie, David Whistler).

MCZ Museum of Comparative Zoology, Harvard Univer-
sity (Raymond Paynter).

PB Pierce Brodkorb.

PPHM Panhandle Plains Historical Museum (James Han-
son, Gerald Schultz, Jackie Wilson).

PU Princeton University, Museum of Natural History

(Donald Baird).

SBU Saint Bonaventure University, Department of Bi-

ology (Stephen Eaton).

TMM Texas Memorial Museum, University of Texas (Ja-
nette Bannan, Ernest Lundelius).

TTU Texas Tech University.

UCMP University of California, Berkeley, Museum of Pa-
leontology (J. Howard Hutchison).

UCMVZ University of California, Berkeley, Museum of Ver-
tebrate Zoology (Ned Johnson, Victoria Dziadosz,
National Science Foundation Grant BMS 2700102).

UF University of Florida, Florida State Museum (S.

David Webb, Chandra Aulsbrook).

UFZA University of Florida, Florida State Museum Zooar-
chaeological Collection (Elizabeth S. Wing).

UK University of Kansas, Museum of Natural History
(Marion Mengel).

UMMP University of Michigan, Museum of Paleontology
(Claude Hibbard).

UNSM University of Nebraska State Museum (C. Bertrand
Schultz).

USNM United States National Museum of Natural History
(John Barber, Storrs Olson, Robert Purdy, Clayton
Ray).

WWR Welder Wildlife Refuge (W. Caleb Glazener).

YPM Yale University, Peabody Museum.

LITERATURE CITED

Brodkorb, P. 1957. New passerine birds from the Pleisto-
cene of Reddick, Florida. Jour. Paleontol. 21(1): 129-138.

. 1959. The Pleistocene avifauna of Arredondo, Flor-
ida. Bull. Florida State Mus., Biol. Sci. 4(9): 269— 2 9 1 .

. 1963. A giant flightless bird from the Pleistocene of

Florida. Auk 80(2): 111—115.

. 1964a. Catalogue of fossil birds: Part 2 (Anseriformes

through Galliformes). Bull. Florida State Mus., Biol. Sci.
8(3): 195-335.

. 1964b. Notes on fossil turkeys. Quart. Jour. Florida

Acad. Sci. 27(3):223— 229.

. 1967. Catalogue of fossil birds: Part 3 (Ralliformes,

Ichthyornithiformes, Charadriiformes). Bull. Florida
State Mus., Biol. Sci. 1 1(3):99— 220.

Campbell, K.E. 1976. An early Pleistocene avifauna from
Haile XVA, Florida. Wilson Bull. 88(2):345-347.

. 1980. A review of the Rancholabrean Avifauna of the

Itchtucknee River, Florida, (this vol.).

Chapman, F.M. 1896. Notes on birds observed in Yucatan.
Bull. Amer. Mus. Nat. Hist. 8( 18):27 1—290.

Cope, E.D. 1871. Synopsis of the extinct Batrachia, Reptilia
and Aves of North America. Trans. Amer. Phil. Soc.
14:1-252.

Coues, E. 1903. Key to North American Birds, 5th ed.
Vol. 2. Pp. 537-1152. Dana Estes and Co., Boston.

Cracraft, J. 1968. First record of the turkey Meleagris gal-
lopavo from the Pleistocene of Mexico. Condor 7 0( 3 ) : 2 7 4 .

Cuvier, G. 1820. Mem. Mus. Hist. Nat. 6:1, 4.

Downs, T., and J.A. White. 1968. A vertebrate faunal
succession in superposed sediments from late Pliocene to
middle Pleistocene in California. Proc. Internat. Geol.
Cong. 10:41-47.

Eaton, S.W., and T.L. Moore. 1965—66. Age and sexual
variation in the skeleton of the wild turkey. Sci. Studies
St. Bonaventure Univ. 23:25-40.

Feduccia, A. 1973. Fossil birds from the Late Pleistocene
Ingleside Fauna, San Patricio County, Texas. Condor
7 5(2):243 — 244.

Gray, G.R. 1840. A List of the Genera of Birds with an
Indication of the Typical Species of Each Genus Compiled
from Various Sources. London. 80 pp.

Hamon, J.H. 1964. Osteology and Paleontology of the pas-
serine birds of the Reddick, Florida, Pleistocene. Florida
Geol. Surv. Geol. Bull. 44. 210 pp.

Hewitt, O.H., Ed. 1967. The Wild Turkey and Its Man-
agement. The Wildlife Society, Washington, D.C. 589 pp.

Hibbard, C.W. 1960. An interpretation of Pliocene and
Pleistocene climates of North America. Rept. Michigan
Acad. Arts, Sci., Letters for 1959-60:5-30.

. 1970. Pleistocene mammalian local faunas from the

great plains and central lowland provinces of the United
States. Pp. 395-433 in Pleistocene and Recent environ-
ments of the central great plains (W. Dort, Jr. and J.K.
Jones, Eds.) Dept. Geol. Univ. Kansas Spec. Publ. 3. 433

pp.

Hibbard, C.W., C.E. Ray, D.E. Savage, D.W. Taylor,
and J.E. Guilday. 1965. Quaternary mammals of
North America. Pp. 509-525 in The Quaternary of the
United States (H.E. Wright and D.G. Frey. Ed.). Prince-
ton Univ. Press, Princeton.

Hibbard, C.W., and W.W. Dalquest. 1966. Fossils from
the Seymour Formation of Knox and Baylor Counties,
Texas, and their bearing on the late Kansan climate of
that region. Univ. Michigan, Contr. Mus. Paleontol.
2 1(1): 1—66.

. 1973. Proneofiber, a new genus of vole (Cricetidae:

Rodentia) from the Pleistocene Seymour Formation of
Texas, and its evolutionary and stratigraphic significance.
Jour. Quat. Res. 3(2):269-274.

Hirschfeld, S.E. 1968. Vertebrate fauna of Nichol’s Ham-
mock, a natural trap. Quart. Jour. Florida Acad. Sci.
3 1(3): 177-189.

Holland, W.J. 1908. A preliminary account of the Pleisto-
cene fauna discovered in a cave opened at Frankstown,
Pennsylvania, in April and May, 1907. Ann. Carnegie
Mus. 4(3— 4):228— 233.

. 1912. A preliminary account of the Pleistocene fauna

discovered in a cave opened at Frankstown, Pennsylva-
nia, in April and May, 1907. Proc. 7th Internat. Zool.
Cong., Boston, 19-24 August, 1907:748-752.

156

Steadman: Turkey Osteology and Paleontology

Holman, J.A. 1959. Birds and mammals from the Pleisto-
cene of Williston, Florida. Bull. Florida State Mus., Biol.
Sci. 5( 1): 1—24.

. 1964. Osteology of gallinaceous birds. Quart. Jour.

Florida Acad. Sci. 27(3):230-252.

Hopkins, M.L., R. Bonnichsen, and D. Fortsch. 1969.
The stratigraphic position and faunal associates of Bison
( Gigantobison ) latifrons in southeastern Idaho, a progress
report. Tebiwa 12(1): 1-8.

Howard, H. 1927. A review of the fossil bird, Parapavo
californicus (Miller), from the Pleistocene asphalt beds of
Rancho La Brea. Univ. California Publ., Bull. Dept.
Geol. Sci. 1 7( 1): 1—56.

. 1928. The beak of Parapavo californicus (Miller).

Southern California Acad. Sci. Bull 2 7(3):90— 9 1 .

. 1936. A new record for Parapavo californicus (Mil-
ler). Condor 38(6):249-250.

. 1962. Bird remains from a prehistoric cave deposit

in Grant County, New Mexico. Condor 64(3): 24 1—2 42 .

. 1963. Fossil birds from the Anza-Borrego Desert.

Los Angeles Co. Mus. Contrib. Sci. No. 73:1-33.

Johnston, C.S., and D.E. Savage. 1955. A survey of var-
ious late Cenozoic vertebrate faunas of the Panhandle of
Texas. Part I. Univ. California Publ. Geol. Sci. 3 1(2): 27 —
50.

Klein, J.G. 1971. The ferungulates of the Inglis IA local
fauna, Early Pleistocene of Florida. M.S. Thesis, Univ.
Florida, Gainesville. 115 pp.

Kooliath, M.R.M. 1975. The 1 7 th century wild turkey in
Kanawha River Valley. M.S. Thesis, St. Bonaventure
Univ. 72 pp.

Leidy, J 1889. Notice and description of fossils in caves and
crevices of the limestone rocks of Pennsylvania. Geol.
Surv. Pennsylvania Ann. Rep. pp. 1—21.

Leopold, A.S. 1948. The wild turkeys of Mexico. Trans.
XIII North Amer. Wildlife Conf.:393-400.

Lindsay, E.H., N.M. Johnson, and N.D. Opdyke. 1975.
Preliminary correlation of North American land mammal
ages and geomagnetic chronology. Pp. 111-119 in Studies
on Cenozoic paleontology and stratigraphy, Claude W.
Hibbard Mem. Vol. 3 (G.R. Smith and N.E. Friedland,
Eds.). Univ. Michigan, Mus. Paleontol., Papers on Pa-
leontol. No. 12.

Linnaeus, C. 1758. Systema Naturae, ed. 10, Vol. 1. 824
PP

Marsh, O.C. 1870a. [Remains of a fossil turkey from New
Jersey]. Proc. Acad. Nat. Sci. Philadelphia 22:11.

. 1870b. New fossil turkey. Amer. Naturalist 4(5):317.

. 1872. Notice of some new Tertiary and Post-Tertiary

birds. Amer. Jour. Sci. 4(22):256-262.

Martin, L.D., and J. Tate. 1970. A new turkey from the
Pliocene of Nebraska. Wilson Bull. 82(2):2 14-2 18.

Martin, R.A. 1974. Fossil mammals from the Coleman IIA
fauna, Sumter County. Pp. 35-99 in Pleistocene Mam-
mals of Florida (S.D. Webb, Ed.). Univ. Florida Press
Gainesville.

McCoy, J.J. 1963. The fossil avifauna of Itchtucknee River,
Florida. Auk 80(3):335-35 1 .

Mercer, H.C. 1899. The bone cave at Port Kennedy, Penn-
sylvania, and its partial excavation in 1894, 1895, and
1896. Jour. Acad. Nat. Sci. Philadelphia 1 1(2):269— 286.

Miller, A.H., and R. Bowman. 1956. Fossil birds of the
Late Pliocene of Cita Canyon, Texas. Wilson Bull.
68( 1 ):38 — 46.

Miller, L.H. 1909. Pavo californicus , a fossil peacock from
the Quaternary asphalt beds of Rancho La Brea. Univ.
California Publ., Bull. Dept. Geol. 5( 19):285— 289.

. 1911. Avifauna of the Pleistocene cave deposits of

California. Univ. California Publ., Bull. Dept. Geol.,
6(16):385-400.

— . 1912. Contributions to avian paleontology from the

Pacific Coast of North America. Univ. California Publ.,
Bull. Dept. Geol. 7(5):61— 115.

. 1916a. A review of the species Pavo californicus.

Univ. California Publ., Bull. Dept. Geol. 9(7):89— 96.

. 1916b. Review of the Pleistocene species, Pavo cal-
ifornicus. (Abstract of paper given at Sixth Ann. Mtg.
Pacific Coast Sect. Paleo. Soc., 27 Feb. 1915). Bull. Geol.
Soc. America 27(1): 171.

. 1925. The birds of Rancho La Brea. Carnegie Inst.

Washington Publ. 349:63-106.

. 1927. The finding of Pleistocene material in an as-
phalt pit at Carpinteria, California: Bird remains. Science
66(1702): 156.

. 1940. A new Pleistocene turkey from Mexico. Condor

42(3): 154—156.

. 1942. A new fossil bird locality. Condor 44(6):283-

284.

. 1943. The Pleistocene birds of San Josecito Cavern,

Mexico. Univ. California Publ. Zool. 47(5):143— 168.

Miller, L.H. , and C.S. Johnston. 1937. A Pliocene record
of Parapavo from Texas. Condor 39(5):229.

Miller, W.E. 1971. Pleistocene vertebrates of the Los An-
geles Basin and vicinity (exclusive of Rancho La Brea).
Bull. Los Angeles Co. Mus. Nat. Hist., Sci., No. 10:1—
124.

Neill, W.T., H.J. Gut, and P. Brodkorb. 1956. Animal
remains from four Preceramic sites in Florida. Amer. An-
tiq. 2 1(4):383— 395 .

Olson, S.L., AND J. Farrand. 1974. Rhegminornis restud-
ied: a tiny Miocene turkey. Wilson Bull. 86(2): 1 14-120.

Olson, S.L., and D.D. Gillette. 1978. Catalogue ot type
specimens of fossil vertebrates. Academy of Natural Sci-
ences, Philadelphia. Part III: Birds. Proc. Acad. Nat. Sci.
Philadelphia 129(7):99— 100.

Opdyke, N.D., E.H. Lindsay, N.M. Johnson, and T.
Downs. 1977. The paleomagnetism and magnetic po-
larity stratigraphy of the mammal-bearing section of Anza
Borrego State Park, California. Quat. Res. 7:316-329.

Pattee, O.H., and S.L. Beasom. 1977. Rio Grande turkey
hens with leg spurs. Auk 94(1): 159.

Patton, T.H. 1969. Miocene and Pliocene artiodactyls, Tex-
as Gulf Coastal Plain. Bull. Florida State Mus., Biol. Sci.
14(2): 1 15 — 226.

Paynter, R.A. 1955. The ornithogeography of the Yucatan
Peninsula. Peabody Mus. Nat. Hist. Yale Univ. Bull. 9.
347 pp.

Peterson, O.A. 1926. The fossils of the Frankstown Cave,
Blair Co., Pennsylvania. Ann. Carnegie Mus. 16(2):249-
315.

Pollock, H.E.D., and C.E. Ray. 1957. Notes on vertebrate

Steadman: Turkey Osteology and Paleontology

157

animal remains from Mayapan. Carnegie Inst. Washing-
ton, Dept. Archaeo., Curr. Reports No. 41:633-656.

Ray, C.E. 1957. A list, bibliography and index of the fossil
vertebrates of Florida. Florida Geol. Surv. Spec. Publ.
No. 3:1-175.

Rea, A. 1980. Late Pleistocene and Holocene turkeys in the
Southwest, (this vol.).

Ridgway, R., and H. Friedmann. 1946. The birds of
North and Middle America. Part X. U.S. Nat. Mus. Bull.
50:1-484.

Robertson, J.S., Jr. 1976. Latest Pliocene mammals from
Haile XV A, Alachua County, Florida. Bull. Florida
State Mus. Biol. Sci. 20(3): 11 1-186.

Schorger, A.W. 1966. The Wild Turkey: Its History and
Domestication. Univ. Oklahoma Press, Norman. 625 pp.

Schultz, C.B., and E.B. Howard. 1935. The fauna of
Burnet Cave, Guadelupe Mountains. New Mexico Proc.
Acad. Nat. Sci. Philadelphia 87:273-298.

Shufeldt, R.W. 1897. On fossil bird-bones obtained by ex-
peditions of the University of Pennsylvania from the bone
caves of Tennessee. Amer. Naturalist 31:645-650.

. 1913. Contributions to avian paleontology. Auk

30(l):29-39.

. 1915. Fossil birds in the Marsh collection of Yale

University. Trans. Connecticut Acad. Arts Sci. 19:1-110.

. 1917. Fossil birds found at Vero, Florida, with de-
scriptions of new species. Florida State Geol. Surv., IX
Ann. Rept.:35-42.

. 1918. Notes on some bird fossils from Florida. Auk

35(3):357-358.

Skinner, M.F. 1942. The fauna of Papago Springs Cave,
Arizona, and a study of Stockoceros\ with three new an-
tilocaprines from Nebraska and Arizona. Bull. Amer.
Mus. Nat. Hist. 80(6): 143-220.

Steadman, D.W. 1975. The Plio-Pleistocene evolution of
turkeys (Aves: Meleagridinae). M.S. Thesis, Univ. Flor-
ida, Gainesville. 143 pp.

Steadman, D.W., J. Stull, and S.W. Eaton. 1979. Nat-
ural History of the Ocellated Turkey. Jour. World Pheas-
ant Assoc. 4: 15-37.

Sushkin, P.P. 1928. On the affinities of Parapavo califor-
nicus (Loye Miller). Ibis 12th Ser. 4(1): 135-138.

Temminck, C.J. 1820. Manuel d’Ornithologie, ed. 2. Paris.
439 pp.

Van Devender, T.R., and R.D. Worthington. 1977. The
herpetofauna of Howell’s Ridge Cave and the paleoecol-
ogy of the northwestern Chihuahuan Desert. Pp. 85-106
in Trans. Symp. Biol. Res. Chihuahuan Desert Region,
U.S. and Mexico (R.H. Wauer and D.H. Riskind, Eds.).
U.S. Dept. Int. , Nat. Park Serv. Trans, and Proc. Series,
No. 3.

Vigors, N.A. 1825. Observations on the natural affinities
that connect the orders and families of birds. Trans. Linn.
Soc. London 14(2 2):3 95— 5 17.

Webb, S.D. 1974. Pleistocene Mammals of Florida. Univ.
Florida Press, Gainesville. 270 pp.

Wetmore, A. 1924. Fossil birds from southeastern Arizona.
Proc. U.S. Nat. Mus. 64(5): 1-18.

. 1931a. The avifauna of the Pleistocene in Florida.

Smithsonian Misc. Coll. 85(2): 1-41.

. 1931b. The fossil birds of North America. Pp. 401-

472 in Check-list of North American Birds. 4th ed. Amer-
ican Ornith. Union, Lancaster, Pennsylvania.

. 1943. Fossil birds from the Tertiary deposits of Flor-
ida. Proc. New England Zool. Club 22:59-68.

. 1944. Remains of birds from the Rexroad fauna of

the Upper Pliocene of Kansas. Univ. Kansas Sci. Bull.
30(9):89-105.

. 1945. Record of the turkey from the Pleistocene of

Indiana. Wilson Bull. 57(3):204.

Wheatley, C.M. 1871a. Notice of the discovery of a cave
in eastern Pennsylvania, containing remains of Post-Plio-
cene fossils, including those of Mastodon, Tapir, Megal-
onyx, Mylodon, etc. Amer. Jour. Sci. and Arts, ser.
3, 1(4): 235— 237 .

. 1871b. The Bone Cave of eastern Pennsylvania.

American Jour. Sci. and Arts, ser. 3, 1:384-385.

Williams, L.E. 1967. Wild turkeys (Meleagris gallopavo)
with super-numerary leg spurs. Auk 84(1): 113-1 14.

Williams, L.E., and D.H. Austin. 1969. Leg spurs on
female wild turkeys. Auk 86(3):56 1—562 .

Wing, E.S., and D.W. Steadman. In press. Faunal re-
mains from the Dzibilchaltun site. In Excavations at Dzi-
bilchaltun, Yucatan, Mexico, by E.W. Andrews IV and
E.W. Andrews V, Publ. No. 48 of Mid. Amer. Res. Inst.,
Tulane Univ.

Woolfenden, G.E. 1959. A Pleistocene avifauna from Rock
Spring, Florida. Wilson Bull. 7 1(2): 183—187.

158

Steadman: Turkey Osteology and Paleontology

Figure 1. Distribution of fossil and living turkeys, omitting late Pleistocene (Rancholabrean) and Holocene records of Meleagris gallopavo and
M. ocellata within their modern range. Numbers correspond with those in Table 1. Only sites from which specimens have been examined by the
author are listed. See Rea (this vol.) for distribution of M. crassipes.

Steadman: Turkey Osteology and Paleontology

159

III

Figure 2. Measurements of the coracoid and scapula. I. Left cora-
coid, dorsal view: A — Head to external end of sternal facet; B —
Head to internal distal angle; C — Head to pneumatic foramen; D —
Head through scapular facet; F — Least width of shaft. II. Left cora-
coid, medial view: E — Depth of head. III. Left scapula, dorsal view:
A — Proximal width; B — Tip of acromion to external tip of glenoid
facet; D — Least width of neck. IV. Left scapula, proximal view: C —
Depth of glenoid facet.

Figure 3. Measurements of the humerus. I. Left humerus, anconal
view: A — Total length; B — Proximal width; C — Width of midshaft;
E — Distal width. II. Left humerus, ventral view: D — Depth of mid-
shaft.

160

Steadman: Turkey Osteology and Paleontology

Figure 4. Measurements of the ulna and radius. I. Left ulna, palmar
view: A — Total length; B — Proximal width; C — Width of midshaft.

II. Left ulna, anconal view: D — Depth of midshaft; E — Distal depth.

III. Left ulna, distal view: E — Distal depth. IV. Left radius, palmar
view: B — Proximal width; D — Least width of shaft; F — Distal width.
V. Left radius, medial view: A — Total length; C — Proximal depth; E —
Least depth of shaft.

FTT

Figure 5. Measurements of the carpometacarpus and femur. I. Left
carpometacarpus, internal view. A — Total length; B — Proximal depth;
C — Length of metacarpal I; E — Least depth of metacarpal II; G —
Distal depth; H— Protrusion of M III beyond knob of M II. II. Left
carpometacarpus, cross section near center of metacarpal II: D — Least
width of metacarpal II. III. Left femur, anterior view: A — Total
length; B — Proximal width; D — Width of midshaft; F — Distal width.
IV. Left femur, lateral view: E — Depth of midshaft; H — Depth of
external condyle; J — Depth of fibular condyle. V. Left femur, proxi-
mal view: C — Depth of head. VI. Left femur, distal view: G —
Depth of internal condyle; H — Depth of external condyle; J — Depth
of fibular condyle. VII. Left femur, medial view: G — Depth of internal
condyle.

Steadman: Turkey Osteology and Paleontology

161

Figure 6. Measurements of the tibiotarsus. I. Left tibiotarsus, an-
terior view: C — Width of midshaft; E — Distal width. II. Left tibio-
tarsus, posterior view: A — Length without cnemial crest. III.
Left tibiotarsus, proximal view; B — Width of head. IV. Left tibio-
tarsus, distal view: F — Depth of internal condyle; G — Depth of
external condyle. V. Left tibiotarsus, medial view: F — Depth of in-
ternal condyle. VI. Left tibiotarsus, lateral view: G — Depth of external
condyle.

Figure 7. Measurements of the tarsometatarsus. I. Left tarsometa-
tarsus, posterior view: A — Total length; B — Proximal width; C — Least
width of shaft; E — Proximal end to middle of spur core; F — Top of
spur core to end of middle trochlea; G — Middle of spur core to end of
middle trochlea; L — Distal width. II. Right tarsometatarsus, cross
section through spur core: H — Width of spur core; J — Length of
spur core; K — Angle of spur core. III. Left tarsometatarsus, me-
dial view: D — Least depth of shaft; M — Depth of inner trochlea;
N — Depth of middle trochlea. IV. Left tarsometatarsus, distal view:
L — Distal width; M — Depth of inner trochlea; N — Depth of middle
trochlea; P — Depth of outer trochlea.

A

B

C

-• D

Figure 8. Lateral views of furculae of males: A. Pavo muticus (PB
19183); B. Meleagris cf. M. leopoldi or M. ansa (Inglis IA; UF 20117);

C. Meleagris ocellata (PB 23S42); D. Meleagris gallopavo osceola (PB
27938).

162

Steadman: Turkey Osteology and Paleontology

Steadman: Turkey Osteology and Paleontology

163

O

o

o

Figure 10. Posterior views of tarsometatarsi of males: A, B. Meleagris cf. M. progenes (Benson; USNM 10551, AMNH 6330); C. Meleagris
leopoldi (Cita Canyon; PPHM 3169); D. Meleagris cf. M. leopoldi or M. anza (Inglis IA; UF 20713); E. Meleagris sp. (Coleman IIA; UF 11601N);
F. Meleagris crassipes (San Josecito Cave; LACM UC-100023).

164

Steadman: Turkey Osteology and Paleontology

CM

(uSjm"o16m“d “mX”' STSSiir* A' “ Mrf"S“ (PB ”819' PB 23U7)' C- Mde‘grh “Uotam “h'sMs

Steadman: Turkey Osteology and Paleontology

165

r~

10

I

o

o

Figure 12. Lateral views of tarsometatarsi of males: A. Meleagris cf. M. leopoldi or M. anza (Inglis IA; UF 20713); B, C. Meleagris sp. (Coleman;
UF 11601N, UF 11603C); D. Meleagris californica (Rancho La Brea; LACM E-5775); E. Meleagris gallopavo silvestris (USNM 501686).

166

Steadman: Turkey Osteology and Paleontology

Figure 13. Medial views (A, B) and lateral view (C) tarsometatarsi of males: A. Meleagris gallopavo osceola (USNM 487663); B. Meleagris
gallopavo (Seminole Field; USNM 120S2 — holotype of Meleagris “ tridens ”); C. Meleagris ocellata (PB 2354 2).

Figure 14. Anterior views of tarsometatarsi of females: A, B, C. Rhegminornis calobates (Thomas Farm; PB 8448, PB 8447, MCZ 2331 — not
certainly females); D. Proagriocharis kimballensis (UNSM Coll. Loc. Ft-40; UNSM 20037); E. Meleagris progenes (Rexroad; UMMP 48189); F.
Meleagris crassipes (San Josecito Cave; LACM UC-100022); G. Meleagris cf. M. leopoldi or M. anza (Inglis IA; UF 20789); H. Meleagris sp.
(Coleman II A; UF 11601Z); I. Meleagris californica (Rancho La Brea; LACM E-7122); J. Meleagris gallopavo osceola (PB 23114); K. Meleagris
ocellata (PB 30884).

168

Steadman: Turkey Osteology and Paleontology

Table 1 Fossil localities discussed in text. Numbered localities are shown in Fig. 1.

Fossil Locality

Age

Species of Turkey

References*

1. Thomas Farm, Florida

early Miocene (Hemingfordian)

Rhegminornis calobates

Olson and Farrand 1974; Patton
1969; Wetmore 1943.

2. Westmoreland State Park,
Virginia

late Miocene (Clarendonian?)

Meleagridinae, cf. Meleagris

Lauck Ward pers. comm.; this
paper.

3. UNSM Coll Loc. Ft-40,
Nebraska

late Miocene or early
Pliocene (Hemphillian)

Proagriocharis kimballensis

Martin and Tate 1970.

4. Buckhorn, New Mexico

late Miocene or early
Pliocene (Hemphillian)

Meleagridinae, cf. Meleagris

This paper.

S. Clifton Country Club,
Arizona

late Miocene or early
Pliocene (Hemphillian)

Meleagridinae, genus and
species indeterminate

This paper.

6. Bone Valley (Palmetto
Mine), Florida

late Miocene or early
Pliocene (Hemphillian)

Meleagridinae, cf. Meleagris

This paper; G. Morgan pers.
comm.

7. University Drive,
California

Pliocene (Hemphillian or
Blancan)

Meleagris sp.

This paper

8. Haile XVA, Florida

late Pliocene (Blancan)

Meleagris sp.

Campbell 1976; Robertson 1976;
this paper.

9. Benson, Arizona

late Pliocene (Blancan)

Meleagris cf. M. progenes

Brodkorb 1964b; Lindsay et al.
1975; Wetmore 1924, 1944; this
paper.

10. Cita Canyon, Texas

late Pliocene (Blancan)

Meleagris leopoldi

Brodkorb 1964b; Lindsay et al.
1975; A. Miller and Bowman
1956; this paper.

11. Rexroad, Kansas

late Pliocene (Blancan)

Meleagris progenes

Brodkorb 1964b; Lindsay et al.
1975; this paper.

12. Gilliland, Texas

early Pleistocene (early
Irvingtonian)

Meleagridinae, genus and
species indeterminate

Brodkorb 1964b; Hibbard and
Dalquest 1973; this paper.

13. Inglis IA, Florida

early Pleistocene (early
Irvingtonian)

Meleagris cf. M. leopoldi or M .
anza

Klein 1971; Webb 1974; this
paper.

14. Vallecito Creek, California

early Pleistocene (Irvingtonian)

Meleagris anza

Howard 1963; Opdyke et al. 1977;
this paper.

IS. Port Kennedy Cave,
Pennsylvania

early Pleistocene (Irvingtonian;
possibly Rancholabrean?)

Meleagridinae, cf. Meleagris

D. Gillette pers. comm.; Mercer
1899; this paper.

16. Haile XVIA, Florida

early Pleistocene (Irvingtonian)

Meleagris sp.

M. Frazier pers. comm.; this
paper.

17. Williston, Florida

middle Pleistocene (late
Irvingtonian)

Meleagris sp.

M. Frazier pers. comm.; this
paper.

18. Coleman IIA, Florida

middle Pleistocene (late
Irvingtonian)

Meleagris sp. (intermediate
between leopoldi-anza and
gallopavo)

Martin 1974; this paper.

19. Sante Fe River IIA,
Florida

middle Pleistocene (late
Irvingtonian)

Meleagris cf. M. gallopavo

Webb 1974; this paper.

20. American Falls Lake Beds,
Idaho

late Pleistocene (Rancholabrean;
Illinoian glacial?)

Meleagris gallopavo

Hopkins et al. 1969.

Papago Springs Cave,
Arizona

late Pleistocene (Rancholabrean)

Meleagris sp.

Skinner 1942; Rea this volume;
this paper.

Arizpe, Sonora, Mexico

late Pleistocene (Rancholabrean)

Meleagris cf. M. gallopavo

Cracraft 1968; Rea this volume;
this paper.

Burnet Cave, New Mexico

late Pleistocene (Rancholabrean)

Meleagris cf. M. gallopavo ; also
M . crassipes

Rea this volume; Schultz and
Howard 1935; this paper.

Howell’s Ridge Cave, New
Mexico

late Pleistocene (Rancholabrean)
and Holocene

Meleagris sp.; also M. crassipes

Rea this volume; Van Devender
and Worthington 1977; this
paper.

Ingleside, Texas

late Pleistocene (Rancholabrean)

Meleagris gallopavo

Feduccia 1973.

Sheffield Gravel Pits,
Texas

late Pleistocene?
(Rancholabrean?)

Meleagris sp.

This paper.

Carlisle Cave,
Pennsylvania

Pleistocene (probably
Rancholabrean)

Meleagris cf. M. gallopavo

Leidy 1889; this paper.

Steadman: Turkey Osteology and Paleontology

169

Table 1.

Continued.

Fossil Locality

Age

Species of Turkey

References*

Frankstown Cave,
Pennsylvania

late Pleistocene (Rancholabrean)

Meleagris cf. M. gallopavo

Hibbard et al. 1965; Peterson
1926; this paper.

Manalapan, New Jersey

late Pleistocene?
(Rancholabrean?)

Meleagris gallopavo

This paper

Bradenton, Florida

late Pleistocene (early
Rancholabrean)

Meleagris sp.

Webb 1974; this paper.

Rock Spring, Florida

late Pleistocene (early
Rancholabrean)

Meleagris cf. M . gallopavo

Webb 1974; this paper.

Withlacoochee River,
Florida

late Pleistocene (early
Rancholabrean)

Meleagris sp.

Webb 1974; this paper.

Haile VIIA, Florida

late Pleistocene (early
Rancholabrean)

Meleagris sp.

Webb 1974; this paper.

Reddick IB, Florida

late Pleistocene (early
Rancholabrean)

Meleagris cf. M. gallopavo

Webb 1974; this paper.

Melbourne, Florida

late Pleistocene (middle
Rancholabrean)

Meleagris cf. M. gallopavo

Webb 1974; this paper.

Arredondo, Florida

late Pleistocene (middle
Rancholabrean)

Meleagris cf. M. gallopavo

Webb 1974; this paper.

Sabertooth Cave, Florida

late Pleistocene (late
Rancholabrean)

Meleagris cf. M . gallopavo

Webb 1974; this paper.

Aucilla River, Florida

late Pleistocene (late
Rancholabrean)

Meleagris gallopavo

Webb 1974; this paper.

Ichetucknee River, Florida

late Pleistocene (late
Rancholabrean)

Meleagris gallopavo

Webb 1974; Wetmore 1931a.

Kendrick IA, Florida

late Pleistocene (late
Rancholabrean)

Meleagris sp.

Webb 1974; this paper.

Vero, Florida

late Pleistocene (late
Rancholabrean)

Meleagris cf. M . gallopavo

Webb 1974; this paper.

Seminole Field, Florida

late Pleistocene (late
Rancholabrean)

Meleagris gallopavo

Webb 1974; Wetmore 1931a.

Bowman IA, Florida

late Pleistocene (Rancholabrean)

Meleagris cf. M. gallopavo

M. Frazier, S.D. Webb pers.
comm.; this paper.

Davis Quarry, Florida

late Pleistocene (Rancholabrean)

Meleagris gallopavo

M. Frazier, S.D. Webb pers.
comm.; this paper.

Econfina River, Florida

late Pleistocene (Rancholabrean)

Meleagris sp.

M. Frazier, S.D. Webb pers.
comm., this paper.

Florida Lime Company,
Florida

late Pleistocene (Rancholabrean)

Meleagris sp.

M. Frazier, S.D. Webb pers.
comm.; this paper.

Haile II A, Florida

late Pleistocene (Rancholabrean)

Meleagris cf. M. gallopavo

M. Frazier, S.D. Webb pers.
comm.; this paper.

Hog Creek, Florida

late Pleistocene (Rancholabrean)

Meleagris sp.

M. Frazier, S.D. Webb pers.
comm.; this paper.

Mefford Cave I, Florida

late Pleistocene (Rancholabrean)

Meleagris cf. M. gallopavo

M. Frazier, S.D. Webb pers.
comm.; this paper.

Oakhurst Quarry, Florida

late Pleistocene (Rancholabrean)

Meleagris sp.

M. Frazier, S.D Webb pers.
comm.; this paper.

St. Mark’s River, Florida

late Pleistocene (Rancholabrean)

Meleagris sp.

M. Frazier, S.D. Webb pers.
comm.; this paper.

Santa Fe River IA, Florida

late Pleistocene (Rancholabrean)

Meleagris cf. M. gallopavo

Webb 1974; this paper.

Santa Fe River IVA,
Florida

late Pleistocene (Rancholabrean)

Meleagris sp.

M. Frazier, S.D. Webb pers.
comm.; this paper.

Steinhatchie River, Florida

late Pleistocene (Rancholabrean)

Meleagris cf. M. gallopavo

M. Frazier, S.D. Webb pers.
comm.; this paper.

Wekiva Run III, Florida

late Pleistocene (Rancholabrean)

Meleagris sp.

M. Frazier, S.D. Webb pers.
comm.; this paper.

Wacissa River, Florida

Holocene

Meleagris gallopavo

This paper. Continued

170

Steadman: Turkey Osteology and Paleontology

Table 1.

Continued.

Fossil Locality

Age

Species of Turkey

References*

Nichol’s Hammock,
Florida

Holocene (pre-1900)

Meleagris gallopavo

Hirschfeld 1968.

Good’s Shellpit, Florida

Holocene (approx. 3500-5000
BP)

Meleagris gallopavo

Neill et al. 1956.

Silver Glen Springs,
Florida

Holocene (approx. 3500-5000

BP)

Meleagris sp.

Neill et al. 1956; this paper.

Buffalo Site, West Virginia

Holocene (approx. AD 1650)

Meleagris gallopavo

Kooliath 1975.

Hartman’s Cave,
Pennsylvania

Holocene (post-European
contact)

Meleagris gallopavo

Leidy 1889; this paper.

21. Rancho La Brea,
California

late Pleistocene (Rancholabrean)

Meleagris californica

Hibbard et al. 1965; this paper.

22. Imperial Highway,
California

late Pleistocene (Rancholabrean)

Meleagris sp.

W. Miller 1971; this paper.

23. Carpinteria, California

late Pleistocene (Rancholabrean)

Meleagris californica

L. Miller 1927; Hibbard et al.
1965.

24. Workman and Alhambra
Streets, California

late Pleistocene (Rancholabrean)

Meleagridinae cf. Meleagris

W. Miller 1971; this paper.

25. La Mirada, California

late Pleistocene (Rancholabrean)

Meleagris cf. M. californica

W. Miller 1971; this paper.

26. Potter Creek Cave,
California

late Pleistocene (Rancholabrean)

Meleagris sp.

Hibbard et al. 1965; L. Miller
1911.

Dzibilchaltun, Yucatan,
Mexico

Holocene (approx. BC 1000-AD
900)

Meleagris ocellata

Wing and Steadman in press.

Mayapan, Yucatan,
Mexico

Holocene (approx. AD 1200-
1500)

Meleagris ocellata

Pollock and Ray 1957.

Canciin Island, Quintana
Roo, Mexico

Holocene (approx. AD 300-900)

Meleagris cf. M . ocellata

This paper.

Tulum, Quintana Roo,
Mexico

Holocene (approx. AD 1300-
1500)

Meleagris cf. M ocellata

This paper.

Barton Ramie site, Belize

Holocene (approx. AD 500-
1000)

Meleagris cf. M . ocellata

Brodkorb 1964b; this paper.

Macanche, Peten,
Guatemala

Holocene (approx. AD 300-900)

Meleagris cf. M. ocellata

This paper.

27. San Josecito Cave, Nuevo
Leon, Mexico

late Pleistocene (Rancholabrean)

Meleagris crassipes

Hibbard et al. 1965; L. Miller
1940.

* Includes only literature from which this table was compiled; see references cited in this table for additional literature on the fossil sites. Earlier
published records of turkeys from these sites, if recorded under a different name, are mentioned in the site by site accounts.

Steadman: Turkey Osteology and Paleontology

171

Table 2. Analysis of similarity. For every character in the comparative osteology section, except those that are similar in all forms, each form
was rated against the others as follows: 100 = complete agreement; 75 = much agreement; 50 = partial agreement; 25 = slight agreement; 0 = no
agreement. The first value in each case is the mean of all such agreement values. The number in parentheses is the number of characters on which
the mean agreement value is based.

Rhegmi-

nornis

calo-

bates

Thomas

Farm,

Florida

Pro-

agrio-

charis

kimbal-

lensis

UNSM

Coll.

Loc.

Ft-40,

Nebra-

ska

Mele-

agris

pro-

genes

Rex-

road,

Kansas

Mele-

agris

leo-

poldi

Cita

Canyon,

Texas

Mele-

agris

anza,

Valle-

cito

Creek,

Cali-

fornia

Mele-
agris cf
M. leo-
poldi
or M.
anza
Inglis
IA

Florida

Mele-
agris sp.,
Cole-
man
IIA,
Florida

Mele-

agris

gallo-

pavo

Recent

speci-

mens

Mele-

agris

cali-

fornica,

Rancho

La

Brea,

Cali-

fornia

Mele-

agris

ocellata

Recent

speci-

mens

Mele-

agris

cras-

sipes

San

Jose-

cito

Cave,

Nuevo

Leon,

Mexico

Proagriocharis
kimballensis, UNSM
Coll. Loc. Ft-40

75 (8)

Meleagris progenes
Rexroad

59 (11)

50 (14)

Meleagris leopoldi
Cita Canyon

71 (13)

84 (11)

92 (12)

Meleagris cf. M.
leopoldi or M. anza
Inglis IA

56 (13)

73 (21)

78 (30)

87 (17)

68 (7)

Meleagris sp.

Coleman IIA

54 (13)

63 (21)

78 (26)

85 (17)

50 (7)

89 (64)

Meleagris g allopavo,
Recent specimens

56 (13)

58 (21)

73 (30)

87 (17)

64 (7)

68 (73)

82 (65)

—

—

—

Meleagris californica,
Rancho La Brea

54 (13)

54 (21)

75 (29)

82 (17)

57 (7)

78 (72)

82 (65)

80 (75)

Meleagris ocellata
Recent specimens

48 (13)

58 (21)

72 (30)

78 (17)

57 (7)

68 (73)

72 (65)

62 (85)

72 (75)

—

—

Meleagris crassipes
San Josecito Cave

52 (13)

46 (21)

73 (24)

69 (17)

57 (7)

74 (47)

72 (47)

73 (47)

79 (47)

62 (47)

—

172

Steadman: Turkey Osteology and Paleontology

Table 3. Analysis of similarity in the tarsometatarsus. See Table 2 for explanation of values.

Rhegmi-

nornis

calo-

bates

Thomas

Farm,

Florida

Pro-

agrio-

charis

kimbal-

lensis

UNSM

Coll.

Loc.

Ft-40,

Nebr-

aska

Mele-

agris

progenes

Rex-

road,

Kansas

Mele-

agris

leopoldi

Cita

Canyon,

Texas

Mele-
agris cf.
leopoldi
or M.
anza ,
Inglis
IA,

Florida

Mele-
agris sp.
Cole-
man
IIA,
Florida

Mele-

agris

gallo-

pavo

Recent

speci-

mens

Mele-
agris
cali-
fornica
Rancho
La Brea,
Cali-
fornia

Mele-

agris

ocellata

Recent

speci-

mens

Mele-

agris

cras-

sipes

San

Jose-

cito

Cave,

Nuevo

Leon,

Mexico

Proagriocharis
kimballensis UNSM
Col! Loc. Ft-40

75 (8)

Meleagris progenes
Rexroad

59 (11)

83 (6)

Meleagris leopoldi
Cita Canyon

71 (13)

84 (11)

92 (12)

Meleagris cf. M .
leopoldi or M. anza
Inglis IA

56 (13)

77 (12)

96 (12)

87 (17)

Meleagris sp.
Coleman IIA

54 (13)

73 (12)

94 (12)

85 (17)

96 (18)

—

Meleagris gallopavo
Recent specimens

56 (13)

71 (12)

98 (12)

87 (17)

89 (18)

93 (18)

—

—

—

—

Meleagris californica
Rancho La Brea

54 (13)

67 (12)

94 (12)

82 (17)

83 (18)

88 (18)

92 (18)

Meleagris ocellata
Recent specimens

48 (13)

79 (12)

83 (12)

78 (17)

79 (18)

78 (18)

76 (18)

72 (18)

—

—

Meleagris crassipes
San Josecito Cave

52 (13)

56 (12)

83 (12)

69 (17)

72 (18)

74 (18)

78 (18)

85 (18)

58 (18)

—

Steadman: Turkey Osteology and Paleontology

173

Table 4. Measurements (in mm) of coracoids of male turkeys, with mean, standard deviation, observed range, and sample size. See Fig. 2 for
explanation of measurements.

Head to
External End

Head to

Head to

Head through

of Sternal

Internal

Pneumatic

Scapular

Depth

Least Width

Facet

Distal Angle

Foramen

Facet

of Head

of Shaft

Proagriocharis

71.7*

67.2*

57.9*

23.1*

9.1

6.4

kimballensis

UNSM Coll. Loc. Ft-40

1

1

1

1

1

1

Meleagridinae, genus

—

—

—

24.5

9.8

~8.2

and species indet.
Clifton Country Club3

1

1

1

Meleagris progenes

79.8*

—

65. 0b

31. 0b

10. 8b

10.1"

Rexroad

1

1

1

1

1

Meleagris cf. M.

101.93

94.26* ± 2.52

76.21 ± 2.54

35.95 ± 1.27

13.62 ± 0.86

10.44 ± 0.45

leopoldi or M. anza

94.2-110.7

80.4-97.4*

72.6-80.2

33.9-37.8

11.8-15.0

9.6-11.2

Inglis IA

6

10

8

13

20

27

Meleagris sp.

103.72*

97.30

79.35

37.80

14.84 ± 0.49

11.12 ± 0.84

Coleman IIA

101.2-105.4*

95.3-99.0

76.0-80.8

37.0-38.7

14.2-15.8

10.0-1 1.9

5

4

4

7

8

8

M . gallopavo

115.2*

—

—

40.8

15.5

10.8

Ingleside

1

1

1

1

M. gallopavo

118.5

—

90.0

43.2

16.2

12.6

Manalapan

1 15c-122d
2

1

1

1

1

M. gallopavo

117.05*

110.10

89.40

41. 18

15.55

11.83

Ichetucknee River

113.9*— 120.2

108.3-111.9

87.1-91.7

39.1-42.7

14. 1-17.0

11.4-12.3

2

2

2

5

6

3

M. gallopavo

—

—

—

38.1

14.6

—

Seminole Field

1

1

M . gallopavo

104.30*

97.25*

77.10*

37.95*

14.55

11.30

Davis Quarry

101.0*— 107.6*

94.8*-99. 7*

75.5-78.7*

36.3-39.6*

13.5-15.6

11.0-11.6

2

2

2

2

2

2

Meleagris cf. M.

—

—

—

41.7*

16.6

—

gallopavo
Mefford Cave I

1

1

M . gallopavo

—

—

—

39.1*

14.2

—

Nichol’s Flammock

1

1

M . gallopavo

106.5

98.6

82.0

37.7

14.3

9.7

Good’s Shellpit

1

1

1

1

1

1

M. gallopavo

112.7*

102.8*

85.8*

40.0*

14.6

10.9

Garfield Site

1

1

1

1

1

1

M. gallopavo

115.19 ± 2.97

108.12 ± 1.98

88.94 ± 3.67

41.68 ± 1.00

16.15 ± 0.85

11.71 ± 0.46

Buffalo Site

110.0-118.5

103.7-11 1.0

83.2-97.5

39.9-43.3

14.1-17.7

10.5-13.3

12

9

16

27

27

81

M. gallopavo

112.3

106.8*

91.0

40.8

16.0

10.4

Flartman’s Cave

1

1

1

1

1

1

M. gallopavo silvestris

112.37 ± 4. 14

104.79 ± 3.89

85.12 ± 3.39

40.71 ± 1.33

15.94 ± 0.64

11.29 ± 0.54

New York, Pennsylvania,

104.2-120.8

97.2-111.5

77.8-90.4

38.4-43.1

14.6-17.2

10.2-12.5

Virginia

26

27

27

27

28

28

M. gallopavo osceola

109.19 ± 4.29

101.19 ± 3.90

83.01 ± 2.94

37.94 ± 1.33

14.50 ± 0.77

10.68 ± 0.45

Florida

104.9-117.0

97.1-107.3

80.0-89.5

35.2-39.9

13.3-15.9

10.0-1 1.2

9

9

9

9

9

9

M. gallopavo intermedia

114.7

—

87.3

38.6

15.1

10.9

Texas

1

1

1

1

1

M. gallopavo mexicana

113.65

104.70

83.15

40.40

16.00

11.35

Chihuahua, Mexico

111.8-115.5

103.0-106.4

81.5-84.8

39.8-41.0

15.6-16.4

10.8-1 1.9

2

2

2

2

2

2

M. gallopavo

111.74 ± 4.25

103.93 ± 4.06

84.59 ± 3.31

40.00 ±1.74

15.60 ± 0.89

11.15 ± 0.57

Total skeletal

104.2-120.8

97.1-111.5

77.8-90.4

35.2-43.1

13.3-17.2

10.0-12.5

specimens

38

38

39

39

40

40

Continued

174

Steadman: Turkey Osteology and Paleontology

Table 4. Continued

Head to
External End
of Sternal
Facet

Head to
Internal
Distal Angle

Head to
Pneumatic
Foramen

Head through
Scapular
Facet

Depth
of Head

Least Width
of Shaft

M . californica

101.01 ± 3.97

94.26* ± 3.06

76.54 ± 3.44

35.94 ± 1.19

13.21 ± 0.85

10.08 ± 0.46

Rancho La Brea

92.0-106.1

85.8*-98.8

68.2-82.4

32.3-38.2

1 1.7-14.6

9.1-11.0

15

29

27

29

24

29

M. californica

99.17

92.42* ± 2.11

74.34

35.16 ± 1.09

13.61 ± 0.72

10.02 ± 0.48

Carpinteria

96.7-101.4

89.4-94.9

72.7-76.4

33.5-37.3

12.4-14.8

9.2-10.6

7

8

7

8

8

8

M . californica

100.42 ± 3.48

93.86* ± 2.96

76.08 ± 3.24

35.78 ± 1.20

13.31 ± 0.82

10.07 ± 0.46

Total specimens

92.0-106.1

85.8*-98.8

68.2-82.4

32.3-38.2

11.7-14.8

9.1-11.0

22

37

34

37

32

37

M . ocellata

93.4*

87.7

70.9

34.0

12.55

9.1

Dzibilchaltiin

1

1

1

1

12.4-12.7

2

1

M. ocellata Mayapan

95.0

88.45

87.3-89.6

70.09 ± 2.23
67.2-73.2

32.74 ± 1.28
30.0-34.9

12.26 ± 0.68
10.9-13.3

9.14 ± 0.56
8.2-10.2

1

2

10

21

21

13

M . ocellata

89.25 ± 3.90

84.08 ± 3.21

68.43 ± 3.17

31.99 ± 1.31

12.03 ± 0.74

8.81 ± 0.61

Yucatan, Mexico and

81.3-93.7

78.9-87.7

62.0-72.2

30.0-34.1

10.6-12.9

7. 9-9. 8

Peten, Guatemala

10

10

10

8

10

10

M. crassipes

79.6*

76.0*

—

29.1

9.8

8.3

San Josecito Cave

1

1

1

1

1

a May possibly represent a female;
(1915).

* Slightly damaged specimens.

b from Brodkorb 1964b; c from Cope 1871; d from Marsh

1872. Probably the same bone measured by Shufeldt

Table 5. Measurements (in mm) of the coracoid of female turkeys, with mean, standard deviation, observed range, and sample size. See Fig. 2
for explanation of measurements.

Head to
External End
of Sternal
Facet

Head to
Internal
Distal Angle

Head to
Pneumatic
Foramen

Head through
Scapular
Facet

Depth
of Head

Least Width
of Shaft

Meleagridinae, genus

72.0**

70.4**

—

—

—

7.9

and species indet.

1

1

1

Gilliland

Meleagris cf. M .

77.90*

75.20*

60.73*

—

10.30

8.40

leopoldi or M . anza

76.9*-78.9*

74. 7*— 75. 7*

60.2*— 61. 1*

10.0-10.6

00

Oo

1

00

O'

Inglis IA

2

2

3

2

5

Meleagris sp.

86.25

80.50

66.95

31.40

11.85

8.95

Coleman IIA

83.7-88.8

78.2-82.8

64.2-69.7

31.0-31.8

11.8-11.9

8. 7-9.2

2

2

2

2

2

2

Meleagris cf.

—

—

—

31.20

11.55

—

M . gallopavo

30.8-31.6

11.5-11.6

Rock Spring

2

2

Meleagris cf.

—

—

—

31.0

12.1

—

M . gallopavo

1

1

Reddick IB

M. gallopavo

86.45*

81.95*

65.35

31.30

11.70

9.20

Ichetucknee River

83.1-89.8*

79.9-84.0*

65.3-65.4

30.4-32.1

11.2-12.2

O'

1

O'

00

2

2

2

3

4

4

M. gallopavo

—

—

64.8

31.05

10.2

8.35

Seminole Field

30.0-32.1

8. 1-8.6

1

2

1

2

Steadman: Turkey Osteology and Paleontology

175

Table 5. Continued.

Head to
External End
of Sternal
Facet

Head to
Internal
Distal Angle

Head to
Pneumatic
Foramen

Head through
Scapular
Facet

Depth
of Head

Least Width
of Shaft

Meleagris sp.

—

77.5*

63.5*

30.1*

11.1*

8.2

Florida Lime Company

1

1

1

1

1

M. gallopavo

81.25

76.00*

62.57*

28.33*

10.57

8.08

Nichol’s Hammock

81.2-81.3

75.8*— 76.2

59.6*— 64. 1

28.0*-28.6

9.8-11.0

8.0-8. 2

2

2

3

3

3

4

M. gallopavo

83.9

77.6

62.3

29.2

10.8

8.0

Good’s Shellpit

1

1

1

1

1

1

M . gallopavo

92.11 ± 2.29

86.22 ± 2.37

70.11 ± 2.27

32.74 ± 1.09

12.44 ± 0.64

9.25 ± 0.45

Buffalo Site

86.1-96.1

80.7-91.5

65.4-74.0

30.2-35.3

10.8-13.7

7.9-10.5

15

16

IS

32

31

56

M. gallopavo silvestris

87.68 ± 1.71

81.57 ± 1.71

66.51 ± 1.88

32.18 ± 0.33

12.08 ± 0.46

8.78 ± 0.48

New York, Pennsylvania,

85.8-90.2

80.0-84.7

64.2-71.0

31.6-32.7

11.3-13.0

8. 1-9.6

Virginia

13

12

13

13

13

13

M . gallopavo osceola

83.40 ± 2.24

77.92 ± 2.18

63.73 ± 2.23

29.73 ± 1.10

10.94 ± 0.76

8.37 ± 0.49

Florida

80.6-88.0

74.7-82.7

60.7-67.6

28.0-31.6

10.0-12.4

7. 8-9. 3

10

10

11

11

11

11

M. gallopavo intermedia

85.8*

80.5*

65.9*

31.1*

12.0

8.2

Texas

1

1

1

1

1

1

M. gallopavo mexicana

87.80

81.77

65.37

31.83

12.57

8.70

Chihuahua, Coahuila,

85.2-92.5

79.3-86.3

63.3-68.4

30.0-34.2

11.6-13.8

8. 2-9.0

Mexico

3

3

3

3

3

3

M. gallopavo merriami

88.67

83.00

66.70

31.53

11.50

8.50

Arizona

85.2-90.6

79.6-85.0

63.3-70.2

31.0-31.8

11.1-11.9

8. 2-8. 7

3

3

3

3

3

3

M. gallopavo

86.10 ± 3.24

80.21 ±3.11

65.16 ± 2.79

31.08 ± 1.52

11.64 ± 0.84

8.56 ± 0.48

Total skeletal

79.8-92.5

73.7-86.3

58.1-71.0

28.0-34.2

10.0-13.8

7. 8-9. 6

specimens3

31

30

32

32

32

32

M. californica

81.64 ± 3.42

76.16 ± 2.50

61.84 ± 2.24

28.85 ± 1.06

10.86 ± 0.60

8.08 ± 0.47

Rancho La Brea

73.1-84.8

68.9-79.2

56.7-66.1

25.4-30.2

9.5-11.8

6. 8-8. 8

11

27

25

24

27

27

M . californica

—

—

—

29.46

11.13

8.25

Carpinteria

28.3-30.0

10.6-11.7

8.0-8. 5

7

7

4

M . californica

81.64 ± 3.42

76.16 ± 2.50

61.84 ± 2.24

28.99 ± 1.00

10.92 ± 0.57

8.10 ± 0.44

Total

73.1-89.1

68.9-79 .2

56.7-66.1

25.4-30.2

9.5-11.8

6. 8-8. 8

specimens

11

27

25

31

34

31

Meleagris sp.

—

73.9*

59.8*

29.4*

10.2

7.8

Potter Creek Cave

1

1

1

1

1

M. ocellata

—

—

7.5

Dzibilchaltun

1

M. ocellata

80.9

73.60

59.38

26.86 ± 2.40

10.14 ± 0.72

7.61 ± 0.74

Mayapan

71.7-75.7

56.9—61.2

22.2-29.1

9.0-11.1

6. 2-8. 7

1

4

5

11

11

11

M. ocellata

77.76 ± 1.94

72.93 ± 1.73

59.04 ± 1.97

26.74 ± 0.71

10.05 ± 0.46

7.39 ± 0.32

Yucatan, Mexico and

74.8-80.8

70.6-75.9

55.5-61.6

25.9-28.0

9.1-11.0

7. 0-7. 9

Peten, Guatemala

10

11

11

9

11

11

M. crassipes

—

67.30*

57.10*

25.60

8.83

6.90

San Josecito Cave

65. 1*— 68. 7*

55.4*-58.8

25.1-25.9

8.5-9. 1

6. 2-7. 7

3

2

3

3

3

a Includes one specimen from northern Florida not identified to subspecies.
* Slightly damaged specimen.

** Moderately damaged specimen.

176

Steadman: Turkey Osteology and Paleontology

Table 6. Measurements (in mm) of the scapula of male turkeys, with mean, standard deviation, observed range, and sample size. See Fig. 2 for
explanation of measurements.

Proximal

Width

Tip of Acromion
to External Tip
of Glenoid Facet

Depth of
Glenoid Facet

Least Width
of Neck

Meleagris cf. M.

22.94 ± 0.54

26.26 ± 0.74

10.33 ± 0.51

11.36 ± 0.54

leopoldi or M . anza

22.1-24.2

25.6-27.9

9.6-11.2

10.5-12.3

Inglis IA

11

8

15

17

Meleagris sp.

25.57

28.83

11.20

12.50

Coleman IIA

2 4.1-26.6

27.4-29.6

10.8-11.8

12.2-12.8

3

3

3

4

M . g allopavo

28.5

31.8

12.1*

14.0

Manalapan

1

1

1

1

Meleagris cf. M.

26.50*

28.95*

11.00

13.5

gallopavo Melbourne

25.6*— 27.4

28. 1 * — 2 9 . 8

10.9-11. 1

2

2

2

1

M. gallopavo

—

—

10.7

12.0

Ichetucknee River

1

1

Meleagris cf. M.

28.6

31.2

11.9

14.7

gallopavo Mefford Cave

1

1

1

1

M. gallopavo

28.34

31.58

12.26 ± 0.64

14.50 ± 0.48

Buffalo Site

26.9-29.2

30.2-32.4

11.2-13.2

13.7-15.4

5

6

13

12

M . gallopavo

27.56 ± 0.93

31.00 ± 1.09

12.38 ± 0.76

13.81 ± 0.90

silvestris, New York

25.4-29.3

29.0-33.0

10.9-13.4

12.1-16.0

Pennsylvania, Virginia

29

27

16

29

M. gallopavo

25.61 ± 0.80

28.82 ± 1.06

11.51 ± 0.55

12.66 ± 0.76

osceola Florida

24.0-26.5

27.1-30.1

10.9-12.2

11.2-13.5

8

8

8

8

M . gallopavo

—

29.3

—

13.0

intermedia, Texas

1

1

M . gallopavo mexicana

27.7

30.1

12.5

14.6

Chihuahua, Mexico

1

1

1

1

M . gallopavo

27.0

29.6

11.1

13.2

merriami, Arizona

1

1

1

1

M. gallopavo

27.15 ± 1.18

30.44 ± 1.38

12.06 ± 0.78

13.56 ± 0.96

Total skeletal

24.0-29.3

27.1-33.0

10.9-13.4

11.2-16.0

specimens

39

38

26

40

M . californica

23.66 ± 0.72

26.17 ± 0.78

10.56 ± 0.50

12.04 ± 0.87

Rancho La Brea

22.4-25.2

24.8-27.9

9.6-11.3

10.8-13.8

16

16

23

14

M . californica

23.80

26.75

10.78

11.85

Carpinteria

23.5-24.4

26.5-26.9

9.6-12.0

11.7-12.0

5

4

6

2

M. californica

23.69 ± 0.66

26.28 ± 0.74

10.60 ± 0.58

12.01 ± 0.82

Total specimens

22.4-25.2

24.8-27.9

9.6-12.0

10.8-13.8

21

20

29

16

M. ocellata

21.75 ± 0.68

24.80 ± 0.73

9.92 ± 0.34

11.14 ± 0.56

Mayapan

20.7-22.8

23.8-26.0

9.3-10.7

10.1-12.2

12

14

28

18

M. ocellata

20.14 ± 1.29

22.96 ± 1.28

9.38

10.23 ± 0.83

Yucatan, Mexico

18.7-22.5

21.1-25.1

8.9-10.2

9.4-11.9

and Peten, Guatemala

10

10

7

10

Slightly damaged specimen.

Steadman: Turkey Osteology and Paleontology

177

Table 7. Measurements (in mm) of the scapula of female turkeys, with mean, standard deviation, observed range, and sample size. See Fig. 2
for explanation of measurements.

Proximal

Width

Tip of Acromion
to External Tip
of Glenoid Facet

Depth of
Glenoid Facet

Least Width
of Neck

Meleagris progenes

IS. 9

17.9*

7.9

—

Rexroad

1

1

1

Meleagris cf. M. leopoldi

17.1

23.8

8.1

9.2

or M. anza Inglis IA

1

1

1

1

Meleagris sp.

2 1. IS*

23.40*

9.45

10.30

Haile XVIA

20. 1*— 22.2

22.3*-24.5

8.7-10.2

9.5-11.1

2

2

2

2

Meleagris sp.

22.2

24.6

9.4

—

Coleman IIA

1

1

1

Meleagris cf. M.

—

—

9.1

—

gallopavo Reddick IB

1

M. gallopavo

20.0

22.0

8.1

—

Seminole Field

1

1

1

M. gallopavo

19.87

22.20

8.33

9.7

Nichol’s Hammock

18. 9-20. S

21.2-23.0

7.9-9. 1

9. 7-9. 9

3

3

3

3

Meleagris sp.

22.0*

—

9.6

-12.0

Howell’s Ridge Cave

1

1

1

M . gallopavo

22.43

2S.23

9.70

11.08

Buffalo Site

22.1-23.0

24.4-26.4

9.1-10.2

10.8-11.3

3

3

7

5

M. gallopavo silvestris,

21.87 ± 0.36

24.30 ± 0.49

9.63

10.32 ± 0.54

New York, Pennsylvania,

21.0-22.3

23.2-2S.O

9.0-10.3

9.4—11.1

Virginia

12

12

6

12

M. gallopavo osceola

20.18 ± 0.64

22.77 ± 0.S4

8.75 ± 0.32

9.97 ± 0.59

Florida

18.9-21.1

22.0-23.8

7.9-9. 1

9.0-11.0

11

11

11

11

M. gallopavo mexicana

21.30

23.60

9.40

10.23

Chihuahua, Coahuila,

20.0-22.6

22.2-24.9

9.0-10.0

9.9-10.6

Mexico

3

3

3

3

M. gallopavo merriami

21.23

23.33

9.53

10.30

Arizona

20.5-21.8

22.3-24. 1

9. 2-9. 9

10.0-10.7

3

3

3

3

M. gallopavo

21.03 ± 1.03

23.46 ± 1.05

9.13 ± 0.60

10.15 ± 0.55

Total skeletal

18.9-22.6

20.8-25.0

7.9-10.3

9.0-11.1

specimens3

30

30

24

30

M. californica

19.25 ± 0.78

21.82 ± 0.94

8.70 ± 0.50

9.40 ± 0.48

Rancho La Brea

17.9-21 .2

20.7-24.6

7.8-10.0

8.7-10.8

19

18

21

22

M . californica

19. S3

22.10

8.98

9.50

Carpinteria

19.0-19.8

22.0-22.2

8. 5-9. 4

8. 9-9. 9

3

4

4

3

M . californica

19.29 ± 0.74

21.87 ± 0.85

8.74 ± 0.49

9.41 ± 0.48

Total specimens

17.9-21.2

20.7-24.6

7.8-10.0

8.7-10.8

22

22

25

25

M. ocellata

17.78

20.42

8.18 ± 0.42

8.72 ± 0.44

Mayapan

16.7-19.0

19.3-21.7

7. 6-8. 8

8. 1-9.2

7

6

10

10

M . ocellata

17.34 ± 0.40

19.51 ± 0.54

8.16 ± 0.27

8.86 ± 0.64

Yucatan, Mexico

16.8-18.9

19.0-20.5

7. 6-8. 5

7. 9-9. 8

and Peten, Guatemala

11

11

9

11

3 Includes one specimen from northern Florida not identified to subspecies.
* Slightly damaged specimen.

178

Steadman: Turkey Osteology and Paleontology

Table 8. Measurements (in mm) of the humerus of male turkeys, with mean, standard deviation, observed range, and sample size. See Fig. 3
for explanation of measurements.

Total Length

Proximal Width

Width of
Midshaft

Depth of
Midshaft

Distal Width

Meleagris cf. M.

136.69 ± 1.97

36.42 ± 0.89

15.00 ± 0.55

11.58 ± 0.42

30.05 ± 0.91

leopoldi or M. anza

132.0-141.0

34.9-38.6

14.2-16.1

10.9-12.3

28.0-32.0

Inglis IA

18

22

30

29

23

Meleagris sp.

—

—

—

—

28.1

Haile XVIA

1

Meleagris sp.

141.72

36.83

15.21

12.04

31.30 ± 0.52

Coleman IIA

136.9-146.8

33.6-37.8

14.0-15.9

11.2-12.4

30.0-31.9

6

7

7

7

14

Meleagris cf. M.

—

38.9

—

—

—

g allopavo, Arizpe

1

Meleagris cf. M.

—

—

16.1

12.3

—

gallopavo Burnet Cave

1

1

Meleagris sp. (juv.)

131

—

—

—

—

North Liberty

1

M. gallopavo

—

42.3

—

—

—

Carlisle Cave

1

Meleagris cf. M .

—

—

—

—

31.4

gallopavo

1

Frankstown Cave

M. gallopavo

1S3.2S

42h

—

—

33b

Manalapan

147. 0a — 1 5 9 . 5 b

2

1

1

Meleagris cf. M.

—

39.3

16.6

13.1

—

gallopavo Reddick IB

1

1

1

Meleagris cf. M.

—

—

—

—

30.1

gallopavo Melbourne

1

Meleagris cf. M.

149.0

—

15.9

12.9

32.7

gallopavo Arredondo

1

1

1

1

M . gallopavo

144.5

39.6

15.9

12.0

32.1

Aucilla River

1

1

1

1

1

M. gallopavo

—

39.30

—

—

33.10

Ichetucknee River

38.0-40.6

32.1-34.6

2

3

M . gallopavo

—

—

—

—

32.5

Seminole Field

1

M . gallopavo

—

36.9

15.95

11.65

—

Good’s Shellpit

15.7-16.2

11.6-11.7

1

2

2

M . gallopavo

157.1 ± 5.3

42.11 ± 0.99

16.63 ± 0.79

12.59 ± 0.59

33.59 ± 1.01

Buffalo Site

147-166

39.4-44.0

14.8-18.3

11.1-14.1

32.0-35.5

11

28

107

108

31

M . gallopavo

157.0

—

17.0

13.7

—

Hartman’s Cave

1

1

1

M. gallopavo silvestris,

150.67 ± 3.38

40.75 ± 1.14

16.17 ± 0.84

12.56 ± 0.66

32.69 ± 0.84

New York, Pennsylvania,

144.0-159.0

39.0-43.2

14.2-17.4

11.2-13.8

31.3-34.3

Virginia

28

28

29

29

27

M . gallopavo osceola

148.31 ± 6.63

38.42 ± 1.54

15.65 ± 0.71

12.31 ± 0.54

31.18 ± 1.10

Florida

138.0-159.5

36.0-40.8

14.9-16.4

11.5-13.1

29.1-33.0

8

8

8

8

8

M . gallopavo intermedia

156.0

39.2

16.2

12.6

31.8

Texas

1

1

1

1

1

M . gallopavo mexicana

152.5

40.5

17.3

13.3

34.1

Chihuahua, Mexico

1

1

1

1

1

M . gallopavo merriami,

149.5

39.3

15.7

12.1

31.4

Arizona

1

1

1

1

1

M. gallopavo

150.33 ± 4.26

40.19 ± 1.52

16.08 ± 0.83

12.52 ± 0.63

32.35 ± 1.11

Total skeletal

138.0-159.5

36.0-43.2

14.2-17.4

11.2-13.8

29.1-34.3

specimens

39

39

40

40

38

Steadman: Turkey Osteology and Paleontology

179

Table 8. Continued.

Total Length

Proximal Width

Width of
Midshaft

Depth of
Midshaft

Distal Width

M . California i

135.99 ± 4.24

35.35 ± 1.21

14.79 ± 0.65

11.70 ± 0.56

29.32 ± 0.96

Rancho La Brea

128.0-142.5

33.3-38.5

13.3-15.9

10.8-12.8

26.7-30.8

34

34

34

34

34

M . californica

137.00

36.32 ± 0.60

14.98

12.02

29.75 ± 0.26

Carpinteria

134. 1-138.5

35.5-37.0

14.7-15.6

11.4-12.7

29.2-30.0

5

8

5

5

8

M. californica

136.12 ± 4.00

35.54 ± 1.18

14.82 ± 0.62

11.74 ± 0.55

29.40 ± 0.88

Total specimens

128.0-142.5

33.3-38.5

13.3-15.9

10.8-12.8

26.7-30.8

39

42

39

39

42

M . ocellata

130.0

32.65

13.7

10.9

27.8

Dzibilchaltun

1

31.7-33.6

2

1

1

1

M . ocellata

127.38 ± 2.17

33.63 ± 0.60

13.24

10.10

27.53

Mayapan

123.0-129.5

32.6-34.6

12.3-14.0

9.4-10.9

26.7-28.2

8

1 1

5

5

6

M. ocellata

123.77 ± 3.98

32.51 ± 1.34

12.68 ± 0.73

10.09 ± 0.86

26.60 ± 1.58

Yucatan, Mexico and

118.0-130.5

30.5-34.9

11.1-13.5

8.7-11.7

24.2-28.9

Peten, Guatemala

9

9

9

9

9

M. crassipes

119.07

30.93

12.10

9.70

25.40

San Josecito Cave

117.1-120.2

30.6-31.2

11.3-12.7

9.2-10.5

25.2-25.6

3

3

3

3

2

a From Shufeldt 1915; h from Marsh

1872.

Table 9. Measurements (in mm) of the humerus of female turkeys, with mean,
3 for explanation of measurements.

standard deviation, observed range, and sample size. See Fig.

Total

Length

Proximal

Width

Width of
Midshaft

Depth of
Midshaft

Distal

Width

Meleagris cf. M . leopoldi

1 12.0

29.18

12.42

9.38

24.38

or M. anza Inglis IA

26.9-30.3

11.8-13.2

9.0-9. 7

24.2-24.7

1

5

4

4

4

Meleagris anza

112.4

-30.3

14.3

8.5

-24.0

Vallecito Creeka

1

1

1

1

1

Meleagris sp.

1 19.8*

31.33*

13.4

10.8

25.70

Coleman IIA

30.1*-32.0

1

24.8-26.2

1

3

1

3

Meleagris sp.

122.0

31.4

13.9

10.5

26.30

Papago Springs Cave1'

1

1

1

1

26.2-26.4

2

Meleagris cf. M . gallopavo

125.50

32.45

13.70

10.65

26.40

Carlisle Cave

124.0-127.0

32.2-32. 7

13.3-14.1

10.3-11.0

26.2-26.6

2

2

2

2

2

Meleagris cf. M. gallopavo

—

—

—

—

24.9

Rock Spring

1

M. gallopavo

121.25*

31.35

12.68

9.87

26.35

Ichetucknee River

118.7*— 123.8

30.9-31.8

11.9-14.0

9.2-10.6

25.2-27.0*

2

2

4

3

4

M. gallopavo

—

—

—

—

25.75

Seminole Field

25.0-26.5

Continued

180

Steadman: Turkey Osteology and Paleontology

Table 9. Continued.

Total

Length

Proximal

Width

Width of
Midshaft

Depth of
Midshaft

Distal

Width

M. gallopavo

120.4

29.4

12.95

9.95

24.85

Good’s Shellpit

12.9-13.0

9.8-10.1

24.5-25.2

1

1

2

2

2

M. gallopavo

127.7 ± 3.4

32.67 ± 0.86

13.26 ± 0.55

10.05 ± 0.48

26.86 ± 0.59

Buffalo Site

123-133

30.9-33.9

11.3-14.2

8.0-11.1

25.9-28.2

12

21

115

115

29

M. gallopavo

131.0

34.1

14.05

10.80

27.9

Hartman’s Cave

13.5-14.6

10.7-10.9

1

1

2

2

1

M. gallopavo silvestris,

121.84 ± 2.29

32.15 ± 0.42

12.51 ± 0.56

9.53 ± 0.51

26.08 ± 0.38

New York, Pennsylvania,

117.0-125.0

31.7-33.3

11.7-13.2

8.9-10.3

25.3-26.8

Virginia

13

13

13

13

13

M . gallopavo osceola,

116.14 ± 2.77

29.79 ± 0.81

12.27 ± 0.45

9.42 ± 0.28

24.53 ± 0.79

Florida

1 1 1.2-120.2

28.2-31.2

11.5-12.9

8. 9-9. 8

22.9-25.8

11

11

11

11

11

M . gallopavo intermedia.

120.8

32.0

12.4

9.9

26.1

Texas

1

1

1

1

1

M. gallopavo mexicana,

124.50

32.47

13.27

9.83

26.17

Chihuahua, Coahuila,

121.0-129.8

31.0-34.1

12.9-13.8

9.4-10.3

24.8-27.8

Mexico

3

3

3

3

3

M . gallopavo merriami,

119.7

31.55

13.35

10.20

26.40

Arizona

30.8-32.3

12.8-13.9

10.1-10.3

2 5.8-27.0

1

2

2

2

2

M . gallopavo

119.62 ± 4.13

31.23 ± 1.36

12.53 ± 0.60

9.56 ± 0.45

25.51 ± 1.03

Total skeletal specimens'

111.2-129.8

28.2-34.1

11.5-13.9

8.9-10.3

22.9-27.8

30

31

31

31

31

M. californica

114.52 ± 3.10

28.83 ± 0.74

12.09 ± 0.55

9.31 ± 0.50

24.42 ± 0.54

Rancho La Brea

106.4-121.4

27.0-29.9

10.7-13.4

8.4-1 1.0

23.4-25.4

31

27

30

31

26

M. californica

115.1

28.68

12.0

9.5

24.74

Carpinteria

27.7-29.4

24.2-25.7

1

5

1

1

5

M. californica

114.54 ± 3.05

28.81 ± 0.73

12.09 ± 0.54

9.32 ± 0.49

24.47 ± 0.56

Total specimens

106.4-121.4

27.0-29.9

10.7-13.4

8.4-11.0

23.4-25.7

32

32

31

32

31

Meleagris sp.

116.9

31.9

-13.0

~9.9

26.2

Potter Creek Cave

1

1

1

1

1

M . ocellata

110.40

27.42

11.33

8.73

22.94 ± 1.68

Mayapan

108.8-112.0

26.1-28.5

11.0-12.0

7. 9-9. 2

19.8-25.0

2

6

3

3

8

M. ocellata

107.59 ± 2.03

27.68 ± 0.54

11.16 ± 0.29

8.94 ± 0.20

22.71 ± 0.78

Yucatan, Mexico and

104.2-110.7

26.9-28.5

10.8-11.6

8. 6-9. 2

21.7-24.1

Peten, Guatemala

10

10

10

10

10

M . crassipes

105.98

27.74

11.26

8.92

22.54

San Josecito Cave

102.1-111. 1

25.8-28.9

10.8-11.8

8. 6-9. 3

21.6-23.4

5

5

5

5

5

a From Howard 1963; b may possibly represent a male; c includes one specimen from northern Florida not identified to subspecies.
* Slightly damaged specimen.

Steadman: Turkey Osteology and Paleontology

181

Table 10. Measurements (in mm) of the ulna of male turkeys, with mean, standard deviation, observed range, and sample size. See Fig. 4 for
explanation of measurements.

Total

Length

Proximal

Width

Width of
Midshaft

Depth of
Midshaft

Distal

Depth

Meleagris cf. M. leopoldi

134.03* ± 2. 85

18.14 ± 0.75

8.70 ± 0.34

9.70 ± 0.38

15.16 ± 0.65

or M. anza Inglis IA

128.0*-139.0*

16.0-19.0

8. 1-9.1

9.0-10.5

14.3-16.1

15

10

32

32

12

Meleagris sp.

137.40

19.15

8.98 ± 0.43

10.39 ± 0.38

15.95 ± 0.30

Coleman IIA

133.4-139.4

18.7-19.8

8. 2-9. 9

9.9-11.0

15.3-16.3

5

6

12

11

11

Meleagris sp.

—

16.5*

—

—

14.7

Sheffield Gravel Pits3

1

1

M . gallopavo

155.0*

20.00*

9.57

10.63

16.05*

Manalapan

153*— 15 7*

19.0*-2 1.0

9.0-10.0

10.3-10.9

15.5* — 16.6

2

2

3

3

2

Meleagris cf. M. gallopavo

—

—

—

—

16.3*

Sante Fe IIA

1

Meleagris cf. M . gallopavo

—

19.3

9.5

10.9

—

Reddick IB

1

I

1

M. gallopavo

141.50*

19.77

9.30

10.73

16.45

Ichetucknee River

139.0*-144.0

19.0-20.3

8. 9-9. 6

10.5-10.9

16.0-16.9

2

3

3

3

2

M. gallopavo

—

—

—

—

16.30

Seminole Field

16.1-16.4

3

Meleagris cf. M. gallopavo

10.4

12.0

Haile IIA

1

1

Meleagris cf. M. gallopavo

150.0

20.0

10.0

11.2

17.4*

Mefford Cave I

1

1

1

1

1

Meleagris sp.

—

—

—

—

16.4*

Santa Fe IVA

1

M. gallopavo

—

—

9.0

10.2

15.8*

Good’s Shellpit

1

1

1

M . gallopavo

152.33

20.18 ± 0.45

9.62 ± 0.42

10.90 ± 0.40

17.26 ± 0.68

Buffalo Site

150.0-157.0

19.7-21.1

9.0-10.7

10.1-11.6

16.3-18.8

3

10

45

45

16

M . gallopavo silvestris

148.54 ± 3.63

19.72 ± 0.88

9.43 ± 0.62

11.00 ± 0.56

16.60 ± 0.57

New York, Pennsylvania,

143.0-158.0

18.1-21.7

8.0-10.4

10.0-12.0

15.2-17.9

Virginia

24

24

23

23

23

M. gallopavo osceola,

146.89 ± 5.75

18.89 ± 0.64

8.99 ± 0.49

10.15 ± 0.49

15.54 ± 0.54

Florida

139.8-156.0

17.9-19.8

8. 2-9.8

9.4-11.0

14.9-16.3

8

8

8

8

8

M. gallopavo intermedia,

157.0

20.5

10.2

10.6

16.3

Texas

1

1

1

1

1

M. gallopavo mexiana

150.0

19.3

10.0

11.2

17.9

Chihuahua, Mexico

1

1

1

1

1

M. gallopavo

148.44 ± 4.36

19.53 ± 0.88

9.36 ± 0.62

10.79 ± 0.64

16.38 ± 0.75

Total skeletal specimens

139.8-158.0

17.9-21.7

8.0-10.4

9.4-12.0

14.9-17.9

34

34

33

33

33

M. californica

132.96 ± 3.81

18.14 ± 0.41

8.73 ± 0.25

9.89 ± 0.28

15.28 ± 0.54

Rancho La Brea

124.6-139.0

17.3-18.7

7. 9-9. 3

9.1-10.4

14.5-16.3

36

18

36

36

22

M. californica

128.50

17.58

8.74 ± 0.30

9.82 ± 0.13

15.14 ± 0.63

Carpinteria

123.6-131.0

16.4-18.3

8. 1-9.1

9.7-10.1

14.3-16.2

7

6

9

9

8

M. californica

132.23 ± 3.98

18.00 ± 0.54

8.73 ± 0.26

9.88 ± 0.26

15.24 ± 0.56

Total specimens

123.6-139.0

16.4-18.7

7.9-9 .3

9.1-10.4

14.3-16.3

43

24

45

45

30

M. ocellata

136.0

17.10*

7.95

9.60

15.3

Dzibilchaltun

16.2-18.0*

7. 3-8. 6

9.2-10.0

1

2

2

2

1

Continued

182

Steadman: Turkey Osteology and Paleontology

Table 10. Continued.

Total

Length

Proximal

Width

Width of
Midshaft

Depth of
Midshaft

Distal

Depth

Meleagris cf. M. ocellata,

127.0*

—

Tulum

1

M. ocellata

125.70 ± 3.22

16.32 ± 1.29

7.31 ± 0.31

8.98 ± 0.63

13.50 ± 1.21

Yucatan, Mexico and

120.5-130.5

14.0-17.8

6. 8-7. 7

7. 7-9.6

11.6-15.3

Peten, Guatemala

8

8

8

8

8

M . crassipes

113.95

16.0*

7.40

8.47

13.05*

San Josecito Cave

112.9-115.0

7. 3-7. 5

8.2-8. 7

13.0-13.1*

2

1

3

3

2

a May possibly represent a female.
* Slightly damaged specimen.

Steadman: Turkey Osteology and Paleontology

183

Table 11. Measurements (in mm) of the ulna of female turkeys, with mean, standard deviation, observed range, and sample size. See Fig. 4 for
explanation of measurements.

Total

Length

Proximal

Width

Width of
Midshaft

Depth of
Midshaft

Distal

Depth

Meleagris cf. M. leopoldi

107.75*

15.47

7.38

8.24

12.90

or M. anza Inglis IA

103.0*— 1 12.5*

15.4-15.6

7. 1-8.1

7. 9-8. 9

12.2-13.6

2

3

6

6

4

Meleagris sp.

121.0

15.47

7.57

8.78

13.38*

Coleman IIA

15.3-15.8

7. 3-8.0

8. 2-9. 3

13.0-13.7*

1

3

3

4

4

Meleagris sp.

—

13.4**

—

—

12.2*

Howell’s Ridge Cave

1

1

M . gallopavo

1 18.90

—

8.9

8.70

13.7

Ichetucknee River

118.2-119.6

2

1

8.7

2

l

M. gallopavo

14.83*

12.9*

Seminole Field

14.5*— 15. 1

3

1

M. gallopavo

112.08

14.38

6.88

7.68

12.10

Nichol’s Hammock

107.7-115.1

13.3-15.0

6.4-7. 1

7. 3-7. 9

11.9-12.3

4

5

4

4

4

M. gallopavo

108.6

13.9

7.0

8.0

11.3

Good’s Shellpit

1

1

1

1

1

M. gallopavo

122.50 ± 3.37

16.39 ± 0.38

7.75 ± 0.30

8.76 ± 0.34

13.86 ± 0.37

Buffalo Site

116.0-127.0

15.8-17.2

7. 0-8. 2

8. 0-9. 5

13.1-14.5

12

10

52

52

17

M . gallopavo silvestris,

118.71 ± 3.32

15.45 ± 0.42

7.14 ± 0.33

8.39 ± 0.37

13.21 ± 0.36

New York, Pennsylvania,

112.9-123.0

15.0-16.2

6. 8-7. 9

8. 0-9.0

12.9-14.0

Virginia

12

13

12

12

12

M. gallopavo osceola,

113.62 ± 3.29

14.99 ± 0.62

7.06 ± 0.17

7.84 ± 0.28

12.45 ± 0.36

Florida

109.8-120.2

13.9-16.1

6. 7-7.2

7. 3-8. 2

11.9-12.8

11

11

1 1

11

11

M . gallopavo mexicana

123.25

16.80

7.75

8.75

14.25

Chihuahua, Mexico

119.1-127.4

16.1-17.5

7. 3-8. 2

8.4-9. 1

14.0-14.5

2

2

2

2

2

M . gallopavo merriami,

125.5

18.1

7.5

9.1

14.0

Arizona

1

1

1

1

1

M . gallopavo

116.82 ± 5.04

15.45 ± 0.86

7.16 ± 0.34

8.18 ± 0.50

12.95 ± 0.69

Total skeletal specimens3

107.8-127.4

13.9-18.1

6. 7-8.2

7.3-9. 1

11.7-14.5

27

28

27

27

27

M. californica

108.95 ± 2.98

14.37 ± 0.42

6.98 ± 0.31

8.03 ± 0.27

12.24 ± 0.36

Rancho La Brea

102.7-115.5

13.6-15.1

6. 2-7.4

7. 6-8. 6

11.7-13.0

21

17

21

21

16

M. californica

110.92

14.98

7.08

8.21

12.61 ± 0.40

Carpinteria

108.1-115.8

14.6-15.6

6. 9-7. 4

8. 0-8. 4

12.1-13.4

4

4

7

7

8

M . californica

109.26 ± 3.06

14.48 ± 0.48

7.0! ± 0.28

8.08 ± 0.26

12.36 ± 0.41

Total specimens

102.7-115.8

13.6-15.6

6. 2-7.4

7. 6-8. 6

11.7-13.4

25

21

28

28

24

M. ocellata

108.39 ± 2.20

13.82 ± 0.73

6.40 ± 0.38

7.88 ± 0.35

11.78 ± 0.37

Yucatan, Mexico and

104. 1 — 112.1

12.8-15.0

5. 8-6. 8

7. 3-8. 3

11.1-12.3

Peten, Guatemala

10

10

10

10

10

M. eras sipes

106.35

13.5*

6.50

7.53

11.63

San Josecito Cave

106.0-106.7

5. 9-6. 8

7. 4-7. 8

11.3-11.9

2

1

6

6

3

3 Includes one specimen from northern Florida not identified to subspecies.
* Slightly damaged specimen.

** Moderately damaged specimen.

184

Steadman: Turkey Osteology and Paleontology

Table 12. Measurements (in mm) of the radius of male turkeys, with mean, standard deviation, observed range, and sample size. See Fig. 4 for
explanation of measurements.

Total

Length

Proximal

Width

Proximal

Depth

Least Width
of Shaft

Least Depth
of Shaft

Distal

Width

Meleagris cf. M. leopoldi

119.93*

9.47

10.64 ± 0.58

5.02 ± 0.18

4.61 ± 0.28

12.72 ± 0.51

or M . anza Inglis IA

1 17.8*— 12 1.5*

9.1-10.2

9.7-11.7

4. 6-5. 3

3. 9-4. 9

11.8-13.7

3

6

10

13

14

1 1

Meleagris sp.

120. SS

10.12

11.02

5.48

4.40

13.40

Coleman IIA

119.3-121.8

9.8-10.4

10.3-11.5

5. 1-5. 7

4. 0-5.0

12.9-13.9

2

4

4

4

3

2

M. gallopavo

142*

—

—

—

—

—

Manalapan

1

Meleagris cf. M. gallopavo

—

10.3

—

—

4.2

13.1

Rock Spring

1

1

1

Meleagris cf. M . gallopavo

—

—

—

—

4.7

13.1

Reddick IB

1

1

M. gallopavo

135. S*

—

—

5.3

4.2

—

Aucilla River

1

1

1

M. gallopavo

137.0

11.4

12.1

5.9

4.9

15.0

Ichetucknee River

1

1

1

1

1

1

Meleagris sp.

—

—

—

—

5.0

13.1*

Florida Lime Co.

1

1

Meleagris cf. M . gallopavo

133.5

11.0

12.9

6.0

5.0

14.9

Mefford Cave I

1

1

1

1

1

1

Meleagris cf. M . gallopavo

137.0

11.0*

11.0*

6.4

5.6

13.9*

Steinhatchie River

1

1

1

1

1

1

M. gallopavo

—

—

—

5.1

4.6

13.9

Nichol’s Hammock

1

1

1

M . gallopavo

138.25

11.51 ± 0.55

12.15 ± 0.40

5.66 ± 0.31

4.82 ± 0.22

14.42 ± 0.34

Buffalo Site

136.0-140.0

10.8-12.3

11.1-12.8

5. 2-6. 2

4. 4-5. 2

14.0-15.0

4

15

17

21

25

13

M. gallopavo silvestris.

134.60 ± 4.32

10.90 ± 0.53

11.83 ± 0.49

5.57 ± 0.38

4.72 ± 0.30

13.77 ± 0.44

New York, Pennsylvania,

127.0-145.0

9.8-11.8

10.7-12.7

5. 0-6. 4

4.0-5. 1

12.8-15.0

Virginia

24

23

23

23

23

23

M. gallopavo osceola.

132.94 ± 5.00

9.98 ± 0.57

10.79 ± 0.36

5.14 ± 0.16

4.32 ± 0.24

12.94 ± 0.37

Florida

126.7-141.9

9.3-10.9

10.2-11.2

4. 9-5. 5

4.0-4. 7

12.3-13.4

9

9

9

9

9

9

M. gallopavo intermedia,

142.0

11.7

11.7

5.6

—

13.6

Texas

1

1

1

1

1

M . gallopavo

134.38 ± 4.63

10.67 ± 0.69

11.54 ± 0.65

5.45 ± 0.38

4.61 ± 0.34

13.54 ± 0.55

Total skeletal specimens

126.7-145.0

9.3-11.8

10.2-12.7

4. 9-6. 4

4.0-5. 1

12.3-15.0

34

33

33

33

32

33

M. californica

117.66 ± 2.62

9.46 ± 0.42

10.58 ± 0.45

4.99 ± 0.24

4.43 ± 0.21

12.55 ± 0.39

Rancho La Brea

1 11.5-121.6

8.7-10.4

9.9-11.8

4. 6-5. 6

4.0-4. 9

11.9-13.3

22

28

28

21

35

30

M . californica

115.98

9.49 ± 0.33

10.62 ± 0.50

5.12

4.60

12.43

Carpinteria

113.8—1 19. 1

8.9-10.0

9.8-11.3

5. 0-5. 4

4. 2-4. 9

12.1-12.8

4

8

8

5

6

6

M. californica

117.40 ± 2.60

9.47 ± 0.40

10.59 ± 0.46

5.02 ± 0.23

4.45 ± 0.23

12.53 ± 0.37

Total specimens

111.5-121.6

8.7-10.4

9.8-11.3

4. 6-5. 6

4. 0-4. 9

11.9-13.3

26

36

36

26

41

36

Meleagris cf. M.

120.5

9.9

10.6

4.9

4.1

12.3

californica

1

1

1

1

1

1

La Mirada

Meleagris sp.

—

—

—

—

—

12.6*

Imperial Highway

1

M. ocellata

112.70 ± 2.67

9.00 ± 0.94

9.54 ± 0.73

4.38 ± 0.52

3.69 ± 0.48

11.21 ± 1.32

Yucatan, Mexico and Peten,

108.1-117.0

7.3-10.1

8.6-10.6

3.4-4. 9

2. 9-4. 2

9.1-13.0

Guatemala

8

8

8

8

8

8

M. crassipes

97.8*

8.9

—

4.7

4.1

11.2*

San Josecito Cave

1

1

1

1

1

Slightly damaged specimen.

Steadman: Turkey Osteology and Paleontology

185

Table 13. Measurements (in mm) of the radius of female turkeys, with mean, standard deviation, observed range, and sample size. See Fig. 4
for explanation of measurements.

Total

Length

Proximal

Width

Proximal

Depth

Least Width
of Shaft

Least Depth
of Shaft

Distal

Width

Meleagris cf. M. leopoldi

100.9

7.70

8.2

3.94

3.55

10.35

or M . anza Inglis IA

7. 2-8. 2

3. 7-4. 2

3. 3-3. 8

10.2-10.5

1

2

1

5

4

2

Meleagris sp.

—

—

—

—

4.0

11.1

Coleman IIA

1

1

Meleagris cf. M. gallopavo

—

—

—

—

3.6

—

Rock Spring

1

M. gallopavo

111.60 ± 3.20

8.88 ± 0.38

9.81 ± 0.41

4.55 ± 0.28

3.94 ± 0.20

11.88 ± 0.34

Buffalo Site

107.0-117.0

8. 1-9.5

9.1-10.7

4. 1-5.1

3. 7-4.6

11.4-12.9

10

16

16

13

18

16

M . gallopavo silvestris,

107.70 ± 3.17

8.61 ± 0.32

9.23 ± 0.26

4.22 ± 0.25

3.65 ± 0.27

11.04 ± 0.40

New York, Pennsylvania,

102.0-112.7

8. 2-9. 2

8. 9-9. 8

3. 9-4. 7

3. 2-4.0

10.3-11.8

Virginia

12

13

13

12

12

12

M. gallopavo osceola,

102.89 ± 3.13

7.84 ± 0.41

8.34 ± 0.40

4.07 ± 0.20

3.38 ± 0.23

10.52 ± 0.41

Florida

99.2-108.6

7. 2-8. 6

7.8-9. 1

3. 8-4. 4

3. 1-3. 7

9.9-11.1

11

11

1 1

11

11

11

M. gallopavo mexicana

111.30

9.00

9.70

4.20

3.95

11.55

Chihuahua, Mexico

108.1-114.5

8. 1-9.9

9. 5-9. 9

4. 1-4.3

3. 7-4. 2

11.2-11.9

2

2

2

2

2

2

M. gallopavo merriami,

112.3

8.9

10.4

4.5

3.7

12.1

Arizona

1

1

1

1

1

1

M. gallopavo

106.14 ± 4.23

8.32 ± 0.59

8.91 ± 0.66

4.16 ± 0.23

3.54 ± 0.31

10.87 ± 0.57

Total skeletal specimens3

99.2-114.5

7. 2-9. 9

7.8-10.4

3.8-4. 7

3. 1-4.2

9.9-12.1

27

28

28

27

27

27

M. californica

97.72 ± 2.60

7.74 ± 0.38

8.51 ± 0.38

4.13 ± 0.20

3.59 ± 0.20

10.41 ± 0.30

Rancho La Brea

93.8-104.9

7. 1-8.4

8. 0-9. 2

3. 9-4. 5

3. 3-4.0

9.8-10.8

13

18

15

13

20

17

M. californica

96.05

8.28

8.62

4.07

3.72

10.50

Carpinteria

95.3-96.8

7. 9-8. 5

8. 1-9.1

3. 9-4. 4

3. 5-4.0

10.3-10.9

2

4

4

3

4

4

M. californica

97.50 ± 2.50

7.84 ± 0.42

8.53 ± 0.40

4.12 ± 0.21

3.61 ± 0.21

10.43 ± 0.29

Total specimens

93.8-104.9

7. 1-8.5

8.0-9. 2

3. 9-4. 5

3. 3-4.0

9.8-10.9

15

22

19

16

24

21

M . ocellata

97.20 ± 2.35

7.66 ± 0.40

8.10 ± 0.30

3.80 ± 0.27

3.27 ± 0.24

9.83 ± 0.58

Yucatan, Mexico and Peten,

92.8-100.7

7. 1-8.3

7. 7-8. 8

3.5-4. 1

2. 9-3. 7

9.2-11.0

Guatemala

10

10

10

10

10

10

M . crassipes

—

—

—

4.0

3.5

—

San Josecito Cave

1

1

Includes one specimen from Florida not identified to subspecies.

186

Steadman: Turkey Osteology and Paleontology

Table 14. Measurements (in mm) of the carpometacarpus of male turkeys, with mean, standard deviation, observed range, and sample size. See
Fig. S for explanation of measurements.

Total

Length

Proximal

Depth

Length
of Meta-
carpal I

Least
Width
of Meta-
carpal II

Least
Depth
of Meta-
carpal II

Greatest
Depth of
Intermeta-
carpal
Space

Distal

Depth

Protrusion
of Meta-
carpal III
beyond
Knob of
Metacarpal
II

Meleagris progenes

66. 6a

20. 0a

10.0

80a

5.7

8.0*

17.2

3. 8a

Rexroad

1

1

1

1

1

1

1

1

Meleagris cf. M .

72.63 ± 1.86

20.72 ± 0.64

10.80 ± 0.54

8.41 ± 0.43

6.26 ± 0.29

7.18 ± 0.47

18.68 ± 0.86

3.30 ± 0.40

leopoldi or M .

69.S-76.3

19.6-22.0

9.7-11.8

7. 8-9. 4

5. 3-6. 9

6. 4-8.0

17.3-20.1

2. 8-4.0

anza, Inglis IA

IS

10

17

27

27

13

12

17

Meleagridinae, cf.

73.6

19.6

12.2

—

6.8

—

18.0

4.1

Meleagris, Port

1

1

1

1

1

1

Kennedy Cave

Meleagris sp.

75.64

22.00

11.00

8.55

6.28

7.48

19.80

3.50

Coleman IIA

73.6-77.0

21.4-22.8

10.3-11.6

8. 2-8. 9

5. 9-6. 9

7. 0-8. 5

18.6-20.8

3. 1-3.9

S

6

6

6

6

4

3

6

Meleagris cf. M.

—

—

—

8.0

6.55

—

—

3.80

g allopavo

6. 3-6. 8

3. 6-4.0

Rock Spring

1

2

2

Meleagris cf. M.

76.7

22.7

—

8.5

6.4

—

—

3.57

gallopavo

3. 3-4.0

Reddick IB

1

1

1

1

3

Meleagris cf. M.

79.0

21.85*

11.60

8.20

6.0

—

—

3.9

gallopavo

21.8*— 21.9*

11.3-11.9

8.2

Melbourne

1

2

2

2

1

1

M. gallopavo

78.28 ± 2.40

23.00

11.98 ± 0.44

8.90 ± 0.58

6.59 ± 0.47

8.28

21.05*

3.83

Ichetucknee

74.0-82.1

21.9-24.9

11.3-12.7

8.2-10.0

6. 1-7.9

7.8-9. 1

20.3-22.6*

3. 4-4. 3

River

8

5

8

8

9

5

4

7

M. gallopavo

—

—

11.4

9.25

6.65

—

—

4.03

Seminole Field

9. 1-9.4

6. 3-7.0

3. 8-4. 5

1

2

2

3

Meleagris sp.

—

—

—

8.2

5.90

—

—

3.35

Florida Lime Co.

5. 8-6.0

3. 1-3.6

1

2

2

Meleagris cf. M.

83.6

—

13.0

9.5

6.5

8.1

21.5

3.8

gallopavo

1

1

1

1

1

1

1

Mefford Cave I

M . gallopavo

—

21.8

12.7

—

—

—

—

—

Nichol’s Hammock

1

1

M . gallopavo

77.1

—

—

7.4

6.1

—

—

3.1

Good’s Shellpit

1

1

1

1

M. gallopavo

82.80 ± 2.13

24.36 ± 0.71

12.31 ± 0.60

8.98 ± 0.54

6.80 ± 0.35

8.22 ± 0.55

21.56 ± 0.68

3.99 ± 0.41

Buffalo Site

79.0-89.5

22.7-26.1

10.8-13.8

7.5-10.1

5. 9-7. 8

6. 5-9. 5

19.8-23.1

3.0-4. 9

59

61

52

115

120

65

34

66

M. gallopavo

—

—

—

8.3

6.6

8.5

—

—

Hartman’s Cave

1

1

1

M. gallopavo

80.29 ± 2.22

23.60 ± 0.88

11.85 ± 0.48

8.97 ± 0.71

6.34 ± 0.48

8.00 ± 0.70

20.64 ± 0.74

4.24 ± 0.49

silvestris,

76.1-83.7

21.9-25.2

11.0-13.1

7.6-10.2

5. 7-7.5

6. 5-9. 3

18.9-22.1

3.3-5. 7

New York,

26

26

26

26

26

25

26

25

Pennsylvania,

Virginia

M. gallopavo

79.60 ± 3.00

22.18 ± 0.78

11.68 ± 0.38

8.14 ± 0.32

6.41 ± 0.46

7.51 ± 0.42

19.41 ± 0.79

3.62 ± 0.37

osceola, Florida

74.3-84.2

20.6-23.2

11.0-12.2

7. 6-8. 6

5. 9-7.0

7. 0-8. 2

18.4-20.9

3. 1-4.1

9

9

9

9

9

9

9

9

M. gallopavo

84.0

23.1

14.5

9.0

6.6

7.4

19.7

4.2

intermedia,

1

1

1

1

1

1

1

1

Texas

M. gallopavo

80.22 ± 2.46

23.23 ± 1.04

11.88 ± 0.64

8.77 ± 0.72

6.36 ± 0.47

7.86 ± 0.66

20.30 ± 0.92

4.08 ± 0.53

Total skeletal

74.3-84.2

20.6-25.2

11.0-14.5

7.6-10.2

5. 7-7.5

6. 5-9. 3

18.4-22.1

3. 1-5.7

specimens

36

36

36

36

36

35

36

35

Steadman: Turkey Osteology and Paleontology

187

Table 14. Continued.

Total

Length

Proximal

Depth

Length
of Meta-
carpal I

Least
Width
of Meta-
carpal II

Least
Depth
of Meta-
carpal II

Greatest
Depth of
Intermeta-
carpal
Space

Distal

Depth

Protrusion
of Meta-
carpal III
beyond
Knob of
Metacarpal
II

M . californica

72.62 ± 2.03

20.58 ± 0.66

11.28 ± 0.56

8.11 ± 0.54

6.25 ± 0.32

7.45 ± 0.50

19.19 ± 0.63

3.38 ± 0.38

Rancho La Brea

67.5-77.1

19.4-21.9

10.3-12.2

7. 1-9.0

5.6-7. 1

6. 5-8.4

17.9-20.4

2.3-4. 1

32

31

32

32

32

32

23

32

M . californica

73.80

20.76

11.36

8.33

6.32

7.50

19.25

3.33

Carpinteria

71.3-76.7

20.2-21.4

10.5-12.4

7. 9-8. 8

6. 1-6.4

7. 2-7. 8

19.0-19.5

3. 1-3.6

3

5

5

3

5

3

2

6

M . californica

72.72 ± 2.08

20.60 ± 0.64

11.29 ± 0.56

8.13 ± 0.53

6.26 ± 0.30

7.45 ± 0.48

19. 19 ± 0.60

3.37 ± 0.35

Total specimens

67.5-77.1

19.4-21.9

10.3-12.4

7. 1-9.0

5.6-7. 1

6. 5-8. 4

17.9-20.4

2.3-4. 1

35

36

37

35

37

35

25

38

M. ocellata

67.47

20.15

10.43

7.13

5.12

6.80

17.3

2.88

Dzibilchaltun

66.7-68.2

20.0-20.3

10.1-11.0

6. 6-7. 8

4. 7-5.6

6.6-7. 1

2.5-3. 1

3

2

3

3

4

3

1

4

Meleagris cf. M .

—

—

—

—

5.4

—

—

—

ocellata,

1

Macanche

M . ocellata

68.03 ± 2.63

20.27 ± 0.54

10.80 ± 0.65

7.17 ± 0.64

5.42 ± 0.34

7.07 ± 0.52

18.8

3.22 ± 0.31

Mayapan

62.2-72.1

19.5-21.3

9.3-11.8

6. 1-8.2

4. 9-5. 9

6. 5-8.0

2. 8-3. 9

23

10

19

18

17

12

1

18

M. ocellata

66.42 ± 2.37

19.79 ± 0.96

10.51 ± 0.56

7.40 ± 0.62

5.27 ± 0.46

6.68

16.86 ± 1.30

3.10

Yucatan, Mexico

63.3-69.4

18.8-21.2

10.0-11.6

6.0-8. 1

4. 5-5. 9

6. 0-7. 2

15.1-18.8

2. 2-3. 8

and Peten,

9

9

9

9

9

7

9

7

Guatemala

M . crassipes

59.60

18.30*

9.25

7.30

4.75

6.60

16.8

3.35

San Josecito

58.8-60.4

18.2-18.4*

9. 1-9.4

7. 1-7.5

4. 6-4. 9

6. 3-6. 9

3. 3-3. 4

Cave

2

2

2

2

2

2

2

2

a From Brodkorb 1964b.

* Slightly damaged specimen.

188

Steadman: Turkey Osteology and Paleontology

Table IS. Measurements (in mm) of the carpometacarpus of female turkeys, with mean, standard deviation, observed range, and sample size.
See Fig. 5 for explanation of measurements.

Total

Length

Proximal

Depth

Length
of Meta-
carpal I

Least
Width
of Meta-
carpal II

Least
Depth
of Meta-
carpal II

Greatest
Depth of
Intermeta-
carpal
Space

Distal

Depth

Protrusion
of Meta-
carpal III
beyond
Knob of
Metacarpal
II

Meleagris progenes

—

—

—

6.05

—

—

—

2.7

Rexroad

5.8-6.3a

2

1

Meleagris cf. M.

61.25 ± 1.71

17.80

9.48

6.48 ± 0.38

5.22 ± 0.20

5.96

15.40

2.46

leopoldi or M.

58.3-63.1

17.7-17.9

9. 1-9.9

6. 2-7. 6

4. 9-5. 6

5. 6-6. 6

14.8-16.0

2. 0-3. 2

anza, Inglis IA

8

2

6

11

12

7

5

7

Meleagris sp.

—

18.03

9.87

7.0

5.6

—

—

—

Coleman IIA

17.5-18.3

9.3-10.2

3

3

1

1

Meleagris cf. M .

63.3

—

—

7.5

5.2

6.3

—

3.1

gallopavo

1

1

1

1

1

Burnet Caveh

Meleagris sp.

—

—

—

7.4

6.0

—

—

—

Haile VIIA

1

1

Meleagris cf. M .

—

—

—

—

5.5

—

—

3.2

gallopavo

1

1

Reddick IB

Meleagris cf. M.

61.3*

—

—

6.9

4.8

—

—

—

gallopavo

1

1

1

Melbourne

M. gallopavo

66.8

18.9*

9.9

7.1

5.2

6.8

16.0*

2.4

Ichetucknee

1

1

1

1

1

1

1

1

River

Meleagris cf. M.

60.0

—

—

6.2

4.7

—

—

3.2

gallopavo , Vero

1

1

1

1

M gallopavo

—

18.7

9.6

7.3

5.10

—

—

2.82

Seminole Field

5.1

2.3-3. 1

1

1

1

2

4

M . gallopavo

61.60

17.20

9.05

7.00

4.85

6.4

—

2.70

Nichol’s Hammock

60.3-62.9

17.1-17.3

8.3-9. 8

6. 8-7. 2

4. 8-4. 9

2. 5-2. 9

2

2

2

2

2

1

2

M . gallopavo

62.1

17.9

10.0

7.0

4.9

—

—

2.2

Good’s Shellpit

1

1

1

1

1

1

M. gallopavo

67.68 ± 1.82

19.50 ± 0.58

10.36 ± 0.47

7.25 ± 0.40

5.29 ± 0.30

7.00 ± 0.45

17.68 ± 0.67

3.16 ± 0.38

Buffalo Site

63.7-72.7

17.6-20.9

9.3-11.5

6. 2-8. 2

4. 5-6. 2

6. 0-8.0

16.1-19.0

2. 2-4. 4

67

76

79

145

149

88

38

94

M . gallopavo

65.48 ± 1.30

19.21 ± 0.52

10.06 ± 0.32

6.94 ± 0.34

4.74 ± 0.26

6.76 ± 0.42

16.87 ± 0.62

3.23 ± 0.35

silvestris,

63.0-68.2

18.5-20.1

9.7-10.9

6. 3-7. 8

4. 3-5. 2

6. 1-7.5

15.8-18.0

2. 8-3. 9

New York,

14

14

14

14

14

14

14

14

Pennsylvania,

Virginia

M. gallopavo

63.16 ± 2.75

17.54 ± 0.72

9.40 ± 0.55

6.44 ± 0.25

4.82 ± 0.13

6.05 ± 0.44

15.52 ± 0.67

2.76 ± 0.41

osceola, Florida

59.7-67.8

16.7-19.1

8.7-10.4

6. 1-6.8

4.6-5. 1

5. 4-6. 9

14.4-16.5

2. 3-3. 6

11

11

11

11

11

11

11

11

M. gallopavo

68.45

19.40

10.35

7.10

5.30

6.75

16.95

3.75

mexicana

66.5-70.4

19.2-19.6

10.2-10.5

6. 8-7. 4

5.3

6. 5-7.0

16.6-17.3

3.3-4. 2

Chihuahua,

2

2

2

2

2

2

2

2

Mexico

M. gallopavo c

64.60 ± 2.64

18.51 ± 1.02

9.76 ± 0.60

6.73 ± 0.41

4.81 ± 0.24

6.45 ± 0.55

16.27 ± 0.95

3.07 ± 0.47

Total skeletal

59.5-70.4

16.7-20.1

8.4-10.9

6. 1-7.8

4. 3-5. 2

5. 4-7. 5

14.4-18.0

2. 3-4. 2

specimens

28

28

28

28

28

28

28

28

M . californica

61.46 ± 1.37

17.34 ± 0.39

9.83 ± 0.30

6.68 ± 0.40

5.18 ± 0.30

6.60 ± 0.23

16.48 ± 0.46

2.68 ± 0.25

Rancho La Brea

59.1-65.5

16.6-18.2

9.2-10.4

6. 0-7. 5

4. 6-5. 9

6. 2-7.0

15.5-17.2

2. 2-3. 2

32

32

31

32

32

32

23

32

Steadman: Turkey Osteology and Paleontology

189

Table IS. Continued.

Total

Length

Proximal

Depth

Length
of Meta-
carpal I

Least
Width
of Meta-
carpal II

Least
Depth
of Meta-
carpal II

Greatest
Depth of
Intermeta-
carpal
Space

Distal

Depth

Protrusion
of Meta-
carpal III
beyond
Knob of
Metacarpal
II

M. californica

61.70

17.80

9.99 ± 0.46

7.07

5.01

6.42

16.58

2.73

Carpinteria

60.8-62.6

17.3-18.7

9.4-10.9

6. 1-7.8

4. 8-5. 3

6. 1-6.8

16.3-17.3

2.5-3. 1

6

7

8

6

7

5

5

7

M. californica

61.50 ± 1.27

17.42 ± 0.45

9.86 ± 0.34

6.74 ± 0.45

5.15 ± 0.28

6.58 ± 0.25

16.50 ± 0.45

2.68 ± 0.24

Total specimens

59. 1-6S.S

16.6-18.7

9.2-10.9

6. 0-7. 8

4. 6-5. 9

6. 1-7.0

15.5-17.3

2. 2-3. 2

38

39

39

38

39

37

28

39

Meleagris cf. M.

56.8*

—

—

5.2

3.8

5.8

13.7

3.1

ocellata, Tuliim

1

1

1

1

1

1

M . ocellata

53.92

17.2

8.40

5.60

4.07

5.73

15.2

2.80

Mayapan

49.3-59.4

8. 0-8. 8

5. 1-6.0

3. 8-4. 2

5. 1-6.9

2. 7-2. 9

4

1

2

3

3

3

1

2

M. ocellata

57.60 ± 1.27

17.09 ± 0.46

9.12 ± 0.62

6.38 ± 0.27

4.79 ± 0.18

6.02 ± 0.44

14.97 ± 0.60

2.54 ± 0.30

Yucatan, Mexico

56.2-60.1

16.6-18.0

8.3-10.6

5. 9-6. 7

4. 4-5.0

5. 6-6. 9

13.9-16.0

2. 0-2. 9

and Peten,

10

10

10

10

10

8

10

8

Guatemala

M. crassipes

56.16*

16.03*

9.13*

6.70

4.48 ± 0.16

6.26

15.75

2.38

San Josecito

54. 4*— 5 7 . 2

15.7-16.7*

8. 6-9. 7*

6. 6-6. 8

4. 3-4. 8

6.0-6. 7

15.7-15.8

2.2-2. 6

Caved

7

3

3

5

8

5

2

7

a From Brodkorb 1964b; b may represent a male; c includes one specimen from northern Florida not identified to subspecies; d may possibly
include one or more specimens which represent males.

* Slightly damaged specimen.

190

Steadman: Turkey Osteology and Paleontology

Table 16. Measurements (in mm) of the femur of male turkeys, with mean, standard deviation, observed range, and sample size. See Fig. S for
explanation of measurements.

Meleagris
progenes a
Rexroad

Meleagris sp.
University
Drive

Meleagris cf.
M. leopoldi
or M . anza,

Inglis IA

Meleagris sp.

Williston

Meleagris sp.
Coleman

IIA

M eleagris
cf. M.
gallopavo
Frankstown
Cave

M . gallopavo
Manalapan

Meleagris
cf. M.
gallopavo,
Santa Fe
River IIA

Meleagris
cf. M.
gallopavo
Reddick IB

M . gallopavo
Ichetuck-
nee River

M . gallopavo
Seminole
Field

M . gallopavo
Davis
Quarry

Meleagris sp.
St. Mark’s
River

M . gallopavo
Good’s
Shellpit

Meleagris sp.
Silver
Glen
Springs

M gallopavo
Buffalo
Site

Depth of Depth of Depth of

Total Proximal Depth Width of Depth of Distal Internal External Fibular

Length Width of Head Midshaft Midshaft Width Condyle Condyle Condyle

— 25.8* — — — — — — —

1

26.7*

1

22.2

1

19.9

1

123.21* ± 2.57 29.96 ± 0.83 11.29 ± 0.41 12.23 ± 0.50 10.90 ± 0.51

1 18.8*— 127.0*
16

130.0*

1

131.24*
129. 1*-133.0
5

29.0-31.2

10.7-12.1

1 1.2-13.3

10.2-11.9

26.95

25.5-27.9

22.16* ± 0.82 22.55 ± 0.94 19.78 ± 0.86
20.2*-23.3* 21.3-24.6 18.7-21.1

147.50

145. 0-150. 0h
2

133.2*

1

138.5*

1

141.75*

139.7*-143.8*

2

124.60*
124.5*— 124. 7*
2

142.0

14

13

34

34

6

14

13

11

31.2

11.3

12.7

10.9

22.8

20.7

1

1

1

1

1

1

33.03*

12.46 ± 0.41

12.83 ± 0.56

11.56 ± 0.52

28.70

23.56*

23.50

20.47

!. 4-34.0*

11.8-13.2

12.1-13.8

10.5-12.2

28.2-29.4

22.3*-24. 7*

22.4-24.8

19.6-21.

7

9

12.9

1

14

14

5

5

7

6

32.6

13.17

13.60

12.53

30.1

25.2

25.1

21.2*

13.0-13.3

13.2-13.9

12.1-13.0

1

3

3

3

1

1

1

1

32.7

12.2

13.1

11.7

29.6

—

24.3

21.5

1

1

1

1

1

1

1

33.7

12.3

14.0

1 1.4

30.55*

24.50

25.0

22.8

30.3-30.8*

24.0-25.0

24.8-25.2

1

1

1

1

2

2

2

1

36.05

13.15

14.07

12.43

30.75*

25.2*

23.88*

20.65=*

). 0-36.1

13.1-13.2

13.3-14.6

11.4-13.0

30. 1*— 3 1.4*

21.0*-25.6*

18.2*— 2 1.

2

2

3

3

2

1

4

4

— -

—

12.90

11.05

28.9*

25.5*

—

—

12.6-13.2

10.3-11.8

1

1

2

2

33.5

12.2

—

—

—

—

—

—

1

1

31.7*

10.3*

—

1

1

_

10.5

12.15

10.70

26.55

22.80

21.60

19.15

12.1-12.2

10.4-11.0

26.4-26.7

22.0-23.6

20.5-22.9

18.9-19.

1

2

2

2

2

3

2

—

—

—

—

25.5

23.2*

22.1

19.0

1

1

1

1

35.37

12.91 ± 0.40

13.59 ± 0.70

11.91 ± 0.55

30.93

24.28

21.45

1.5-36.9

12.1-13.5

12.0-15.1

11.2-13.4

29.7-31.6

23.4-25.3

20.0-22.

7

11

15

15

3

4

4

1

Steadman: Turkey Osteology and Paleontology

191

Table 16. Continued.

Total

Length

Proximal

Width

Depth
of Plead

Width of
Midshaft

Depth of
Midshaft

Distal

Width

Depth of
Internal
Condyle

Depth of
External
Condyle

Depth of
Fibular
Condyle

M. gallopavo

139.30 ± 4.37

34.65 ± 1.20

12.79 ± 0.37

13.60 ± 0.77

11.99 ± 0.64

29.91 ± 0.85

24.65 ± 0.85

24.21 ± 0.79

21.06 ± 0.78

silvestris,

127.0-150.0

32.6-37.1

12.0-13.4

12.0-15.2

10.9-13.3

27.9-32.0

23.1-26.0

22.6-25.5

19.7-22.5

New York,

28

28

27

28

28

28

27

27

27

Pennsylvania,

Virginia

M . gallopavo

135.99 ± 5.32

32.59 ± 1.59

11.59 ± 0.49

13.38 ± 0.58

11.07 ± 0.53

27.76 ± 0.84

23.73 ± 1.20

23.06 ± 0.96

20.17 ± 0.92

osceola,

128.7-146.7

30.8-34.9

10.7-12.2

12.2-14.1

10.0-11.8

26.0-29.3

21.6-25.4

21.4-24.8

19.0-21.8

Florida

9

9

9

9

9

9

9

9

9

M . gallopavo

144.0

—

—

13.5

11.5

27.2

—

—

20.1

intermedia,

1

1

1

1

1

Texas

M. gallopavo

145.0

35.6

12.6

14.8

12.9

31.1

25.4

24.7

22.3

mexicana,

1

1

1

1

1

1

1

1

1

Chihuahua,

Mexico

M . gallopavo

143.0

33.8

12.8

14.4

12.3

29.6

24.5

24.2

21.2

mem ami

1

1

1

1

1

1

1

1

1

Arizona

M gallopavo

138.91 ± 4.81

34.18 ± 1.54

12.50 ± 0.64

13.60 ± 0.74

11.80 ± 0.73

29.38 ± 1.29

24.45 ± 1.00

23.93 ± 0.98

20.87 ± 0.90

Total

127.0-150.0

30.8-37.1

10.7-13.4

12.0-15.2

10.0-13.3

26.0-32.0

21.6-26.0

21.4-25.5

19.0-22.5

skeletal

40

39

38

40

40

40

38

38

39

specimens

M. cali-

124.19* ± 3.08

31.47 ± 1.25

11.45 ± 0.50

12.18 ± 0.45

10.25 ± 0.43

27.82 ± 0.85

22.28 ± 0.88

22.21 ± 0.87

19.95 ± 0.97

fornica

1 17. 7*— 132. 2*

28.6-33.8

10.4-12.6

11.4-13.2

8.9-11.1

25.8-29.3

21.1-23.7

20.0-23.7

18.1-21.8

Rancho

35

44

48

43

43

34

10

26

33

La Brea

M. cali-

127.20*

31.70

11.40

12.38

10.57

27.80

22.2

22.43

19.83

fornica

1 24. 7* — 128.8*

31.4-32.0

11.2-11.7

11.8-13.1

10.2-10.9

27.1-28.2

22.0-22.9

18.8-20.4

Carpinteria

3

3

5

6

6

3

1

3

3

M. cali-

124.43* ±3.11

31.48 ± 1.21

11.44 ± 0.47

12.20 ± 0.45

10.29 ± 0.43

27.82 ± 0.82

22.27 ± 0.83

22.23 ± 0.84

19.94 ± 0.96

fornica

117.7*— 132.2*

28.6-33.8

10.4-12.6

11.4-13.2

8.9-11.1

25.8-29.3

21.1-23.7

20.0-23.7

18.1-21.8

Total

38

47

53

49

49

37

11

29

36

specimens

M. ocellata

110.84 ± 3.37

25.96 ± 1.15

9.36 ± 0.44

10.40 ± 0.66

9.49 ± 0.66

23.63 ± 1.05

19.04 ± 1.22

18.86 ± 0.78

16.42 ± 0.95

Yucatan,

105.6-116.9

24.1-27.8

8.9-10.2

9.3-1 1.7

8.3-10.2

22.1-25.2

17.2-20.6

18.0-20.1

15.3-18.1

Mexico,

10

10

8

10

10

10

8

8

8

and Peten,

Guatemala

M. crassipes

107.8*

26.7

10.2

10.1

9.1

23.10*

19.1

18.5

14.9

San Jose-

23.0-23.2*

cito Cave

1

1

1

1

1

2

1

1

1

a From Brodkorb 1964b; b larger measurement from Marsh 1872, and Shufeldt 1915.
* Slightly damaged specimen

192

Steadman: Turkey Osteology and Paleontology

Table 17. Measurements (in mm) of the femur of female turkeys, with mean, standard deviation, observed range, and sample size. See Fig. 5
for explanation of measurements.

Depth of

Depth of

Depth of

Total

Proximal

Depth of

Width of

Depth of

Distal

Internal

External

Fibular

Length

Width

Head

Midshaft

Midshaft

Width

Condyle

Condyle

Condyle

Meleagris cf.

102.47*

23.35

8.88

9.73

8.45

21.70

17.72*

17.50

15.04

M. leopoldi

99.4*-105.2*

23.0-23.7

8.6-9 .2

9.1-10.2

8. 2-8. 9

21.3-22.0

17.2*— 18.7*

17.0-18.3

14.4-16.1

or M. anza,

Inglis IA

3

2

4

6

6

3

4

4

5

Meleagridi-

—

25.5*

9.2*

—

—

—

—

—

—

nae, gen.
and sp.
indet.

Gillilandab

1

1

Meleagris sp.

—

27.4

9.6

11.2

10.6

23.7

21.2*

20.10

16.9*

Coleman

1

19.7-20.5

IIA

1

1

1

1

1

2

1

Meleagris sp.

-110

-26.2

—

—

—

—

—

—

—

Papago

Springs

Caveb

1

1

Meleagris

—

—

—

10.7

9.6

—

—

—

—

cf. M.
gallopavo
Burnet
Cave

1

1

Meleagris sp.

114.4

26.3

9.9

10.3

9.6

24.0

20.2

19.7

17.0

Haile VIIA

1

1

1

1

1

1

1

1

1

Meleagris

—

—

—

—

—

23.5

—

19.1

16.8

cf. M
gallopavo

Reddick IB

1

1

1

M. gallopavo,

—

—

- —

10.65

9.30

22.5*

18.6*

18.0*

16.0*

Ichetuck-

10.3-11.0

8.9-9. 7

nee River

2

2

1

1

1

1

M. gallopavo

—

26.1

9.2

—

—

22.65

—

18.25

15.70

Seminole

22.3-23.0

17.8-18.7

15.2-16.2

Field

1

1

2

2

2

Meleagris sp.

—

26.8

9.8

—

—

—

—

—

—

Hog Creek

1

1

Meleagris

1 IS. 4*

25.8*

—

11.8

9.9

23.3*

—

18.9*

17.4*

cf. M.
gallopavo
Sante Fe IA

1

1

1

1

1

1

1

M. gallopavo

104. 10*

24.70

8.80

9.75

8.65

21.00

—

17.65

15.45

Nichol’s

101.4-106.8*

24.0-25.4

8. 6-9.0

9.S-10.0

8.2-9. 1

20.2-21.8

17.5-17.8

15.3-15.6

Hammock

2

2

2

2

2

2

2

2

M. gallopavo

—

—

8.5

10.2

9.1

—

—

—

—

Good’s

Shellpit

1

1

1

M. gallopavo

117.27 ± 3.17

27.42 ± 0.57

10.14 ± 0.41

11.22 ± 0.50

9.63 ± 0.42

23.85 ± 0.61

20.00

19.46 ± 0.44

17.02 ± 0.57

Buffalo

112.9-122.0

26.1-28.4

9.5-11.0

10.3-11.8

8.9-10.3

22.9-25.0

19.2-20.8

18.4-20.0

16.3-18.3

Site

9

20

24

29

29

10

5

14

12

M . gallopavo

113.57 ± 3.16

26.28 ± 0.64

9.81 ± 0.31

10.68 ± 0.43

9.31 ± 0.52

22.87 ± 0.50

19.04 ± 0.54

18.52 ± 0.51

15.95 ± 0.51

silvestris,

108.9-119.0

25.1-27.3

9.3-10.2

9.9-11.2

8.6-10.1

22.1-23.5

17.9-19.9

18.0-19.3

15.1-16.7

New York,
Pennsyl-

12

12

12

12

12

12

12

12

12

vania,

Virginia

Steadman: Turkey Osteology and Paleontology

193

Table 17. Continued.

Depth of

Depth of

Depth of

Total

Proximal

Depth of

Width of

Depth of

Distal

Internal

External

Fibular

Length

Width

Head

Midshaft

Midshaft

Width

Condyle

Condyle

Condyle

M. gallopavo

108.50 ± 2.49

25.33 ± 0.97

8.81 ± 0.34

10.46 ± 0.41

8.72 ± 0.48

21.16 ± 0.89

18.65 ± 0.83

17.78 ± 0.62

15.55 ± 0.79

osceola,

104.0-113.9

23.9-27.2

8. 2-9. 2

9.8-11.2

8. 1-9. 7

20.0-23.5

17.8-20.5

16.9-19.1

14.8-17.3

Florida

11

11

11

11

11

11

11

11

1 1

M. gallopavo

116.90

26.80

10.03

11.30

9.83

23.67

19.57

19.20

16.93

mexicana,

112.0-124.2

25.4-28.5

9.7-10.5

10.9—1 1.6

9.3-10.2

22.5-25.0

18.1-21.4

17.4-20.6

15.1-18.6

Chihuahua,

Coahuila,

Mexico

3

3

3

3

3

3

3

3

3

M. gallopavo

115.80

27.17

10.03

11.20

9.97

23.67

18.73

18.47

16.23

mem ami

112.4-119.3

26.9-27.5

9.8-10.4

11.0-11.4

9.9-10.1

23.3-23.9

18.1-19.1

18.0-18.8

15.9-16.8

Arizona

3

3

3

3

3

3

3

3

3

M. gallopavo

112.01 ± 4.54

26.04 ± 1.04

9.43 ± 0.65

10.69 ± 0.49

9.18 ± 0.65

22.36 ± 1.23

18.92 ± 0.80

18.26 ± 0.83

15.90 ± 0.86

Total

104.0-124.2

23.9-28.5

8.2-10.5

9.8-11.6

8.1-10.2

20.0-25.0

17.8-21.4

16.9-20.6

14.8-18.6

skeletal

specimens0

30

30

30

30

30

30

30

30

30

M. cali-

104.66* ± 2.92

25.35 ± 0.96

9.28 ± 0.42

10.03 ± 0.48

8.47 ± 0.42

22.03 ± 0.89

17.45

17.06 ± 0.70

15.28 ± 0.79

fornica

99.4-111.8*

22.9-27.3

8.3-10.1

9.1-10.8

7. 7-9. 2

20.4-23.4

16.8-18.4

15.7-17.8

13.8-16.0

Rancho
La Brea

23

33

33

23

23

14

4

17

16

M . cali-

103.90*

25.60

9.40

10.32

8.52

22.20

17.8

17.58

15.38

fornica

103.1-104.7*

24.3-27.1

8. 7-9.8

9.9-10.8

8.3-8. 9

21.8-22.4

17.2-18.1

15.1-16.0

Carpinteria

2

6

6

4

4

4

1

4

4

M. cali-

104.60* ± 2.82

25.39 ± 0.97

9.29 ± 0.42

10.07 ± 0.48

8.48 ± 0.40

22.07 ± 0.79

17.52

17.16 ± 0.68

15.30 ± 0.72

fornica

99.4*-l 11.8*

22.9-27.3

8.3-10.1

9.1-10.8

7. 7-9. 2

20.4-23.4

16.8-18.4

15.7-18.1

13.8-16.0

Total

25

39

39

27

27

18

5

21

20

specimens

M. ocellata

98.0*

—

—

9.3

9.4

—

—

—

—

Dzibil-

chaltun

1

1

1

M. ocellata

97.54 ± 1.84

22.26 ± 0.81

8.11 ± 0.43

9.30 ± 0.46

8.37 ± 0.51

20.07 ± 0.47

16.33 ± 0.79

16.34 ± 0.46

14.26 ± 0.51

Yucatan,

94.9-101.1

20.9-23.7

7. 6-8. 9

8.7-10.1

7. 4-9. 3

19.3-21.0

15.5-17.8

15.6-17.0

13.8-15.3

Mexico,
and Peten,
Guatemala

11

11

9

11

11

11

9

9

9

M. eras sipes

94.05*

23.05

8.70

9.39 ± 0.55

8.51 ± 0.34

20.9*

—

16.0*

13.6

San Jose-

91.3*-96.8*

22.5-23.6

8. 5-8. 8

8.6-10.3

8. 1-8.9

cito Cave

2

2

4

8

8

1

1

1

a From Brodkorb 1964b; b may possibly represent a male; c includes one specimen from northern Florida not identified to subspecies.
* Slightly damaged specimen.

194

Steadman: Turkey Osteology and Paleontology

Table 18. Measurements (in mm) of the tibiotarsus of male turkeys, with mean, standard deivation, observed range, and sample size. See Fig.
6 for explanation of measurements.

Length With-
out Cnemial
Crestf

Width
of Head

Width of
Midshaft

Depth of
Midshaft

Distal

Width

Depth of
Internal
Condyle

Depth of
External
Condyle

Meleagridinae

—

—

—

—

18.9*

19. 1*

—

cf. Meleagris

1

1

Buckhorn

Meleagris sp.

—

—

-12.4

-9.9

18.4*

19.5*

17.9*

Haile XVA

1

1

1

1

1

Meleagris cf. M .

20S.2

23.43 ± 0.72

11.74 ± 0.47

9.10 ± 0.40

19.58 ± 0.83

19.24* ± 0.58

17.92* ± 0.75

leopoldi or M.

199. 5-208. S

22.5-24.8

10.3-12.4

8. 3-9. 8

18.3-21.7

18.6-20.5*

16.8-19.8*

anza, Inglis IA

4

10

24

25

22

11

20

Meleagris sp.

—

—

—

—

—

18.5*

—

Haile XVIA

1

Meleagris sp.

208.5

25.10 ± 0.76

12.34 ± 0.52

9.34 ± 0.44

21.09 ± 0.47

21.08

18.95 ± 0.25

Coleman IIA

207.0-210.0

24.2-26.1

11.5-13.0

8. 7-9. 8

20.3-21.6

20.6-21.8

18.6-19.5

4

8

9

8

10

6

12

M. gallopavo

—

—

12.1

9. 1

21.50*

20.30*

18.85*

Ingleside

1

1

2 1.2*— 2 1.8

19.9*-20.7*

18.7 * — 19.0

2

2

2

Meleagris cf.

—

—

—

—

21.90

21.05*

19.35*

M. gallopavo ,

21.9

20.9*— 2 1. 2*

19.2-19.5*

Frankstown Cave

2

2

2

M . gallopavo

245

30b

12.5

10.5

17c

—

17.3

Manalapan’1

1

1

1

1

1

1

M. gallopavo

243.75*

—

—

—

Igc.d

—

—

Manalapan

243‘‘-244.5*e

2

1

M. gallopavo

235.5*

26.3

11.75*

9.95

20.0*

19.4*

17.9*

Manalapan1

233*-238*

1 1.4*— 12. 1

9.8-10.1

2

1

2

2

1

1

1

Meleagris cf.

213.0

24.2

12.6

9.1

20.9

19.8*

18.7*

M . gallopavo

1

1

1

1

1

1

1

Sante Fe IIA

Meleagris cf.

—

—

11.8

9.0

—

—

—

M . gallopavo

1

1

Rock Spring

Meleagris cf.

233.5

25.30

13.0

10.3

21.9

22.4

20.2

M . gallopavo

24.1-26.5

Reddick IB

1

2

1

1

1

1

1

M. gallopavo

—

—

—

—

21.0*

20.8*

19.0*

Aucilla River

1

1

1

M . gallopavo

230.5

26.33

12.54 ± 1.10

9.85 ± 0.87

21.81* ± 1.06

21.18* ± 1.04

19.92 ± 0.86

Ichetucknee

226.5-233.5

25.7-27.0

10.7-14.1

8.8-11.4

20.4*-23.6*

19.8-22.3*

18.5-21.2

River

4

3

1 1

11

12

9

11

M . gallopavo

—

—

—

—

21.43

—

19.10

Seminole Field

21.0-22.1

19.0-19.2

3

2

Meleagris sp.

—

—

—

—

21.0*

—

19.9*

Oakhurst Quarry

1

1

Meleagris cf.

—

—

—

—

19.3**

19.8**

17.5**

M . gallopavo

1

1

1

Sante Fe IA

Meleagris sp.

—

—

-12.2

-9.2

19.7*

—

—

Sante Fe IVA

1

1

1

Meleagris sp.

—

—

—

—

20.6*

20.9*

18.1*

Wekiva Run III

1

1

1

M . gallopavo

—

—

—

—

20.8

—

19.7

Nichol’s Hammock

1

1

M. gallopavo

—

23.2

—

—

—

—

—

Good’s Shellpit

1

Steadman: Turkey Osteology and Paleontology

195

Table 18. Continued.

Length With-

Depth of

Depth of

out Cnemial

Width

Width of

Depth of

Distal

Internal

External

Crestf

of Head

Midshaft

Midshaft

Width

Condyle

Condyle

M. gallopavo

230.0

26.25

12.33 ± 0.63

9.98 ± 0.52

22.50 ± 0.91

22.07

19.80 ± 0.64

Buffalo Site

226.0-234.0

25.5-27.0

10.7-13.2

9.0-11.1

20.2-24.8

21.4-23 1

18. 1-21.3

2

2

15

15

22

6

34

M. gallopavo

220.3 ± 6.9

25.10 ± 0.59

12.15 ± 0.63

9.63 ± 0.46

21.93 ± 0.85

21.63 ± 0.82

19.53 ± 0.40

silvestris, New

200.0-233.0

24.0-26.2

10.9-13.1

8.8-10.4

20.3-23.3

20.1-22.9

18.8-20.1

York, Pennsyl-
vania, Virginia

26

23

26

26

26

25

25

M. gallopavo

231.9 ± 10.6

23.20 ± 0.92

11.79 ± 0.69

9.28 ± 0.40

20.70 ± 0.88

20.58 ± 1.29

18.79 ± 1.18

osceola, Florida

221.5-253.0

21.6-24.5

10.9-13.0

8.8-10.1

19.0-22.2

18.1-22.6

17.0-20.2

9

9

9

9

10

10

10

M. gallopavo

231.0

—

12.2

10.0

21.5

20.7

18.8

intermedia, Texas

1

1

1

1

1

1

M . gallopavo,

227.0

26.2

12.8

10.3

23.5

23.1

20.4

mexicana

1

1

1

1

1

1

1

Chihuahua,

Mexico

M. gallopavo

225.5

24.8

12.5

10.0

22.6

22.7

19.6

merriami,

1

1

1

1

1

1

1

Arizona

M . gallopavo

223.6 ± 9.1

24.62 ± 1.11

12.09 ± 0.64

9.59 ± 0.47

21.66 ± 1.03

21.38 ± 1.10

19.29 ± 0.74

Total skeletal

200.0-253.0

21.6-26.2

10.9-13.1

8.8-10.4

19.0-23.5

18.1-23.1

17.0-20.4

specimens

38

35

38

38

39

38

38

M. californica

203.5 ± 6.2

23.27 ± 0.61

11.26 ± 0.53

8.96 ± 0.36

20.00 ± 0.54

19.79 ± 0.74

18.03 ± 0.72

Rancho La Brea

192.0-212.0

22.6-24.5

10.9—1 1.9

8. 0-9. 4

19.3-20.9

18.8-21.2

17.2-19.9

15

9

18

18

23

14

21

M. californica

202.4

23.56

11.22

9.14

19.70 ± 0.36

19.76 ± 0.52

17.98 ± 0.30

Carpinteria

199.0-204.5

22.8-24.3

11.0-11.3

8. 6-9. 5

19.2-20 .3

19.3-20.7

17.6-18.6

4

5

5

5

11

8

9

M . californica

203.2 ± 5.6

23.37 ± 0.60

11.25 ± 0.47

9.00 ± 0.36

19.90 ± 0.50

19.78 ± 0.65

18.02 ± 0.62

Total specimens

192.0-212.0

22.6-24.5

10.9-11.9

8. 0-9. 5

19.2-20.9

18.8-21.2

17.2-19.9

19

14

23

23

34

22

30

M . ocellata

176.5

—

10.15

8.35

18.35

—

16.40

Dzibilchaltiin

9.6-10.7

8. 1-8.6

17.6-19.1

15.8-17.0

1

2

2

2

2

Meleagris cf.

—

—

—

—

19.0*

—

16.2**

M. ocellata
Macanche

1

1

M. ocellata

183.50 ± 5.46

19.86 ± 1.11

10.08 ± 0.69

7.92 ± 0.49

17.86 ± 0.98

17.87

16.30

Yucatan, Mexico

175.0-193.0

18.4-21.8

8.9-10.9

7. 3-8. 6

16.6-19.2

16.9-19.0

15.3-17.6

and Peten,
Guatemala

9

10

9

9

9

7

7

M. crassipes

164.0

20.0

10.4

8.0

18.1

17.1*

15.45*

San Josecito

163.5-164.5

15.2-15.7*

Cave

2

1

1

1

1

1

2

a From Cope, 1871; b abnormally large probably because of a difference in position of specimen during measurement; c abnormally small, probably
because of misinterpretation of described measurement or fragmentary nature of specimen; 11 from Marsh 1872; e from Shufeldt 1915. This is
probably the specimen which Marsh measured; f specimens measured by the author,
t Measurements accurate only to within 0.5 mm.

* Slightly damaged specimen.

** Moderately damaged specimen.

196

Steadman: Turkey Osteology and Paleontology

Table 19. Measurements (in mm) of the tibiotarsus of female turkeys, with mean, standard deviation, observed range, and sample size. See Fig.
6 for explanation of measurements.

Length With-
out Cnemial
Crestf

Width
of Head

Width of
Midshaft

Depth of
Midshaft

Distal

Width

Depth of
Internal
Condyle

Depth of
External
Condyle

Meleagridinae

—

18.5*

—

15.0*

16.1

14.0*

cf. Meleagris

1

1

1

1

Westmoreland, Parka,b

Meleagridinae cf.

—

—

—

16.8*

16.5*

14.9

Meleagris, Bone

1

1

1

Valley (Palmetto Mine)

Meleagris progenes

—

—

—

14.5*

15.2*

13.0*

Rexroad

1

1

1

Meleagris cf. M .

—

18.72

9.85

7.80

16.07

15.85*

14.38*

leopoldi or M . anza

18.4-19.4

9.5-10.2

7. 7-7. 9

16.0-16.1

15.7*— 16.1*

13.9*— 15.0

Inglis IA

4

2

2

3

4

4

Meleagris sp.

—

—

10.45

8.40

17.88

17.80*

16.05

Coleman IIA

10.2-10.7

8. 0-8. 8

17.0-18.8

17.2-18.7*

15.6-16.5

2

2

5

4

2

Meleagris cf. M.

—

—

10.3

8.3

18.4*

18.0*

16.3*

gallopavo

1

1

1

1

1

Burnet Cave

M . gallopavo

—

—

—

—

17.9*

15.9

Ingleside

1

1

M . gallopavo

183

19

9.6

—

16.5

16

Manalapanc

1

1

1

1

1

M. gallopavo

180*

—

—

17.0

17.4*

15.9

Manalapan'1

1

1

1

1

M. gallopavo

—

—

—

—

15.2*

17.0*

Aucilla River

1

1

M . gallopavo

178.0

21.4

10.17

8.27

17.42* ± 1.11

17.00* ± 1.03

15.39* ± 0.89

Ichetucknee River

175. 5-180. 5

9.5-11.0

7. 7-8.8

15.2*— 18.8

14.8*— 18.2

13.7*— 16.5

2

1

6

6

12

11

12

Meleagris sp.

—

—

—

17.0

18.0*

15.8

Kendrick

1

1

1

M. gallopavo

—

—

—

—

17.0

16.2

15.30

Seminole Field

14.7-16.0

Meleagris sp.

1

1

3

Econfina River

176.0*

—

—

—

14.9*

—

14.9*

1

1

1

Meleagris sp.

—

—

—

—

18.3*

—

—

Hog Creek

1

Meleagris sp.

—

—

9.7

7.8

—

—

—

Santa Fe IVA

1

1

M. gallopavo

172.3

17.30

8.54

7.04

15.67

15.93

14.55

Nichol’s Hammock

169.5-174.0

17.1-17.7

8. 2-9.0

6.9-7. 1

15.1-16.8

15.8-16.1

14.1-15.0

3

3

5

5

3

3

4

M . gallopavo

—

—

—

—

16.6*

16.1*

14.5

Good’s Shellpit

1

1

1

M. gallopavo

185.7

20.92

10.17 ± 0.42

8.17 ± 0.22

18.03 ± 0.55

18.29 ± 0.84

15.97 ± 0.56

Buffalo Site

182.0-189.0

20.0-22.0

9.5-10.9

7. 9-8. 6

16.9-19.0

17.0-20.1

15.0-17.0

3

6

16

15

28

11

32

M. gallopavo

—

—

—

—

18.00

16.45

15.45

Hartman’s Cave

17.9-18.1

16.1-16.8

15.1-15.8

2

2

2

M . gallopavo silvestris,

175.7 ± 6.3

19.48 ± 0.45

9.24 ± 0.41

7.42 ± 0.33

17.42 ± 0.40

17.23 ± 0.39

15.58 ± 0.43

New York, Pennsylvania,

164.0-187.0

18.9-20.2

8.8-10.0

7. 0-8.0

16.9-18.1

16.5-17.9

15.1-16.4

Virginia

12

9

12

12

12

12

12

M. gallopavo osceola,

178.1 ± 4.4

18.04 ± 0.65

9.17 ± 0.24

7.35 ± 0.30

16.44 ± 0.56

16.40 ± 0.62

14.92 ± 0.31

Florida

171.0-186.5

17.3-19.6

8.9-9. 7

6. 8-7. 8

15.7-17.6

15.4-17.5

14.4-15.5

10

11

11

11

10

10

10

M. gallopavo mexicana,

183.5

20.07

10.17

7.90

18.40

18.17

16.13

Chihuahua, Coahuila,

178.0-190.0

19.5-21.2

9.8-10.5

7.6-8. 1

17.2-19.3

17.0-19.6

14.8-17.3

Mexico

3

3

3

3

3

3

3

Steadman: Turkey Osteology and Paleontology

197

Table 19. Continued.

Length With-
out Cnemial
Crestf

Width
of Head

Width of
Midshaft

Depth of
Midshaft

Distal

Width

Depth of
Internal
Condyle

Depth of
External
Condyle

M. gallopavo merriami,

175.5

19.73

9.83

8.13

17.60

17.55

15.30

Arizona

170.0-181.0

19.1-20.3

9.5-10.1

7. 5-8. 9

17.6

17.2-17.9

15.3

2

3

3

3

2

2

2

M . gallopavo

177.0 ± 6.1

18.94 ± 1.01

9.35 ± 0.47

7.51 ± 0.43

17.12 ± 0.86

17.00 ± 0.86

15.43 ± 0.66

Total skeletal

164.0-190.0

17.3-21.2

8.8-10.5

6. 8-8. 9

15.7-19.3

15.4-19.6

14.2-17.3

specimens6

28

27

30

30

28

28

28

M. californica

168.4 ± 3.1

19.25 ± 0.54

9.25 ± 0.30

7.69 ± 0.37

16.74 ± 0.43

16.60 ± 0.38

15.11 ± 0.40

Rancho La Brea

164.0-173.0

18.4-19.9

8. 8-9. 8

6. 9-8. 3

15.8-17.7

15.6-17.3

14.1-15.9

8

8

16

16

26

25

27

M. californica

168.7

18.8

9.70

7.50

16.66

16.66

15.00

Carpinteria

165.0-174.0

9.5-10.0

7.4-7. 7

16.1-17.3

15.6-17.8

14.3-15.5

3

1

3

3

5

5

6

M. californica

168.4 ± 3.4

19.20 ± 0.53

9.32 ± 0.33

7.66 ± 0.35

16.73 ± 0.43

16.61 ± 0.46

15.09 ± 0.40

Total specimens

164.0-174.0

18.4-19.9

8.8-10.0

6. 9-8. 3

15.8-17.7

15.6-17.8

14.1-15.9

11

9

19

19

31

30

33

Meleagridinae cf.

—

—

—

—

16.3

—

14.5

Meleagris, Workman

1

1

and Alhambra Streets

M. ocellata

—

—

7.9

7.0

15.4

14.6

13.2

Dzibilchaltun

1

1

1

1

1

M. ocellata Yucatan,

158.2 ± 4.7

17.14 ± 0.52

8.81 ± 0.23

7.11 ± 0.41

15.02 ± 0.46

14.95 ± 0.72

13.66 ± 0.50

Mexico, and Peten,

150.5-167.0

16.5-18.1

8.5-9. 1

6. 4-7. 6

14.2-16.0

14.1-16.1

13.1-14.5

Guatemala

9

11

9

9

10

8

8

M. crassipes

150.7*

18.6

9.45

7.60

17.45

15.90*

14.18

San Josecito Cave

150.0-152.0*

9. 0-9. 9

7. 2-8.0

17.4-17.5

IS .8*— 16.0*

13.7-14.6

3

1

2

2

2

2

4

a Based on two specimens which may represent opposite ends of the same bone; b may represent a male; c from Marsh 1872; 11 measured by the
author — possibly the same as the specimen measured by Marsh; e includes one specimen from northern Florida not identified to subspecies,
t Measurements accurate only to within 0.5 mm.

* Slightly damaged specimen.

198

Steadman: Turkey Osteology and Paleontology

Table 20. Measurements (in mm) of the tarsometatarsus of male turkeys, with mean, standard deviation, observed range and sample size. See
Fig. 7 for explanation of measurements.

Total

Length

A

Proximal

Width

B

Least
Width of
Shaft
C

Least
Depth of
Shaft
D

Proximal
End to
Middle of
Spur Core
E

Top of Spc
Core to
End of
Middle
Trochlea
F

Proagriocharis kimballensis

98c

14c

—

—

58

—

UNSM Coll. Loc. Ft-40

1

1

1

Meleagris progenes,

—

—

—

4.9

—

53.6

Rexroad

1

1

Meleagris cf. M . progenes,

113. S

18.7

7.45

5.05

64.9

51.90

Benson

7. 1-7.8

5.0-5. 1

50. 0-53. S

1

1

2

2

1

2

M . leopoldi

139. S

21.0

8.7

6.0

88.8

~56.9

Cita Canyon

1

1

1

1

1

1

Meleagris cf. M. leopoldi

151.88 ± 3.88

22.06 ± 0.59

9.24 ± 0.44

5.95 ± 0.25

96.14 ± 2.92

59.82 ± 2.

or M . ansa, Inglis IA

146.6-158.0

21.1-23.7

8. 4-9. 9

5. 3-6. 5

91.6-101.1

5 2. 8-63. £

8

26

26

30

14

13

Meleagris sp.

153.06

23.98 ± 0.30

9.60 ± 0.48

6.32 ± 0.21

90.87

67.04

Coleman IIA

149.4-157.0

23.6-24.5

9.1-10.5

6.0-6. 7

86.5-96.3

59.7-70.^

5

8

11

13

6

7

M. gallopavo

165.5*

—

9.8

6.2

95.5*

74.4

American Falls

1

1

1

1

1

M . gallopavo

161.5

24.3*

9.4

6.2

94.50

72.5

Ingleside

93.0-96.0

1

1

1

1

2

1

M. gallopavo

176. 5e

23.35

—

—

110e

—

Manalapan

1

23e-23.7f

2

1

Meleagris cf. M. gallopavo.

8.9

6.1

98.3

67.8

Santa Fe River IIA

1

1

1

1

Meleagris sp.

—

23.8

—

—

—

—

Bradenton

1

Meleagris sp.

—

—

7.8

5.1

—

-65.8

Withlacoochee River

1

1

1

Meleagris cf. M. gallopavo

166.25*

24.8

9.20

6.23

97.75*

73.55

Reddick IB

161.5-171.0*

9. 0-9. 4

6. 2-6. 3

97.2 * — 98 . 3

O'

00

00

1

OO

2

1

2

3

2

2

Meleagris cf. M. gallopavo

—

23.9

—

—

—

—

Melbourne

1

Meleagris cf. M. gallopavo

—

—

—

—

—

-

Sabertooth Cave

M. gallopavo

—

—

8.2

6.8

—

77.2

Aucilla River

1

1

1

M. gallopavo

173.75

23.99 ± 0.89

8.90 ± 0.83

5.87 ± 0.40

103.23

73.16 ± 2.

Ichetucknee River

170.0-177.5

22.3-25.3

7.5-10.1

5.2-6. 7

101.0-106.7

68.7-77.:

2

8

11

19

3

8

Meleagris sp.

—

22.2

—

—

—

—

Kendrick IA

1

M. gallopavo

—

23.9*

9.0

5.75

—

—

Seminole Field

5. 7-5. 8

1

1

2

Meleagris cf. M. gallopavo

—

—

8.6

—

—

—

Bowman IA

1

Meleagris cf. M. gallopavo,

—

—

8.1

5.2

—

—

Santa Fe River IA

1

1

M. gallopavo

—

24.1*

9.2

—

103.1

—

Wacissa River

1

1

1

Steadman: Turkey Osteology and Paleontology

199

Table 20. Continued.

Middle of

Spur

Core Angle

to End of

of Spur

Depth of

Depth of

Depth of

Middle

Width of

Length of

Core (in

Distal

Inner

Middle

Outer

Trochlea

Spur Core

Spur Core

Degrees)

Width

Trochlea

Trochlea

Trochlea

G

Ha

J"

K

L

M

N

P

40c

6.2

15.7*

38"

—

—

—

—

1

1

1

1

48.0

7.1

—

49

19.0

8.4

9.1

9.9*

1

1

1

1

1

1

1

48.95

—

—

56

18.9

8.7

9.45

10.0

48.6-49.3

1

9. 0-9. 9

2

1

1

2

1

50.7

—

—

55.75"

21.0

9.5

10.2

11.2

53.0-58.5

1

2

1

1

1

1

55.13 ± 3.12

7.70 ± 0.86

31.21* ± 2.78

47.4 ± 3.4

21.68 ± 0.52

9.56 ± 0.38

10.58 ± 0.37

10.88 ± 0.52

48.3-59.6

6.1-10.3

26.2*— 35.5*

43-56

20.9-22.7

8. 9-9. 9

10.0-11.1

10.1-11.8

13

37

15

28

15

11

20

8

61.86

7.53 ± 0.50

28.82*

52.9 ± 4.0

23.75 ± 1.08

10.68

12.04 ± 0.30

12.98

54.9-65.1

7.0-8 .2

21. 1*— 32.4*

45-58

22.0-24.8

10.0—1 1 . 5

11.7-12.5

12.0-13.9

7

9

4

10

8

5

8

5

70.0

7.0

17.4

50

26.0

—

12.1

13.1

1

1

1

1

1

1

1

68.5

6.70

—

55.0

23.5

10.0

10.8

12.0

6. 5-6. 9

55

1

2

2

1

1

1

1

66. 5e

1

—

—

—

—

—

—

—

62.9

5.8

58

10.6

1

1

1

1

-60.3

1

—

—

—

—

—

—

—

68.20

65

24.03

10.83

11.77

12.75

63.2-73.2

23.8-24.3

10.4-11.1

11.1-12.1

12.5-13.0

2

1

3

3

2

—

6.5

1

—

—

—

—

—

—

—

6.4

19.0*

—

—

—

—

—

1

1

68.2

6. 1

20.0*

63

24.8

—

11.9

13.7*

1

1

1

1

1

1

1

68.55 ± 2.94

6.70

19.53*

57.1 ± 5.0

23.44 ± 1.50

10.13

11.38 ± 0.69

12.85

63.9-72.5

5. 9-7. 6

16. 1 * — 2 1.7*

46-66

21.1-26.2

9.3-11.0

10.2-12.4

12.3-13.5

8

5

3

12

9

6

12

4

6.32 ± 0.68

16.60*

10.30

11.30

5. 1-7.5

15.0*— 17.9*

10.0-10.6

11.0-11.6

12

5.9

3

58

2

2

68.0

1

1

22.0

10.3*

1

1

1

—

6.1

17.3*

67

—

—

—

—

1

1

1

Continued

200

Steadman: Turkey Osteology and Paleontology

Table 20. Continued.

Total

Length

A

Proximal

Width

B

Least
Width of
Shaft
C

Least
Depth of
Shaft
D

Proximal
End to
Middle of
Spur Core
E

Top of Spur
Core to
End of
Middle
Trochlea
F

M. g allopavo

168.0

24.72 ± 0.91

9.13 ± 0.47

6.25 ± 0.26

98.8

76.0

Buffalo Site

22.9-26.8

8.0-9. 9

5. 7-6.8

1

22

27

27

1

1

M . g allopavo

—

—

—

—

—

—

Hartman’s Cave

M. gallopavo silvestris,

160.55 ± 5.92

24.49 ± 1.13

9.34 ± 0.58

6.04 ± 0.39

91.48 ± 4.44

73.14 ± 3.59

New York, Pennsylvania,

146.0-172.5

22.3-27.3

8.0-10.8

5. 2-6. 9

81.3-101.3

65.4-78.0

Virginia

31

32

33

32

25

25

M . gallopavo osceola,

176.55 ± 8.51

23.11 ± 0.99

8.96 ± 0.68

5.94 ± 0.44

105.96

72.92

Florida

166.5-192.5

21.7-24.9

8. 1-9.9

5. 5-7.0

98.2-117.2

70.6-74.0

10

10

10

10

7

5

M. gallopavo intermedia,

172.0

23.8

9.1

6. 1

103.2

72.7

Texas

1

1

1

1

1

1

M. gallopavo mexicana.

160.5

26.2

9.6

6.6

92.6

71.8

Chihuahua, Mex.

1

1

1

1

1

1

M. gallopavo

164.53 ± 9.40

24.20 ± 1.25

9.26 ± 0.61

6.03 ± 0.40

94.84 ± 7.70

73.05 ± 3.22

Total skeletal specimens

146.0-192.5

2 1.7-27.3

8.0-10.8

5. 2-7.0

81.3-117.2

65.4-78.0

43

44

45

44

34

32

M . californica

140.16 ± 4.79

22.05 ± 0.84

8.81 ± 0.34

5.81 ± 0.26

79.41 ± 3.81

64.62 ± 2.89

Rancho La Brea

130.6-149.5

20.5-24.0

8. 1-9. 7

5. 3-6. 3

69.7-86.9

58.7-70.4

50

34

45

49

49

48

M. californica

141.60

22.32

8.74 ± 0.21

5.79 ± 0.19

81.26

64.77

Carpinteria

140.5-144.0

22.1-22.5

8.5-9. 1

5. 6-6.0

80.3-82.7

62.8-66.5

5

5

8

8

5

7

M. californica

140.29 ± 4.60

22.08 ± 0.78

8.80 ± 0.32

5.80 ± 0.25

79.58 ± 3.68

64.64 ± 2.72

Total specimens

130.6-149.5

20.5-24.0

8. 1-9. 7

5. 3-6. 3

69.7-86.9

58.7-70.4

55

39

53

57

54

55

M. ocellata

140.0

20.0

7.8

5.10

84.6

60.9

Dzibilchaltiin

4. 9-5. 3

1

1

1

3

1

1

Meleagris cf. M . ocellata

—

—

—

4.6

—

—

Cancun Island

1

M . ocellata

136.38 ± 4.96

22.56 ± 0.53

8.42 ± 0.50

6.05 ± 0.35

84.1

65.2

Mayapan

128.0-145.5

21.9-23.6

7. 5-9.0

5. 3-6. 8

17

12

15

17

1

1

M. ocellata

138.12 ± 4.41

20.36 ± 0.88

7.77 ± 0.57

5.30 ± 0.37

85.08 ± 3.86

58.20 ± 2.47

Yucatan, Mexico and

130.2-146.0

19.1-21.9

6. 3-9.0

4. 6-5. 8

78.6-91.0

53.2-61.7

Peten, Guatemala

14

14

14

12

13

13

M. crassipes

108.30*

—

7.83

4.70

59.83*

54.15*

San Josecito Cave

103.7-114.4

7. 5-8. 3

4. 5-4. 9

57.8*— 61.3*

49.9*-58.4

(spurred specimens only)8

3

3

3

3

2

M. crassipes

102.68 ± 5.18

17.85

7.51 ± 0.42

4.60 ± 0.21

59.83*

54.15*

San Josecito Cave

96. 1*— 1 14.4

17.5-18.2

6. 6-8. 3

4. 0-4. 9

57.8*— 61 .3*

49.9*-58.4

(all specimens)11

10

2

16

15

3

2

Steadman: Turkey Osteology and Paleontology

201

Table 20. Continued.

Middle of
Spur
Core

to End of
Middle
Trochlea
G

Width of
Spur Core
Ha

Length of
Spur Core

Jb

Angle
of Spur
Core (in
Degrees)
K

Distal

Width

L

Depth of
Inner
Trochlea
M

Depth of
Middle
Trochlea
N

Depth of
Outer
Trochlea
P

70.15

6.62 ± 0.49

21.97*

60.8 ± 4.4

24.57 ± 0.80

11.27 ± 0.55

12.18 ± 0.54

12.91 ± 0.33

69.2—71. 1

5. 8-7. 9

19.2*— 25.8*

56-69

23.1-25.8

10.2-11.9

11.2-12.9

12.2-13.3

2

18

7

13

13

9

18

8

—

—

—

—

23.6

1

—

12.0

1

—

68.90 ± 3.12

6.64 ± 0.84

20.84 ± 2.52

61.7 ± 5.1

24.53 ± 1.08

10.90 ± 0.69

12.01 ± 0.44

12.85 ± 0.60

62.1-73.2

5. 4-8. 7

16.8-24.9

54-73

21.9-26.6

9.5-12.4

10.8-13.0

11.5-14.0

27

21

14

26

33

32

32

32

70.04

6.32

22.48

63.4

22.46 ± 1. 19

10.18 ± 0.36

11.11 ± 0.55

12.08 ± 0.85

65.8-74.8

6. 1-6.5

22.2-23.0

57-68

20.2-24.1

9.7-10.9

10.1-11.9

11.3-14.1

7

4

4

5

10

10

10

10

68.8

7.5

22.5

55

23.4

—

—

—

1

1

1

1

1

67.9

6.4

—

56

27.8

12.2

12.3

13.1

1

1

1

1

1

1

1

69.09 ± 3.05

6.61 ± 0.76

21.27 ± 2.27

61.6 ± 5.2

24.12 ± 1.49

10.76 ± 0.73

11.81 ± 0.60

12.68 ± 0.73

62.1-74.8

5.4-8. 7

16.8-24.9

54-73

20.2-27.8

9.5-12.4

10.1-13.0

11.3-14. 1

36

27

19

33

45

43

43

43

60.77 ± 2.82

5.94 ± 0.58

18.60* ± 2.16

61.2 ± 4.9

22.53 ± 0.74

10.00 ± 0.48

10.88 ± 0.38

11.74 ± 0.56

55.7-66.5

4. 9-7. 6

14.9*-22.8*

54-70

20.9-24.6

9.0-11.1

10.1-11.6

10.4-12.7

49

47

31

47

46

37

43

35

60.60

6.11

18.87*

61.3

22.63

10.00

11.20

12.03

58.3-62.4

5. 5-6. 9

15.3*— 2 1.8

56-67

21.3-23.3

9.8-10.5

10.8-11.4

11.6-12.4

7

7

7

6

7

6

7

6

60.75 ± 2.67

5.96 ± 0.58

18.65* ± 2.13

61.2 ± 4.9

22.54 ± 0.73

10.00 ± 0.45

10.93 ± 0.37

11.79 ± 0.54

55.7-66.5

4. 9-7. 6

14.9*-22.8*

54-70

20.9-24.6

9.0-11.1

10.1-11.6

10.4-12.7

56

54

38

53

53

43

50

41

55.4

6.70

—

45

21.0

—

9.8

—

6. 2-7. 2

1

2

1

1

1

61.4

55

21.63 ± 0.87

10.90

20.2-22.9

10.7-11.1

1

1

15

2

54.41 ± 2.65

6.16 ± 0.67

27.25 ± 2.68

50.2 ± 4.6

19.91 ± 1.17

9.47 ± 0.64

9.81 ± 0.43

10.83 ± 0.72

SO.2-57.4

4. 8-7. 3

24.1-32.2

43-58

18.0-21.8

8.1-10.4

9.0-10.8

9.1-12.0

13

11

11

13

14

11

12

12

48.47*

7.0

12.6*

36

19.8

9.9

9.40*

11.0

45.5*-54.0

8.8*-10.0

3

1

1

1

1

1

2

1

48.47*

7.0

12.6*

36

18.70

9.33

8.93

10.38

45.5*-54.0

18.2-19.8

8. 8-9. 9

8.3-10.0

9.9-11.0

3

1

1

1

4

3

6

4

a Only relatively smooth adult spur cores considered; b only fully pointed spur cores considered; c from Martin and Tate 1970; (l from A.H. Miller
and Bowman 1956; e from Marsh 1872. Undoubtedly the same specimen as measured by Shufeldt 1915; f specimens measured by the author; s may
include a specimen which represents a female; h probably includes specimens which represent females.

* Slightly damaged specimen.

202

Steadman: Turkey Osteology and Paleontology

Table 21. Measurements (in mm) of the tarsometatarsus of female turkeys, with mean, standard deviation, observed range, and sample size. See
Fig. 7 for explanation of measurements.

Total

Length

Proximal

Width

Least
Width of
Shaft

Least
Depth of
Shaft

Distal

Width

Depth
of Inner
Trochlea

Depth
of Middle
Trochlea

Depth
of Outer
Trochlea

Rhegminomis

—

8.1*

3.9

2.55

9.5

—

4.40*

—

calobatesa

2.2- 2.9

4.2*-4.6

Thomas Farm

1

1

2

1

2

Proagriocharis

~ 78

12.95

5.05

3.5

—

—

—

—

kimballensis,

12.8-13.1

4. 9-5. 2

UNSM Coll. Loc.

1

2

2

1

Ft-40

Meleagris

—

—

—

3.9

—

—

8.1

—

progenes,

1

1

Rexroad

Meleagris cf. M .

124.83*

17.38*

7.60

5.08

17.85

8.30

8.93

9.2

leopoldi or M.

123.0*-126.0

16.4*— 18. 1*

7. 2-8.0

4. 8-5. 5

17.8-17.9

7. 9-8. 7

8. 9-9.0

anza, Inglis IA

3

5

2

6

2

2

3

1

Meleagris sp.

130.8

19.67

8.0

5.8

19.55

8.1

9.65

10.5

Coleman IIA

18.4-20.5

19.2-19.9

9.6-9. 7

1

3

1

1

2

1

2

1

Meleagris cf.

—

18.1

6.8

4.8

18.1

8.6

8.9

9.7

M. gallopavo

1

1

1

1

1

1

1

Carlisle Cave

M. gallopavo

—

19b

—

—

—

—

—

—

Manalapan

1

M . gallopavo

136.80

20.76 ± 0.68

7.94

5.30

19.62

8.64

10.12

10.82

Ichetucknee

13S. 1-139.7

20.0-22.1

7. 7-8. 8

4. 9-5. 9

18.0-21.1

8. 2-8. 9

9.6-10.7

10.3-11.7

River

3

8

5

7

5

5

5

5

M . gallopavo

—

18.94

7.17

4.80

18.9

8.77

9.25

—

Seminole Field

18.1-20.2

7. 1-7.3

4. 6-4. 9

8. 5-9.0

9. 1-9.4

5

3

5

1

3

2

Meleagris cf. M. gallo-

—

—

7.8

5.0

—

—

9.9

10.6

pavo

1

1

1

1

Haile IIA

M . gallopavo

—

—

6.77

4.57

17.50

7.70

8.70

8.4

Nichol’s Hammock

6. 5-7.0

4.4-4. 7

17.0-18.1

7. 2-8. 2

8. 5-9.0

3

3

3

2

3

1

M . gallopavo

136.25

19.91 ± 0.99

7.42 ± 0.44

5.00 ± 0.29

19.87 ± 0.88

8.92 ± 0.58

9.68 ± 0.55

10.88 ± 0.50

Buffalo Site,

133.0-140.0

18.0-22.7

6.5-8. 1

4. 2-5. 7

18.2-21.2

8.2-10.0

9.1-10.5

9.9-11.9

West Virginia

4

15

21

31

15

10

17

13

M . gallopavo

—

—

—

—

—

—

9.8

—

Hartman’s Cave

1

M. gallopavo

126.90 ± 5.08

19.46 ± 0.63

7.32 ± 0.45

4.94 ± 0.19

19.72 ± 0.78

8.70 ± 0.51

9.60 ± 0.30

10.15 ± 0.54

silvestris, New

115.2-134.0

18.3-20.7

6. 6-8. 2

4. 6-5. 3

18.2-21.0

8. 0-9. 8

9.1-10.1

9.3-11.1

York, Pennsyl-

19

20

20

19

20

20

20

20

vania, Virginia

M gallopavo

131.68 ± 4.00

18.30 ± 0.67

6.96 ± 0.32

4.84 ±0.11

17.76 ± 0.98

8.02 ± 0.46

8.98 ± 0.46

9.64 ± 0.54

osceola

125.0-137.0

17.9-20.2

6. 3-7. 3

4. 6-5.0

15.8-19.0

7. 2-8. 8

8. 2-9. 7

9.0-10.7

Florida

11

11

1 1

11

11

11

11

11

M. gallopavo

134.90

21.55

7.65

5.40

21.85

9.90

10.20

11.15

mexicana

130.3-139.5

20.6-22.5

7. 4-7. 9

5. 1-5.7

21.5-22.2

9.4-10.4

9.9-10.5

10.9-11.4

Chihuahua,

2

2

2

2

2

2

2

2

Mexico

M. gallopavo

128.74 ± 5.58

19.14 ± 1.09

7.21 ± 0.44

4.92 ± 0.24

19.15 ± 1.42

8.51 ± 0.70

9.41 ± 0.50

10.02 ± 0.64

Total skeletal

115.2-139.5

17.2-22.5

6. 3-8. 2

4.3-5. 7

15.8-22.2

7.2-10.4

8.2-10.5

9.0-11.4

specimens0

33

34

34

33

34

34

34

34

M. californica

114.85 ± 3.92

18.36 ± 0.64

7.05 ± 0.32

4.84 ± 0.22

18.75 ± 0.61

8.44 ± 0.40

9.28 ± 0.38

9.98 ± 0.45

Rancho La Brea

105.7-122.3

17.5-19.8

6. 5-7. 6

4. 5-5. 3

17.6-19.9

7. 9-9. 4

8.4-10.0

9.2-10.9

32

25

32

32

32

23

32

24

M. californica

116.00

18.88

7.15

4.95

19.20

8.62

9.62

10.12

Carpinteria

112.3-119.7

18.6-19.2

7. 1-7.2

4.8-5. 1

19.0-19.5

8. 4-9.0

9. 4-9. 8

9.8-10.6

2

4

2

4

4

4

4

4

Steadman: Turkey Osteology and Paleontology

203

Table 21. Continued.

Least

Least

Depth

Depth

Depth

Total

Proximal

Width of

Depth of

Distal

of Inner

of Middle

of Outer

Length

Width

Shaft

Shaft

Width

Trochlea

Trochlea

Trochlea

M. californica

114.92 ± 3.91

18.43 ± 0.63

7.05 ± 0.31

4.85 ± 0.21

18.80 ± 0.59

8.47 ± 0.38

9.32 ± 0.37

10.00 ± 0.44

Total specimens

105. 7-122.3

17.5-19.8

6. 5-7. 6

4. 5-5. 3

17.6-19.9

7. 9-9. 4

8.4-10.0

9.2-10.9

34

29

34

36

36

27

36

28

Meleagris cf.

—

17.4

—

—

—

—

—

—

M. ocellata
Barton Ramie
Site

1

M. ocellata

121.50

20.60*

8.15

5.75

20.65

—

—

—

Mayapan

120.5-122.5

20.4*-20.8*

8. 0-8. 3

5. 5-6.0

20.1-21.2

2

2

2

2

2

M. ocellata

116.86 ± 4.14

16.98 ± 0.56

6.31 ± 0.35

4.56 ± 0.36

16.33 ± 0.97

7.97 ± 0.47

8.31 ± 0.35

9.32 ± 0.45

Yucatan, Mexico

111.9-124.5

16.0-18.0

5. 8-7.0

3. 9-4. 9

14.5-17.8

7. 1-8.9

7. 9-9.0

8.7-10.0

and Peten,
Guatemala

11

11

11

9

11

9

9

9

M. crassipes

100.27*

17.85

7.44 ± 0.41

4.58 ± 0.22

18.33

9.05

8.70

10.17

San Josecito

96. 1*-103.6

17.5-18.2

6. 6-8. 2

4. 0-4. 8

18.2-18.6

8. 8-9. 3

8. 3-9.0

9.0-10.4

Cave (unspurred
specimens only)'1

7

2

13

12

3

2

4

3

a May represent a male;

b from Marsh

1872; c includes

one specimen

from northern

Florida not

identified to su

bspecies; d may

include some

specimens which represent males.
* Slightly damaged specimen.

204

Steadman: Turkey Osteology and Paleontology

Table 22. Ratios (in percent) of the measurements of the tarsometatarsus of male turkeys, with mean, standard deviation, observed range, and
sample size. See Fig. 7 for explanation of measurements.

B/A

B/E

Cl A

D/A

D/C

F/A

Proagriocharis

14. 3a

24. la

—

—

—

kimballensis, UNSM
Coll. Loc. Ft-40

1

1

Meleagris cf. M. progenes,

16.5

28.8

6.2

4.5

67.95

44.0

Benson

64.1-71.8

1

1

1

1

2

1

M. leopoldi

15.0

23.6

6.2

4.3

69.0

— 40.8

Cita Canyon

1

1

1

1

1

1

Meleagris cf. M. leopoldi

14.33

22.68 ± 0.85

6.00 ± 0.28

3.82 ± 0.28

64.40 ± 3.92

40.20

or M . anza, Inglis IA

13.8-14.8

21.1-23.6

5. 7-6.4

3.2-4. 1

56.8-71.4

39.2-41.8

7

12

8

8

18

7

Meleagris sp.

15.72

26.37

6.30

4.06

65.58 ± 2.12

44.10

Coleman IIA

15.2-16.2

25.4-27.9

6. 0-6. 6

3. 9-4. 4

61.9-68.5

39.5-46.6

4

3

4

5

9

3

M . gallopavo

—

—

5.9*

3.7*

63.3

45.0*

American Falls

1

1

1

1

M. gallopavo

14.1*

26.1*

5.8

3.8

66.0

44.9

Ingleside

1

1

1

1

1

1

M. gallopavo

13.0

20.9

—

—

—

—

Manalapan1'

1

1

Meleagris cf. M . gallopavo,

—

—

—

—

68.5

—

Santa Fe River IIA

1

Meleagris sp.

—

—

—

—

65.4

—

Withlacoochee River

1

Meleagris cf. M. gallopavo

15.4

25.2

5.55*

3.75*

68.00

44.20*

Reddick IB

5.3*-5.8

3.7*-3.8

66.0-70.0

42.6-45.8*

1

1

2

2

2

2

Meleagris cf. M . gallopavo

—

—

—

—

—

—

Sabertooth Cave

M . gallopavo

—

—

—

—

—

—

Aucilla River

M. gallopavo

14.00

23.47

5.50

3.55

66.74 ± 4.68

42.25

Ichetucknee River

13.4-14.6

22.2-24.4

5. 2-5. 8

3.4-3. 7

63.0-77.3

42.1-42.4

2

3

2

2

9

2

M . gallopavo

—

—

—

—

—

—

Seminole Field

Meleagris cf. M. gallopavo.

64.2

Santa Fe River IA

1

M . gallopavo

—

23.4*

—

—

—

—

Wacissa River

1

M. gallopavo

14.8

25.2

5.4

3.8

69.11 ± 3.24

—

Buffalo Site

64.6-73.8

1

1

1

1

11

M . gallopavo silvestris,

15.27 ± 0.72

27.11 ± 1.23

5.82 ± 0.41

3.76 ± 0.25

64.70 ± 3.06

45.59 ± 1.24

New York, Pennsylvania,

13.8-16.7

23.9-29.1

5. 0-6. 6

3. 2-4. 3

58.8-74.7

43.2-48.1

Virginia

31

25

31

30

32

24

M. gallopavo osceola,

13.10 ± 0.57

21.71

5.08 ± 0.38

3.40 ± 0.31

66.51 ± 5.55

41.78

Florida

12.4-14.2

19.8-23.9

4. 5-5. 7

3. 1-4.0

59.1-77.8

39.6-43.5

10

7

10

10

10

5

M. gallopavo intermedia.

13.8

23.1

5.3

3.5

67.0

42.3

Texas

1

1

1

1

1

1

M . gallopavo mexicana,

16.1

28.3

5.7

4.1

68.7

44.7

Chihuahua, Mexico

1

1

1

1

1

1

M. gallopavo

14.75 ± 1.16

25.91 ± 2.59

5.64 ± 0.50

3.68 ± 0.31

65.25 ± 3.76

44.84 ± 1.94

Total skeletal specimens

12.4-16.7

19.8-29.1

4. 5-6. 6

3. 1-4.3

58.8-77.8

39.6-48.1

43

34

43

42

44

31

M . californica

15.70 ± 0.62

27.85 ± 1.58

6.26 ± 0.21

4.13 ± 0.18

66.04 ± 2.84

46.11 ± 1.62

Rancho La Brea

14.5-17.1

24.0-32.0

5. 9-6. 7

3. 8-4. 6

60.2-74.1

41.8-49.5

34

33

45

49

45

48

Steadman: Turkey Osteology and Paleontology

205

Table 22. Continued.

G/A

H/J

J/A

K/A

L/A

M/A

N/A

P/A

40. 8a

39.5*

16 o*a-b

38. 8a

—

—

—

—

1

1

1

1

42.8

—

—

-

16.6

7.7

7.9

8.8

1

1

1

1

1

36.3

—

—

—

15.0

6.8

7.3

8.0

1

1

1

1

1

36.74

25.68* ± 1.83

20.35*

31.98

14.26

6.24

6.86 ± 0.31

6.96

35.6-38.5

23.2-2 9.0

18.5*— 22.2*

28.9-35.1

13.7-14.8

5. 9-6. 7

6. 4-7. 3

6. 7-7.3

7

14

6

6

7

5

8

5

40.47

27.10*

17.85*

36.45*

15.50

6.7

7.75

8.1

36.3-43.1

22.0*-34. 1*

14.0* — 2 1.7*

33.7*-38.4

14.8-15.8

7. 6-7. 9

3

4

2

4

4

1

2

1

42.3*

40.2

10.5*

30.2*

15.7*

—

7.3*

7.9*

1

1

1

1

1

1

1

42.4

—

—

34.0

14.6

6.2

6.7

7.4

1

37.7

1

—

—

1

1

1

1

1

40.95*

—

—

38.0*

14.40*

6.45*

7.00*

7.65*

39.1-42.8*

13. 9* — 14.9

6. 4-6. 5*

6.9-7. 1*

7.3*-8.0

2

33.7*

1

2

2

2

2

—

1

30.5*

1

—

—

—

—

—

—

39.95

36.5*

9.1*

36.85

13.65

6.00

6.95

7.40

39.9-40.0

36.5-37.2

13.3-14.0

5. 5-6. 5

6. 6-7. 3

6. 9-7. 9

2

1

1

2

2

2

2

2

38.40*

—

—

—

—

—

—

36. l*-40.0*

3

—

35.3*

1

—

—

—

—

—

—

41.2

30.90*

_

33.3

_

7.0

7.4

27.0*-33.8*

1

7

1

1

1

42.92 ± 1.29

32.09 ± 3.08

12.82 ± 1.35

38.57 ± 3.54

15.32 ± 0.69

6.81 ± 0.37

7.51 ± 0.31

8.02 ± 0.40

40.8-45.2

28.7-40.0

10.4-14.7

33.3-45.3

13.6-16.9

6. 2-7. 5

6. 8-8. 2

6. 8-8. 8

25

13

14

25

31

30

30

30

39.81

28.15

13.05

36.28

12.72 ± 0.57

5.77 ± 0.25

6.28 ± 0.20

6.85 ± 0.51

37.3-43.2

27.5-28.4

13.0-13.2

32.8-40.0

11.8-13.4

5.2-6. 1

5. 9-6. 5

6.4-8. 1

7

4

4

5

10

10

10

10

40.0

33.3

13.1

32.0

13.6

—

—

—

1

1

1

1

1

42.3

—

—

34.9

17.3

7.6

7.7

8.2

1

1

1

1

1

1

42.18 ± 1.92

31.28 ± 3. 12

12.88 ± 1.15

37.89 ± 3.67

14.72 ± 1.35

6.58 ± 0.58

7.22 ± 0.61

7.74 ± 0.66

37.3-45.2

27.5-40.0

10.4-14.7

32.0-45.3

11.8-17.3

5. 2-7. 6

5. 9-8. 2

6. 4-8. 8

34

18

19

32

43

41

41

41

43.35 ± 1.60

33.18* ± 3.76

13.18* ± 1.48

43.68 ± 3.59

16.11 ± 0.42

7.10 ± 0.30

7.76 ± 0.27

8.35 ± 0.37

39.2-46.6

27.6*-46.0*

10.2* — 15.3*

37.4-52.1

15.0—16.9

6.4-7. 7

7.2-8 .3

7. 5-9.0

49

31

31

47

46

37

43

35 Continued

206

Steadman: Turkey Osteology and Paleontology

Table 22. Continued.

B/A

B/E

C/A

D/A

D/C

F/A

M. californica

15.78

27.58

6.25

4.16

66.31

45.62

Carpinteria

15.5-16.0

27.3-27.8

6. 0-6. 5

4.0-4. 3

63.6-69.8

44.5-46.2

4

4

4

5

7

5

M . californica

15.71 ± 0.59

27.82 ± 1.49

6.26 ± 0.21

4.13 ± 0.17

66.08 ± 2.74

46.06 ± 1.56

Total specimens

14.5-17.1

24.0-32.0

5. 9-6. 7

3. 8-4. 6

60.2-74.1

41.8-49.5

38

37

49

54

52

53

M . ocellata

14.3

23.6

5.6

3.8

67.9

43.5

Dzibilchaltun

1

1

1

1

1

1

M . ocellata

16.71 ± 0.61

—

6.15 ± 0.33

4.44 ± 0.18

71.87 ± 2.76

44.8

Mayapan

15.6-17.5

5. 5-6. 6

3. 9-4. 7

68.2-76.2

12

15

17

15

1

M . ocellata

14.74 ± 0.59

23.99 ± 1.28

5.61 ± 0.31

3.82 ± 0.18

68.30 ± 3.83

41.97 ± 1.99

Yucatan, Mexico

13.9-15.9

22.3-2 6.6

4. 8-6. 2

3.5-4. 1

61.1-73.4

38.2-45.2

and Peten, Guatemala

14

13

14

12

12

13

M . crassipes

—

—

7.20*

4.33*

60.00

49.55*

San Josecito Cave

7.0*-7.4

4.2*-4.5

59.0-61.0

48. l*-5 1.0

(spurred specimens only)

3

3

3

2

M. crassipes

17.55

—

7.40* ± 0.33

4.55* ± 0.20

61.55 ± 3.15

49.55

San Josecito Cave

16.9-18.2

7.0*-8.0*

4.2*-4.9*

54.8-66.7

48. l*-5 1.0

(all specimens)

2

10

10

15

2

Steadman: Turkey Osteology and Paleontology

207

Table 22. Continued.

G/A

H/J

J/A

K/A

L/A

M/A

N/A

P/A

42.58

32.57*

13.12*

44.80

16.22

7.13

7.88

8.62

41.3-43.3

26.9*-35.9*

10.8*-15 .4*

39.6-47.7

16.0-16.5

7. 0-7. 3

7. 7-8.0

8. 3-8. 8

5

7

4

4

4

3

4

4

43.28 ± 1.56

33.07 ± 3.63

13.17* ± 1.51

43.77 ± 3.56

16.12 ± 0.40

7.10 ± 0.29

7.77 ± 0.26

8.38 ± 0.36

39.2-46.6

26.9*-46.0*

10.2*— 15.4*

37.4-52.1

15.0-16.9

6.4-7. 7

7. 2-8. 3

7. 5-9.0

54

38

35

51

50

40

47

39

39.6

—

—

32.1

15.0

—

7.0

—

1

1

1

1

42.2

—

—

37.8

15.84 ± 0.66

—

7.75

—

14.5-16.8

7.4-8. 1

1

1

15

2

38.67 ± 1.67

22.90 ± 2.64

19.64 ± 1.80

36.22 ± 3.89

14.40 ± 0.56

6.82 ± 0.38

7.08 ± 0.21

7.82 ± 0.43

36.1-41.4

19.2-27.9

17.3-23.1

29.4-43.6

13.3-15.1

6. 2-7. 3

6. 8-7. 5

7. 0-8. 6

13

9

11

13

14

11

12

12

44.67*

55.6*

11.0*

31.5

17.3

8.6

8.60

9.6

42.6*-47.2

8. 5-8. 7

3

1

1

1

1

1

2

1

44.67*

55.6*

11.0*

31.5

17.83

8.70

8.62

9.80

42.6*-47.2

17.3-18.6

8. 6-8. 8

8. 4-9.0

9.6-10.2

3

1

1

1

3

2

5

3

a Based on at least one measurement of Martin and Tate 1970; b based on different individuals (UNSM 20036, 20038); c from measurements of
Marsh 1872; 11 may include one or more specimens which represent females.

* Slightly damaged specimen.

LATE PLEISTOCENE AND HOLOCENE TURKEYS
IN THE SOUTHWEST

By Amadeo M. Rea1

ABSTRACT: Late Quaternary turkey remains from 17 southwestern sites are analyzed. All pre-ag-

ricultural turkeys, except those from northern Sonora and one cave in southern New Mexico, are found
to be Meleagris crassipes L. Miller, an extinct species not closely related to the modern M. gallopavo
Linnaeus, which inhabits much of the Southwest today. M. gallopavo is found associated with sedentary
agriculturalists with a subsistence base of two or three crops at all other archaeological sites and time
horizons. The major southwestern Indian cultures are herein delimited in time and space, emphasizing
Mesoamerican components (particularly the four cultivars — maize, squash, gourd, beans — and macaws).
It is proposed that the living turkey M. g. merriami Nelson is a parallel Mesoamerican component that
was imported and became feral with the breakdown of southwestern cultures that had occurred by A.D.
1450, if not before.

It is currently believed that the Common Turkey, Meleagris
gallopavo Linnaeus, occurred in the Southwest in the Pleisto-
cene as well as the Holocene ( = Recent) Epochs, evolving from
local precursors (AOU 1957, Brodkorb 1964a; Schorger 1966;
Steadman this vol.). The late Lyndon L. Hargrave long main-
tained (1970a: 16, 25) thatM. g. merriami Nelson, the subspe-
cies of Common Turkey found today throughout most of the
Southwest (Fig. 1), was derived from Pueblo Indian domes-
ticated turkeys that became feral at the time of or following
the breakdown and dispersal of the Anasazi Culture in the late
thirteenth century. Hargrave’s reasoning was that there were
no Pleistocene or Pre-Basket Maker II cultural horizon Com-
mon Turkeys known from the Southwest. The independent
discovery by Storrs Olson of the National Museum of Natural
History, Smithsonian Institution, and by me, of the quite dis-
tinct paleospecies Meleagris crassipes L. Mi'iler, in cave de-
posits in New Mexico, provided an opportunity to test Har-
grave’s hypothesis. Dr. Olson kindly placed the cave materials
he was studying at my disposal because of the cultural (ethno-
biological) ramifications of this problem. To resolve the ques-
tion of the origin of M. gallopavo, a re-examination of all re-
puted pre-agricultural specimens of M. gallopavo throughout
the range of modern M. g. merriami was necessary, particu-
larly since both Olson and I found both M. crassipes and M.
gallopavo, in the same cave deposits.

In this paper I shall attempt to present arguments that will
provide answers to the following questions: (1) What were the
geographic and temporal ranges of M. crassipes and M. gal-
lopavo in the Southwest? (2) When did M. crassipes become
extinct? (3) What was the relationship between the evolution
of sedentary agricultural people and the domestication of the
turkey in the area? (4) Did Puebloan peoples capture and do-

1 Curator of Birds and Mammals, San Diego Natural History Mu-
seum, San Diego, CA 92112.

mesticate birds from preexisting local wild populations, or
were domestic turkeys imported along with Mesoamerican cul-
tigens (maize, beans, pumpkins) and the macaw?

On the basis of the examination of existing fossil and ar-
chaeological remains, the first two questions can be answered
with some degree of certitude, though carefully dated, strati-
fied excavations might modify these answers. The third ques-
tion will be sketched only in its broadest outlines, with the
details left to the study by Charmion R. McKusick (this vol.).
The final question cannot be answered directly, but a sugges-
tion can be offered on the basis of available remains.

MATERIALS AND METHODS

Fossil and modern specimens cited in this report are distin-
guished by the following initials: AMNH (American Museum
of Natural History), AMR (A.M. Rea Collection), FM (Field
Museum), LACM (Natural History Museum of Los Angeles
County), LLH (L.L. Hargrave Collection, Museum of North-
ern Arizona), MALB (Museum of Arid Land Biology, Uni-
versity of Texas, El Paso), MWU (Midwestern University),
SC (Stanton’s Cave, Museum of Northern Arizona), SD (San
Diego Natural History Museum), TMM (Texas Memorial
Museum, University of Texas, Austin), UA (Department of
Ecology, Llniversity of Arizona), UAPL (Paleontology Labo-
ratory, University of Arizona), USNM (National Museum of
Natural History, Smithsonian Institution), and WAC (Western
Archeological Center, Tucson).

All fossil and archaeological specimens were compared with
the type series of M. crassipes (LACM), M. californica
(LACM), and Recent wild M. gallopavo (total 31; M. g. mer-
riami, M. g. intermedia Sennett, and M. g. silvestris Vieillot;
AMR, UA, SD, LLH). Most of the cave material is not min-
eralized, except where noted. For identifications I used both
qualitative and quantitative characters with a heavier reliance

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 330:209-224.

210

Rea: Southwestern Turkeys

Figure 1. Historic ( 1 9th— 20th Century) distributions of southwestern subspecies of Meleagris gallopavo. After European settlement, native turkeys
were extirpated from the border ranges of southeast Arizona, and the areas were restocked with M . g. merriami starting early this century. Some
forested areas of Utah and northwest Arizona, lacking historic populations, have been stocked with M. g. merriami. The Colorado and New
Mexico range has been greatly reduced. (After Aldrich and Duvall 1955; Bailey 1929; Bailey and Neidrach 1965; Phillips et al. 1964; Schorger
1966; and other sources.)

on qualitative differences between the two species where pos-
sible. Most qualitative characters are given by Steadman (this
vol.), although a few additional characters are mentioned be-
low. Osteometric data are also given by Steadman (this vol.).
Radiocarbon dates are followed by the standard deviation,
with the laboratory and sample number in parentheses. All
measurements are in millimeters. I have used two systems of
dating throughout: years before present (B.P.) for dates older
than 2000, and A.D. designations of the Gregorian calendar
for those younger. Paleontologists more frequently use the for-

mer, prehistorians the latter. Breaking at A.D. 1/2000 B.P.
appears to be the least awkward compromise.

SYSTEMATICS

Meleagris crassipes has been known since its discovery and
description (Miller 1940, 1943) only from San Josecito Cavern
in southern Nuevo Leon, Mexico, on the east flank of the
Sierra Madre Oriental. The species is represented by over 50
elements, mostly limb bones. The scapula, one character of

Rea: Southwestern Turkeys

211

which is most important in the evolutionary history of turkeys
(Steadman this vol.), is unknown. The associated fauna of San
Josecito Cavern includes the Rancholabrean land mammals
Canis divas, Nothrotherium sp., Equus sp., Tetrameryx sp.,
Felix atrox, and Smilodon sp. Extinct birds include Coragyps
occidentalis (L. Miller) (=C. atratus ? (Bechstein)), Teratornis
merriami L. Miller, Spizaetus grinnelli (L. Miller), Neogyps
errans L. Miller, Neoplirontops americanus L. Miller, Wet-
moregyps daggetti (L. Miller), and Polyborus prelutosus How-
ard (Miller 1943). Meleagris crassipes is a distinctive species,
showing little similarity to the two living species of turkeys,
M. gallopavo and M. ocellata Cuvier, or to their immediate
precursors. Instead, it appears to have been a dead-end side
branch in the evolutionary history of turkeys (Steadman this
vol.). M. crassipes was a small turkey with relatively large
legs, and little sexual dimorphism in size. San Josecito Cavern
had surface evidence of human occupancy, but all the bird
bones were recovered from “below the zone of human activity”
(Miller 1943:144).

I am convinced that the osteological and external differences
between the two living species of turkeys are insufficient to
warrant placing them into separate monotvpic genera, Agri-
ocharis Chapman and Meleagris Linnaeus. The supposed ge-
neric characters are almost exclusively a matter of secondary
sexual characteristics, such as the position, angle, and length
of metatarsal spur, and the male head and chest ornamenta-
tion. The structure and color pattern of the wing, mantle,
rump, and breast feathers of the two living species are strik-
ingly similar. Osteologically, Holocene M. gallopavo and M.
ocellata more closely resemble each other than either does M.
crassipes. Ridgway (in Ridgway and Friedmann 1946:458)
noted, “ Agriocharis is, in fact, so closely related to Meleagris
that I am somewhat doubtful as to the expediency of recog-
nizing it as a genus.” Brodkorb (1964a, b) removed the fossil
species M. leopoldi A. Miller and Bowman and M. crassipes
from Meleagris, leaving in that genus the three species M. alta
Marsh, M. tridens Wetmore, and M. gallopavo Linnaeus. I
regard Agriocharis as a synonym of Meleagris, thus returning
the species M. ocellata Cuvier, M. leopoldi A.H. Miller and
Bowman, and M. crassipes L.H. Miller to Meleagris, along
with M. progenes (Brodkorb) and M. anza (Howard). I also
strongly doubt that M. californica (L. Miller) is sufficiently
distinct to merit being placed in the separate genus Parapavo
L. Miller, and prefer considering it as well a species of Mele-
agris. Steadman (this vol.) has independently reached similar
conclusions.

Finally, the so-called New World family Meleagrididae
seems to me unjustifiable. The two living species and their
paleo-antecedents are merely medium to large pheasants. I
recommend placing them in the family Phasianidae, together
with the spurred fowl of the Old World.

FOSSIL AND EARLY ARCHAEOLOGICAL
RECORDS OF MELEAGRIS IN
THE SOUTHWEST

STANTON’S CAVE, Coconino Co., Arizona. Grand Can-
yon, 51 river km below Lee’s Ferry. Distal end of tarsometa-
tarsus (SC 76). This bone was obtained from a packrat nest
in the cave. Individual nests have not been dated, but the cave
floor deposit ranges from 38,000 B.P. to present, with the bulk

of the material being Pleistocene (Euler 1978). Bird bones re-
covered from the cave include Teratornis merriami, Gymno-
gyps ampins L. Miller (=G. californianus ? Shaw), G. califor-
nianus, and Centrocercus urophasianus (Bonaparte). All of
these species are absent from the area today. Mammals include
Oreamnos harringtoni and Bison sp. The packrat nest was
burned by vandals before the cave was excavated, so the tar-
sometatarsus is slightly calcined but not distorted. The speci-
men is clearly from M . crassipes on the basis of characters and
lies within the upper size range of the specimens from San
Josecito Cavern. The only evidence of human activity in Stan-
ton’s Cave was the presence of caches of split willow twig
figurines, dated 3000-4000 B.P. (Euler 1978:158). The cave
was never a habitation site.

LAGUNA SALADA, Apache Co., Arizona. Distal end of
tibiotarsus (FM uncatalogued). Martin and Rinaldo (1960:115)
reported the tibiotarsus of a turkey, M. gallopavo, taken at a
playa camp site on the Upper Colorado, radiocarbon dated
3280 ± 60 (Gro. 1614) B.P. I have re-examined this bone and
find it to be from a Sandhill Crane, Gras canadensis (Lin-
naeus). Certain bones of Gras and Meleagris are superficially
similar (Hargrave and Emslie 1979).

PAPAGO SPRINGS CAVE, Santa Cruz Co., Arizona.
Complete humerus (AMNH 8683, 8687); distal half of right
humerus (AMNH 8693); proximal half of right femur (AMNH
8684); proximal two thirds of right femur (AMNH 2685); all
mineralized. Steadman (pers. comm.) found these specimens
indistinguishable from those of female M. gallopavo or male
M. crassipes. I cleaned these fossils of some heavy matrix to
expose critical characters for identification. The complete hu-
merus is from M. crassipes. It differs from M. gallopavo in:
much greater protrusion of head anconad, especially mediad;
narrower capital groove; and very much smaller and differ-
ently shaped impression of M. brachialis anticus. Its length is
about 121.8 mm. The scar of M. latissimus dorsi is not ex-
posed. The partial humerus is from M. cf. crassipes. The
impression of M. brachialis anticus is very short and broad,
not extending up the shaft as in M. gallopavo. Its distal width
is 26.5 mm. The femora are from M. crassipes. They differ
from M. gallopavo in: very much wider neck, not pinched off;
thicker ridge on posterior view separating trochanter from
head and neck (2.5 mm vs. a fine line in equivalent-sized M.
gallopavo)', and general configuration of obdurator ridge area.
The fauna of Papago Springs Cave is late Pleistocene, with
abundant fossils of the extinct pronghorn, Stockoceros onus-
crosagris, as well as Camelops sp., Bison sp., Platygonus sp.,
and two species of Equus (Skinner 1942).

NORTH PAPAGO (SONOITA) CAVE, Santa Cruz Co.,
Arizona. Complete tarsometatarsus lacking spur (AMNH
8686). This specimen is from M. crassipes. This cave is an
extension of the above and is presumably the same age.

ARIZPE, Sonora, Mexico. Rio San Miguel drainage 97 km
southeast of Cananea. Head of humerus (AMNH 6823), min-
eralized. This bone was identified by Cracraft (1968) as M.
gallopavo. I can find no characters to distinguish it from a
large male of that species. It is far too large (width of head,
38.9 mm) for M. crassipes, but it could be a very large M.
californica. Steadman (this vol.) considers the character dif-
ferences of the head of the humerus noted by Cracraft (1968)
too inconsistent for a specific identification between M. gal-
lopavo and M. californica. I refer the humerus to M. gallopavo

212

Rea: Southwestern Turkeys

on geographic probability. This deposit from northeastern So-
nora yielded Bison cf. alleni and Equus cf. tau, suggesting a
late Pleistocene age.

LA BRISCA, Sonora, Mexico. Rio San Miguel drainage
about 25 km northwest of Arizpe. Distal end of ulna (IGCU-
2546), mineralized. Size is not diagnostic in this specimen (dis-
tal width, 13.0 mm). Steadman’s measurements (this vol., Ta-
bles 10, 1 1) indicate that the specimen lies within the size range
of female M. gallopavo, female M. californica, or male M.
crassipes. The ulna of M. crassipes is distinguishable from
both M. gallopavo and M. californica on the basis of the shape
of the internal condyle (short, almost squared in the former;
broadly rounded in the latter two). The fossil ulna is the latter
type. On the basis of this character and geographic probability
I refer the La Brisca fossil to M . gallopavo.

TULAROSA CAVE, Catron Co., New Mexico. FM
73,2504 and 73,648, tarsometatarsi; 73,647, coracoid lacking
part of head; 73,2247, head of coracoid; all from the pre-pottery
occupational phase, radiocarbon dated 2300 (±200) to ca. 2150
years B.P. (Additional turkey bones recovered from younger
cultural levels not re-examined.) All four elements are M. gal-
lopavo. The tarsometatarsi are smaller and appear more gra-
cile than in the wild female M. g. merriami living in the area
today. The pathological coracohumeral surface of one cora-
coid suggests captivity.

SAN ANTONIO SITE, Socorro Co., New Mexico. Rio
Grande south of Socorro, 5.6 km NE San Antonio. USNM
14690: distal end of humerus; radius, lacking distal end; nearly
complete ulna. Steadman (pers. comm.) examined these bones
and could find no qualitative characters. Measurements are:
humerus, distal width 23.8 mm; ulna: length 101.8+ mm,
distal depth 12.3 mm; radius: proximal width 7.9 mm, prox-
imal depth 8.3 mm. On the basis of measurements (Steadman
this vol., Tables 8 through 13), I refer these specimens to M.
crassipes. The fossils are from the base of a pumicite bed and
there were no other associated bones. They are presumed to
be Pleistocene, perhaps Blancan (Needham 1936).

HOWELL’S RIDGE CAVE, Grant Co., New Mexico.
There have been three excavations of this cave deposit, two
producing bones of Meleagris. Zeller-Howard pit: coracoid
head (LACM 33890), scapula (LACM 33889), proximal and
distal ends of an ulna (LACM 33891, 33892). Van Devender
pit: tarsometatarsus, rodent gnawed on trochleae and most of
head (SD uncataloged). No stratigraphic data were recorded
on materials recovered from the Zeller-Howard pit, but How-
ard (1962) presumed that they were of late Pleistocene or Ho-
locene age. Associated (Howard 1962) were Equus sp., Cam-
elops sp., abundant Gymnogyps amplus ( =G . calif or ni anus?)
(including young), Coragyps occidentalis (=C. atratus?), and
single elements of Spizaetus sp. and Anabernicula sp. Howard
(1962:242) considered only the possibility that the turkey was
either M. gallopavo or M. californica. She very tentatively
referred the turkey elements to M. gallopavo. I identified the
proximal end of the ulna (33891) as from M. crassipes on the
basis of the short, distinctly squared shape of the internal con-
dyle. After further cleaning of the head of the coracoid, I am
inclined to consider it a specimen of M. crassipes rather than
M. gallopavo or M. californica on the basis of the shape of the
coracohumeral surface (broader, less triangular). The How-
ell’s Ridge scapula has an almost obsolete foramen, unlike the

known M. gallopavo and M. californica where the foramen is
well developed, and it well may represent M. crassipes.

The large galliform bones recovered from the Harris pit
(UTEP) are Centrocercus urophasianus . Van Devender (Van
Devender and Worthington 1978) excavated additional parts
of this cave with careful stratigraphic controls. Several large
galliforms were obtained from the 90-100 cm level, midway
between radiocarbon dates of 3330 ± 170 B.P. (A-1354) on
the 70-80 cm level, and 6697 ± 324 B.P. (average of A-1429
and A-1430) on the 110-112 cm level Both dates are based
on endocarps of Celtis reticulata (netleaf hackberry). The 90-
100 cm level includes the nearly complete and excellently pre-
served tarsometatarsus of M. crassipes, apparently a mature
female (no evidence of spur attachment). The other galliforms
from this stratum are Centrocercus urophasianus. Zeller
(Howard 1962:241) suspected human association on the basis
of a few flint chips and charcoal, but Van Devender and Harris
(pers. comm.) found no suggestion of human occupation.

SHELTER CAVE, Dona Ana Co., New Mexico. Incom-
plete humerus (LACM 1010/653), two pedal phalanges (LACM
1010/657), distal ends of right and left ulnae (LACM 1010/556,
557). The humerus more closely resembles topotypes of M.
crassipes than Hargrave’s extensive series of M. gallopavo mer-
riami in such characters as: (1) general greater curvature of
the shaft; (2) size (least depth of shaft 9.4 mm); (3) very small
brachial depression; and (4) small, slender depression for M.
latissimus dorsi. The pedal phalanges cannot be identified to
species. The ulnae are identical to those of M. crassipes, except
one measures slightly larger (16.0 mm) than the largest of six
ulnae (14.1 to 15.6 mm) of M. crassipes from San Josecito
Cavern. Associated with these specimens are Gymnogyps cal-
ifornianus ? and the extinct species Urubitinga fragilis, Geo-
coccyx conklingi Howard, and Pyelorhamphus molothroides
A. Miller, all indicating Late Pleistocene age, but with evi-
dence of Basket-Maker-like (Archaic) material culture over-
lying the fore part of the cave (Howard and A. Miller 1933).
This cave and the following are on the west and east sides,
respectively, of Pyramid Peak in the southern Organ Moun-
tains, north of El Paso, Texas.

CONKLING CAVERN, Dofia Ana Co., New Mexico.
Shaft of humerus (LACM 1009/21), carpometacarpus (LACM
1009/22), distal end of radius (LACM 1009/23). The humerus,
from the 20-23 foot level, is far too large (least diameter of
shaft, 11.9 mm; Steadman this vol., Tables 8, 9) and too
straight for M. crassipes, but agrees with that of male M.
gallopavo. The brachial depression is not deep, but its size and
shape resemble those found in M. gallopavo. Surface striations
indicate immaturity. The carpometacarpi of M. crassipes and
M. gallopavo can be distinguished by the outer curvature of
the distal end of metacarpal III, but this area is broken on
specimen 1009/22. Measurements of the carpometacarpus are:
proximal depth 23.4 mm, minimum axial length 75.6 mm, and
maximum axial length (78.2+ mm). These measurements are
all too large for M. crassipes and greater than the largest M.
californica measured by Steadman (this vol., Tables 14, 15).
Its provenience is marked “dump” (presumably material out-
side the cave disturbed by treasure hunters, lacking strati-
graphic data). The radius, also lacking provenience data, mea-
sures 12.8 mm in distal width. I consider all three elements to
be M. gallopavo. Associated specimens from Conkling Cavern

Rea: Southwestern Turkeys

213

are Coragyps occidentalis (=C. atratus ?), Gymnogyps calif or-
nianus, Geococcyx conklingi , as well as Centrocercus
urophasianus and Cyanocephalus cyanocephalus, two species
living today in more northern arid regions (Howard and Miller
1933). Mammals include extinct Equus sp., an ursid, Hem-
iauchenia sp., Camelops sp., and Nothrotherium. There were
human remains (presumably Paleo-Indian) at the 10 and 2 1
foot levels, but no Late Archaic material culture. At the 20
foot level there was a hard, water-deposited, horizontal lens,
precluding the possibility of intrusive burial (Bryan 1929).
Howard and A. Miller (1933:17) consider the better stratified
lower levels of Conkling Cavern to be Pleistocene.

DRY CAVE, Eddy Co., New Mexico. Guadalupe Moun-
tains. Distal end of tarsometatarsus (trochlea gnawed by ro-
dent) (MALB 5-239). This specimen of M. crassipes was ex-
cavated by A.H. Harris from the Sabertooth Camel Maze
section of Dry Cave. The fossil remains are dated 25,160 ±
1730 B.P. This portion of the cavern was sealed so that it was
not contaminated with younger materials (Harris pers.
comm.). Associated animals include Breagyps clarki, Canis
dims, Camelops sp., Tapirus sp., and Equus sp.

BURNET CAVE, Eddy Co., New Mexico. East slope of
Guadalupe Mountains, about 80 road km west of Carlsbad.
Shaft of humerus (ANSP 14161); carpometacarpus (ANSP
13495); femur, lacking distal end (ANSP 14134); tarsometa-
tarsus with spur (ANSP 13492); tibiotarsus, lacking proximal
end (ANSP 14133). The upper 0.5 to 1 meter of the open cave
deposit have Basket-Maker-like burials, but lack corn cobs
and pottery (Schultz and E. Howard 1935:273-274). Beneath
1 meter, extending to a depth of 3 meters, were numerous
hearths and “Folsom-like” artifacts associated with extinct
mammals ( Bison antiquus, Preptoceras sinclairi, Stockoceros
onusrosagris , Euceratherium collinum) and the condor Gym-
nogyps calif ornianus . The spurred tarsometatarsus and the
carpometacarpus, both recovered in the 5 ft 9 in. stratum, are
from M. crassipes. The femur (stratum unknown) also appears
to be from M. crassipes (muscle lines and other characters)
and is buff-tan colored like the previous two elements. The
tibiotarsus (14133) is too large and too long for M. crassipes
(width across distal cotyla 18.4 mm vs. maximum 17.4 mm in
M. crassipes ; distance from distal end of scar for attachment
of fibula to distal end 122.7+ mm vs. maximum 104.3 mm in
M. crassipes; Steadman this vol., Tables 18, 19). The shaft of
the humerus (14161) is too straight and too large for M. cras-
sipes. Both the tibiotarsus and humerus fit M. gallopavo, the
humerus being in the size range of a male. These latter two
bones, as Olson (pers. comm.) has pointed out, are quite white
in color, in contrast to the brownish (and probably older) M.
crassipes from this deposit. These two elements were presum-
ably associated with the upper 0.5 meter of deposit containing
Archaic cultural remains.

DARK CANYON CAVE, Eddy Co., New Mexico. About
25 km southwest of Carlsbad. Nearly complete coracoid
(AMR, uncataloged). Howard (1971:237-240) reported on a
large portion of this enormous deposit, although thousands of
bird bones still remain to be identified. She reported two tur-
key bones, listing them as “ Meleagris gallopavo ?” These and
certain other galliform bones were returned to the late L.L.
Hargrave, and cannot now be found in his collection. The
cave contained abundant teeth of Equus sp. , and many extinct

avian species, including Anabernicula sp., Gymnogyps ampins
(=G. calif ornianus?), Coragyps occidentalis (=C. atratus?),
Neophrontops sp., Geococcyx conklingi, and others. The de-
posit containing the extinct species was overlain with Basket-
Maker-like material culture.

A coracoid, lacking part of the sternal facet, was obtained
during subsequent excavations at approximately the 15 foot
level. The distinctive shape of the coracohumeral surface and
additional characters mentioned by Steadman (this vol.) in-
dicate that it is from M . crassipes. Measurements of the spec-
imen are: head to internal distal angle 70.5 mm; head through
scapular facet 26.3 mm; least width of shaft 7.6 mm. It was
associated with the coracoid of Gymnogyps sp.

PRATT CAVE, Culberson Co., Texas. McKittrick Canyon,
south of Carlsbad, Guadalupe Mountains. Fragmental distal
end of humerus (WAC 38A1); fragmental distal end of shaft
of tibiotarsus (WAC 35 A 1); distal end of tibiotarsus (WAC
2599/34A1). There is some question of the antiquity of this
deposit. I have re-examined the extralimital pigeon, Columba
flavirostris Wagler (distal end of humerus), and the extinct
roadrunner, Geococcyx conklingi (proximal end of tibiotarsus),
identified from the deposit by Hargrave. I find the humeri of
C. flavirostris and C. fasciata Say, the species of pigeon to be
expected at the site, to be indistinguishable, even using the
comparative materials from the Hargrave Collection (all the
C. flavirostris specimens being from captive birds). The bone
from Pratt Cave is undoubtedly C . fasciata. Rob McKenzie
and I have compared the roadrunner tibiotarsus with the type
series (LACM) of Conkling’s Roadrunner, finding that it in-
deed appears to be G. conklingi, though it is slightly larger
and somewhat different in characters. This cave lacks an ex-
tinct megafauna, and the herpetofauna is also modern. The
entire deposit appears to be less than 6000 B.P. (Gehlbach and
Holman 1974:191, 195). The late Archaic remains are radio-
carbon dated 2320 ± 70 to 1420 ± 60 B.P. A. Schroeder (pers.
comm.) provided stratigraphic data on the cave showing that
all the turkey bones are from cultural levels in the cave. The
fragmentary humerus is the size of a large male M . gallopavo
(depth through internal condyle 12.5 mm; depth through ex-
ternal condyle 27.6 mm). The shaft of a tibiotarsus is immature
and burned at the distal end. It cannot be identified to species.
The distal end of tibiotarsus from a test pit is the size of a
large female M. gallopavo. Steadman (this vol.) finds no di-
agnostic characters in distal ends of tibiotarsi

LUBBOCK LAKE, Lubbock Co., Texas. Llano Estacado
(southern High Plains), north of the Edwards Plateau. Re-
stored fragments of distal end of tibiotarsus (TTU-A1399) and
humerus (TTU-A1391); two pedal phalanges (TTU-A1390,
1443). These several fragmentary elements, from the same ho-
rizon in a Clovis Man level, are referred to a single individual.
They are not identifiable to species. The date is late Pleistocene
(12,000-11,000 B.P.). The associated megafauna includes
Mammuthus sp., Equus sp., Camelops sp., Bison sp., and
Tapirus sp. (Johnson 1977:65).

KLEIN CAVE, Kerr Co., Texas. South-central Texas,
northwest of San Antonio. Distal end of femur (MWU 9110).
The cave deposit is dated at 8000 B.P. (Roth 1972). The lack
of an extinct megafauna is attributed to the shallowness of the
cave (Roth 1972:77). Mammals from the cave that require a
mesic habitat and are no longer found on the Edwards Plateau

214

Rea: Southwestern Turkeys

Table 1. Late Pleistocene and early Holocene paleontological remains from the southwest.

Cave, Location

Species

Element

Provenience

Association

Age

Original

Source

STANTON’S CAVE
Coconino Co., Ariz.

Meleagris

crassipes

tarsometa-

tarsus

pack rat
midden

Teratornis ,
Gymnogyps,
modern fauna

38,000
B.P. to
Recent

Rea and
Hargrave MS

LAGUNA SALADA
Apache Co., Ariz.

Grus canadensis

tibiotarsus

camp site

cultural

3280 ± 60
B.P.

Martin and
Rinaldo 1960

PAPAGO SPGS.
CAVE

Santa Cruz Co., Ariz.

M . crassipes
M. cf.
crassipes
M. crassipes

humerus

humerus

2 femora

none

Pleistocene

megafauna

Late

Pleistocene

Skinner 1942

NORTH PAPAGO
CAVE

Santa Cruz Co., Ariz.

M. crassipes

tarsometa-

tarsus

none

Pleistocene

megafauna

Late

Pleistocene

ARIZPE
Sonora, Mexico

M cf.
gallopavo

humerus

head

none

Bison, Equus

Late

Pleistocene

Cracraft

1968

LA BRISCA
Sonora, Mexico

M. cf.
gallopavo

ulna

surface

Strix brea, Equus

Late

Pleistocene

TULAROSA CAVE
Catron Co., N. Mex.

M. gallopavo

2 tarso-
metatarsi
2 cora-
coids

pre-pot-
tery level

cultural

2300-2150

BP.

Martin
et al. 1952

SAN ANTONIO
SITE

Socorro Co., N. Mex.

M. cf.
crassipes

humerus

ulna

radius

base of

pumicite

bed

none

Early?

Pleistocene

Needham 1936

HOWELL’S
RIDGE CAVE
Grant Co., N. Mex.

M. crassipes
M. crassipes ?
M. crassipes

ulna

scapula

coracoid

none

(Zeller-

Howard

pit)

Equus, Camelops,
Gymnogyps,
Coragyps,
Spizaetos

presumed

late

Pleistocene

Howard 1962

M. crassipes

tarsometa-

tarsus

90-100 cm
level

Centrocercus

urophasianus

3330-6697

B.P.

Van

Devender
and Worth-
ington 1978

SHELTER CAVE
Doha Ana Co., N. Mex.

M. crassipes
M. crassipes

humerus
2 ulnae

none

Gymnogyps,
Pyelorhamphus,
Geococcyx conklingi

Late

Pleistocene
to Late
Archaic
“Basket
Maker”

Howard and
A. Miller

1933

LUBBOCK LAKE
Lubbock Co., Texas

Meleagris sp.

tibio-

tarsus

humerus

fragments

Clovis Man
level

Mammuthus, Tapirus,
Equus, Camelops

12,000-

11,000

B.P.

Johnson
1977, Rea
in press

KLEIN CAVE
Kerr Co. , Texas

M. gallopavo

femur

?

Synaptomys cooperi,
Mustela erminea,
Tamias striatus

8000 B.P.

(noncul-

tural)

Roth 1972,

Feduccia

1972

SAN JOSECITO
CAVERN

Nuevo Leon, Mexico

M. crassipes

>50

elements

below

cultural

zone

Teratornis, Gymno-
gyps, Smilodon,
Tetrameryx, Equus,
N othrotherium

Late

Pleistocene

Miller 1943

INGLESIDE

PIT

San Patricio Co.,
Texas

M. gallopavo

3 tibio-
tarsi
2 tarso-
metatarsi
1 coracoid

Ciconia maltha,

Gopherus

hexagonata,

Geochelone,

Mammuthus,

Mammut, Camelops

Late

Pleistocene

Feduccia

1973,

Steadman
(this vol.),
Lundelius
1972

CONKLING CAVERN
Doha Ana Co., N. Mex.

M . gallopavo
M. gallopavo

radius

humerus

shaft

none
6 to 7 m
level

Coragyps,

Gymnogyps,

Camelops,

Equus

Paleo-
Indian
13,000-
9000 B.P.

Howard and
A. Miller
1933

M . gallopavo

carpometa-

carpus

“dump”

Rea: Southwestern Turkeys

215

Table 1 Continued.

Cave, Location

Species

Element

Provenience

Association

Age

Original

Source

DRY CAVE
Eddy Co. , N. Mex.

M. crassipes

tarsometa-

tarsus

Breagyps, Canis
dints, Camelops,
Tapirus, Equus

25,000

B.P.

Harris 1978

BURNET CAVE
Eddy Co., N. Mex.

M . crassipes
M. crassipes
M . crassipes

tarsometa-

tarsus

carpometa-

carpus

femur

5 ’9"

(1.75 m)

Folsom-like arti-
facts; Bison,
Tetrameryx,
Gymnogyps

Late

Pleistocene

Schultz and
E. Howard
1935, Wetmore
1932

M . gallopavo
M . gallopavo

tibio-

tarsus

humerus

none
(upper
0.5 m?)

presumably
Late Archaic

DARK CANYON CAVE
Eddy Co., N. Mex.

? (lost)

unknown

(2)

unknown

Gymnogyps,

Cora gyps,
Anabernicula,
Neophrontops, Equus

Late

Pleistocene

Howard 1971

M. crassipes

coracoid

4.5 m,
new pocket

Gymnogyps , Equus

Late

Pleistocene

PRATT CAVE MKA-1
Eddy Co., N. Mex.

M. gallopavo
Meleagris sp.

humerus

tibiotarsi

cultural

Geococcyx
conklingi,
and modern fauna,
Late Archaic

232 0-1420
B.P.

Unpublished

report

include Tamias striatus, Microtus pennsylvanicus, Synapto-
mys cooperi, Mustela erminea, Myotis lucifugus, and Myotis
evotis. This indicates a climatic shift from cool, moist condi-
tions to warm, dry conditions on the plateau. Roth (1972)
found no evidence of human habitation in Klein Cave. Fed-
uccia ( 1972) identified the distal end of a femur as M. gallopavo
within the size range of a female. Steadman (this vol.) consid-
ers the depth of the intercondylar fossa to be a diagnostic
character between M. crassipes and M. gallopavo, and we
could not distinguish the Klein Cave specimen from three re-
cent wild female M. gallopavo at hand. This is the westernmost
specimen of M. gallopavo from the late Pleistocene/earlv Ho-
locene, exclusive of Sonora and southwest New Mexico, not
directly associated with man.

INGLESIDE PIT, San Patricio Co., Texas. About 1.5 km
inland from the Gulf of Mexico. Feduccia (1973:143) reported
six bones (all mineralized): coracoid (TMM 30967-1741), tib-
iotarsi (TMM 30967-1139, 30967-1063B, 30967-1564), tarso-
metatarsi (TMM 30967-1169, 30967-1467) as M. gallopavo.
Steadman (this vol.) has re-examined the material and agrees
with Feduccia’s identification. The fauna is late Pleistocene.

In summary, this critical evaluation of turkey remains from
the Southwest (Table 1; Fig. 3) shows that M . crassipes was
widespread from Nuevo Leon, Mexico, to the Grand Canyon,
Arizona, in the late Pleistocene. In most caves it is associated
with other extinct birds and Rancholabrean land mammals.
The earliest radiocarbon date for M. crassipes is 25,000 B.P.
(Dry Cave, New Mexico), and it persisted at least until some
time between 3300 and 6600 B.P. (Howell’s Ridge Cave, New
Mexico). Its presence in greater numbers in southern New
Mexico reflects the large number of cave deposits with bones
of smaller vertebrates in that area. M. crassipes does not ap-
pear temporally sympatric with M. gallopavo in any deposit.

So far there is no direct evidence that early man played a role
in the extinction of this medium-sized turkey.

The late Pleistocene to early Holocene turkey specimens re-
ferred to M. gallopavo are restricted to southern New Mexico
(Conkling Cavern), northern Sonora (Arizpe, La Brisca), the
south-central and Gulf portions of Texas (Klein Cave and In-
gleside Pit), and points east (Steadman this vol.). Throughout
the remainder of the Southwest, M. gallopavo occurs only after
Paleo-Indian times in association with the remains of sedentary
agriculturalists.

There are three hypotheses that might explain the unusual
distributions of the two turkey species from the Southwest:

1. M. crassipes and M. gallopavo occurred sympatrically in
the Southwest from the late Pleistocene onward, but only M.
crassipes was taken into caves by predators; M. crassipes be-
came extinct, but M. gallopavo persisted. I consider differen-
tial predation unlikely. The male of M. crassipes was as large
as a female M. gallopavo. If both species were present there
should have been an equal chance of either species being de-
posited in caves by predators.

2. M. crassipes was the only turkey present in the South-
west. When it became extinct M. gallopavo extended its range
from the east or the south to fill the vacated area. There is
nothing intrinsically wrong with this hypothesis and it cannot
be proven or disproven. However, climatic factors would seem
to discredit the idea of a natural invasion by this large and
relatively sedentary species. During the Holocene the South-
west underwent a general trend toward a warmer, dryer cli-
mate, with woodlands giving way to grasslands and deserts
and xeric-adapted plants from Mexico migrating northward at
about the time M. gallopavo would have been invading (Van
Devender 1976, 1977; Van Devender and Wiseman 1977;
Wells 1966; Wells and Berger 1967). Forests became more and
more restricted to mountain islands in the Southwest and the

216

Rea: Southwestern Turkeys

juniper-pinon-oak woodlands retreated upslope 260 to 1000
meters (Van Devender and Spaulding 1979). The modern
species of turkey would have been invading as its preferred
habitat of woodlands was shrinking.

3. A third hypothesis is that (a) only M. crassipes occupied
the niche of a large phasianid bird during the late Pleistocene,
but became extinct during the Holocene; (b) man imported M.
gallopavo, along with other domesticated plants and animals;
(c) these turkeys escaped, forming a feral population north of
M. g. mexicana and west of M. g. intermedia. Any serious
consideration of this hypothesis requires a look at the evolution
of sedentary cultures in the Southwest, the development of
Mesoamerican cultivars, the development of trade routes be-
tween Mesoamerica and the Southwest, the transmission of
cultural items, and finally, the present distribution and char-
acteristics of the various subspecies of M. gallopavo.

THE EVOLUTION OF SEDENTARY
CULTURES IN THE SOUTHWEST

Paleo-Indians and Desert Culture

The waves of early man sweeping across the North Amer-
ican continent from Asia are called Paleo-Indians. These cul-
tures are usually characterized by their distinctive lithic arti-
facts; Clovis and Folsom people are well-known examples.
Paleo-Indians are usually characterized as megafauna (big
game) hunters who exploited many now extinct mammal
species, beginning sometime between 15,000 and 10,000 B.P.
They also hunted smaller game (Jennings 1974:90-92; Johnson
1977). The somewhat younger Desert Culture of the Archaic
stage not only included the Southwest and Great Basin
(McGregor 1965: 124-125) but extended from southeast Oregon
to the Valley of Mexico (Martin and Plog 1973:69). The Co-
chise Complex, a part of this widespread Desert Culture, oc-
cupied much of the arid Southwest, lasting in places as late as
2000 B.P. Regardless of the names and times, these cultures,
until at least late Cochise times (Dick 1965:100; Martin and
Schoenwetter 1960:33-34), were pre-ceramic and pre-agricul-
tural, dependent entirely on hunting and gathering (especially
seeds) for subsistence.

Anasazi Culture (Colorado Plateau)

Some knowledge of three subsequent cultures, the Anasazi,
Mogollon, and the Hohokam (Fig. 2) are critical to under-
standing the pre-history of M. gallopavo in the Southwest. The
Archaic phenomenon was a general cultural stage that spread
across North America from 8-9000 to 2-4000 B.P. It was best
developed in the Southwest (or at least is best known there
because of the circumstances of preservation in this arid re-
gion) (Amsden 1949; A. Morris 1933:39-55; McGregor
1965:170-186; 206-217; Wormington 1978:27-57; Jennings
1974:134; Martin and Plog 1973:81). The late Archaic San
Juan peoples from the Anasazi or Four Corners region (San
Juan and Little Colorado drainages) developed through several
cultural stages (Basket Maker II and III; Pueblo I through IV),
ending in the post-conquest pueblos of today such as Hopi,
Acoma, and Zuni. (The cultural stage Basket Maker I is a
hypothetical proto-agricultural stage that has not been discov-
ered.)

Basket Maker II people, dating approximately A.D. 300-

500 (McGregor 1965:471) are characterized by finely woven
sandals and other textiles, atlatls for hunting, and corn/squash
agriculture. The turkey appears in deposits from this time pe-
riod only as feathers or feathered artifacts (Hargrave 1970a;
Emslie and Hargrave 1978), not as bones. Toward the end of
Basket Maker III time (McGregor (1965:215] puts this around
A.D. 600-700), the turkey was certainly domesticated in the
regions occupied by the Anasazi and evidence of whole turkey
skeletons (almost never individual elements) appears (A. Mor-
ris 1933:196-197; E. Morris 1939:120). At approximately the
same time, Basket Maker III culture underwent three critical
advances, one technical (development of fired pottery), and
two agricultural (introduction of beans [ Phaseolus sp ] and the
appearance of newer and larger varieties of corn [ Zea mays]).
The bow and arrow appeared about this time (Amsden
1949:133; Wormington 1978:55) or somewhat earlier.

Early Basket Makers used mammal fur cordage for twining
blankets and for clothing. Later in the culture, feathers, par-
ticularly from turkeys, were used for this purpose (Hough
1914:5-6, 71-73; Wormington 1978:55, 89; A. Morris
1933:197; Amsden 1949; McGregor 1965:181, 215; Lange
1950.) Amsden pointed out that the fur-cord robes were as-
tonishingly heavy but that turkey-feather material had the ad-
vantage of lightness. At this time turkey feathers were also
starting to be used for ceremonial purposes. The intensive use
of turkey feathers continues in the complex religious ceremo-
nies of today’s Pueblo Indians.

Pueblo II and III peoples had their greatest expansion be-
tween A.D. 900 and 1100 (see map, McGregor 1965:279).
There is evidence (turkey bone awls and individual elements,
occasionally charred, disposed of in trash mounds) that, be-
ginning in this period, the turkey was used for food, at least
locally. But at earlier sites of these peoples only whole birds
occur as carefully interred burials (A. Morris 1933: 196-197;
Reed 195 1:197, 200-201; McKusick this vol.). Stiger
(1979:140) hypothesizes the use of turkeys for grasshopper con-
trol by Pueblo III times.

Compared to the adjacent Mogollon Culture, the Anasazi
were late in overall cultural development. In the acquisition
of genetic materials whose ultimate origin was Mesoamerica,
they were centuries behind (see discussion beyond and Table
2).

An exceptionally early Anasazi turkey (M. gallopavo) is a
mummy from Canyon del Muerto in Canyon de Chelly, es-
timated to date to A.D. 250 (Schorger 1961, 1966:20, 1970).
It is from a race unknown in the wild. McKusick (pers. comm.)
believes the dating is correct.

Turkey bones are found in late Archaic Basket-Maker-like
deposits from southern New Mexico (Burnet Cave, Pratt
Cave, and perhaps the dump material from Conkling Cavern).
Presumably these people had a good agricultural base and
were capable of maintaining captive turkeys, but little ethno-
biological material was salvaged from these earlier excava-
tions.

Mogollon Culture (Mountains)

Evidence for the domestication and use of turkeys in the
mountainous areas of the Southwest occupied by the Mogo-
llon peoples (Fig. 2) is equivocal. Reed (1951:202) states cat-
egorically that “the turkey was certainly not domesticated or

Rea: Southwestern Turkeys

zi7

Figure 2. Generalized southwestern cultural areas, ca. A.D. 700-1100. (Modified from various sources.) Four major cultural areas are shown.
The Mimbres is a late local development of Mogollon Culture. The Sinagua sub-area resulted from influences of Anasazi, Hohokam, and Mogollon.
Also late was a Puebloid influenced sub-culture (not shown) along the Hohokam-Mogollon interface, often considered a distinct culture, the Salado.

kept by the Mogollon Pueblo groups of the forested uplands.”
(Such statements rely heavily on the presumption that wild
turkeys were available; e.g., Haurv 1936:93.) However, tur-
keys clearly were being raised in Tularosa Cave where Hough
(1914:5) reported finding desiccated chicks, eggs, and great
quantities of droppings. Schorger (1961) reported finding the
crop of an adult mummy filled with a variety of colored flint

corn grown by the Mogollon people between A.D. 500 and 700.
Martin et al. (1952:499) noted that turkeys appear in the Mo-
gollon record (Pine Lawn Phase, 2150 B.P. to A.D. 500) sev-
eral centuries earlier than in the Anasazi area.

The oldest evidence for a primitive maize north of Mexico
is from Bat Cave, Catron Co., New Mexico, at levels even
older than the Mogollon Culture (Dick 1965:100). This is a

218

Rea: Southwestern Turkeys

Figure 3. Localities where fossil to early cultural Meleagris bones have been reported from the Southwest. See Steadman (this vol.) for details
on San Josecito Cavern and Ingleside Pit. 1, Stanton’s Cave. 2, Laguna Salada. 3, North Papago Cave. 4, Papago Springs Cave. 5, La Brisca.
6, Arizpe. 7, Tularosa Cave. 8, Howell’s Ridge Cave. 9, San Antonio Site. 10, Shelter Cave. 11, Conkling Cavern. 12, Burnet Cave. 13, Dry
Cave. 14, Dark Canyon Cave. IS, Pratt Cave. 16, Lubbock Lake. 17, Klein Cave 18, San Josecito Cavern. 19, Ingleside Pit.

pod popcorn brought into the Southwest as a cultivated va-
riety, presumably from the south at least by 5500 B.P. (Dick
1965:93; Mangelsdorf 1954:409). There is a continuous and
copious record of corn and its varietal progression in Bat Cave.
Abundant vertebrate bones were recovered, but the bird and
smaller mammal bones are still unanalyzed (Dick pers.
comm.). In nearby Tularosa Cave, a primitive pod corn ap-
pears around 2000 B.P. Pottery and turkey feathers also occur
about this time, their earliest occurrence in the Southwest.
Some caves produced large numbers of turkey bones (Martin
and Rinaldo 1950:492; Martin et al. 1952:204, 1954:155), oth-

ers few or none. The presence of very few turkey bones from
Mimbres sites (sometimes considered a separate culture, A.D.
700-1150) has been interpreted as absence of the bird as a
domesticate. However, Mimbres people were familiar with the
bird and painted it realistically on pottery.

Hohokam (Sonoran Desert)

Paralleling in time the early Basket Maker developments to
the north was the Hohokam Culture in the Lower Sonoran
Desert regions of southern Arizona. Like the Mogollon Cul-

Rea: Southwestern Turkeys

219

Table 2. Comparison of subspecific characters of male Common Turkey, Meleagris gallopavo.'

Subspecies

Rectrix Tips

Upper Tail Coverts

Rump

Size2

M. g. merriami

buff or grayish white

buff to white

bluish-black

large

M. g. gallopavo

white

pale pinkish
buff to white

blue-black

medium

M. g. mexicana

white

pinkish white

coppery and greenish

large

Eastern races3

cinnamon to chestnut

dark chestnut

glossy black (intermedia)

variable (large
to small)

1 After Schorger (1966) and Ridgway and Friedmann (1946).

2 As determined by wing measurements given in Ridgway and Friedmann (1946); there is relatively little size difference between the tarsometatarsi
of these subspecies (means of males range from 162 to 173.8 mm). Sample sizes are small, ranging from 1 to 11. Osteological measurements
should be more indicative of actual body size of subspecies.

3 M. g. sylvestris, M. g. osceola, and M. g. intermedia. These three subspecies are richly colored, dark-rumped races.

ture, the Hohokam had corn agriculture and ceramics by at
least 2000 B.P. (Ventana Cave, Haurv 1950:164-165; Snake-
town, Haury 1976:117-118; Cutler and Blake 1976:365-366;
Bohrer 1970), as well as an elaborate water control technology.

Turkey bones are virtually or completely absent from Ho-
hokam sites. Several reputed bones require verification, but
Haury (pers. comm.) has been unable to relocate the remains
reported from the first Snaketown excavation (Haury
1937:156). The second, more thorough excavation produced
no turkey bones (McKusick 1976). A supposed turkey bone
necklace (Arizona State Museum, Carpenter 1977) from a San
Pedro site proves to be lagomorph long bones (pers. obs.).
I have identified a coracoid of a female-sized Small Indian
Domestic from the Las Colinas Site (Phoenix, Arizona) that
is too small and gracile for either a wild M. g. merriami or an
intrusive barnyard female. However, it dates from the late
period of Salado-Publoid intrusion, A.D. 1180-1450. I know
of no good evidence of any turkey element from a pure Ho-
hokam horizon (i.e. , one without Salado cultural influence).

ORIGINS AND DEVELOPMENT OF
MESOAMERICAN CULTIVARS

At least five Mesoamerican domesticates or semi-domesti-
cates are known to have entered southwestern cultures by dif-
fusion or direct trade. These are corn, squash of several
species, gourds, beans of several species, cotton, and macaws.
Another animal domesticate, the dog, Canis familiaris, is more
ancient, and probably of Old World origin. It is known from
late Archaic times (4500 B.P. at Ventana Cave, Haury
1950: 158; beginning of Snaketown, Haury 1976: 1 15, 120; from
the beginning of Basket Maker II, Amsden 1949:62-65; Wor-
mington 1978:46-47).

CORN. Various hypotheses have been advanced to
explain the origin of this unique cereal. The present best ar-
chaeological evidence from the Tehuacan Valley, Puebla,
Mexico, indicates a succession from wild pod corn starting ca.
7200 B.P. (Mangelsdorf, MacNeish, and Galinat 1964:541-
543). It is believed that corn diffused from Mesoamerica to the
north as a pod corn, with subsequent infusions of genetic traits
(Mangelsdorf 1950, 1954, 1974; Mangelsdorf and Smith
1949:2 13-247; Dick 1965:92-98). The oldest southwestern
corn is from Bat Cave (late Cochise Culture dated 5500 B.P.).
Both the Hohokam and Basket Maker peoples had corn from
their beginnings.

SQUASH. Squash species (Cucurbita spp.) are thought to be

as ancient as maize, again being derived from the south (Whit-
aker and Bemis 1975). Squashes were domesticated in Me-
soamerica at least 9000 B.P. (Cutler and Whitaker 1961). The
earliest domesticated species in the Southwest, C. pepo, is
known from Cordova and Tularosa Caves and at Bat Cave
(5500 B.P.) in the Mogollon region (Martin et al. 1952; Dick
1965). The earliest verifiable remains of C . mixta and C. mos-
chata in the southwest are from Pueblo II times (A.D. 900-
1050), many centuries after the first appearance of C. pepo.

BOTTLE GOURD. The one species of bottle gourd,
Lagenaria siceraria, is common to the Old and New World
(Cutler and Whitaker 1961). It appears in the Mexican ar-
chaeological record at 9000 B.P. (Tamaulipas, MacNeish
1958). The bottle gourd appears in remains dated at around
2300 B.P. in Tularosa and Cordova Caves in the Mogollon
area (Martin et al. 1952:475). Verifiable remains from the An-
asazi area are from A.D. 608-683 and 610 (Cutler and Whit-
aker 1961). The Hohokam had the gourd by the Sacaton Phase
(A.D. 900-1100; Haury 1976:183).

BEANS. Two species of beans were particularly important
in Southwestern cultures. The oldest known, a kidney bean
(one variety of Phaseolus vulgaris), appears in deposits dated
between 3000 and 2500 B.P. in Bat Cave (Dick 1965:98-99).
In Tularosa Cave it appears in remains from 2300 B.P., along
with maize and pepo squash (Kaplan 1956:2 18). The Hohokam
had the kidney bean before 2000 B.P. (Haury 1976:118, 346;
Bohrer 1970:425), but the Anasazi probably acquired it later
in Basket Maker III times (Amsden 1949:132; Wormington
1978:55). It appears in the archaeological record in Tamaulipas
and the Tehuacan Valley of Mexico 5000 to 7000 B.P. (Kaplan
1967).

A second bean species important in Southwestern cultures
is the tepary (tepari) ( Phaseolus acutifolius). The tepary ap-
pears among the Hohokam remains in the Sacaton Phase of
Snaketown (A.D. 900-1100; Haury 1976:118, 338; Kaplan
1956:2 19) and in deposits near Tucson dating from about A.D.
900-1200 (Bohrer, Cutler, and Sauer 1969:4-5). Carlson (1963)
records teparies from a Basket Maker III site dated ca. A.D.
700. The tepary was apparently absent from Bat Cave (Smith
1950; Dick 1965:98-99) and Tularosa and Cordova Caves
(Martin et al. 1952:474-475). Kaplan (1956:218) suggests the
tepary was introduced to the Mogollon area after A.D. 1 100.
MacNeish (1964:534) recorded it as a domesticate in the Cox-
catlan Phase of the Tehuacan Valley, Puebla, between 7200
and 5400 B.P. (see also Kaplan 1967:208-210).

220

Rea: Southwestern Turkeys

Two other species of beans, the lima, Phaseolus lunatus,
and the jack bean, Canavalia ensiformis , were grown in the
Southwest (Kaplan 1956). The lima appears in late deposits
(A.D. 1200-1400 in the Verde Valley, Arizona). Canavalia, a
tropical derivative, appears in Hohokam, Anasazi, and Salado
ruins dated about a century later (Sauer and Kaplan 1969).

COTTON. Cotton, Gossypium sp., was a cultivar in the
Tehuacan Valley, Puebla, as early as 5400 to 4300 B.P.
(MacNeish 1964:536). The Hohokam probably arrived with
this crop (Haury 1976:1 18, 346; Bohrer 1970:425). Some cotton
fiber is present in Tularosa Cave (Martin et al. 1952:475).
Cotton cloth made its initial appearance here at the beginning
of the San Francisco Phase (ca. A.D. 700), as did the bow and
arrow. McGregor (1965:246) and Wormington (1978:69-70)
place the introduction of cotton to the Anasazi region during
Pueblo I period (A.D. 700-900). This cultivar is a Meso-
american derivative (see Kent 1957 for a discussion of origins).

MACAW. The Scarlet Macaw, Ara macao, and the Military
Macaw, A. militaris, were kept in captivity by early South-
western peoples. The Scarlet Macaw was an important trade
item to all the Southwestern cultures beginning about A.D.
1100 (Hargrave 1970b). Of the two, the more southerly and
the more brightly colored Scarlet Macaw accounts for most of
the identifiable records. In southern Utah, the northern pe-
riphery of Anasazi culture, only macaw feathers have been
recovered from archaeological sites (Hargrave 1970b:29; Em-
slie and Hargrave 1978; Hargrave in press).

Di Peso ( 1 974c: 2 72—2 73) discusses use and trade in Mesoam-
erica. In northern Chihuahua, 322 Scarlet Macaw skeletons
were recovered from Casas Grandes ruins. The presence of all
age stages (including eggs, nestlings, and juvenals) as well as
adobe breeding pens indicates that this city was an important
breeding center for the late macaw trade to the north (Di Peso
1974b: 182-185; Di Peso 1974c:267, 269, 272-273; McKusick

1974) . Macaws were a source of feathers for ceremonial pur-
poses. Lifelike macaws, some eating corn, are depicted in late
pueblo kiva (ceremonial room) murals (Smith 1952; Hibben

1975) . The descendants of the Anasazi still use macaw and
other parrot feathers ceremonially.

CULTURAL EXCHANGE BETWEEN
MESOAMERICA AND THE SOUTHWEST

The advanced civilizations of Mexico and Central America
profoundly modified the southwestern cultures through time.
Their influence included not only the exchange of ideas, raw
materials, and manufactured items, but also the direct trans-
mission of cultivars (corn, squash, gourds, beans, and cotton)
and live birds (macaws). Although this exchange extended over
two millenia, dynamic periods of especially strong influence
can be detected. Early in the proto-agricultural and early ag-
ricultural record there were considerable temporal differences
as to when elements were acquired by various recipient cul-
tures (contrast Mogollon with Anasazi, for instance). But by
A.D. 700 a definable constellation of derivatives, mostly Me-
soamerican, arrived cross-culturally in the Southwest (Table
2). Many of these elements represent genetic modifications of
crops from the advanced cultures to the south.

Di Peso (1974a: 104) perceives A.D. 700 ± 50 as a period of
great transition between Mesoamerica and the northern fron-
tier: “Something occurred which stirred some of the northern

frontiersmen. Perhaps it was a motivation which emanated
from the great cities located south of the Tropic of Cancer.
Here the famed Teotihuacan culture of the Mesa Central had
just come to a disastrous end. ... In the Tehuacan Valley,
it was the time of the Venta Salada Phase, when full-time
agriculturalists irrigated their fields and lived in large com-
munities associated with separate ceremonial cities. It is
thought that in certain areas south of the Tropic of Cancer,
the population increased 5000-fold over the original number.
Many of these people were engaged not only in agricultural
pursuits but in commerce, salt-making, cotton processing, and
other industries which raised their living standards.”

Di Peso’s Puchteca (merchant) class may have arisen at this
time. For earlier periods, actual routes are less well known.
At any rate, in the centuries following A.D. 700, an enormous
network in trade in turquoise and other minerals, ceramics,
birds, feathers, hides, textiles, shell, and slaves developed be-
tween the city of Casas Grandes in Chihuahua and the sur-
rounding northern frontier, extending throughout Hohokam,
Mogollon, and Anasazi country (Di Peso 1974b: note especially
pp. 129, 144, 171, and 193).

As with the various cultivars, turkeys, a small breed of M.
gallopavo, were present in the Mogollon area by 2000 B.P.
(Tularosa Cave). There were several breeds imported (Mc-
Kusick this vol.), most likely from several source areas. About
A.D. 700, whole turkeys, not just feathers, appear in the
southern Anasazi archaeological record. Complete turkeys did
not arrive in southern Utah until Pueblo II time, around A.D.
900 (Emslie and Hargrave 1978). The turkey spread among
agricultural people (except the Hohokam) raising three subsis-
tence crops (corn, squash, and beans). The addition of the
third crop (beans) apparently gave a sufficient caloric base for
maintaining turkeys as a domestic animal, not generally used
as food. Another cultigen (or perhaps only a cultivar), the
sunflower (Helianthus spp ), may have been an additional
source of turkey feed.

From around A.D. 1350 to 1450, the entire Southwest
underwent a period of population decline and areal contraction
known to archaeologists as “The Great Abandonment.” Al-
though the decline was most conspicuous in the Anasazi on
the Colorado Plateau because of the great number of large
masonry sites, the Hohokam also disappeared at this time, and
the Mogollon/Mimbres Culture a little earlier (Haury
1976:351-357; Wormington 1978:107, 144, 161, 166; Mc-
Gregor 1965:420-421, 426, 428, 433). Thus, the demographic
sequence consisted of a slow (±1000 years) population build-
up and expansion, followed by a rapid decline. Though inten-
sively studied and debated by archaeologists, ecologists, den-
droclimatologists, geologists, and palynologists, no simplistic
solutions satisfactorily explain the causality of the Great Aban-
donment. Perhaps the combined demographic/ecologic model
of Martin and Plog (1974:318-333) comes closest.

Regardless of causal factors, the results were the same:
Pueblo and puebloid-influenced population centers almost
completely disappeared. These included the villages with their
great herds of turkeys (McKusick this vol.). Flocks were aban-
doned to fend for themselves. I propose that at this time, if
not before, the turkey became feral locally throughout what
is the modern range of the subspecies M. g. merriami (Fig. 1).
It filled the niche in suitable habitats left vacant by the Pleis-
tocene M. crassipes north of the range of the Sierra Madre

Rea: Southwestern Turkeys

221

Occidental subspecies, M. g. mexicana. Meleagris g. mexicana
probably occurred as far north as the border ranges of south-
west New Mexico and southeast Arizona, below the Mo-
gollon Rim (Aldrich and Duvall 1955; Phillips et al. 1964).
Such a distributional pattern is shared by a number of other
essentially Mexican montane or encinal vertebrates, including
Crotalus lepidus, C. pricei, C. willardi, Otus trichopsis, Peu-
cedramus taeniatus, Junco phaeonotus , Nasua narica, and
Mephitis macroura. Native M. g. mexicana has been extir-
pated from these border ranges (Chiricahuas, Huachucas, Ba-
boquivaris, Santa Catalinas, Santa Ritas, Pelonciltos, and
probably the San Luis Mountains). Some of these areas have
been restocked with M. g. merriami. Only a few specimens of
M. g. mexicanus were collected and preserved prior to this
restocking, and these records suggest the former northern limit
of the Sierra Madrean subspecies.

MELEAGRIS G ALLOP A VO MERRIAMI:

A FERAL POPULATION?

The presumption has been that the turkey, M. g. merriami,
occurred wild throughout suitable parts of the Southwest and
that it was taken captive by the Basket Maker peoples and
eventually domesticated. I suggest exactly the reverse of the
above assumption. First, that Mesoamerican turkey feathers
were brought to the Anasazi area by trade. Next, the live bird
was imported as a domesticate, and later it became feral
throughout the range of M. g. merriami. S. Emslie, C.
McKusick, and B. Wright are of the opinion that turkeys es-
caped from domestication quite early, long before the Great
Abandonment. The details of when the various domestic
breeds and the wild form appear in the different cultural areas
are discussed by McKusick (this vol.).

Some areas of the Southwest — parts of Utah, and the North
Kaibab Plateau and the Hualapai Mountains of Arizona — lack
native populations of M. g. merriami yet have suitable conifer
or pine-oak habitat to support turkeys. Some of these areas
(Fig. 3) have been successfully stocked in recent decades
(Schorger 1966:438-439, 459). The historical absence of tur-
keys in these ranges tends to support the feral turkey hypoth-
esis. M. g. meiriami historically occurred in habitats where
aboriginal peoples carried on a turkey industry or in imme-
diately contiguous pine-oak habitats. Parmalee (this vol.)
found no turkey bones among the numerous Galliformes re-
covered from 16 archaeological sites (5 Archaic and 11 Fre-
mont) in Utah. Sinaguan, Patayan, Virgin River Anasazi, and
Fremont peoples were not turkey raisers (McKusick this vol.).
The disjunct forests in these cultural areas historically lacked
turkeys. But other, similarly discontinuous habitats within the
turkey-raising Anasazi area (e.g., the Lukachukai and Chuska
Mountains) did host native populations.

A number of plants have become locally feral, self-main-
taining populations in the arid West, after the abandonment
of Anglo-European mining or ranching sites: various mus-
tards, Brassica spp.; horehound, Marrubium vulgare ; iris, Iris
sp.; tree of heaven, Ailanthus altissima ; to mention a few.
Some southerners brought the opossum, Didelpltis v. virgini-
anus, to California early in the twentieth century and it rapidly
spread throughout suitable parts of the state (Grinned, Dixon,
and Linsdale 1937). The Old World honey bee, Apis mellifera,
is now widespread in the feral state. The escaped burro, Equns

asinus, maintains stable populations on the Lower Colorado
River drainage, partially filling a niche vacated by several
Pleistocene species of horses, Equus spp. These examples dem-
onstrate how easily feral populations may be established.

VARIATION IN EARLY CULTURAL
TURKEYS

The early M. gallopavo that appeared in the Southwest were
hardly uniform in characters. Schorger (1961) reported a rel-
atively small mummified adult male turkey from Tularosa
Cave, implying that it was a captive bird. Its distinguishing
characteristic was a neck feathered to the base of the skull.
Later Schorger (1970) formally described this anomaly as M.
g. tularosa, based on two specimens from different localities.
The original site description (Hough 1914:5-6) leaves little
doubt that it was a domestic form: “A desiccated adult bird,
parts of other individuals, desiccated chicks, and a number of
eggs were found in a portion of the cave which was evidently
a pen where turkeys were kept in captivity, there being great
quantities of the droppings of the birds in the debris.” Mc-
Kusick (this vol.) finds two basic size varieties in prehistoric
sites from the Southwest. Her Small Indian Domestic is the
same as Schorger’s (1970) M. g. tularosa “subspecies,” and her
Large Indian Domestic corresponds to present-day wild M. g.
merriami.

Hargrave (1970a) examined feathers of a small brown-toned
turkey in Sand Dune Cave (Basket Maker II, A.D. 700 or
earlier), and gave these the formal name M. g. coltoni. All the
feathers appeared to be from a single individual. According to
McKusick (pers. comm.) these are juvenal-plumaged. Since
normal black and white feathers occurred in the site, Hargrave
reasoned that his new “subspecies” was not the result of post-
mortem color changes (“foxing”). Emslie and Hargrave (1978)
reported additional “M. g. coltoni" feathers from Westwater
Ruin, San Juan Co., southeastern Utah (Basket Maker III/
Pueblo I, ca. A.D. 700). I do not advocate formal nomencla-
tural recognition of the individual strains “ coltoni ” and "tu-
larosa" because they were undoubtedly domestic birds.

CHARACTERISTICS OF THE
SUBSPECIES OF COMMON TURKEY

If the turkeys that now make up the subspecies M . g. mer-
riami did not evolve locally, then what was their source? There
are two possibilities: from the east or from the south. The
eastern woodland cultures had a subsistence base of maize,
squash, and beans, but did not domesticate the turkey; they
hunted it in the wild (Schorger 1966:137). In the eastern part
of the United States the turkey has a long continuous fossil
record, extending back to the Miocene (Steadman this vol ).
The eastern races (M. g. sylvestris, M. g. intermedia, and M.
g. osceola) do not have subspecific characters that would sug-
gest that M. g. merriami was derived from them (Table 2).
The eastern races are strongly refuscent, intermediate to small
in size, and lack the whitish-tipped lower rump. The races to
the south appear to me to be better candidates. M. g. merriami
may have been derived from the Large Indian Domestic breed,
and this, in turn, from M. g. gallopavo and/or M. g. mexicana.
The southernmost M. g. gallopavo is colored almost like M.
g. merriami on the rump, tail coverts, and the tips of the

222

Rea: Southwestern Turkeys

rectrix, but it is considerably smaller. M. g. mexicana, which
ranges geographically between M. g. gallopavo and M. g. mer-
riami, is the same size as M. g. merriami. The bimodal size
variation within each sex of puebloid turkeys (McKusick this
vol.) suggests at least two parental stocks. Schorger’s obser-
vation (1970:170) that his “M. g. tularosa" had little white in
the wings and no apparent white in the rump suggest a non-
Mexican source for the Small Indian Domestic breed. Its col-
oring, as well as its small size and odd neck feathering, do not
match any wild population living today. Modern M. g. mer-
riami is apparently homogeneous throughout its extensive
range, with no evidence of “M. g. tularosa ” influence. With
at least five centuries of selection in the wild, loss of the great
variability would be expected. McKusick (pers. comm.) be-
lieves it is futile to speculate on the origin of races since the
Common Turkey has been domesticated and transported in
Mesoamerica for over 4000 years.

Little is known of archaeological turkey distribution and
domestication in Mexico. Flannery (1967) found turkey in re-
mains from about A.D. 180 (Palo Blanco phase) in the Te-
huacan Valley. This is south of the range of wild M. gallo-
pavo, in habitat that was then, as now, highly xeric (cactus-
thornscrub desert). Both the domestic turkey and the dog
were eaten with increasing frequency in the Tehuacan area
until the time of conquest. In the Valley of Oaxaca, Flannery
(pers. comm, to Hargrave) found that domestic M. gallopavo ap-
peared around 2400 B.P. These are still farther from the wild
range. Apparently agricultural peoples took this domesticate
both to the north and to the south. Archaeological turkey re-
mains dating A.D. 700-1300 have been found on the coastal
lowlands of northwestern Mexico, below the known range of
M. gallopavo (Stuart Scott and Elizabeth Wing pers. comm.).
This is further evidence for importation and domestication.

CONCLUSIONS

Two hypotheses are suggested by the present osteological
data. The first is that only M. crassipes was present in the
Southwest (north of the Sierra Madrean outliers) during the
late Pleistocene and the early Holocene. The available data
thus far support this idea. No fossil M. gallopavo are known
from within the modern range of M. g. merriami. Meleagris
crassipes and M. gallopavo have not been found together. Nor
hasM. crassipes been found associated with proto-agricultural
or agricultural peoples (Archaic stage or later).

The second hypothesis is that at least two stocks of M. gal-
lopavo were imported into the Southwest from the south, the
east, or both, by peoples with a multi-crop subsistence base
after the extinction ofM. crassipes. The spread ofM. gallopavo
appears to be directional, following the diffusion of Meso-
american cultigens. Turkeys from late Archaic deposits in
southern New Mexico are undated. The oldest dated turkeys
are from the Mogollon cultural area. Trade in turkey feathers
preceded the trade in whole birds. After several centuries tur-
keys appeared in the southern Anasazi area, then in the north-
ern Anasazi regions. Peripheral cultural areas to the west have
little or no evidence of turkey. Although the Fremont culture
had corn agriculture and apparently suitable habitat for tur-
key, the bird itself was probably not taken that far north.
However, lacking any barriers, feral turkeys may have spread
from the Mesa Verde area of southwestern Colorado through-

out much of that state, resulting in the present range of M.
gallopavo.

A third hypothesis, that domestic strains of M. gallopavo
preceded the wild form, M . g. merriami, in each cultural area,
is a part of McKusick’s study (this vol.). The idea of the im-
portation of Common Turkey stocks by Indians comes from
Hargrave (1970) as well as Schorger (1961, 1966, 1970).

ACKNOWLEDGMENTS

Many people contributed ideas that became incorporated
into this study. Credit must first go to the late ethnobiologist,
Lyndon Lane Hargrave, who suggested a critical look at fossil
turkeys. Storrs Olson and David Steadman both turned over
to me their data on M. crassipes from a number of south-
western sites, and Steadman assisted with many of the iden-
tifications. Charmion McKusick assisted with the cultural as-
pects from her years of study of this species in archaeological
contexts. Both Steadman and McKusick offered extensive
comments on earlier drafts of this paper. Steve Emslie and
Barton Wright read the paper and shared their expertise in
southwestern archaeology.

I thank all the curators who loaned paleontological and ar-
chaeological remains of turkey. Evelyn Burnett and Deanne
Demere typed drafts of the manuscript. Eric Lichtwardt pre-
pared the maps. Portions of this study were supported by Na-
tional Science Foundation Grant NSF-BNS-77-08S8S.

LITERATURE CITED

Aldrich, J.W., and A.J. Duvall. 1955. Distribution of
American gallinaceous game birds. Fish and Wildlife Ser-
vice circular No. 34:1-30.

American Ornithologists’ Union. 1957. Check-list of
North American birds. Fifth ed. American Ornithologists’
Union. Baltimore, Maryland. 691 pp.

Amsden, C.A. 1949. Prehistoric southwesterners from Bas-
ketmaker to Pueblo. Southwest Museum, Los Angeles,
California. 163 pp.

Bailey, A.M., and R.J. Niedrach. 1965. The birds of Col-
orado. Denver Museum of Natural History, Denver, Col-
orado. Vol. 1, 454 pp.

Bailey, F.M. (with W.W. Cooke). 1929. Birds of New
Mexico. New Mexico Department of Game and Fish. 807

pp.

Bohrer, V.L. 1970. Ethnobotanical aspects of Snaketown,
A Hohokam village in southern Arizona. Amer. Antiquity
35:413-430.

Bohrer, V.L., H.C. Cutler, and J.D. Sauer. 1969. Car-
bonized plant remains from two Hohokam sites, Arizona
BB:13:41 and Arizona BB:13:50. Kiva 35:1-10.
Brodkorb, P. 1964a. Catalogue of fossil birds: Part 2 (An-
seriformes through Galliformes). Bull. Florida State
Mus., Biol. Sci. 8:195-335.

. 1964b. Notes on fossil turkeys. Quart. J. Florida

Acad. Sci. 27:223-229.

Bryan, W.A. 1929. The recent bone-cavern find at Bishop’s
Cap, New Mexico. Science 70:39-41.

Carlson, R.L. 1963. Basketmaker III sites near Durango,
Colorado. The Earl Morris Papers, No. 1, Univ. of Col-
orado Studies, Series in Anthropology, No. 8:1-82.

Rea: Southwestern Turkeys

223

Carpenter, A.H. 1977. A prehistoric shell and bone neck-
lace exposed in a bank of the San Pedro River, Arizona.
Kiva 43: 19-25.

Cracraft, J. 1968. First record of Turkey Meleagris gallo-
pavo from the Pleistocene of Mexico. Condor 70:274.

Cutler, H.C., and L.W. Blake. 1976. Corn from Snake-
town. Pp. 365-366 [Appendix) in The Hohokam: desert
farmers and craftsmen: excavations at Snaketown, 1964-
1965 (E.W. Haury, Ed.). Univ. Arizona Press, Tucson.
412 pp.

Cutler, H.C., and T.W. Whitaker. 1961. History and
distribution of the cultivated cucurbits in the Americas.
Amer. Antiquity 26:469-485.

Dick, H.W. 1965. Bat Cave. School of American Research
Monograph No. 27. 114 pp.

Di Peso, C.C. 1974a. Casas Grandes: a fallen trading center
of the Gran Chichimeca. Vol. 1. Preceramic, Plainware,
and Viejo Periods (G.J. Fenner, Ed.). Northland Press,
Flagstaff, Arizona. 287 pp.

. 1974b. Commerce. Pp. 123-232 in Casas Grandes,

a fallen trading center of the Gran Chichimeca, Vol. 8
(C.C. Di Peso, J.B. Rinaldo, and G. Fenner, Eds.).
Northland Press, Flagstaff, Arizona. 424 pp.

. 1974c. Bird remains. Pp. 267-273 in Casas Grandes,

a fallen trading center of the Gran Chichimeca, Vol. 8
(C.C. DiPeso, J.B. Rinaldo, and G. Fenner, Eds.).
Northland Press, Flagstaff, Arizona. 424 pp.

Emslie, S.D., and L.L. Hargrave. 1978. Feathers from
the Westwater Ruin, southeastern Utah. Pp. 235-246
[Appendix] in Excavations of Westwater Ruin (42SA14),
San Juan Co., Utah: First Field Session. Report submit-
ted to Utah Navajo Development Council, Blanding,
Utah (L.W. Lindsay, Ed ). Division of State History,
Utah. 246 pp. (unpublished)

Euler, R.C. 1978. Archeological and paleobiological studies
at Stanton’s Cave, Grand Canyon National Park, Arizona
— a report of progress. National Geographic Society Re-
search Report, 1969 Projects. Pp. 141-162.

Feduccia, A. 1972. The Pleistocene avifauna of Klein Cave,
Kerr County, Texas. Southwest. Naturalist 17:295-296.

. 1973. Fossil birds from the Late Pleistocene Ingleside

Fauna, San Patricio County, Texas. Condor 75:243-244.

Flannery, K.V. 1967. Vertebrate fauna and hunting pat-
terns. Pp. 132-177 in Prehistory of the Tehuacan Valley,
Vol. 1. Environment and Subsistence (D.S. Byers, Ed.).
Univ. Texas Press, Austin and London. 331 pp.

Gehlbach, F.R., and J. A. Holman. 1974. Paleoecology of
amphibians and reptiles from Pratt Cave, Guadalupe
Mountains National Park, Texas. Southwestern Natural-
ist 19:191-198.

Grinnell, J., J.S. Dixon, and J.M. Linsdale. 1937. Fur-
bearing mammals of California: their natural history, sys-
tematic status, and relations to man. Vol. 1. Univ. Calif.
Press, Berkeley. 375 pp.

Hargrave, L.L. 1970a. Feathers from Sand Dune Cave: a
Basketmaker Cave near Navajo Mountain, Utah. Mu-
seum of Northern Arizona Tech. Series No. 9:1-52.

. 1970b. Mexican macaws: comparative osteology and

survey of remains from the Southwest. Anthro. Pap.,
Univ. Arizona, No. 20:1-67.

. (in press). A macaw feather artifact from southeast-
ern Utah. Southwestern Lore.

Hargrave, L.L., and S.D. Emslie. 1979. Osteological
identification of Sandhill Crane vs. turkey. Amer. Antiq-
uity 44(2):295-299.

Harris, A.H. 1977[= 1978], Wisconsin Age environments in
the northern Chihuahuan Desert: Evidence from the
higher vertebrates. Pp. 23-52 in Trans. Symp. Biol. Res.
Chihuahuan Desert Regions, U.S. and Mexico (R.H.
Wauer and D.H. Riskind, Eds.). Nat. Park Serv. Trans,
and Proc., Ser. No. 3:1-658.

Haury, E.W. 1936. The Mogollon Culture of southwestern
New Mexico. Medallion Papers, No. 20:1-146.

. 1937. Food. Pp. 156-158 in Excavations at Snake-

town: material culture (H.S. Gladwin, E.W. Haury, E.B.
Sales, and N. Gladwin, Eds.). Medallion Papers, No. 25.
305 pp.

— . 1950. The stratigraphy and archaeology of Ventana

Cave, Arizona. Univ. New Mexico, Albuquerque, and
Univ. Arizona, Tucson. 599 pp.

. 1976. The Hohokam: desert farmers and craftsmen:

excavations at Snaketown, 1964-1965. Univ. Arizona
Press, Tucson. 412 pp.

Hibben, F.C. 1975. Kiva art of the Anasazi at Pottery
Mound. KC Publications, Las Vegas, Nevada. 145 pp.
Hough, W. 1914. Culture of the ancient Pueblos of the Up-
per Gila region, New Mexico and Arizona. U.S. National
Museum Bulletin 87:1-139.

Howard, H. 1962. Bird remains from a prehistoric cave de-
posit in Grant Co., New Mexico. Condor 64:241-242.

. 1971. Quaternary avian remains from Dark Canyon

Cave, New Mexico. Condor 73:237-240.

Howard, H., and A.H. Miller. 1933. Bird remains from
cave deposits in New Mexico. Condor 35:15-18.
Jennings, J D. 1974. Prehistory of North America, 2nd Ed.

McGraw-Hill Book Company, New York. 437 pp.
Johnson, E. 1977. Animal food resources of Paleoindians.

The Museum Journal 17:65-77.

Kaplan, L. 1956. The cultivated beans of the prehistoric
Southwest. Annals, Missouri Bot. Garden 43:189-251.

. 1967. Archaeological Phaseolus from Tehuacan. Pp.

209-210 in The prehistory of the Tehuacan Valley, Vol.
1. Environment and subsistence (D.S. Byers, Ed.). Univ.
Texas Press, Austin, Texas. 331 pp.

Kent, K.P. 1957. The cultivation and weaving of cotton in
the prehistoric southwestern United States. Trans, of the
Amer. Phil. Soc., n.s., 47(3):457-732.

Lange, C.H. 1950. Notes on the use of turkeys by Pueblo
Indians. El Palacio 5 7:204-209.

Lundelius, E.L., Jr. 1972. Fossil vertebrates from the late
Pleistocene Ingleside fauna, San Patricio County, Texas.
Texas Univ. Bur. Econ. Geol., Report of Investigations
No. 77:1-74.

MacNeish, R.S. 1958. Preliminary archaeological investi-
gations in the Sierra de Tamaulipas, Mexico. Trans.
Amer. Phil. Soc., n.s., 48(6): 1-2 10.

. 1964. Ancient Mesoamerican civilization. Science

143(3606):53 1—537.

Mangelsdorf, P.C. 1950. The mystery of corn. Sci. Amer.
183:20-24.

224

Rea: Southwestern Turkeys

. 1954. New evidence on the origin and ancestry of

maize. Amer. Antiquity 19:409-410.

. 1974. Corn: its origin, evolution, and improvements.

Harvard Univ. Press, Cambridge. 262 pp.

Mangelsdorf, P.C., R.S. MacNeish, and W.C. Gali-
nat. 1964. Domestication of corn. Science 143(3606):538-
545.

Mangelsdorf, P.C., and C.E. Smith, Jr. 1949. New ar-
chaeological evidence on the evolution of maize. Botanical
Museum Leaflet Harvard Univ. Cambridge, Massachu-
setts. No. 13:213— 260.

Martin, P.S., and F. Plog. 1973. The archaeology of Ar-
izona: a study of the southwest region. Doubleday/Nat-
ural History Press, Garden City, New York. 422 pp.

Martin, P.S., and J.B. Rinaldo. 1950. Turkey Foot Ridge
Site, a Mogollon village, Pine Lawn Valley, Western New
Mexico. Fieldiana: Anthropology 38(2):237-396.

. 1960. Excavations in the upper Little Colorado

drainage, eastern Arizona. Fieldiana: Anthropology 51:
1-127.

Martin, P.S., J.B. Rinaldo, and E.A. Bluhm. 1954.
Caves of the Reserve area. Fieldiana: Anthropology 42:1—
227.

Martin, P.S., J.B. Rinaldo, E. Bluhm, H.C. Cutler,
and R. Grange, Jr. 1952. Mogollon Cultural continuity
and change: the stratigraphic analysis of Tularosa and
Cordova Caves. Fieldiana: Anthropology 40:1-528.

Martin, P.S., and J. Schoenwetter. 1960. Arizona’s old-
est cornfield. Science 132:33-34.

McGregor, J.C. 1965. Southwestern archaeology, 2nd. Ed.
Univ. Illinois Press, Urbana. 511 pp.

McKusick, C.R. 1974. The Casas Grandes avian report. Pp.
273-284 in Casas Grandes, a fallen trading center of the
Gran Chichimeca, Vol. 8 (C.C. Di Peso, J.B. Rinaldo,
and G. Fenner, Eds.). Northland Press, Flagstaff, Ari-
zona. 424 pp.

. 1976. Avifauna. Pp. 374-377 [Appendix] in The

Hohokam: desert farmers and craftsmen (E.W. Haury,
Ed.). Univ. Arizona Press. 412 pp.

. 1979. Three groups of turkeys from southwestern

archaeological sites, (this vol.).

Miller, L. 1940. A new Pleistocene turkey from Mexico.
Condor 42:154-156.

. 1943. The Pleistocene birds of San Josecito Cavern,

Mexico. Univ. California Publ. Zool. 47:143-168.

Morris, A. A. 1933. Digging in the Southwest. Doubleday,
Doran, and Co., Inc., Garden City, New York. 301 pp.

Morris, E.H. 1939. Archaeological studies in the La Plata
district, southwestern Colorado and northwestern New
Mexico. Carnegie Inst. Washington. 248 pp.

Needham, C.E. 1936. Vertebrate remains from Cenozoic
rocks. Science 84:537.

Parmalee, P. W. LTilization of birds by the Archaic and Fre-
mont cultural groups of Utah, (this vol ).

Phillips, A., J. Marshall, and G. Monson. 1964. The
birds of Arizona. Univ. Arizona Press, Tucson. 220 pp.

Rea, A.M. (in press.) The paleoavifauna of Lubbock Lake,
Texas. In Lubbock Lake: late Quaternary studies on the
Llano Estacado (Eileen Johnson, Ed.).

Reed, E.K. 1951. Turkeys in southwestern archaeology. El
Palacio 58:195-205.

Ridgway, R., and H. Friedmann. 1946. The birds of

North and Middle America. Part 10. U.S. Nat. Mus.
Bull. No. 50. 484 pp.

Roth, E.L. 1972. Late Pleistocene mammals from Klein
Cave, Kerr County, Texas. Texas J. Sci. 24:75-84.

Sauer, J , and L. Kaplan. 1969. Canavalia beans in Amer-
ican prehistory. Amer. Antiquity 34:417-424.

Schorger, A.W. 1961. An ancient pueblo turkey. Auk
78:138-144.

. 1966. The wild turkey: its history and domestication.

Univ. Oklahoma Press, Norman. 625 pp.

. 1970. A new subspecies of Meleagris gallopavo. Auk

87:168-170.

Schultz, C.B., and E.B. Howard. 1935. The fauna of
Burnet Cave, Guadalupe Mountains, New Mexico. Proc.
Natl. Acad. Sci. Phil. 87:273-298.

Skinner, M.F. 1942. The fauna of Papago Springs Cave,
Arizona, and a study of Stockoceros\ with three new an-
tilocaprines from Nebraska and Arizona. Bull. Amer.
Mus. Nat. Hist. 80:143-220.

Smith, C.E., Jr. 1950. Prehistoric plant remains from Bat
Cave. Bot. Mus. Leaflets, Harvard Univ. 14(7): 157-180.

Smith, W. 1952. Kiva mural decorations at Awatovi and
Kawaika-a. Papers Peabody Mus., Amer. Archaeology
and Ethnology, Harvard Univ. No. 37:1-363.

Steadman, D. A review of the osteology and paleontology of
turkeys (Aves: Meleagridinae) (this vol.).

Stiger, M.A. 1979. Mesa Verde subsistence patterns from
Basketmaker to Pueblo III. Kiva 45:133-134.

Van Devender, T.R. 1976. The biota of the hot deserts of
North America during the last glaciation: the packrat
midden record. Abstracts, Amer. Quat. Assoc. 1976:62—
67.

. 1977. Holocene woodlands in the southwestern des-
erts. Science 198:189-192.

Van Devender, T.R., and W.G. Spaulding. 1979. De-
velopment of vegetation and climate in the southwestern
United States. Science 204:701-710.

Van Devender, T.R., and F. Wiseman. 1977. A prelimi-
nary chronology of bioenvironmental changes during the
Paleoindian Period in the monsoonal Southwest. The
Museum Journal 17:13-27.

Van Devender, T.R., and R.D. Worthington. 1977
[=1978]. The herpetofauna of Howell’s Ridge Cave and
the Paleoecology of the northwestern Chihuahuan Desert.
Pp 85-106 in Trans. Symp. Biol. Res. Chihuahuan Des-
ert Regions, U.S. and Mexico (R.H. Wauer and D.H.
Riskind, Eds.). Nat. Park Serv. Trans, and Proc., Ser.
No. 3:1-658.

Wells, P.V. 1966. Late Pleistocene vegetation and degree
of climatic change in the Chihuahuan Desert. Science
153:970-975.

Wells, P.V., and R. Berger. 1967. Late Pleistocene his-
tory of coniferous woodland in the Mohave Desert. Sci-
ence 155: 1640-1647.

Wetmore, A. 1932. Additional records of birds from cavern
deposits in New Mexico. Condor 34:141-142.

Whitaker, T.W., and W.P. Bemis. 1975. VIII. Origin and
evolution of the cultivated Cucurbita. Bull. Torrey Bot.
Club 102:362-368.

Wormington, H.M. 1978. Prehistoric Indians of the South-
west. Denver Mus. Nat. Hist., Popular Series No. 1.
(14th ed.) 191 pp.

THREE GROUPS OF TURKEYS FROM SOUTHWESTERN
ARCHAEOLOGICAL SITES

By Charmion R. McKusick1

ABSTRACT: The study of a minimum of 6713 whole or partial turkey skeletons from 95 southwestern

archaeological sites dating between ca. 300 B.C. and A.D. 1723 has resulted in the differentiation of three
groups of turkeys that are separable from each other and from all other turkeys known to have occurred
in the Southwest:

(1) The Small Indian Domestic, Meleagris gallopavo tularosa, first appeared between ca. 300 and 150
B.C. in the Mogollon Culture Area of west central New Mexico coincidentally with the establishment of
a stable agricultural food supply, peaked in the Eastern Periphery Pueblos, and became extinct with the
fall of the Pueblos in 1672.

(2) The Large Indian Domestic, Meleagris gallopavo merriami, appeared in the Anasazi Culture Area
in northeastern Arizona about A.D. 400 along with the beginnings of agriculture. The Large Indian
Domestic is the predominant race of turkey in southwestern archaeological collections dating from its first
appearance about 400 A.D. until 1723, the last known date for large flocks of Indian turkeys.

(3) Merriam’s Wild Turkey, also Meleagris gallopavo merriami, may have been present as a feral form

of the Large Indian Domestic as early as A.D. 500,
Wild Turkey existed shortly before A.D. 600.

The study of turkeys from southwestern archaeological sites
summarized herein was undertaken as the result of the dis-
covery in 1967 that the remains of over 900 turkeys recovered
from the excavation of Mound 7, Gran Quivira National Mon-
ument, New Mexico, were unlike any previously studied from
southwestern archaeological sites. This unusually homoge-
neous group of small, gracile-boned turkeys with humped
backs and short tarsi raised the question of the time and place
of their domestication, and the identity of the wild progenitors
of southwestern Indian turkeys in general. As is often the case
when several investigators are working on the same problem
without each other’s knowledge, they tend to arrive at the
same conclusion simultaneously, but from different directions.
The late A. W. Schorger, an outstanding authority on the wild
turkey, wrote to T.W. Mathews of the Southwest Archaeo-
logical Center, Globe, Arizona, in the summer of 1969 seeking
information relative to a desiccated short-shanked turkey with
a fully-feathered neck from a southwestern archaeological site
that he was studying. Mathews supplied him with the series
of measurements I had taken from the Mound 7 turkeys. From
a comparison of characters and measurements it became ob-
vious to all three of us that the Tularosa mummy was a spec-
imen of the small breed known only from skeletal elements at
Gran Quivira. Based on our combined data, Schorger
(1970:168-170) described the Tularosa Turkey as a new extinct
subspecies, Meleagris gallopavo tularosa, in January of 1970.

Although Schorger’s enquiry resulted in a concerted effort
to solve the turkey problem, the problem itself had been
around a long time. I had discussed the Indian domestic tur-
keys, their origin and development with the late Lyndon L.

1 Southwest Bird Laboratory, Route 1, Box 35-D, Globe, Arizona
85501.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 330:225— 235

and feral turkeys clearly identifiable as Merriam’s

Hargrave, Erik Reed, and Mathews in 1963, but no plan of
research was proposed. The subject came up again when Al-
den C. Hayes submitted his avian collection from the exca-
vation of Mound 7 at Gran Quivira National Monument for
identification. To complete this study, I found it necessary to
isolate diagnostic characters of the three groups of Indian tur-
keys. Mathews confirmed the validity of these characters in
producing reliable identifications. A.W. Schorger reviewed the
entire project before his death, and wrote to us that our find-
ings agreed with his. Small Indian Domestic Turkeys, Large
Indian Domestic Turkeys, and Merriam’s Wild Turkeys may
be distinguished from each other by comparison of characters
of the mandible, pelvis, coracoid, scapula, humerus, carpo-
metacarpus, femur, and tibiotarsus, as well as by means of mea-
surements of sample populations. Illustrations of the age stages
of the three groups of turkeys (which are pertinent to the ar-
chaeologist’s reconstruction of the yearly round of human ac-
tivity), the diagnostic osteological characters of the three
groups, and the means of measurements of the sample popu-
lations are found elsewhere (McKusick 1980).

The capability of segregating the various groups of turkeys
that occur at prehistoric habitational sites in the Southwest
made it possible for the first time to test the various hypotheses
of turkey origins and distribution.

1. Reed (1951), citing Hargrave’s observation that turkey
pens are found in areas where wild turkeys do not now exist
and where there is no evidence of there ever having been any,
hypothesized that turkeys were hunted in areas where they
were available as wild birds and were raised in areas where
they were not.

2. Reed (1951) further hypothesized that turkeys were eaten
mostly by the northern Anasazi and their cultural descendents.

3. Johnson (1965) hypothesized that hunting turkeys rather

226

McKusick: Turkey Domestication

Figure 1. Locations of dated southwestern archaeological sites yielding turkey remains.

than raising them was a diagnostic trait of Western Pueblo
Culture.

4. Hargrave hypothesized (pers. comm. 1963) that aberrant
turkey bones recovered from southwestern human habitational
sites represented remnants of Pleistocene forms. He did not
attempt to explain how the Indians had obtained these forms,
or why they had supposedly persisted into the fourteenth cen-
tury A.D.

5. Hargrave discussed (pers. comm. 1963), and Rea (this
vol.) tested, the hypothesis that modern Merriam’s Wild Tur-
keys are descendents of feral domestic turkeys, rather than
that Indian domestic turkeys are the descendents of indigenous
wild forms.

6. McKusick hypothesized in 1968 (Western Archaeological
Center Avian Files) that Indian domestic turkeys were intro-
duced into the Southwest as part of the Formative Level Cul-
tural Complex or Complexes (corn, beans, squash, permanent
housing, pottery, social stratification) from some place or
places outside the Southwest.

Recently, A.M. Rea’s interest in Meleagris crassipes re-
opened the issue of the southwestern specimens identified as
Pleistocene and pre-agricultural turkeys. This project made it
necessary for me to re-examine all turkey bones from south-

western archaeological human habitational sites that had been
set aside as Pleistocene remnants, just as Rea was re-exam-
ining all pre-Formative Stage specimens that had been for-
merly identified asM. gallopavo. Rea did not provide me with
a hypothesis to test in this project, rather he asked for a review
of the specimens and an objective evaluation of their identi-
fication.

METHODS

In this study, variables such as age stages and sexual di-
morphism were determined before subspecific differences or
differences between wild and domestic forms of the same sub-
species were considered. I raised poults of the modern table
breed, Meleagris gallopavo gallopavo, to adulthood to secure
data on differential rates of growth of males and females
through their developmental stages. I obtained skeletons to
represent the desired osteological age classes from free-ranging
domestic turkeys. In addition, C.A. Thomas of the Southwest
Archaeological Center enlisted the aid of Gila County, Ari-
zona, turkey hunters. I examined turkeys from both spring
and fall hunting seasons, collecting head, feet, and feather
samples, noting the color of vane, rachis, and legs, recording

McKusick: Turkey Domestication

227

the sex, and tagging the carcass for matching of the bones and
samples after the birds were roasted and eaten. Seven age
stages were defined to answer specific questions for the ar-
chaeologist about aboriginal turkey usage, turkey culture, and
the yearly round of human activity (McKusick 1979).

Information on age stages and sexual dimorphism also
proved useful in the proper identification of a series of speci-
mens that Hargrave had accumulated at the Southwest Ar-
chaeological Center, Globe, Arizona, from collections studied
between the early 1930’s and 1968. These had been referred
to Pleistocene forms in spite of the fact that they were found
in collections from human habitational sites in the Southwest,
some of which have been dated as late as A.D. 1340.

Data on the Small Indian Domestic Turkey, Meleagris gal-
lopavo tularosa, are based on an unmixed population of 923
individuals from Mound 7, Gran Quivira National Monu-
ment, New Mexico. Data for the Large Indian Domestic Tur-
key, Meleagris gallopavo merriami, are based on an early, un-
mixed, classic population of 32 individuals from Tse-ta’a,
Canyon de Chelley, Arizona. Archaeological samples of Mer-
riam’s Wild Turkey, also Meleagris gallopavo merriami, are
too fragmentary to be useful in establishing ranges of mea-
surements; therefore, modern wild specimens were employed.

As the study progressed, it seemed advisable to test methods
and criteria against as many collections from as many areas
of the Southwest and as many time periods as possible. Ac-
cordingly, every turkey specimen available for loan from
southwestern archaeological sites was borrowed, examined,
and where possible, measured.

SOUTHWESTERN TURKEYS IN
TIME AND SPACE

I derived the time data and geographical locations found in
this section from file records and publications of the institu-
tions that loaned the specimens I examined, and from personal
communication with individuals who performed or were fa-
miliar with the excavations of the sites from which the speci-
mens were recovered. Site locations are indicated on Fig. 1.

For the purpose of easy comparison, I shall discuss the sites
by the time periods established by Willey (1966:188) for cul-
tural designations of the Anasazi Culture, since in no other
cultural area did turkeys assume such an important role in the
economy and life of the people. It is recognized that the pop-
ulations of some areas, such as Chaco Canyon, reached a given
cultural level earlier than the main stream of Anasazi devel-
opment, while others, such as the western Basket Maker,
lagged considerably. To the non-archaeologist reader, cultural
development at some sites may seem slightly out of phase with
the main stream of southwestern development, but this is a
well-recognized phenomenon that in our modern day culture
is illustrated by the contrast between the cultural manifestation
in Manhattan, a medium-sized midwestern town, and a back-
woods farm in Appalachia. What is important in this study is
the date of occurrence of the turkeys, their breeds, and their
relationships to the inhabitants of the site at which they were
found.

Proto-Agricultural and Basket Maker II

300 B.C. to A.D. 400, Fig. 2

The only occurrence of turkey feathers in the Southwest that
has been referred to a Proto-Agricultural context is that re-

ported from the excavation of Fresnal Cave, LA 10 101, near
Cloudcroft, New Mexico, which dates from ca. 2500 B.C. to
ca. A.D. 1 (Vorsila Bohrer pers. comm.). Turkey feathers from
this cave are mainly the irridescent tips of body feathers; only
a few specimens are from wing coverts or the tail. The collec-
tion includes rachis sections stripped of their vanes in a manner
associated elsewhere with the production of feather cordage.
Therefore, I believe these turkey specimens represent a level
of cultural development no earlier than Basket Maker II, and
I have included them here despite the greater antiquity of the
major portion of the collection.

In undisputed Basket Maker II levels, turkey feather blan-
kets occur in addition to rabbit fur blankets in such areas as
Grand Gulch, Utah; Durango, Colorado; and Canyon del
Muerto, Arizona (Morris 1939:18). Excavations at Canyon del
Muerto have also produced from a Basket Maker II level the
headless mummy of a Small Indian Domestic (a “Tularosa
Turkey”) with vegetal cordage around its neck. It was dated
at ca. A.D. 250 on the basis of associated cultural material by
R.G. Vivian (Southwest Archaeological Center Photo Files).
R. Richert, who is familiar with the circumstances of the find,
believes this is a conservative date; i.e., the bird may be even
older (Richert pers. comm.).

Bones of four Small Indian Domestics from Tularosa Cave,
New Mexico, occur in Pre-Pottery Phase levels dated by Mar-
tin et al. (1952:483) at ca. 300 B.C. ± 150-200 years.

The more western Basket Maker II peoples of the Kayenta
area of northeastern Arizona apparently were behind their
eastern counterparts in the use of turkeys, since Guernsey and
Kidder (1921:111) did not find any feather-string in these sites.
I re-examined the feather collection from Woodchuck Cave,
a Basket Maker II burial cave dated at A.D. 200 ± 100 years
(Locket and Hargrave 1953). There are two specimens of pas-
serine feather cordage, one a Z-twisted skin strip and the other
a skin strip with feathers Z-wrapped on a two-ply Z-twist
vegetal core, but only one specimen of whole-feather turkey
wrapping (Specimen No. 3112/B6.30). No turkey bones were
recovered from this cave, and turkey feathers are so few that
it is probable that they were trade items, since wild turkeys
are not known from this area at any time level. Du Pont Cave
in Utah, dated by tree rings at A.D. 217 (Lockett and Har-
grave 1953:31), contained no turkey cordage, only the type of
cordage wrapped with strips of skin of small birds (Nusbaum
et al. 1922:104).

Evidence of cultural lag is found at Sand Dune Cave near
Navajo Mountain, Utah, where the Basket Maker II period
is dated A.D. 300 to A.D. 700 (Hargrave 1970). Here are found
only bundled turkey feathers, and no feather cordage even in
a time period that was elsewhere Basket Maker III. In this
case, far beyond the known range of wild turkeys, feathers are
presumably those of domestic birds traded in from another
area, probably Canyon de Chelley or Canyon del Muerto, Ar-
izona.

Basket Maker III

A.D. 400-700, Fig. 3

Both at Mesa Verde, Colorado, and at Tse-ta’a in Canyon
de Chelley, Arizona, the Large Indian Domestic Breed was
predominant. Skeletal remains from this period show no in-
dications of food use, and presumably the birds were used only
for their feathers. Morris (1941:197, 201, 202) reported turkey
pens between the Basket Maker III slab and pole houses ex-

228

McKusick: Turkey Domestication

cavated at Tseahatso and the surrounding cave wall. Approx-
imately 300 (Lee Abel pers. comm.) natural mummies of tur-
keys that had apparently died of old age were excavated from
this cave. In addition, one young turkey that had suffered a
broken leg was found. The injured limb had been set, bound
with soft fiber, and splinted. In spite of this careful attention,
the bird died and was buried with the splints still in place.

W.D. Lipe excavated the bones of Large Indian Domestic
Turkeys from Site GG70-187, about 3 km east of Grand
Gulch, Utah, associated with Basket Maker III cultural ma-
terials that probably date into the A.D. 600’s. These are classic
specimens that compare well with those from Tse-ta’a in Ar-
izona.

Bones of Merriam’s Wild Turkey were found at AZ P: 16: 1 ,
Bear Ruin, in east central Arizona, with remains that date
from the seventh century A.D. Two tubes cut from the tar-
sometatarsi of males, including the spur cores (Haury 1940:14,
116), represent specimens of extremely early worked turkey
bone.

In the Point of Pines area of Arizona, a turkey bone was
recovered from an early Circle Prairie Phase pit house that
probably dates before A.D. 600 (Wheat 1954:179). I have re-
ferred it to Merriam’s Wild Turkey.

Pueblo I

A.D. 700-900, Fig. 4

Information on Pueblo I turkeys is scanty. The only material
examined is from Site 1205, Site 1678, and Badger House, all
on Mesa Verde, Colorado; Tse-ta’a in Canyon de Chelley,
Arizona; La 3427, the Favorino Site on the San Juan River,
New Mexico; and La 3320, southwest of Duke, New Mexico.
Only the Large Indian Domestic Breed is represented. There
is no reliable evidence of the use of turkeys other than for
feathers in this period.

Guernsey (1931:92-93) reported finding feather cord robes
in Pueblo I burials at Cave 1, Segi Canyon, in northeastern
Arizona. Because fur cord robes were included in the same
burials, it is probable that the feather cord examples were
items of trade and that live turkeys were not yet known in the
area.

Pueblo II

A.D. 900-1100, Fig. 5

During Pueblo II time the Large Indian Domestic Breed
spread from its Four Corners homeland as far west as NA

McKusick: Turkey Domestication

229

8604 near Kiet Siel, Arizona, east to TA 32 near Ranchos de
Taos, New Mexico, and south to Casas Grandes, Chihuahua,
Mexico, forming a T-shaped distribution.

In contrast to earlier levels that lacked turkey remains,
Pueblo II and later levels at Tseh-So in Chaco Canyon, New
Mexico, contained many turkey bones (Brand et al. 1937:101,
106). Burials of headless female turkeys were also found in the
kivas at this site.

Tularosa Cave, although now included in the area where
Large Indian Domestics were predominant, still yielded a
feathered mummy of the Tularosa Turkey, or Small Indian
Domestic. The occurrence of a Small Indian Domestic at this
site may have cultural significance. Four desiccated poults of
erythristic coloration and a specimen of bone from Merriam’s
Wild Turkey were also recovered from Tularosa Cave at levels
dated at A.D. 1100 (P.S. Martin pers. comm.).

The one piece of stripped-vane cordage recovered from NA
863, Medicine Cave, in the area of Flagstaff, Arizona, may
have been a trade item (Bartlet 1934:46); however, a turkey
bone has also been found at a nearby site at the Grand Falls
of the Little Colorado. I have referred the latter to the Large
Indian Domestic Breed. These occurrences in the Sinagua Cul-
tural Area are out of the main region of turkey raising, and
turkeys never assumed any real importance.

Pueblo III

A.D. 1100-1300, Fig. 6

By Pueblo III times the Large Indian Domestic was gener-
ally known throughout the Southwest except in the Sinagua
and Hohokam Culture Areas of Arizona. The Sinagua peoples
had domestic turkeys available, but did not include turkey
raising in their cultural complex. Turkey remains found in
Sinagua sites were probably traded in from the north and east.
The Hohokam peoples appear to have been even more disin-
terested, as no turkey specimens have been found in Hohokam
sites except for two or three specimens from Salado cultural
deposits, such as those at Casa Grande, Arizona.

The butchered and broken condition of Merriam’s Wild
Turkey bones recovered from sites along the Mogollon Rim
during this time period indicates that wild turkeys were hunted
to a limited extent along this upland area from the Galaz Site
in southwestern New Mexico to Walnut Canyon in northern
Arizona.

The Small Indian Domestic appeared at Gran Quivira as a
homogeneous, well-standardized breed about A.D. 1275. It
has been found in much smaller numbers in Pueblo III deposits
at Mesa Verde, Colorado, and at Casas Grandes, Chihuahua,
Mexico.

230

McKusick: Turkey Domestication

Pueblo IV

A.D. 1300-1700, Fig. 7

The large settlements in the Mesa Verde, Colorado; Canyon
de Chelley, Arizona; and Chaco Canyon, New Mexico, areas
were vacant by the Pueblo IV period. The center of turkey
raising had moved south and east to the Rio Grande and the
Eastern Periphery of the Southwest Cultural Area. Although
the Large Indian Domestic was predominant, wild turkeys
were hunted in the Point of Pines area and at Grasshopper in
Arizona. The two specimens from the Verde Valley, Arizona,
are also probably Merriam’s Wild Turkey, which had been
present just to the east as a feral form of the Large Indian
Domestic since before A.D. 600.

The Small Indian Domestic Breed reached its peak at this
time in central New Mexico at the Tompiro Pueblos of Gran
Quivira, Pueblo Pardo, and Tabira. It was still present at
Casas Grandes, Chihuahua, Mexico, in small numbers, along
with the much more numerous Large Indian Domestics, until
the fall of the city ca. A.D. 1340. Single specimens of the Small
Indian Domestic have been found at the University Ruin and
Reeve Ruin in Arizona (which date to ca. A.D. 1350), but
turkey raising never assumed any importance in this area. An

immature specimen from Casa Grande, Arizona, is definitely
from a domesticated turkey but is too young to assign to breed.
A turkey specimen from Gila Pueblo, Arizona, is so young that
it cannot be assigned to breed, wild or domestic.

DISCUSSION OF HYPOTHESES

1 . Reed’s hypothesis that turkeys were hunted where they were
available wild, and raised in areas where they were not.

Reed’s discussion (1951) indicated that his thinking centers
on the obvious evidence of domestication in the Anasazi Cul-
tural Area. Certainly domestic turkeys were present in the
Anasazi Area, but data indicate that they were not domesti-
cated there, but were brought in as a well-established domestic
breed from elsewhere.

However, Small Indian Domestics were known in the Mo-
gollon Cultural Area about 500 years before they were known
in the Anasazi Area, and about 700 years before the favored
Anasazi breed, the Large Indian Domestic, appeared in that
area.

Turkeys were hunted along the Mogollon Rim, but not early
(indeed, there were none to hunt) or in any great numbers. A
few were hunted by about A.D. 600 after the Large Indian

McKusick: Turkey Domestication

231

Domestics were introduced and had an opportunity to become
feral. More were hunted during the 12GG’s and 1300’s than at
any other period, but their remains never approached the
number of those of domestics at the excavations in which they
were found.

2. Reed’s hypothesis (1951) that turkeys were eaten mostly by
the northern Anasazi and their cultural descendents.

This is generally true for the Large Indian Domestic Breed.
It is interesting to note that no Small Indian Domestics turn
up as food refuse anywhere in the Southwest, with the possible
exception of one sample at Antelope House, Canyon del Muer-
to, Arizona. Greatest food use of turkeys occurs on the Mesa
Verde, Colorado, and at some of the Rio Grande Pueblos in
New Mexico, and involves only Large Indian Domestics. Mi-
nor food use of Merriam’s Wild Turkeys occurred late, in the
120Q’s and 1300’s, and only along the Mogollon Rim of Arizona
and New Mexico.

3. Johnson’s hypothesis (1965) that hunting turkeys rather than
raising them was a diagnostic trait of Western Pueblo Culture.

Of 17 Western Pueblo avian collections examined, domestic
turkeys were present in 16 and wild turkeys in only 6. Thus,
turkey hunting does not appear to be diagnostic for Western

Pueblo Culture, and is probably a function of geographical
location.

4. Hargrave’s hypothesis (1963) that Pleistocene forms were
present in southwestern archaeological sites.

I have discarded this hypothesis in light of present evidence.
The specimens he set aside as not conforming to the general
population of turkeys from archaeological sites have subse-
quently proved to be either immature forms of the Large In-
dian Domestic, or adults of the Small Indian Domestic.

5. The Hargrave/Rea hypothesis (1963) that southwestern wild
turkeys are the descendents of feral domestics, rather than that
Indian domestics were the descendents of indigenous wild
forms.

Available data support this hypothesis. Rea (this vol.) has
demonstrated that a different species of turkey was the indig-
enous inhabitant of most of the Southwest. Further, Merriam’s
Wild Turkey appeared in the Southwest subsequent to the
spread of the Large Indian Domestic to which it is clearly
related.

6. McKusick’s hypothesis (1968) that turkeys were brought
into the Southwest already domesticated from somewhere else
as part of the Formative Cultural Complex.

232

McKusick: Turkey Domestication

Figure 6. Turkey distribution in the Southwest from A.D. 1100 to 1300.

This hypothesis has also held up well. All early turkeys are
identifiable as members of two well-standardized domestic
breeds. The only evidence of local domestication in the South-
west is at Point of Pines, Arizona, where feral Large Indian
Domestics (i.e., Merriam’s Wild Turkeys) were apparently re-
domesticated.

No forms of Meleagris gallopavo, domestic or wild, are
found in the Southwest proper before the advent of Formative
Level Culture, first in the Mogollon Area, and much later, in
the Anasazi Area. This is reasonable if one considers that the
only way domestic turkeys could have been maintained in the
circumstances in which they are known to have occurred is if
there were a reliable agricultural surplus (Rea this vol.).

Where the domestics came from is still unknown. The Large
Indian Domestics in their feral form did well in the mountains
of the Mogollon Rim, but the Small Indian Domestics do not
appear to have gone feral. Rea (pers. comm.) has suggested
that this may indicate physiological as well as morphological
differences between the Small and Large Indian Domestics, a
possible result of the former having originally come from an
area that was ecologically very different from the Southwest.
The Small Indian Domestics were a small, dark-plumaged
breed that, following Bergman’s Rule and Allen’s Rule (Van

Tyne and Berger 1959:358), one would expect to inhabit a
warm, moist environment. The arid Southwest was apparently
inappropriate for their survival in the wild.

CONCLUSIONS

A review of the specimens of turkeys known to date from
archaeological sites in the Southwest indicates that research
into the origin and distribution of turkeys has been adversely
affected by a basic misconception. The preconception that In-
dian turkeys were domesticated from birds native to the area
has stood in the way of fruitful research for more than 40
years. Steadman (this vol.) and Rea (this vol.) have now cor-
rected this misconception. The Indian domestics came from
elsewhere, probably relatively far away, because the turkeys
presently surrounding the area — the Rio Grande Turkey,
Meleagris gallopavo intermedia ; Gould’s Wild Turkey, M. g.
mexicana', and the South Mexican Turkey, M. g. gallopavo —
are clearly not involved in the ancestry of the Small Indian
Domestic, M. g. tularosa, and I am unable to find characters
connecting any of them with the ancestry of the Large Indian
Domestic, M. g. merriami.

As introduced domestics, Indian turkeys take on a different

McKusick: Turkey Domestication

233

Figure 7. Turkey distribution in the Southwest from A.D. 1300 to 1700.

role than that to which the southwestern archaeologist has
become accustomed. They are as characteristic of a people as
are their pottery designs or their ceremonial paraphernalia.
They are useful aids in tracing trade routes and movements
of people. Their propagation and use is determined by cultural
factors, not by ecological considerations.

The study of avian remains from southwestern sites began
as an attempt to reconstruct prehistoric climate. Time and
experience, however, have made it evident that certain por-
tions of the avifauna are poor indicators of prehistoric climate.
Southwestern Indians persistently went great distances, in
some cases many hundreds of kilometers, to procure the birds
they desired while ignoring locally available species. Therefore
it is not surprising that the southwest Indians also brought
domesticated turkeys from somewhere beyond the general area
of their habitation.

The keeping of domestic turkeys presupposed a Formative
State of Culture, when agriculture was already well enough
established to provide a year-round surplus of food. In the
Southwest, the Formative Stage is known earliest, about 300
B.C., in the Mogollon Area. As to be expected, the earliest
turkeys in any southwest archaeological sites are also found in
the Mogollon Area, as exemplified by the Small Indian Do-
mestics at Tularosa Cave, that date from 300 B.C. ± ISO to

200 years. The next record of the Small Indian Domestic is at
Canyon del Muerto, ca. A.D. 2S0, but the breed was not gen-
erally favored by the Anasazi, though a few apparently per-
sisted at Antelope House through the Pueblo III occupation.
On Mesa Verde, the few Small Indian Domestics present at
Long House and Mug House date to ca. A.D. 1275 to 1300.
At Casas Grandes, Chihuahua, Small Indian Domestics ap-
parently interbred with Large Indian domestics. Why this hy-
bridization should have occurred here is uncertain, but the
peculiar nature of the site may shed some light on the problem.
Casas Grandes, Chihuahua, was a trading outpost of Mesoam-
erica. Manufactured goods were produced and traded for raw
materials and regional specialties. Turkeys from the American
Southwest were apparently traded south in exchange for ma-
caws. In the Mogollon and Anasazi Culture Areas, Small In-
dian Domestics and Large Indian Domestics co-existed for long
periods without any discernable mixing, perhaps as the result
of a cultural factor: that is, turkey strains may have been the
property of kinship groups (in the same manner as, for ex-
ample, seed corn), and may have been maintained as separate
property. It may be that turkeys traded south to Casas
Grandes entered a different cultural configuration, where they
were merely merchandise, and where cultural factors that may
have kept the strains separate in the north simply did not exist.

234

McKusick: Turkey Domestication

Small Indian Domestics were most numerous at the Tom-
piro Pueblos of central New Mexico. They were present there
from ca. A.D. 1275 until the breed disappeared in A.D. 1672
with the fall of Las Humanas Pueblo at Gran Quivira National
Monument.

The Large Indian Domestic appeared in the Southwest some
time during the Basket Maker III Period, between A.D. 400
and 700. The vast number of turkey burials in Canyon del
Muerto (Morris 1941), plus the rapidly increasing numbers of
feather cord robes present in remains from the end of the pre-
vious period, suggest that a date of A.D. 400 to 500 for intro-
duction of the breed is conservative. By A.D. 600 the Large
Indian Domestic was already the most numerous turkey race
in the Southwest. Its greatest areal expansion is coincidental
with the Pueblo II Anasazi expansion that took place between
A.D. 900 and 1 100 (see Willey 1966:207, Fig. 5). Large Indian
Domestics make up the greatest percentage of turkey remains
at nearly all sites from A.D. 600 through 1672, when large
flocks of domestic turkeys were last noted (Schroeder
1968:102-103).

Merriam’s Wild Turkey was present by A.D. 600; it is not
known before the time at which the Large Indian Domestic
became the predominant southwestern breed. Analysis of plant
remains from Tularosa Cave (Martin et al. 1952:469) outlines
a regression of Mogollon Culture in the Georgetown Phase,
A.D. 500 to 700, in which the Mogollon people dealt with
some crisis in their way of life by retreating for 200 years into
the Archaic Cultural Stage. Since we know that Indian do-
mestic turkeys were already well established in their culture,
it is reasonable to suppose that turkeys that may not have been
adequately tended during this more mobile hunting and gath-
ering period either died or became feral. No feral Tularosa
Turkeys have ever been found, and the Small Indian Domes-
tics that were not cared for probably died. However, feral
Large Indian Domestics seem to have survived rather well,
and are still with us today as Merriam’s Wild Turkey.

The only experiment in turkey domestication that can be
demonstrated in the entire Southwest, at any period, took
place at Point of Pines, Arizona. There, at the base of Nantak
Ridge, a classic population of Large Indian Domestics shows
late admixtures of wild characters and an unprecedented in-
crease in size. These fine big birds have been found at Casas
Grandes, Chihuahua, and perhaps were traded for macaws,
since macaws from AZ W: 10:50 and Chih. D:9:l show simul-
taneous identical abnormalities (McKusick 1974).

Thus, we have the Small Indian Domestic that may have
persisted in small numbers for 1900 years, from as early as
300 B.C. to A.D. 1672; the Large Indian Domestic that was
present in the Southwest for only about 1200 years, although
in much greater numbers than the Small Indian Domestic; and
the feral descendent of the Large Indian Domestic, Merriam’s
Wild Turkey, which has been around for 1400 years, and ap-
pears likely to persist given modern game management prac-
tices.

One factor that has become evident, but that was not con-
sidered in any of the hypotheses tested, is the relationship of
the turkey to Mesoamerican socio-religious practices. While
the occurrences of the macaw in the Southwest, both at Casas
Grandes, Chihuahua, Mexico, and in the United States, are
recognized as a function of the ebb and flow of the popularity
of the Quetzalcoatl Cult, little attention has been paid to the

place of the turkey in this complex. Burland and Forman
(1975:55-56) explain Quetzalcoatl, the Feathered Serpent, as
the manifestation of the intellectual-conscious side of the hu-
man mind. However, Quetzalcoatl has a Dark Twin, Tezca-
tlipoca, The Smoking Mirror. The Smoking Mirror is made
of polished obsidian and used for scrying (crystal-gazing as an
aid to clairvoyance), thus Tezcatlipoca represents the intuitive-
subconscious side of the human mind.

During the review of desiccated specimens of turkey remains
for this paper, I found that there are several occurrences of
the desiccated feet of turkeys that were tucked into dark cor-
ners of rock shelters, particularly in the area of Canyon de
Chelley and Canyon del Muerto. Just as the macaw is the sign
of Quetzalcoatl, so the turkey leg with claws is the sign of
Tezcatlipoca (Burland and Forman 1975:61). These desiccated
turkey feet date to ca. A.D. 1100, the point in time of the
greatest frequency of macaw remains in the northern South-
west (Di Peso 1974). Certainly it would seem desirable to note
occurrences of desiccated turkey feet from future excavations,
like those from Tularosa Cave and the area of Canyon de
Chelley and Canyon del Muerto to determine if other such
parallels are present.

This reassessment of turkey remains from archaeological
sites in the southwestern United States and some areas of Mex-
ico indicates three separate centers of turkey breeding:

1. The South Mexican, where domestic turkeys, Meleagris
gallopavo gallopavo, were known in the Tehuacan Valley be-
tween A.D. 200 and 700 (McNeish 1964).

2. The Mogollon, where Small Indian Domestics, Meleagris
gallopavo tularosa, were present between 300 B.C. ± 150 to
200 years.

3. The Anasazi, where Large Indian Domestics were prob-
ably present by A.D. 400 to 500.

Breeding per se was not successfully accomplished at all
sites where turkey remains were found, and it was probably
not even attempted at all sites. The frequency of turkey egg-
shells and small poults is highest at sites that could be consid-
ered trade centers. These are the same sites that yield such
faunal remains as macaws, mountain lion and bear bones in
ceremonial contexts, and human-bone artifacts. In most cases,
neighboring smaller sites do not have eggshells or small poults.
Presumably the birds were brought in as immatures from areas
of specialization in turkey culture, at least from ca. A.D. 100
on.

ACKNOWLEDGMENTS

Modern turkey specimens were loaned by Thomas W. Ma-
thews and the late Lyndon L. Hargrave of the Southwest Ar-
chaeological Center, Charles C. Di Peso of the Amerind Foun-
dation, and Richard Zusi of the Smithsonian Institution.
Turkey remains from archaeological sites were loaned by
Thomas W. Mathews, Charles C. Di Peso, Emil W. Haury,
Raymond Thompson, and William Longacre of the University
of Arizona; Edward B. Danson and William D. Lipe of the
Museum of Northern Arizona; Bertha P. Dutton of the Mu-
seum of Navajo Ceremonial Art; Stuart Peckham of the Uni-
versity of New Mexico; Alan R. Philips of the Instituto Na-
cional de Anthropologfa y Historia, Mexico, D.F., Mexico; the
Department of Anthropology, University of Minnesota; and
the Chicago Natural History Museum.

McKusick: Turkey Domestication

235

Thomas W. Mathews supervised the project for a period of
10 years, until the closing of the Southwest Archaeological
Center, and contributed much to the analysis of the data.
Chester A. Thomas provided time, funding, and travel ar-
rangements during this period, and enlisted the assistance of
the Gila County, Arizona, turkey hunters, without the coop-
eration of whom the project could not have been completed.

LITERATURE CITED

Bartlett, K. 1934. The Material Culture of Pueblo II in
the San Francisco Mountains, Arizona. Museum of
Northern Arizona, Bull. 7:1-76.

Brand, D.D., F.M. Hawley, and F.C. Hibben. 1937. Tse
So, A Small House Ruin, Chaco Canyon, New Mexico.
Preliminary Report. University of New Mexico Anthro-
pological Series, Bull. 2(2): 1—1 74.

Burland, C., and W. Forman. 1975. Feathered Serpent
and Smoking Mirror. Orbis Publishing, London. 128 pp.
Di Peso, C.C. 1974. Casas Grandes, Vol. 8, Bones-Econo-
my-Burials. The Amerind Foundation, Inc. Northland
Press 8(9): 1-424.

Guernsey, S.J. 1931. Explorations in Northeastern Arizona.
Papers of the Peabody Museum, Harvard University
12(1): 1-123.

Guernsey, S.J., and A.V. Kidder. 1921. Basket-Maker
Caves of Northeastern Arizona. Papers of the Peabody
Museum, Harvard University 8(2):1 — 121.

Hargrave, L.L. 1970. Feathers from Sand Dune Cave: A
Basketmaker Cave Near Navajo Mountain, Utah. Mu-
seum of Northern Arizona Technical Series 9:1-52.
Haury, E.W. 1941. Excavations in the Forestdale Valley,
East-Central Arizona. University of Arizona Social Sci-
ence Bull. 12:1-147.

Johnson, A.E. 1965. The Development of Western Pueblo
Culture. Doctoral dissertation, University of Arizona.
University Microfilms, Inc. Ann Arbor, Mich. 95 pp.
Lockett, H.C., and L.L. Hargrave. 1953. Woodchuck

Cave, A Basketmaker II Site in Tsegi Canyon, Arizona.
Museum of Northern Arizona Bull. 26:1-33.

Martin, P.S., J.B. Rjnaldo, E. Bluhm, H.C. Cutler,
and R. Grange, Jr. 1952. Mogollon Cultural Conti-
nuity and Change. Fieldiana: Anthropology 40:1-528.

MacNeish, R.S. 1964. Ancient Mesoamerican Civilization.
Science 143(3606):53 1-537.

McKusick, C.R. 1974. The Casas Grandes Avian Report.
Pp. 273-307 in Casas Grandes (C.C. Di Peso, Ed.). Vol.
8(9). The Amerind Foundation, Inc. Northland Press
8(9): 1-424.

. 1980. In Press, The Comparative Osteology of South-
west Indian Turkeys. Association of Universal Philosophy
Press, Globe, Arizona.

Morris, A. A. 1933. Digging in the Southwest. Doubleday,
Doran & Co., Inc. Garden City. 301 pp.

Morris, E.H. 1939. Archeological Studies in the La Plata
District, Southwestern Colorado and Northwestern New
Mexico. Carnegie Institution of Washington Publ. No.
19:1-298.

Nusbaum, J.L., A.V. Kidder, and S.J. Guernsey. 1922.
A Basket-Maker Cave in Kane County, Utah. Museum
of the American Indian, Heye Foundation, Indian Notes
and Monographs 29:1-153.

Reed, E.K. 1951. Turkeys in Southwestern Archaeology. El
Palacio 58(7): 195-205.

Schorger, A.W. 1970. A New Subspecies of Meleagris gal-
lopavo. Auk 87(1): 168—170.

Schroeder, A.H. 1968. Birds and Feathers in Documents
Relating to Indians of the Southwest. Pp. 95-114 in Col-
lected Papers in Honor of Lyndon Lane Hargrave, Papers
of the Archaeological Society of New Mexico 1(1): 1-1 70.

Wheat, J.B. 1954. Crooked Ridge Village. Social Science
Bulletin 24, University of Arizona Bull. 25(3): 1—1 83 .

Willey, G.R. 1966. An Introduction to American Archae-
ology. Vol. 1: North and Middle America. Prentice-Hall,
Inc., Englewood Cliffs, New Jersey. 530 pp.

Van Tyne, J., and A.J Berger. 1959. Fundamentals of
Ornithology. John Wiley & Sons, Inc. New York. 624 pp.

UTILIZATION OF BIRDS BY THE ARCHAIC AND
FREMONT CULTURAL GROUPS OF UTAH

By Paul W. Parmalee1

ABSTRACT: Approximately 5050 bird bones recovered from 5 Archaic and 11 Fremont sites in

northern and western Utah were identified. Remains of a minimum of 1029 individuals, representing at
least 21 families and 75 species, occurred in these prehistoric sites. Sixty-six percent of the elements
identified were those of swans, geese, and ducks, thus indicating that species of waterfowl were the major
supplemental avian food resources taken by these aboriginal people. Remains of aquatic and semi-aquatic
birds comprised 90 percent of all elements recovered from these 16 sites. Some major wing and leg
elements, especially those of large species (cranes, eagles, geese), were modified, suggesting a secondary
use of birds as a bone resource for the manufacture of artifacts. Except for the Passenger Pigeon, Ectopistes
migratorius (Linnaeus), all of the species represented in the archaeological samples still occur in Utah.

The prehistoric avifauna of Utah is poorly known, although
large quantities of bird remains have been recovered during
intensive archaeological investigations over the past four de-
cades. This void has not resulted because of a lack of interest
by archaeologists in this material, but rather because there are
limited available comparative osteological collections and or-
nithologists/osteologists with time to devote to such studies. As
has often been, and still is, the case, the archaeologist is faced
with the problem of “farming out” much of the faunal material
recovered during excavations. Archaeologically derived bird
bones have not, for the most part, received much attention.
However, there have been a few exceptions, notable among
them the studies dealing with feather remains from Sand Dune
Cave (Hargrave 1970) and Danger Cave (Sperry 1957), and
bones from Hogup Cave (Parmalee 1970), the Levee and Knoll
sites (Parmalee 1979), and the Bear River No. 2 site (Lay-
bourne 1967).

The present study involves the analysis of approximately
5300 bird specimens from the collections of the Department of
Anthropology, The University of Utah, Salt Lake City. These
avian remains were recovered during site excavations by an-
thropology students and faculty at The University of Utah
from the late 1930’s to 1973 (Table 1). For various reasons,
such as a lack of diagnostic ceramics or lithics, or because of
occupation by two or more aboriginal groups, placement of
each site within an exact cultural time sequence was not al-
ways possible. Based on all available cultural data, however,
five of these sites have been determined by the archaeologists
as being Archaic (ca. 7500-1000 B.C.) and 11 as having been
occupied primarily by peoples of the Fremont culture (ca. A.D.
350-1450). Locations of the 16 sites are plotted in Figure 1.

The diversification of hunting practices among and within

1 Professor of Zooarchaeology, Department of Anthropology, The
University of Tennessee, Knoxville, Tennessee 37916.

historic Indian tribes of North America has been well docu-
mented in the ethnographic literature. Hunting activities were
carried out either by individuals, by family groups, or as a
communal effort, depending upon the types of game sought,
its availability at a particular season (e.g., waterfowl migra-
tions and major bison herd movements), and/or the use of
appropriate techniques that would provide the greatest yield.
The species of birds hunted also varied considerably within
and among tribes. Weisel (1952:348) stated that “The Flathead
ate all the birds and their eggs,” although individuals of most
tribes for which there are subsistence data exhibited distinct
preferences for certain species while refusing to eat others.
Judd (1954:266), in commenting on a list of 13 species of birds
identified from osteological remains recovered at Pueblo Bo-
nito, New Mexico, suggested that “Presumably these were
killed or kept captive for their feathers alone, since the Pueblos
have always shunned winged creatures as a source of food.”
The southern Paiute were reported to have eaten “many kinds”
of birds (Kelly 1964:53, 54), but they would not eat crows,
certain woodpeckers, and meadowlarks. Mandelbaum
( 1940: 199) presented a list of avian species, compiled from data
obtained from tribal informants, that were and were not hunt-
ed for food by the Plains Cree. It is of interest to note that
these people would eat the young of some species, e.g., crows
and ravens, but not the adults.

Kelly (1964:53) stated that “most birds [were] taken from
blind . . . usually shot,” [southern Paiute], “The boys [Hidat-
sa, North Dakota] practice themselves in the use of the bow
by shooting at marmots and small birds, and in winter they
set horse-hair snares for snow-buntings” (Matthews 1877:58).
Lowie (1909:185), in discussing the northern Shoshone, men-
tioned that “Sage-hens were driven into an enclosure, or
trapped with nooses.” He also described (Lowie 1924:197) the
elaborate communal hunt for ducks and “mud-hens.” In con-
trast, Steward (1933:255) makes the following comment on the

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 330:237-250.

238

Parmalee: Prehistoric Utah Avifauna

Table 1 Data relative to Utah archaeological sites from which avifaunas were recovered.

Site Name and
Designation

Utah

County

Year

Excavated

Cultural
Designation
(C14 Dates)

Locale

Published

References

Deadman Cave
42T064

Tooele

1938

Archaic

NE Slope
Oquirrh Mts.
Great Salt Lake

Smith, E.R. 1952

Stansbury I
42TOl

Tooele

1947

Archaic

N. Facing Slope,
Stansbury Is.
Great Salt Lake

Jameson, S.J.S. 1958

Stansbury II
42T02

Tooele

1947

Archaic

N. Facing Slope,
Stansbury Is.
Great Salt Lake

Jameson, S.J.S. 1958

Sandwich Shelter
42TOI08

Tooele

1969

Archaic

7040 ± 280 BP

Alcove, Flank of
Stansbury Mts.
Great Salt Lake

Marwitt, J.P.,

G.F. Frye, and J.M.
Adovasio 1971

Black Rock II
42T029

Tooele

1938

Archaic

N. Slope
Oquirrh Mts.
Great Salt Lake

Black Rock III
42T03

Tooele

1939

Fremont

N. Slope
Oquirrh Mts.
Great Salt Lake

Enger, W.D., Jr. 1942

Bear River No. 1
42B055

Box Elder

1964

Fremont
AD 885 ± 120

Marshy River Bank
Bear River

Aikens, C.M. 1966

Bear River No. 3
42B098

Box Elder

1967

F remont
AD 500 ± 110

Marshy River Bank
Bear River

Shields, W.F., and
G.F. Dailey 1978

Levee

42BO107

Box Elder

1969

Fremont
AD 700 ± 140

Marshy Lake Side
Klondike Lake/
Bear River

Parmalee, P.W. 1979

Knoll

42BO109

Box Elder

1969

Fremont
AD 1310 ± 110

Marshy Lake Side
Klondike Lake/
Bear River

Parmalee, P.W. 1979

Warren

42WB-

Weber

1946

Fremont

Near Mouth of
Warren River

Enger, W.D., Jr., and
W. Blair 1947

Injun Creek
42WB34

Weber

1965

Fremont
AD 1605 ± 100
AD 1365 ± 90

Alluvium, Weber
River Delta,

Injun Creek Marsh

Aikens, C.M. 1966

Unnamed

42SL19

Salt Lake

1961

Fremont

Foothills
Wasatch Mts.
Jordon River

Nephi

42JB2

Juab

1965-1966

Fremont
AD 780 ± 85
AD 1670 ± 80

Alluvial Fan;
Salt Creek
Drainage

Sharrock, F.W., and
J P Marwitt 1967

Evans Mound
42IN40

Iron

1970-1973

Fremont
AD 1095 ± 90

Parowan Valley
Alluvial Fan,
Summit Creek

Berry, M.S. 1972

Pharo Village
42MD180

Millard

1967

Fremont
AD 460 ± 80
AD 1260 ± 90

Alluvial Fan,

Base of Pavant
Mts. Pharo Creek

Marwitt, J.P. 1968

manner of taking waterfowl by the Owens Valley Paiute (Ne-
vada): “Killed in early morning by hunters concealed in blinds
resembling wickiups or summer houses. Decoys, nets, and
communal hunts were unknown.” It is apparent from these
few ethnographic accounts that great variability in hunting
practices and the species of birds used did exist among con-
temporaneous aboriginal groups during the early historic pe-
riod. Therefore, interpretation of osteological avian remains
relative to the possible methods of capturing birds and the
preference for or use of certain species by prehistoric peoples
is basically speculative.

Hayward et al. (1976:25) pointed out that “Since birds are
among the most mobile of vertebrates, it is difficult to define
many of them in terms of their confinement to any special
community.” Although a large portion of Utah is desert, re-
ceiving rarely more than 25 cm of annual precipitation, the
state possesses both salt and freshwater lakes as well as a series
of high mountain ranges, plateaus, and major river systems
(e.g., the Bear, Provo, Green, and Colorado) that provide di-
verse avian habitats. Aboriginals occupying camp sites and
villages established along the larger rivers, mountain streams,
or lakes had available to them not only the aquatic species

Parmalee: Prehistoric Utah Avifauna

239

Figure 1. Location of Archaic (triangles) and Fremont (circles) sites
from which avian specimens were obtained and examined for this
study. Deadman Cave (1), Stansbury I and II (2), Sandwich Shelter
(3), Black Rock II and III (4), Bear River No. 1 and No. 3 (S), Levee
(6), Knoll (7), Warren (8), Injun Creek (9), Unnamed (10), Nephi (11),
Evans Mound (12), Pharo Village (13).

that inhabited these bodies of water and the adjacent marsh-
lands or riparian habitat, but also upland birds (e.g., Sage
Grouse) that occurred in the bordering dry brushlands and
desert. Faunal assemblages from archaeological sites in Utah
reflect the Indian’s exploitation of these varied habitats or
biomes.

Environmental changes that have occurred since portions
of northern and western Utah were covered with ancient Lake
Bonneville are difficult to define and, thus far, none of major
impact have been reflected in the animal remains recovered
from archaeological sites. Durrant (1970:241, 245) comments
on this problem in describing the mammalian fauna from the
Archaic Hogup Cave site (Box Elder County, northwestern
Utah) and discusses factors that affect interpretation of ar-
chaeologically derived faunal material: “Based upon osteolog-
ical remains representing 3,440 individual animals, it is evi-
dent that the mammalian fauna of the Hogup Cave area
throughout Neothermal time was remarkably uniform and
similar to that found there today. This indicates that, with
some fluctuations, there existed a certain similarity in envi-
ronmental factors throughout these past nine millenia. The
data lead me to consider that the climate of the Hogup area
during Neothermal time was somewhat cooler in the early
Anathermal period, then became gradually warmer, reaching
a moderately high temperature during the Altithermal, and
then gradually cooled through the Medithermal to the present.

In addition to the lack of significant qualitative differences
within the fauna over time, another factor contributing to un-
certainty in interpretation is that, although it is known that
the cave vicinity was an autumnal harvesting area for its ab-
original occupants, no one knows either the range of these
people in their hunting forays and migrations or the number
of persons involved. Moreover, no data are available on the
cyclic patterns of the mammals of that time, and certain mam-
mals occur together in the cave deposits that occupy somewhat
discrete ranges at present. Inferences concerning past climatic
conditions in the area of Hogup Cave cannot be made easily
from the data currently at hand. The tendency of man to en-
gage in selective hunting and gathering and to transport items
long distances greatly complicates interpretation. Certainly the
deposits offer little evidence for sudden and dramatic changes
in either vegetation or fauna with the onset of the Altithermal.”

In the faunal assemblage from Hogup Cave and other Ar-
chaic sites, e.g., Danger Cave (Jennings 1957), remains of un-
gulates, especially the Pronghorn (Antilocapra americana
(Ord)), and rabbits (Lepus Linnaeus and Sylvilagus Gray) pre-
dominated. Although the Bison, Bison bison (Linnaeus), was
taken by Archaic peoples, it was not until the later Fremont
cultural period that it became the dominant protein source.
Also, as Jennings (1978:233) has pointed out, . . all the local
variations in favored game do not obscure what seems to be
a Fremont preference for mule deer where ever it is available.
The adaptability in choice of game may then be toward sub-
stitution when deer is rare — a reasonable and expectable ad-
justment.” The Archaic populations of the Great Basin appear
to have been geared to a mobile hunting-collecting way of life,
while the Fremont people, an apparent cultural derivative
from the earlier Desert Archaic, were more sedentary and part-
ly or perhaps predominantly agricultural. Although hunting
and trapping techniques may have varied between the Archaic
and Fremont peoples, both groups relied heavily on the en-
demic fauna, especially mammals, as a primary food resource.
From a strictly “pounds of derived meat” point of view, birds
must be considered as a supplemental food resource in the
total food economy of these people, but undoubtedly one that
was of periodic significance.

MATERIAL AND METHODS

A combined total of about 5300 bird bones from 16 archae-
ological sites were examined and, of this number, 5043 or
about 95 percent were identifiable to family, genera, and/or
species. At least 75 species from 21 families occurred in the
combined samples. Avian osteological collections housed in the
Zoology Section, Illinois State Museum, Springfield, and in
the Zooarchaeology Section, Department of Anthropology,
The University of Tennessee, Knoxville, were used in connec-
tion with this study. Utah Binds by Behle and Perry (1975)
and Birds of Utah by Hayward et al. (1976) are cited as the
latest comprehensive authorities on the known distribution
and abundance of birds in the state.

Analysis of archaeologically derived avian bone samples
often must be conservative because of the innumerable vari-
ables and unknown factors affecting each sample and each
site. In attempting to compare past avian assemblages and
their use by aboriginal man, in this case between Archaic and
Fremont peoples, the inequality of sample size may be a sig-

240

Parmalee: Prehistoric Utah Avifauna

Table 2. Birds identified from Archaic and Fremont Sites in Utah, with number of specimens and minimum number of individuals (given in
parentheses).

Species

Deadman

Cave

42T064

Stans-

bury

I

42TOl

Stans-

bury

II

42T02

Sand-

wich

Shelter

42TO

108

Black

Rock

II

42T029

Black

Rock

III

42T03

Bear
River
No. 1
42B055

Bear
River
No. 3
42B098

Family Podicipedidae — Grebes
Eared Grebe, Podiceps nigricollis
Western Grebe, Aechmorphorus occidentalis
Pied-billed Grebe, Podilymbus podiceps
Grebe sp.

2 (1)

24 (13)

30 (6)

287 (46)
122

16 (10)

1 (1)
1 (1

1 (1)

Family Pelecanidae — Pelicans
White Pelican, Pelecanus erythrorhynchos

3 (1)

6 (1)

13 (4)

133 (10)

F amily Phalacrocoracidae — Cormorants
Double-crested Cormorant, Phalacrocorax auritus

1 (1)

Family Ardeidae — Herons and Bitterns

Great Blue Heron, Ardea herodias

2 (1)

1 (1)

3 (1)

Little Blue Heron, Egretta caerulea

1 (1)

Common Egret, Egretta alba

1 (1)

3 (1)

8 (2)

Snowy Egret, Egretta thula

1 (1)

Black-crowned Night Heron, Nycticorax nycticorax

1 (1)

3 (1)

American Bittern, Botaurus lentiginosus

9 (4)

15 (5)

Heron sp.

3 (1)

Family Plataleidae — Ibises and Spoonbills

White-faced Ibis, Plegadis chilli

Family Anatidae — Swans, Geese and Ducks

Whistling Swan, Olor columbianus

3 (1)

Trumpeter Swan, Olor buccinator

2 (1)

Swan, Olor sp.

2 (1)

6 (3)

Canada Goose, Branta canadensis

103 (13)

1 (1)

2 (1)

1 (1)

3 (1)

18 (3)

59 (8)

Snow Goose, Chen caerulescens

244 (26)

1 (1)

2 (2)

2 (1)

25 (5)

Ross’ Goose, Chen rossii

Goose, sp.

146

3

20

37

Mallard, Anas platyrhynchos, and/or Black Duck,

Anas rubripes

79 (24)

1 (1)

1 (1)

9 (3)

2 (1)

35 (9)

97 (28)

Gad wall, Anas strepera

2 (1)

4 (4)

2 (1)

Pintail, Anas acuta

7 (3)

1 (1)

8 (6)

9 (4)

Mallard/Black Duck/Pintail group, Anas spp.

1 (1)

8 (2)

2 (1)

60 (12)

48 (7)

Green-winged Teal, Anas crecca

2 (2)

16 (9)

1 (1)

16 (10)

22 (9)

Blue-winged Teal, Anas discors, and/or

Cinnamon Teal, Anas cyanoptera

1 (1)

2 (1)

5 (3)

3 (2)

Teal, Anas sp.

3 (2)

20 (5)

22 (5)

24(3)

Wigeon, Anas cf. americana

2 (1)

6 (5)

2 (2)

Shoveler, Anas clypeata

2 (1)

7 (2)

4 (2)

PWood Duck, Aix sponsa

Redhead, Aythya americana

6 (2)

1 (1)

Ring-necked Duck, Aythya collaris, and/or

Lesser Scaup, Aythya affinis

13 (2)

7 (2)

Canvasback, Aythya valisineria

3 (1)

6 (2)

Duck, Aythya sp.

Goldeneye, Bucephala sp.

2 (1)

Bufflehead, Bucephala albeola

2 (1)

2 (1)

2 (1)

Ruddy Duck, Oxyura jamaicensis

1 (1)

1 (1)

3 (1)

1 (1)

Duck sp.

85

2

17

3

79

183

Hooded Merganser, Lophodytes cucullatus

cf. Common Merganser, Mergus merganser

1 (1)

15 (4)

cf. Red-breasted Merganser, Mergus serrator

3 (2)

Merganser, Mergus sp. 18 (3)

Family Accipitridae — Hawks and Harriers
cf. Red-tailed Hawk, Buteo jamaicensis
Swainson’s Hawk, Buteo swainsoni, and/or

Rough-legged Hawk, Buteo lagopus 7 (2) 2 (1) 1 (1)

Parmalee: Prehistoric Utah Avifauna

241

Table 2. Continued.

Per-

cent

Levee

42BO107

Knoll

42BO109

Warren

42WB-

Injun

Creek

42WB34

Un-

named

42SL19

Nephi

42JB2

Evans Pharo
Mound Village
42IN40 40MD180

Total No.
Specimens

of

Speci-

mens

500 (86)

9.92

2 (2)

2 (1)

1 (1)

1 (1)

366 (82)

7.26

1 (1)

0.02

10 (2)

11 (3)

0.22

122

2.42

242 (27)

4.80

69 (5)

IS (3)

1 (1)

1 (1)

1 (1)

242 (27)

4.80

22 (5)

0.44

20 (3)

1 (1)

22 (5)

0.44

115 (37)

2.29

6 (3)

0.12

1 (1)

0.02

11 (2)

23 (6)

0.46

8 (5)

9 (6)

0.18

4 (2)

0.08

43 (7)

1 (1)

68 (17)

1.35

1 (1)

4 (2)

0.08

1 (1)

0.02

1 (1)

1 (1)

0.02

3347 (611)

66.40

1 (1)

2 (1)

6 (3)

0.12

2 (1)

1 (1)

5 (3)

0.10

8 (4)

0.16

29 (7)

12 (3)

13 (3)

69 (11)

2 (2)

312 (54)

6.19

114 (18)

38 (7)

8 (2)

6 (2)

1 (1)

441 (65)

8.74

3 (2)

10 (5)

13 (7)

0.26

28

6

6

54

2

302

5.99

94 (27)

16 (5)

10(3)

23 (5)

4 (1)

11 (3)

2 (1)

21 (4)

405 (116)

8.03

11 (4)

5 (2)

1 (1)

1 (1)

26 (14)

0.52

25 (9)

11 (4)

2 (2)

7 (3)

70 (32)

1.39

61 (10)

6 (2)

13 (7)

31 (8)

2 (1)

9 (2)

241 (53)

4.78

45 (13)

1 (1)

4 (3)

1 (1)

4 (1)

112 (50)

2.22

13 (7)

2 (1)

2 (1)

1 (1)

2 (1)

9 (4)

40 (22)

0.79

117 (14)

2 (1)

4 (2)

192 (32)

3.81

5 (4)

1 (1)

1 (1)

1 (1)

18 (15)

0.36

22 (9)

4 (2)

1 (1)

1 (1)

41 (18)

0.81

2 (2)

2 (2)

0.04

12 (4)

1 (1)

1 (1)

2 (1)

23 (10)

0.46

20 (5)

3 (1)

59 (12)

4 (2)

1 (1)

107 (25)

2.12

5 (2)

2 (1)

16 (6)

0.32

24 (10)

1 (1)

2 (1)

27 (12)

0.54

2 (1)

3 (3)

7 (5)

0.14

7 (3)

6 (4)

19 (10)

0.38

14 (4)

1 (1)

1 (1)

1 (1)

1 (1)

24 (12)

0.48

223

11

122

5

8

22

760

15.07

1 (1)

1 (1)

0.02

2 (2)

18 (7)

0.36

2 (2)

1 (1)

6 (5)

0.12

71 (17)

4 (2)

4 (1)

2 (1)

3 (1)

2 (2)

1 (1)

105 (28)

2.08

67 (34)

1.34

1 (1)

1 (1)

0.02

4 (1)

7 (2)

21 (7)

0.42

Continued

242

Parmalee: Prehistoric Utah Avifauna

Species

Hawk, Buteo sp.

Hawk, sp.

Golden Eagle, Aquila chrysaetos
Bald Eagle, Haliaeetus leucocephalus
Eagle sp.

Marsh Hawk, Circus cyaneus

Family Falconidae — Falcons
Prairie Falcon, Falco mexicanus
Peregrine Falcon, Falco peregrinus
Falcon, Falco sp.

Kestrel, Falco sparverius

Family Tetraonidae — Grouse and Ptarmigan
Blue Grouse, Dendragapus obscurus
Ruffed Grouse, Bonasa umbellus
Sharp-tailed Grouse, Pedioecetes phasianellus
Sage Grouse, Centrocercus urophasianus
Grouse sp.

Family Gruidae — Cranes
Sandhill Crane, Grits canadensis

Family Rallidae — Rails, Gallinules & Coots
Sora, Porzana Carolina
cf. Purple Gallinule, Porphyntla martinica
American Coot, Fulica americana

Family Charadriidae — Plovers and Turnstones
Black-bellied Plover, Pluvialis squatarola

Family Scolopacidae — Snipe and Sandpipers
Common Snipe, Capella gallinago
Long-billed Curlew, Numenius americanus
Willet, Catoptrophorus semipalmatus
cf. Greater Yellowlegs, Tringa melanoleucus

Family Recurvirostridae — Avocets and Stilts
American Avocet, Recurvirostra americana
Black-necked Stilt, Himantopus mexicanus

Family Stercorariidae — Jaegers and Skuas
cf. Parasitic Jaeger, Stercorarius parasiticus

Family Laridae — Gulls and Terns
cf. California Gull, Larus californicus
Gull, Larus sp.

cf. Caspian Tern, Hydroprogne caspia

Family Columbidae — Pigeons and Doves
Passenger Pigeon, Ectopistes migratorius
Mourning Dove, Zenaidura macroura

Family Strigidae — Owls
Great Horned Owl, Bubo virginianus
Burrowing Owl, Speotyto cunicularia
Spotted Owl, Strix occidentalis
Long-eared Owl, Asio otus, and/or
Short-eared Owl, Asio flammeus

Family Picidae — Woodpeckers
Common Flicker, Colaptes auratus

Family Corvidae — Jays, Magpies and Crows
Black-billed Magpie, Pica pica
Common Raven, Counts corax
Clark’s Nutcracker, Nucifraga Columbiana

Table 2. Continued.

Stans-

Stans-

Sand-

wich

Black

Black

Bear

Bear

Deadman

bury

bury

Shelter

Rock

Rock

River

River

Cave

I

II

42TO

II

III

No. 1

No. 3

42T064

42TOl

42T02

108

42T029

42T03

42B055

42B098

2(1) 1(1)

1(1) 1(1)

1 (1)

7(2) 2(2) 1(1)

6 (2) 1 (1)

3 (2)

5(2) 9(2) 1(1)

4 (1)

3(2) 3(1)

10(2) 1(1) 9(3) 1(1)

7 2

30 (4)

1 (1)

4 (3)

1 (1)

1(1) 1(1) 6(3) 25(5)

2 (1)
1 (1)

4 (1)

1 0)

4(1) 4(1) 4(1)

1 (1)

4 (1)

2(1) 6(2) 3(1) 1(1)

1 0)

1 (1)
1 (1)

7 (2)
1 (1)

1 (1)

3 (2)

2 (1)

2 (1)

3 (1)

1 (1)

5 (3)

5 (3)

1 (1)

1 (1)

1 (1)

12 (5)

1 (1)

7 (3)

8 (2)

1 (1)
22 (4)

1 (1)
4 (2)

2 (1)

Parmalee: Prehistoric Utah Avifauna

243

Table 2. Continued.

Per-

cent

Levee

2BO107

Knoll

42BO109

Warren

42WB-

Injun

Creek

42WB34

Un-

named

42SL19

Nephi

42JB2

Evans Pharo
Mound Village
42IN40 40MD180

Total No.
Specimens

of

Speci-

mens

1 (1)

4 (3)

4 (2)

9 (6)

0.18

1 (1)

1 (1)

2 (1)

7 (5)

0. 14

1 U)

2 (1)

1 (1)

6 (5)

0.12

1 (1)

0.02

1

1

0.02

5 (1)

1 (1)

1 (1)

3 (1)

1 (1)

21 (9)

0.42

23 (14)

0.46

1 (1)

1 (1)

2 (2)

0.04

2 (2)

9 (5)

0.18

1 U)

3 (2)

2 (1)

1 (1)

10 (7)

0.20

2 (1)

2 (1)

0.04

287 (50)

5.70

17 (4)

2 (1)

76 (10)

110 (20)

2.18

9 (4)

13 (5)

0.26

1 (1)

1 (1)

5 (1)

1 (1)

3 (1)

12 (3)

29 (11)

0.58

4 (1)

8 (2)

11 (2)

6 (2)

50 (14)

0.99

3

2

11

2

58

85

1.69

37 (7)

0.73

1 (1)

6 (2)

37 (7)

0.73

146 (37)

2.90

1 (1)

1 (1)

0.02

6 (2)

6 (2)

0.12

76 (11)

4 (1)

2 (2)

2 (1)

4 (2)

17 (6)

139 (34)

2.76

2 (1)

0.04

2 (1)

0.04

21 (15)

0.42

1 (1)

0.02

2 (1)

2 (2)

1 (1)

1 (1)

14 (9)

0.28

2 (1)

1 (1)

3 (2)

0.06

2 (2)

3 (3)

0.06

32 (12)

0.63

16 (S)

2 (2)

31 (11)

0.61

1 (1)

1 (1)

0.02

1 (1)

0.02

1 (1)

0.02

28 (12)

0.56

4 (1)

0.08

2 (1)

1 (1)

7 (2)

22 (9)

0.44

1 (1)

2 (2)

0.04

4 (4)

0.08

1 (1)

0.02

2 (2)

3 (3)

0.06

54 (26)

1.08

4 (1)

1 (1)

S (1)

3 (1)

28 (11)

0.56

1 (1)

2 (1)

4 (3)

0.08

1 (1)

1 (1)

0.02

4 (1)

2 (1)

1 (1)

21 (11)

0.42

4 (4)

0.08

1 (1)

4 (4)

0.08

96 (33)

1.91

2 (1)

1 (1)

2 (1)

7 (5)

0.14

3 (1)

2 (1)

9 (3)

6 (2)

11 (2)

87 (27)

1.73

2 (1)

2 (1)

0.04

Continued

244

Parmalee: Prehistoric Utah Avifauna

Table 2. Continued.

Sand-

Stans-

Stans-

wich

Black

Black

Bear

Bear

Deadman

bury

bury

Shelter

Rock

Rock

River

River

Cave

I

II

42TO

II

III

No. 1

No. 3

Species

42T064

42T01

42T02

108

42T029 42T03

42B055

42B098

Family Icteridae — Meadowlarks, Orioles

and Blackbirds

Western Meadowlark, Sturnella neglecta
Yellow-headed Blackbird, Xanthocephalus
xanthocephalus

Red-winged Blackbird, Agelaius phoeniceus

1 (1)

1 (1)

1 (1)

1 0)

Order Passeriformes — Perching Birds

Indet. Passerines

5 (4)

TOTAL

792 (116)

28 (16)

59 (20)

443 (64)

170 (62)

21 (8)

387 (98)

763 (123)

nificant factor. Little or nothing is known of early hunting
techniques. One group may have devised a method of captur-
ing grebes while another learned how to efficiently hunt peli-
cans, yet both birds may have been present as a potentially
abundant food resource. The role birds played in the total
subsistence economy is often problematical; Sharrock and
Marwitt (1967:39) and others have commented that, as a
group, birds were at best of only secondary importance as a
food resource in comparison to mammals. It would require a
large number of grebes or ducks to equal the actual number
of kilos of usable meat derived from an adult deer or bison,
for example. Nevertheless, the value of birds as a constant or
seasonal supplemental food resource cannot be discounted.
Although all species of birds are edible, individual or tribal
preference for or against the taking of a particular species or
group of birds is yet another factor affecting an interpretative
analysis of any given faunal sample.

The bird bones examined were well preserved and in a great
many instances complete. However, intraspecific osteological
variation because of sex, age, or individual variation often
make species determinations uncertain or impossible. For this
reason, and because some elements were broken and/or non-
diagnostic, many identifications could not be accurately car-
ried beyond a general group level (e.g., Duck sp.; Hawk sp.;
Duck, Aythya sp.; Table 2). Identification of similar sized
specimens of closely related species within a particular genus
is also difficult and often limited, depending on the elements
with which one must work.

Remains of swans, geese, and ducks totaled 3347, 66 percent
of all elements identified (Table 3). In addition to the “usual”
problems of identification, the high incidence of hybridization
among members of the Anatidae and other families (e.g., the
Parulidae) may further complicate attempts to arrive at some
species determinations. Johnsgard (1960:25) has commented
that . . waterfowl of the family Anatidae have provided the
greatest number and variety of bird hybrids originating from
both natural and captive conditions.” Not only have fertile
hybrids resulted between species within the same genus (e.g.,
Mallard x Pintail), but also between species of different genera
(Mallard X Common Merganser). I know of no study on the
osteology of hybrid ducks and geese. It is not inconceivable

that some “problem” waterfowl elements from aboriginal sites
could well have come from hybrids. In spite of certain basic
identification problems and the use of tentative determinations
in some instances, interesting and useful data have come to
light concerning the overall use of birds in the food economy
of aboriginal man in Utah.

ACCOUNTS OF SPECIES
Family Podicipedidae — Grebes

The contrast in the utilization of grebes between Archaic
and Fremont peoples who occupied sites bordering the Great
Salt Lake is striking. About 32 percent of all bird remains
from the Archaic sites were those of grebes, the majority of
elements occurring in Sandwich Shelter (Table 2). A total of
287 bones (46 individuals) were identified as the Eared Grebe,
PocLiceps nigricollis Brehm, a common summer resident in
marshes along the east side of Great Salt Lake (Behle and
Perry 1975). The 122 indeterminate grebe elements, which are
probably also those of P. nigricollis, bring the total number
of grebe bones from this one site to slightly over 400. In con-
trast, only 19 grebe elements were recovered from all 11 Fre-
mont sites. The reason(s) for this apparent differential use of
grebes between cultural groups is unclear, as is the paucity of
remains of the Pied-billed Grebe, Podilymbus podiceps (Lin-
naeus), and the Western Grebe, Aechmorphorus occidentalis
(Lawrence) — two species that are also common summer resi-
dents in the Great Salt Lake.

Family Pelecanidae — Pelicans

Elements of the White Pelican, Pelecanus erythrorhynchos
Gmelin, comprised 6.5 percent of all remains from the Fre-
mont site samples, but less than 1 percent of those from the
Archaic sites. The large size of this bird would presumably
have made it a desirable food resource, yet less than 30 indi-
viduals are represented in the combined faunal assemblages.
It is currently a common summer resident of the Great Salt
Lake with a breeding colony at Gunnison Island (Behle and
Perry 1975). None of the elements were from nestlings or fledg-
lings.

Parmalee: Prehistoric Utah Avifauna

245

Table 2. Continued.

Per-

cent

Injun Un- Evans Pharo of

Levee Knoll Warren Creek named Nephi Mound Village Total No. Speci-

42BO107 42BO109 42WB- 42WB34 42SL19 42JB2 42IN40 40MD180 Specimens mens

9 (8)

0.18

4 (4)

0.08

3 (2)

0.06

2 (2)

0.04

5 (4)

0.10

5 (4)

0.10

124 3 (233) 154 (48) 82 (37) 449(82) 33 (17) 101(35) 60 (20) 258 (51) 5043 (1030) 100.1

F amily Phalacrocoracidae — Cormorants

Behle and Perry (1975) list the Double-crested Cormorant,
Phalacrocorax auritus (Lesson), as an uncommon summer res-
ident in northern Utah, and a transient and rare winter visitant
throughout the state. If this species occurred in greater num-
bers in prehistoric times, the Indian made little use of it. Re-
mains of P. auritus occurred in only three sites: one element
each at the Deadman Cave and Injun Creek sites and 20 (three
individuals) at the Levee site.

Family Ardeidae — Herons and Bitterns

Six species representative of this family were identified from
the faunal samples; 90 percent of the remains occurred in Bear
River Nos. 1 and 3 and Levee, sites once located on the marshy
shore of Great Salt Lake. Specimens of the Great Blue Heron
(Ardea herodias Linnaeus), Snowy Egret (Egretta thula (Mo-
lina)), Black-crowned Night Heron (Nycticorax nycticorax
(Linnaeus)), and American Bittern ( Botaurus lentiginosus
(Rackett)), reported as common summer residents in northern
Utah by Behle and Perry (1975), are not unexpected at sites
once located in habitat well suited for these wading birds. A
proximal right humerus from Black Rock II compared closely
with the Little Blue Heron, Egretta caerulea (Linnaeus), a
species of only occasional occurrence in Utah. The Common
Egret, E. alba (Linnaeus), is considered a rare transient, but
remains of this large, showy species were recovered at four
sites and represented a minimum of six individuals. Laybourne
(1967) reported four elements of E. alba from Bear River No.
2. Although of potential food value, these birds may also have
been prized especially for their plumage. Three bones from a
nestling heron or bittern occurred in the faunal sample from
Deadman Cave.

Family Plataleidae — Ibises and Spoonbills

The White-faced Ibis, Plegadis chihi (Vieillot), today is a
“Common summer resident in Great Salt Lake marshes”
(Behle and Perry 1975), and in view of this the recovery of
only one specimen, a complete right humerus from the Levee
site, is surprising. Hayward et al. (1976:46) mentioned a com-
ment by Allen (1872:172), who stated that the White-faced Ibis

was reported to have “. . . become numerous only during the
last two or three years,” but no reason for its apparent increase
was offered. It is evident that the Indian hunted the marshes
for herons, bitterns, and other semi-aquatic species, and it
seems unlikely that there would have been a taboo against
taking this ibis, so the single record may suggest that P. chihi
was a rare species in the vicinity of the Great Salt Lake during
early prehistoric times.

Family Anatidae — Swans, Geese, and Ducks

At least 22 species of waterfowl were represented in the
faunal samples, and their remains made up about 66 percent
of the total sample. Elements of these birds constituted ap-
proximately 51 percent of all avian remains from the Archaic
sites and 73 percent from the Fremont sites (Tables 2 and 3).
Bones of waterfowl constituted 66 percent of the avifauna re-
ported by Parmalee (1970) from Hogup Cave, an Archaic site
located about 25 km west of the Great Salt Lake, and 85
percent of the bird remains identified from the Bear River No.
2 site (Fremont) by Laybourne (1967) were those of waterfowl.
It can be presumed, on the basis of these percentages and the
variety and number of species they represent, that waterfowl,
especially geese and ducks, were often hunted and formed a
valuable supplement in the food economy of these people.

Although the Whistling Swan, Olor columbianus (Ord), oc-
casionally occurs in large concentrations in marshes adjacent
to the Great Salt Lake, and the Trumpeter Swan, Olor buc-
cinator Richardson, was formerly more common (now occa-
sional) in northern Utah (Behle and Perry 1975), the Indians
who occupied these areas rarely took either species. Remains
of one or both swans were identified from five of the Fremont
sites, and both are recorded by Laybourne (1967) from Bear
River No. 2, but no more than six elements were identified
from any one site. Elements of geese, however, were especially
numerous and the number of geese specimens represented 21
percent of the total. A large subspecies of the Canada Goose,
Branta canadensis moffitti Aldrich, is a common resident of
the Great Salt Lake (Behle and Perry 1975) and at least three
other races occur in Utah as transients. With the possible ex-
ception of the giant Canada Goose, B. c. maxima Delacour,
and Hutchins’ Goose, B. c. hutchinsii (Richardson), it is im-

246

Parmalee: Prehistoric Utah Avifauna

Table 3. Families of birds represented in avian samples from 16 Utah archaeological sites.

Family

No. of
Species

No. of
Specimens

Percent of
Specimens

Minimum No.
of Individuals

Archaic

Fre-

mont

Archaic

Fre-

mont

Archaic

Fre-

mont

Archaic

Fre-

mont

Podicipedidae: Grebes

2

3

481

19

32.24

0.54

76

10

Pelecanidae. Pelicans

1

1

9

233

0.60

6.56

2

25

Phalacrocoracidae: Cormorants

1

1

1

21

0.07

0.59

1

4

Ardeidae: Herons, Bitterns

3

6

9

106

0.60

2.99

4

31

Plataleidae: Ibises, Spoonbills

—

1

—

1

—

0.03

—

1

Anatidae: Swans, Geese, Ducks

12

22

765

2582

51.27

72.71

107

504

Accipitridae: Hawks, Eagles, Harriers

4

4

19

48

1.27

1.35

8

26

Falconidae. Falcons

1

3

9

14

0.60

0.39

4

10

Tetraonidae: Grouse

4

4

48

239

3.22

6.73

13

37

Gruidae: Cranes

1

1

30

7

2.01

0.20

4

3

Rallidae: Rails, Gallinules, Coots

1

3

3

143

0.20

4.03

3

34

Charadriidae: Plovers, Turnstones

—

1

—

2

—

0.06

—

1

Scolopacidae: Snipe, Sandpipers

1

3

8

13

0.54

0.37

4

11

Recurvirostridae: Avocets, Stilts

1

2

5

27

0.34

0.76

2

10

Stercorariidae: Jaegers, Skuas

1

—

1

—

0.07

—

1

—

Laridae: Gulls, Terns

2

2

17

11

1.14

0.31

7

5

Columbidae: Pigeons, Doves

2

1

2

2

0.13

0.06

2

2

Strigidae: Owls

3

4

23

31

1.54

0.87

12

14

Picidae: Woodpeckers

1

1

3

1

0.20

0.03

3

1

Corvidae: Jays, Magpies, Crows

2

3

51

45

3.42

1.27

16

17

Icteridae: Meadowlarks, Blackbirds

1

3

3

6

0.20

0.17

3

5

Passeriformes: Family Indeterminate

2?

—

5

—

0.34

—

4

—

TOTALS

46

69

1492

3551

100.00

100.02

276

751

possible to separate these forms or races osteologically. Eight
elements (3 to 4 individuals) from an extremely large race of
B. canadensis (Linnaeus) compared closely with those of B.
c. maxima (Injun Creek site). Seven other specimens of geese
from this site and one from Unnamed site are probably B. c.
hutchinsii. It is of interest to note that elements of the Snow
Goose, Chen caerulescens (Linnaeus), were more numerous
than those of the Canada Goose (441 versus 312); the former
species is presently considered an uncommon transient in the
state. Thirteen bones of Ross’ Goose, Chen rossii (Cassin), re-
ported by Hayward et al. (1976) as a casual although regular
migrant through L>tah, were identified from the Levee and
Knoll sites.

Approximately 45 percent of all identified avian remains,
representing a minimum of 17 species, were those of ducks.
Identification problems involving duck elements have been
discussed. Osteological similarities between species such as the
Cinnamon Teal (Anas cyanoptera Vieillot) and Blue-winged
Teal ( Anas discors Linnaeus), Mallard (Anas platyrhynchos
Linnaeus) and Black Duck (Anas rubripes Brewster), and
Lesser Scaup (Aythya ajfinis (Evton)) and Ring-necked Duck
(Aythya collaris (Donovan)), to cite just a few examples,
prompted the combination of some elements of certain closely
related species under general categories (e.g., Mallard-Black
Duck-Pintail group). None of the identified species of ducks
is unusual with regard to current abundance or distribution in
the Great Salt Lake and Bear River Refuge marshes.

Family Accipitridae — Hawks,

Eagles, and Harriers

Raptors belonging to this family were poorly represented.
Although remains of at least five species were identified, the

number of hawk and eagle elements accounted for less than
2 percent of the total sample. The significance of these birds
to prehistoric aboriginal groups that once occupied this region
is unknown, although ethnographic data indicate that rapto-
rial birds were of considerable symbolic and ceremonial sig-
nificance to historic groups (e.g., the Hopi: Fewkes 1900). Ea-
gle trapping was a well established tradition among most tribes
of the Great Plains as well as several others in the Southwest.
During such hunts, where the birds were grabbed by hand by
a concealed hunter when the hawk or eagle attempted to take
strategically placed bait, numerous hawks and eagles were
usually captured. In a study of approximately 3100 avian re-
mains from 51 South Dakota Arikara sites, a total of nearly
1300 elements (around 43 percent) were identified as those of
hawks and eagles (Parmalee 1977). Although several species
of Buteo hawks, the Marsh Hawk, Circus cyaneus (Linnaeus),
and the Golden Eagle, Aquila chrysaetos (Linnaeus), occur
commonly over much of Utah, for whatever reason they ap-
pear to have rarely been captured by the prehistoric inhabit-
ants of the area.

Family Falconidae — Falcons

Osteological similarities between the Prairie Falcon, Falco
mexicanus Schlegel, and the Peregrine Falcon, Falco peregri-
nus Tunstall, often limit the ability reach accurate species de-
terminations; incomplete elements further complicate the
problem. For these reasons several elements of these falcons
were recorded as Falco sp. (Table 2). Nine of the 2 1 falcon
specimens compared most closely with F. peregrinus, yet only
two could be identified as Prairie Falcon. Behle and Perry
(1975:15) list the Prairie Falcon as a common permanent res-
ident in Utah today, and the Peregrine Falcon as “formerly a

Parmalee: Prehistoric Utah Avifauna

247

permanent resident . . . but present status essentially a rare
transient.” Hayward et al. (1976:67) comment that some of the
early investigators in Utah considered the Peregrine Falcon to
be “rather common.” As was the case with representatives of
the Accipitridae, few falcons were taken by aboriginals inhab-
iting these sites.

Family Tetraonidae — Grouse and Ptarmigan

Bones of four species of grouse made up approximately 6
percent of all identifiable avian remains. Specimens of grouse,
not unlike the broken elements of ducks, often defy species
determination. Although all four species are now considered
by Behle and Perry (1975) as uncommon permanent residents
in Utah, the Sharp-tailed Grouse, Pedioecetes phasianellus
(Linnaeus), and Sage Grouse, Centrocercus uropliasianus (Bo-
naparte), were formerly more widespread and abundant. Ele-
ments of grouse occurred in 11 of the 16 sites, but were most
numerous in those sites such as Nephi and Pharo Village that
were located in more open desert areas.

Family Gruidae — Cranes

Remains of the Sandhill Crane, Grus canadensis (Linnaeus),
occurred in only three sites. The paucity of specimens is sur-
prising since this bird was reported as formerly a common
summer resident in northern Utah (Hayward et al. 1976:74).
The distal end of a left humerus and proximal ulna of a Sand-
hill Crane had been cut from the shaft by the “groove-and-
snap” technique (Parmalee 1976:152, Fig. 76); the shafts of
major wing elements from large birds were often modified for
the manufacture of whistles and other bone tube instruments.
Except for an incomplete coracoid from Deadman Cave that
is referable to the Little Brown Crane, Grus c. canadensis, all
other elements were from birds of the large race, G . c. tabida
(Peters).

Family Rallidae — Rails, Gallinules and Coots

Specimens of the American Coot, Fulica americana Gmelin,
were the most numerous and occurred in 11 of the 16 sites.
Only at the Levee site, however, can elements of this species
be considered numerous (76 bones from a minimum of 11 in-
dividuals). In light of the summer abundance of this species
at the Great Salt Lake and the relative ease with which it can
be taken, 1 1 individuals appears to be a small number for such
a potential food resource.

Behle and Sperry (1975) note only two verified records of
the Purple Gallinule, Porphynda martinica (Linnaeus), for
Utah and consider its status as accidental. It is of interest,
therefore, that two individuals of this species were identified
on the basis of six elements (two incomplete humeri, paired
distal ends of tibiotarsi, complete left femur, and coracoid)
recovered at the Levee site.

Family Charadriidae — Plovers and Turnstones

A complete left tarsometatarsus and distal one-third of a
tibiotarsus of the Black-bellied Plover, Pluvialis squatarola
(Linnaeus), a common transient through Utah during spring
and fall migrations, were the only elements recovered (Bear
River No. 3) of species belonging to this family.

Family Scolopacidae — Snipe and Sandpipers

The rather large group of birds generally termed “shore-
birds” appear to have been of little importance to prehistoric
Indian groups inhabiting northern Utah. Although elements
of four species, the Common Snipe ( Capella gallinago (Lin-
naeus)), Long-billed Curlew ( Numenius americanus Bech-
stein), Greater Yellowlegs (Tringa melanoleucus (Gmelin)), and
the Willet ( Catoptrophorus semipalmatus (Gmelin)), were iden-
tified from nine sites during this study, no more than three
individuals were represented at any one site (Table 2). Behle
and Perry (1975) state that the Long-billed Curlew was for-
merly a common summer resident and transient. In view of
its large size and apparent availability, one might surmise that
this species was a valuable supplemental food resource, but
such was not the case.

Family Recurvirostridae — Avocets and Stilts

Both the American Avocet, Recurvirostra americana Gme-
lin, and Black-necked Stilt, Himantopus mexicanus (Muller),
are considered by Behle and Perry (1975) as common summer
residents in northern Utah and transient throughout the state.
Remains of the American Avocet occurred at six sites and the
maximum number of individuals represented at any one site
(Levee) was five. Like species belonging to the Charadriidae
and Scolopacidae, these birds were seldom or rarely taken by
the Indian. The Black-necked Stilt was represented in the fau-
nal samples by only a single element, a nearly complete right
humerus from the Levee site.

Family Stercorariidae — Jaegers and Skuas

A complete left carpometacarpus, tentatively identified as
a parasitic Jaeger, Stercorarius parasiticus (Linnaeus), oc-
curred in the faunal sample from Black Rock II. This species
is reported by Behle and Perry (1975) as an occasional visitor
in late summer and early fall; the majority of specimens that
have been observed or collected occurred in the vicinity of the
Bear River Refuge.

Family Laridae — Gulls and Terns

Elements of gulls were recovered from eight sites. One
species, the California Gull, Larus californicus Lawrence, was
identified from four bones from Deadman Cave. A furculum
and scapula of a gull (Larus californicus ? or L. delawarensis ?
Ord) from Sandwich Shelter exhibited butchering cuts, indi-
cating that the bird had been processed by the inhabitants.
The Caspian Tern, Hydroprogne caspia (Pallas), an uncom-
mon summer resident in northern Utah, was represented by
a single element at Deadman Cave and the Knoll site. As a
group, gulls and terns appear to have been of only minor im-
portance; all elements combined amounted to less than 1 per-
cent of the total.

Family Columbidae — Pigeons and Doves

Doves appear to have been of little or no value to the Indians
who occupied northern Utah, judging by the paucity of their
remains encountered at archaeological sites. Although the
Mourning Dove, Zenaidura macroura (Linnaeus), is common
throughout Utah during the summer months, it was repre-

248

Parmalee: Prehistoric Utah Avifauna

sented by only three elements in the 16 faunal samples, two
humeri from the Nephi site and one from Stansbury II. Har-
grave (1970) reported a single humerus from the Sand Dune
Cave collections. The recovery of a partial left humerus (miss-
ing distal end) of a Passenger Pigeon, Ectopistes migratorius
(Linnaeus), from the Stansbury II site is especially noteworthy.
Neither Behle and Perry (1975) nor Hayward et al. (1976) list
the Passenger Pigeon as a former inhabitant of Utah. Although
Schorger (1973: Fig. 22) provides casual or accidental records
for several western states (Wyoming, Idaho, Montana, Ne-
vada), there are none for Utah. The specimen from the Stans-
bury II site apparently represents the first record of Passenger
Pigeon from the state.

Family Strigidae — Owls

As a group, owls were of special significance to a large num-
ber of aboriginal people in North America, not particularly as
a food resource but as symbols of the supernatural, of strength
and other desirable qualities, and of death and as group to-
tems. Sperry (1957) lists owl feathers from Danger Cave, Har-
grave (1970) records them from Sand Dune Cave, and elements
of several species of owls have been reported from Hogup
Cave (Parmalee 1970), Bear River No. 2 (Lavbourne 1967),
and other archaeological sites in Utah. Remains of at least four
species of owls were encountered in 12 of the 16 avian samples
examined during this study. The Great Horned Owl, Bubo
virginianus (Gmelin), was the most numerous (28 pieces, a
minimum of 1 1 individuals). Elements of the Short-eared Owl,
Asio flammeus (Pontoppidan), and/or Long-eared Owl, Asio
otus (Linnaeus), both common permanent residents in Utah,
were also numerous (21 specimens representing 11 individu-
als). Of interest was the recovery of a nearly complete left
femur of the Spotted Owl, Strix occidentalis (Xantus), from
Pharo Village, a bird reported by Behle and Perry (1975) as
being a rare permanent resident in Utah. Two elements of the
Great Horned Owl exhibited butchering marks: the distal end
of a humerus from Stansbury II (removal of the outer wing)
and the shaft and intercotylar process of a tarsometatarsus
(removal of the lower leg) from Sandwich Shelter. Removal of
these outer limb elements suggests their possible use in cere-
monial functions or as decorative items, as evidenced by finds
of these bones as human burial accouterments in other regions
(Parmalee 1967).

Family Picidae — Woodpecker

Only one species of woodpecker, the Common Flicker, Co-
laptes auratus (Linnaeus), was represented in the avifaunas
from four of the 16 sites, and then only one element from each.
One Common Flicker element was recovered from Hogup
Cave (Parmalee 1970); Hargrave (1970) records six feathers of
this bird from Sand Dune Cave and mentions a find of eight
rectrices from Cave Dupont, Kane County, Utah, by Nus-
baum (1922). The significance of woodpecker feathers for var-
ious forms of decoration and the use of “stuffed” skins (possibly
as symbolic objects) has been demonstrated by the recovery of
such remains from Lovelock Cave, Nevada (Loud and Har-
rington 1929) and other archaeological sites in the Southwest
(Hargrave 1970). The incorporation of feathers, including
those of woodpeckers, as decoration in Porno basketry is well
known (Barrett 1908). The paucity of osteological remains of

woodpeckers from the sites studied suggests that they were of
little importance to these people, although 10, or possibly 11
species of woodpeckers are known to occur in Utah (Behle and
Perry 1975).

Family Corvidae — Jays, Magpies, and Crows

Although three species of corvids were represented in the
samples, those of the Common Raven, Corvus corax Linnaeus,
were the most numerous and occurred in 12 of the 16 sites.
The raven was esteemed or considered by many aboriginal
groups as a bird possessing certain supernatural powers or
symbolic traits and consequently it often served as a clan to-
tem. Feathers, various body parts, and whole skins were worn
or carried. These artifacts, occasionally buried with their
owner, have been encountered (skulls, wing and leg elements)
as burial accouterments (Ubelaker and Wedel 1975). In a study
of bird remains from Arikara sites in South Dakota, I reported
that elements of corvids made up approximately 15 percent of
the 3100 bones examined (Parmalee 1977); those of ravens
amounted to 10 percent of the total. Clark’s Nutcracker, Nu-
cifraga Columbiana (Wilson), and the Black-billed Magpie,
Pica pica (Linnaeus), both common permanent residents
throughout most of Utah, were represented by only six indi-
viduals. The paucity of remains of these two species and the
total lack of Common Crow, Corvus brachyrhynchos Brehm,
elements, a common winter visitant, suggests that the Com-
mon Raven was the only corvid of significance to these people.

Family Icteridae — Meadowlarks, Orioles,
and Blackbirds

Remains of three species belonging to this family were iden-
tified from the avian samples, but the small number of bones
recorded (nine pieces, eight individuals) suggests that, as a
group, these birds were taken only occasionally. Whether they
represent a minor food supplement in the diet or perhaps a
source of decorative feathers is a matter of speculation. Har-
grave (1970) reported finding pieces of skin and feathers (the
red, buff, and black wing coverts) of an adult male Red-
winged Blackbird, Agelaius phoeniceus (Linnaeus), at Sand
Dune Cave.

Family Indeterminate

Five indeterminate passerine elements, two of which were
incomplete right humeri of a small fringilid (?), were recovered
at Sandwich Shelter. The significance of these, or any small
passerine birds to aboriginal groups who once inhabited this
region, is difficult to evaluate. The apparent scarcity of osteo-
logical remains in sites might imply a general disregard for
these small birds, yet their poor representation in most avi-
faunas may also be attributed to archaeological field excava-
tion/recovery techniques. Until fairly recent times, 'A-inch
(approximately 0.64-cm) hardware cloth was used to screen
the soil being removed during excavation, and elements of
small vertebrates simply passed through such coarse screens
and were lost. There are exceptions, of course, and one of the
most notable is based on the analysis of feathers from Hogup
Cave reported by Baldwin (1970). Of the 13 species of birds
identified from feathers recovered at Hogup Cave, those of the
Gray-crowned Rosy Finch, Leacosticte tephrocotis (Swain-

Parmalee: Prehistoric Utah Avifauna

249

son), were the most numerous (58 percent of the total) and
occurred in 15 of the 16 excavated strata.

The fact that various passerines were taken for their skins
or plumage has been well documented in the ethnographic
literature. The strip of skin and feathers of a Rufous-sided
Towhee, Pipilo erythrophthalmus (Linnaeus), found as a
“choker” at the neck of a human infant in a burial taken from
the Catfish Canyon Site, Glen Canyon area (Hargrave 1960)
provides an interesting example. However, unmodified re-
mains of small birds recovered in cave site deposits, for ex-
ample those of Horned Lark, Eremophila alpestris (Linnaeus),
reported by Sperry (1957) from Danger Cave and others iden-
tified from Hogup Cave, may represent prey individuals taken
by raptors, which also periodically use cave sites as roosts
(Parmalee 1970).

SUMMARY AND CONCLUSIONS

In order to reach what might be considered an accurate
interpretation of any archaeological faunal sample, the almost
limitless number of variables that may have affected that sam-
ple— such as preservation factors, length of site occupation,
number of occupants, season(s) of occupation, occupants’
hunting methods and preferences for certain animals, per-
centage of the site sampled, and field recovery techniques —
must be evaluated. Ethnographic accounts of hunting tech-
niques and food preparation methods, for example the com-
munal duck and coot hunt of the Shoshone described by Lowie
(1924), can provide a useful correlation with the osteological
record. However, the applicability of ethnographic data in
interpreting prehistoric archaeological faunas may indeed be
questionable. Therefore, interpretation of an osteological sam-
ple must take into account as many of the variables as might
be considered applicable. In spite of the unknowns, the iden-
tification of species, the number of remains of each, and an
estimate of the number of individuals represented can prove
indicative of the relative importance of a species or group of
animals in the social and/or economic life of a people.

The identification of approximately 5050 bird bones from 5
Archaic and 11 Fremont sites located in northern and western
Utah has shown a relatively consistent utilization of some
avian groups (e.g., geese and ducks) over several thousand
years. In contrast, other species were apparently not consis-
tently used. Grebe elements made up nearly 33 percent of all
avian remains from the Archaic sites, but less than 1 percent
of the Fremont samples. Bones of the White Pelican, on the
other hand, totaled about 7 percent of the remains from the
Fremont sites, but less than 1 percent from the Archaic. Al-
though there were more than twice the number of elements
from the Fremont sites, it is doubtful that this difference in
sample size is a factor in explaining such discrepancies. Nor
are differences in seasonal occupation of the sites: both the
Eared Grebe and White Pelican are common summer residents
and, therefore, would have been available to both cultural
groups. Perhaps hunting methods or a preference for other
species were factors affecting the taking of one species and not
another.

All of the species represented in the 16 archaeological sam-
ples, except the Passenger Pigeon, either still occur in Utah or
represent birds previously reported from the state. Thirteen of
the 16 sites were located along or near the Great Salt Lake

and Bear River marshes. As one might expect, remains of
aquatic and semi-aquatic birds made up the majority of ele-
ments (approximately 90 percent in both Archaic and Fremont
samples). When concentrations of waterfowl occurred on the
Great Salt Lake, its marshes, and the rivers draining into it,
the aboriginal peoples occupying these areas realized a valu-
able food resource that could be harvested with minimum ef-
fort for maximum return. The Indian had to be an opportunist
in obtaining food; although preference was certainly a factor,
abundance and availability of any given species or group of
animals would have affected his procurement efforts as would
a sufficient return for the amount of energy expended.

Elements of fledgling or juvenile birds were, as a whole,
rare in the avian samples. Two young Ravens were repre-
sented in the material from Deadman Cave and several im-
mature duck bones occurred in the Levee site sample. The
largest number of elements of juvenile birds (15) were re-
covered at the Evans Mound site and included bones of Sage
Grouse, Coot, hawk (Buteo sp. ?), and Raven. This indicates
a late spring/early summer occupation of this site. Osteological
evidence from some aboriginal sites, for example the Emery-
ville Shellmound, San Francisco Bay, California, has shown
that the inhabitants purposely hunted nestling birds, in this
instance, cormorants (Howard 1929). In addition to the
“groove-and-snap” Sandhill Crane wing bone ends described
from Pharo Village, two others removed by the same method
(distal right ulna of Bald Eagle, Haliaeetus leucocephalus (Lin-
naeus); proximal end of a right ulna of a Great Horned Owl)
were found in the Deadman Cave material. Three limb bone
shafts (goose humerus and tibiotarsus, eagle ulna) from which
the ends had been removed were recovered at the Injun Creek
site. These and other examples of scored or otherwise modified
elements, special utilization of feathers and skins, and the in-
ternment of body parts with the dead are evidence that birds
played an important role, in addition to subsistence, in the
social and ceremonial activities of prehistoric Utah inhabit-
ants.

ACKNOWLEDGMENTS

I express my appreciation to Jesse D. Jennings for the loan
of the avian materials examined during this study; and to him
and his Research Assistant, Frank Hull, I also owe a debt of
gratitude for their patience in supplying site information and
other pertinent data as well as reference materials. Examina-
tion of several of the faunal samples was initiated in 1972
when I was associated with the Illinois State Museum, Spring-
field; during this time the late Alexander Wetmore generously
provided species determinations for several problem elements.
I am also indebted to Amadeo M. Rea for identification assis-
tance with certain bones and for the loan of comparative spec-
imens.

For her assistance in the recording of data and the prepa-
ration of Table 2, I gratefully acknowledge my wife Barbara.
Special appreciation is extended to Betty Creech for typing the
manuscript and to Arthur E. Bogan for the preparation of
Figure 1.

LITERATURE CITED

Aikens, C.M. 1966. Fremont-Promontory- Plains relation-
ships including a report of excavations at the Injun Creek

250

Parmalee: Prehistoric Utah Avifauna

and Bear River Number 1 sites, northern Utah. Univ.
Utah Anthro. Papers 82:1-102.

Allen, J.A. 1872. Notes of an ornithological reconnaissance
of portions of Kansas, Colorado, Wyoming, and Utah.
Bull. Mus. Comp. Zool., Harvard 3(6): 1 13-183.
Baldwin, D. 1970. Bird Feathers from Hogup Cave. Pp.
267-269 (Appendix VI) in Hogup Cave by C.M. Aikens,
Univ. Utah Anthro. Papers 93.

Barrett, S.A. 1908. Porno Indian basketry. Univ. Calif.

Publ. Amer. Arch, and Ethnol. 7(3): 133-308.

Behle, W.H., and M.L. Perry. 1975. Utah birds: check-
list, seasonal and ecological occurrence charts and guides
to bird finding. Utah Mus. Nat. Hist., Univ. Utah, Salt
Lake City. 142 pp.

Berry, M.S. 1972. The Evans site. Spec. Report Dept. An-
thro., Univ. Utah: 1-299.

Durrant, S.A. 1970. Faunal remains as indicators of neo-
thermal climates at Hogup Cave. Pp. 241-245 (Appendix
II) in Hogup Cave by C.M. Aikens. Univ. Utah Anthro.
Papers 93.

Enger, W.D., Jr. 1942. Archeology of Black Rock 3 Cave,
LTtah. Univ. Utah Anthro. Papers 7:84-104.

Enger, W.D., Jr., and W. Blair. 1947. Crania from the
Warren mounds and their possible significance to north-
ern periphery archaeology. Amer. Antiq. 13(2): 142-146.
Fewkes, J.W. 1900. Property-right in eagles among the
Hopi. Amer. Anthropologist 2(4):690— 707 .

Hargrave, L.L. 1960. Identification of archaeological feath-
ers from Glen Canyon, Utah. Pp. 239-241 (Appendix II)
in 1958 excavations, Glen Canyon area by W.D. Lipe.
Univ. Utah Anthro. Papers 44.

. 1970. Feathers from Sand Dune Cave: a Basket-

maker cave near Navajo Mountain, Utah. Mus. N. Ariz.,
Tech. Ser. 9:1-55.

Hayward, C.L., C. Cottam, A.M. Woodbury, and H.H.
Frost. 1976. Birds of Utah Great Basin Naturalist
Memoirs No. 1. Brigham Young Univ. Press, Provo. 229

pp.

Howard, H. 1929. The avifauna of Emeryville shellmound.

Univ. Calif. Publ. Zool. 32(2):301— 394.

Jameson, S.J.S. 1958. Archeological notes on Stansbury Is-
land. Univ. Utah Anthro. Papers 34:1-46.

Jennings, J.D. 1957. Danger cave. Univ. Utah Anthro. Pa-
pers 27: 1-328.

. 1978. Prehistory of Utah and the eastern Great Ba-
sin. Univ. Utah Anthro. Papers 98:1-263.

Johnsgard, P.A. 1960. Hybridization in the Anatidae and
its taxonomic implications. Condor 62( 1):2 5— 33 .

Judd, N.M. 1954. The material culture of Pueblo Bonito.

Smithsonian Misc. Collections, Vol. 124. 239 pp.

Kelly, I T. 1964. Southern Paiute ethnography. Univ. Utah
Anthro. Papers 69 (Glen Canyon Ser. No. 21): 1-194.
Laybourne, R.C. 1967. Unworked bird bone identification.
P. 58 (Table 6) in Excavations at Snake Rock village and
the Bear River No. 2 site by C.M. Aikens. Univ. Utah
Anthro. Papers 87.

Loud, L.L., and M R. Harrington. 1929. Lovelock Cave.
Univ. Calif. Publ. Amer. Arch, and Ethnol. 25(1): 1-183.

Lowie, R.H. 1909. The northern Shoshone. Anthro. Papers
Amer. Mus. Nat. Hist. 2(2): 165-306.

— . 1924. Notes on Shoshonean enthnography. Anthro.

Papers Amer. Mus. Nat. Hist. 20(3): 185—3 14.

Mandelbaum, D.G. 1940. The Plains Cree. Amer. Mus.
Nat. Hist. Anthro. Papers 37(2): 155-316.

Marwitt, J.P. 1968. Pharo Village. Univ. Utah Anthro.
Papers 91:1-84.

Marwitt, J.P. , G.F. Fry, and J.M. Adovasio. 1971. Sand-
wich Shelter. Univ. Oreg. Anthro. Papers 1:27-36.

Matthews, W. 1877. Ethnography and philology of the
Hidatsa Indians. U.S. Geol. and Geograph. Survey,
Misc. Publ. No. 7, Washington. 239 pp.

Nusbaum, J.L. 1922. A Basket Maker Cave in Kane County
Lfiah. Indian Notes and Monographs, Mus. Amer. Indi-
an, Heye Foundation 29:1-153.

Parmalee, P.W. 1967. Additional noteworthy records of
birds from archaeological sites. Wilson Bull. 79(2): 155—
162.

. 1970. Birds from Hogup Cave. Pp. 263-266 (Appen-
dix Vj in Hogup Cave by C.M. Aikens, Univ. Utah An-
thro. Papers 93.

. 1976. A general summary of the vertebrate fauna

from Cahokia. Pp. 137-155 in Perspectives in Cahokia
archaeology (James Brown, Ed.). 111. Arch. Survey Bull.
Llrbana. No. 10.

. 1977. The avifauna from prehistoric Arikara sites in

South Dakota. Plains Anthropologist 22(77): 189-222.

. 1979. A preliminary report on the birds from the

Levee and Knoll sites, Box Elder county, Utah. Pp. 1 10—
113 (Appendix III) in The Levee site and the Knoll site
by G.F. Fry and G.F. Dailey, Univ. Utah Anthro. Papers
100.

Schorger, A.W. 1973. The passenger pigeon; its natural
history and extinction. Univ. Okla. Press, Norman. 424

pp.

Sharrock, F.W., and J.P. Marwitt. 1967. Excavations at
Nephi, Utah, 1965-1966. Univ. Utah Anthro. Papers
88:1-60.

Shields, W.F., and G.F. Dalley. 1978. The Bear River
No. 3 site. Misc. Paper 22, Univ. Utah Anthro. Papers
99:57-99.

Smith, E.R. 1952. The archaeology of Deadman Cave,
Utah. Univ. Utah Anthro. Papers 10:1-41.

Sperry, C.C. 1957. Feathers recovered from Danger Cave.
Pp. 305-306 (Appendix C) in Danger Cave by J.D. Jen-
nings, Univ. Utah Anthro. Papers 27.

Steward, J.H. 1933. Ethnography of the Owens Valley
Paiute. Univ. Calif. Publ. Amer. Arch, and Ethnol.
33:233-340.

Ubelaker, D. , and W.R. Wedel. 1975. Bird bones, burials
and bundles in Plains archaeology. Amer. Antiq.
40(4):444-452.

Weisel, G.F. 1952. Animal names, anatomical terms, and
some ethnozoology of the Flathead Indians. Jour. Wash.
Acad. Sciences 42(1 1):345 —3 5 5 .

A FOSSIL PLAIN WANDERER (AVES: PEDIONOMIDAE) FROM
FIRE-HOLE DEPOSITS, MORWELL, SOUTHEASTERN
VICTORIA, AUSTRALIA

By Pat Vickers Rich1 and A.R. McEvey2

ABSTRACT: Lacustrine deposits previously thought to be of mid-Miocene age from the Morwell

open cut mine, southeastern Victoria, Australia, have produced the partial skeleton of a Plain Wanderer
(Aves: Pedionomidae) indistinguishable from the living Pedtonomus torquatus Gould. This indicates either
a very young age for the Morwell Fire-hole No. 2 sediments (contrary to the mid-Miocene dating based
on pollen analysis) or a very slow rate of evolution within the Pedionomidae. The former hypothesis is
favored.

Lacustrine sediments of disputed age from Fire-hole No. 2,
State Electricity Commission Open Cut Coal Mine at Morwell,
Victoria (Australia) have produced a partial skeleton of a Plain
Wanderer (Aves: Pedionomidae) indistinguishable from the
living Pedionomus torquatus.

Both avian and marsupial skeletons were contained in finely
laminated dark grey clays that formed a lenticular body, prob-
ably the remains of a small pond or lake. The depression in
which deposition took place apparently formed when the early
Miocene (Douglas and Ferguson 1976) brown coal of the Mor-
well Formation (Morwell 1A Seam) caught fire and burned in
a restricted area. This steep-sided basin must have filled with
water and served as a natural trap for animals that chanced
to fall in, possibly, in the case of the kangaroos, through a
vegetal mat that may have covered part of the pond’s perim-
eter (T. Rich pers. comm.).

Analysis of the pollen (including Triporopollenites bellus)
collected from these lacustrine sediments suggests a middle to
late Miocene age (A. Partridge, ESSO, Sydney; pers. comm ),
distinctly younger than the Morwell Formation, also palyno-
logicallv dated. The marsupial fossils, on the other hand, in-
cluding two species of kangaroos (Macropus titan ( =giganteus )
and Protemnodon anak\ T. Rich and T. Flannery pers.
comm.), are typical of Pleistocene-aged assemblages. Macro-
pus titan might possibly extend into the Pliocene of western
Victoria (Buninyong; T. Rich pers. comm.), but this would be
the maximum age documented for this species. The following
paper evaluates the partial skeleton of the Plain Wanderer
from Morwell in light of this conflicting evidence.

Abbreviations used below are as follows: NMV, National
Museum of Victoria, Melbourne; SAM, South Australian Mu-
seum, Adelaide.

1 National Museum of Victoria and Earth Sciences Department, Mon-
ash University, Clayton, Victoria.

2 National Museum of Victoria, Melbourne, Victoria, Australia.

SYSTEMATICS

Order Ralliformes (Reichenbach)

Family Pedionomidae Gadow

DIAGNOSIS: The Morwell fossil bird was assigned to the
Pedionomidae because it exhibits the following combination
of characters: Sternum with (1) a single sternal notch either
side of the midline that extends about half the length of this
element; (2) straight posterior lateral processes of equal width
over their entire length.

Synsacrum with (1) foramina between vertebrae not well
developed; (2) morphology broad, flat and not elongate, only
slightly longer than wide; (3) anterior iliac crest, particularly
near anterior end, prominent and separate from anterior blade
of ilium, although the two nearly meet; (4) sacral vertebrae
broadly expanded in comparison to remainder of vertebrae
associated with synsacrum; (S) small anti-trochanter; (6) three
or four vertebrae fused into synsacral complex posterior to
sacrals (i.e., the synsacral caudals); (7) illioischiatic fenestra
forming a small oval, not greatly elongate, but decidedly larger
(at least three times the area) than acetabulum; (8) in ventral
view only four parapophvses attaching to ilium anterior to
sacral vertebrae, including one pair on anteromost vertebra,
which is not completely fused into synsacral complex; (9) only
one sacral parapophysis quite prominent, attaching onto ven-
tral part of vertebral column; (10) no narrow ridge or distinct
haemal processes on ventral parts of synsacral thoracic ver-
tebrae; (11) most prominent pair of parapophvses of sacral
vertebrae forming large acute angle with vertebral column in-
stead of right angle.

Ulna (1) elongate and slender; with (2) shaft curved, not
straight, particularly at proximal end; (3) shaft not distinctly
compressed, but triangular in cross section with rounded
edges, particularly at midpoint; (4) secondary papillae not
prominent; (5) olecranon small and short, not prominent with
palmar borders of internal and external cotyla not existing far

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 330:251-255.

252

Rich and McEvey: Fossil Plain Wanderer

H

Figure 1. Pedionomus torquatus (NMV P39121), Fire-hole deposits, Morwell, Victoria: A partial skeleton in matrix including sternum (st),
scapula (sc), and fragments of vertebrae and ribs. Stereographic pairs: B, E, tibiotarsus (lateral view); C, F, ulna (proximal fragments, palmar
view); D, G, ulna (distal fragment, anconal view); I, H, femur (anterior view). See Table 1 for scale.

palmad of shaft; (6) distinct proximal radial depression absent,
as is any great pneumatization in this area; (7) shaft surface
near proximal end low and rounded palmarly, with only slight
indication of median ridge; (8) prominence for anterior artic-
ular ligament low, not pronounced; (9) carpal tuberosity not
prominent; (10) carpal tuberosity merging with shaft at about
level where external condyle originates; (11) lateral margin of
external condyle nearly parallel with lateral margin of shaft.

Femur with (1) trochanter well developed, deep; (2) prox-
imal end not wide, but laterally compressed; (3) anterior and

posterior borders of proximal end concave in proximal view,
not straight and/or parallel; (4) proximal margin of trochanter,
especially posterior half, recurved and overhanging iliac facet;
(5) trochanter protruding anteriad of head; (6) trochanter
slightly convex laterally, not highly convex, in proximal view;
(7) trochanter not expanding much beyond anterior margin of
shaft, lying nearly parallel to shaft in anterior or posterior
views; (8) trochanter rising to marked peak, in lateral view,
rather than being smoothly rounded; (9) trochanter extending
farthest proximad just slightly anterior of its midpoint, in lat-

Rich and McEvey: Fossil Plain Wanderer

253

B C

Figure 2. Pedionomous torquatus (NMV P39121), Fire-hole deposits, Morwell, Victoria: A, left lateral view of synsacrum; B, C, stereographic
pair, dorsal view, synsacrum. See Table 1 for scale.

eral view; (10) iliac facet and head highly concave proximally,
not flat; (11) proximal end lacking any projection anteriad
along margin between trochanter and head; (12) shaft lacking
deep excavation just internal to trochanteric ridge; (13) shaft
lacking pneumatization on proximal end.

Tibiotarsus poorly preserved; (1) shaft compressed antero-
posteriorly; (2) cnemial crests and rotular crests of moderate
proximal extension; (3) interarticular area between external
articular surface and rotular crest with relatively deep exca-
vation; although there are no characters that exclude it from
the Pedionomidae, the tibiotarsus has few diagnostic charac-
ters.

Genus Pedionomus Gadow

GENERIC DIAGNOSIS: As for family, only genus in fam-
ily.

Pedionomus torquatus Gould

MATERIAL: NMV P39121, partial skeleton including: par-
tial sternum (left half), ribs, distal fragments of left scapula,
vertebral fragments, fused synsacral vertebrae, synsacral frag-
ments, proximal end of right ulna, proximal end of left femur
and midsection of left femur (not articulated), proximal end of
left tibiotarsus (see Figs. 1-2). Found by Thomas Darragh.
For measurements see Table 1.

LOCALITY: Fire-hole No. 2, State Electricity Commission
Morwell Open Cut Coal Mine, Southeastern Victoria.

STRATIGRAPHIC HORIZON: Lacustrine sediments
overlying Morwell Seam lA (early Miocene) and below the
Haunted Hill Gravels (Jenkin 1968). Age uncertain, lying be-
tween early Miocene and the pre-late Pleistocene.

DESCRIPTION AND COMPARISON: Sternum: Sternal
notch slightly deeper, or posterior lateral process more elon-
gate, or possibly both, than in P. torquatus ; posterior lateral
processes elongate and narrow with nearly straight lateral
margin; intermuscular line prominent (variable in modern Pe-
dionomus).

Scapula: Elongate, narrow, and parallel-sided over much of
its length; of same size as in P. torquatus, except more robust,
especially at the distal end.

Synsacrum: Two pairs of transverse processes anterior to
main sacral vertebral attachment to synsacrum present in ven-
tral view; prominent groove present along ventral midline be-
tween last lumbar and first sacral transverse processes; median
dorsal ridge quite prominent; ventral border of synsacrum
nearly straight in lateral view, not concave ventrally; synsa-
crum deepest at anterior end, narrowing posteriorly; arrange-
ment of obturator foramen, acetabulum and ilio-ischiatic fe-
nestrae as in P. torquatus, with the latter largest and ovoid in
shape; synsacrum broad posterior to acetabulum as in Recent
P. torquatus.

254

Rich and McEvey: Fossil Plain Wanderer

Table 1. Measurements (in mm) of Recent and fossil Pedionomidae.

Measurements

Pedionomus
torquatus
Morwell Victoria
NMV P39121

Recent

Pedionomus torquatus
(n = 2)

STERNUM

Length of sternal notch measured along internal side of left
posterior lateral process

10.8 +

9.6-11.2

Anterior width of posterior lateral process

-2.8

2. 8-3. 5

SYNSACRUM

Total length of vertebral component of synsacrum, measured
along dorsal surface

6.6

6. 3-6. 5

Maximum width across vertebral column just posterior to
transverse process of first sacral vertebra

4.2

4. 2-4. 7

Diameter of right acetabulum

-2.4

1.8-1. 9

Maximum measurement across right ilioischiatic fenestra

~4.6

5. 1-5.5

Maximum measurement across right obturator foramen

-1.2

1.6-1. 7

ULNA

Proximal width

4.2

3. 6-3. 9

Proximal depth

2.9

2. 0-2. 9

Length from proximal end of olecranon to distal end of
proximal radial depression

4.2

4. 6-4. 7

Depth of external condyle

2.4 +

2. 4-2. 7

Distal width

3.4

3.1

FEMUR

Proximal width

4.5

4. 4-4. 6

Depth of trochanter

3.2

3. 1-4.1

Depth of head

1.9

2.0

TIBIOTARSUS

Distance from external articular surface to distal end of
fibular crest

10.8

9. 2-9. 4

Depth across external articular surface to base of
external cnemial crest

4.1

3.9-4. 1

Width of shaft at base of fibular crest

2.5

2.4

Depth of shaft at base of fibular crest

1.6

1.5-2. 5

Ulna: No appreciable differences from P. torquatus, al-
though distal end fragmentary.

Femur: Although differences exist between the proximal
ends of some femora of Recent P. torquatus and the Morwell
specimen, it lies within the range of variability found in living
P. torquatus ; the shaft shape compares closely with that of the
living species, the only visible difference being a slightly great-
er anteroposterior flexure in lateral view, being convex ante-
riorly rather than straight; direct connection between the prox-
imal and distal segments of the femur cannot be established,
but the two fragments are very probably from the same bone.

Tibiotarsus: Comparison very limited because of erosion of
proximal end, but appears similar to P. torquatus. See Bock
and McEvey (1969) for a thorough, complete description of
the skeleton of the living P torquatus.

In summary, then, the Morwell Pedionomus is the same size
as the living/5, torquatus and differs only in that (1) the sternal
notch is slightly deeper, (2) the ventral border of the synsacrum
is not as curved (concave ventrally), and (3) the shaft of the
femur may have a slightly greater flexure. Because our sample

of living P. torquatus is so small and the differences noted
above only slight, we believe there is no reason to propose a
new species for the fossil material. The reasoning is strength-
ened by the fact that several skeletal elements are represented,
and all show only minor differences, if any, from/5, torquatus.

DISCUSSION AND CONCLUSIONS

Detailed comparisons of several skeletal elements of the fos-
sil Plain Wanderer from Morwell with those of Recent Pe-
dionomus torquatus show few differences, either qualitative
or quantitative. The differences noted are insufficient to define
a new species. The close similarity between the fossil specimen
and Recent P. torquatus suggests two possibilities. The first
is that the lacustrine sediments at Morwell are very young.
This is because within avian groups with known lengthy rec-
ords in Australia, such as the Aegothelidae (Rich and McEvey
1977) and Phoenicopteridae (Miller 1963), pre-Pleistocene
forms show significant differences from extant birds. This is
also true in areas where the record of fossil birds is far better

Rich and McEvey: Fossil Plain Wanderer

255

known, especially in North America and Europe. The second
possibility is an extremely slow rate of evolution within the
Pedionomidae in comparison with other avian groups. No liv-
ing species of bird anywhere in the world is known to extend
farther back in time than the late Pliocene, approximately
three million years ago.

In summary, we favor the hypothesis that the fossil Plain
Wanderer from Morwell is of a late Pliocene or younger age.
The presence of the fossil pedionomid is, likewise, suggestive
of nearby grasslands during the time of deposition of the Fire-
hole sediments as this is the ecological zone occupied by the
living P. torquatus.

Pedionomus torquatus (SAM P126718) has previously been
reported from Pleistocene deposits of Victoria Cave in South
Australia (van Tets and Smith 1974). Although this synsacral
fragment is similar to P. torquatus, it is likewise similar to
several species of Charadriiformes. It differs from/5, torquatus
in having a narrow, prominent ridge on the ventral surface of
the two anterior synsacral thoracic vertebrae and in being
slightly more concave ventrallv over the posterior half of the
synsacral vertebrae. We hesitate to assign it to any taxonomic
group until more material is available, because the specimen
lacks so many of the characters diagnostic for the Pedionom-
idae.

ACKNOWLEDGMENTS

We thank T.H. Rich, T. Flannery, A. Partridge, and J. van
Tets for discussing aspects of this manuscript; and the M.A.

Ingram Trust, Utah Mining, and Danks Trust for their sup-
port of the osteological research leading to this publication; P.
Thomas, U. Gawronski, C. Armstrong, and K. Mott for typ-
ing; and F. Coffa for expertly photographing a difficult spec-
imen.

LITERATURE CITED

Bock, W.J., and A.R. McEvey. 1969. Osteology of Pe-
dionomus torquatus (Aves: Pedionomidae) and its allies.
Proc. Roy. Soc. Victoria 82(2): 18 2-232.

Douglas, J.G., and J. A. Ferguson. (Eds.). 1976. Geology
of Victoria. Geol. Soc. Australia, Sp. Publ. 5. 528 pp.
Jenkin, J.J. 1968. The geomorphology and upper Cainozoic
geology of southeast Gippsland, Victoria. Geol. Sur. Vic-
toria. Mem. 27:1-147.

Miller, A. 1963. The fossil flamingoes of Australia. Condor
65(4):289 — 299.

Rich, P.V., and A. McEvey. 1977. A new owlet-nightjar
from the early to mid-Miocene of eastern New South
Wales. Mem. Nat. Mus. Victoria 38:247-253.
van Tets, G.F., and M.J. Smith. 1974. Small fossil ver-
tebrates from Victoria Cave, Naracoorte, South Australia
III. Birds (Aves). Trans. Roy. Soc. So. Australia
98(4): 2 25 — 2 2 8 .

PASSENGER PIGEON BONES FROM ARCHAEOLOGICAL

SITES IN NEW MEXICO

By L.L. Hargrave1 and S.D. Emslie2

ABSTRACT: Three bones of the Passenger Pigeon, Ectopistes migratorius (Linnaeus), were recovered

from archaeological sites in Taos and San Juan Counties, New Mexico. The elements represent at least
two individuals and could date as early as A.D. 975. These records are the first Holocene records of E.
migratorius in New Mexico and increase this species’ known distribution in prehistoric western United
States.

The extinct Passenger Pigeon, Ectopistes migratorius (Lin-
naeus), was a species whose preferred habitat was the forested
areas of eastern North America (Ridgway 1916; Schorger
1955). There are only a few known records of this pigeon from
western states; these include Idaho (Burleigh 1972), Nevada
(Linsdale 1951), Washington (Jewett et al. 1953), and Wyo-
ming (Ridgway 1916).

Howard (1937) described six elements representing at least
two individuals of Ectopistes migratorius from the late Pleis-
tocene deposits of Rancho La Brea. These specimens repre-
sented the first record of the Passenger Pigeon from California,
and the first fossil record of this species from the western
United States. A second fossil record from the western United
States, also reported by Howard (1971), consisted of one bone
from Pleistocene deposits in Dark Canyon Cave, Eddy Coun-
ty, New Mexico. Regarding the date of deposits in this cave,
Howard (pers. comm.) states, “The Avifauna from Dark Can-
yon Cave certainly suggests a Late Pleistocene date of depo-
sition. The three best represented species are Coragyps occi-
dentalis (Miller), Gymuogyps ampins Miller, and Caracara
prelutosa (Howard), which are characteristic of the Rancho La
Brea avifauna, and have been found at other Late Pleistocene
localities in the United States and Mexico.”

MATERIAL

The specimens of Ectopistes migratorius reported on here
were recovered from the archaeological sites of Una Vida in
Chaco Canyon, San Juan County, and Picuris Pueblo (San
Lorenzo), Taos County (Fig. 1). The specimens are:

Una Vida: right humerus complete, No. C265 (length:
43.0 mm), Room 46, floor. Left ulna with proximal end
fragmented, No. C271 (approximate length: 47.3 mm),
Room 65.

Picuris Pueblo: left tibiotarsus with the ends fragmented,
No. 1192, TA III, Area III Test Pit C, bottom layer.

1 Former professor of ethnobiology, Prescott Center College, Prescott,
Arizona 86301, now deceased.

2 Department of Biological Sciences, Northern Arizona University,
Flagstaff, Arizona 86001.

The Passenger Pigeon bone from Picuris Pueblo is currently
housed at Adams State College, Alamosa, Colorado 81102 (in
the care of Herbert Dick); the two bones from Una Vida are
housed at the Chaco Center, University of New Mexico, P.O.
Box 26176, Albuquerque, New Mexico 87125 (in the care of
James Judge).

Una Vida is a large classic Chaco town of the Anasazi cul-
ture, Rooms 46 and 65 are contemporaneous and were dated
by tree rings, architectural style, and ceramic typology to with-
in the period A.D. 950-1030, with a more probable range of
A.D. 975-1030 (Gordon Vivian pers. comm.). Picuris Pueblo
has been continuously occupied since A.D. 1250. Ceramic evi-
dence indicates the provenience in which the bone was found
dates at A.D. 1300-1350 (Herbert Dick pers. comm.).

IDENTIFICATION

Pigeons other than Ectopistes migratorius whose remains
might occur in archaeological sites in the western United States
include the Band-tailed Pigeon, Columba fasciata Say, the
Red-billed Pigeon, C. flavirostris Wagler, and the domestic
Rock Dove, C. livia Gmelin. The latter species was introduced
into North America but may occur intrusively in a prehistoric
site and is especially important to consider here as Picuris
Pueblo is also a historic site that is still inhabited. In addition,
Hargrave identified bones of the Band-tailed Pigeon at Picuris
Pueblo.

Comparisons were made with 26 adult skeletons of Ecto-
pistes migratorius , 17 of Columba fasciata, 20 of C. livia, and
six of C. flavirostris. Five skeletons of the White-crowned Pi-
geon, C. leucocephala Linnaeus, were also included in the
comparisons. E. migratorius is distinguished from the species
of Columba by having humerus with stockier shaft; deltoid
crest more rounded, ectepicondylar papilla higher on shaft;
and external condyle more elongate.

Ulna with a small depression on proximal intercotylar ridge
(this depression is absent in Columba flavirostris and there is
a small projection on the ridge in C. leucocephala)', and bicip-
ital attachment placed closer to impression of M. brachialis
anticus.

Contrib. Sci. Natur. Hist. Mus. Los Angeles County. 1980. 330:257-260.

258

Hargrave and Emslie: Passenger Pigeons

ll

10

9

8

7

6

Figure 1. Archaeological specimens of Ectopistes migratorius from
New Mexico (left to right): left tibiotarsus from Picuris Pueblo (No.
1192), right humerus from Una Vida (No. C265), left ulna from Una
Vida (No. C271). (Scale in cm.)

Tibiotarsus with shaft narrower; and fibular crest less prom-
inent.

In addition to the generic characters, the humerus of Ec-
topistes migratorius may be distinguished from those of Co-
lumba livia and C. fasciata by size (Table 1). It is probable
that all wing long bones of E. migratorius are smaller, and do
not equal or exceed the length of the same elements of C . livia
and C. fasciata. The specimens from Una Vida and Picuris
agree with characters listed for£. migratorius.

DISCUSSION

Locations of Pleistocene and Holocene records of Ectopistes
migratorius in the western United States are illustrated in Fig-
ure 2. Ectopistes migratorius was known to occur regularly
only as far west as “along the Missouri River to eastern Mon-
tana and to western Texas (Frio Canyon, Tom Green County,
1881)” (Ridgway 1916:336). The western records are:

California: Six elements representing at least two individuals
from Pleistocene deposits at Rancho La Brea, Los Angeles
County (Howard 1937).

Idaho: One skin (USNM No. 22006) collected at Pack River
by C.B. Kennedy on 17 June 1860. This record was mistak-
enly listed by Ridgway (1916) as being from eastern Oregon
(Burleigh 1971).

Nevada: One skin (USNM No. 53650) collected in West
Humboldt Mountains by Ridgway on 10 September 1867
(Ridgway 1916).

Table 1. Comparison of archaeological humeri of Ectopistes mig-
ratorius with Recent species of Columba. All measurements are in
mm.

Species

N

Mean

Range

Stan-

dard

Devi-

ation

Stan-

dard

Error

E. migratorius

Length

25

41.4

38.8-43.6

1.12

0.22

Proximal breadth

18

14.4

12.3-16.5

0.90

0.21

Distal breadth

26

9.5

9.0-10.4

0.37

0.07

C. fasciata

Length

17

46.2

43.9-48.7

1.56

0.38

Proximal breadth

13

16.3

15.3-17.2

0.56

0.15

Distal breadth

17

11.2

10.6-12.3

0.46

0.11

C. livia

Length

20

46.0

43.7-50.9

1.74

0.39

Proximal breadth

18

17.9

15.5-19.3

1.18

0.28

Distal breadth

20

10.9

9.8-12.0

0.53

0.12

C. leucocephala

Length

5

42.7

40.3-44.5

1.83

0.82

Proximal breadth

5

13.1

12.0-14.0

0.65

0.29

Distal breadth

5

10.1

9.4-10.5

0.40

0.18

C. flavirostris

Length

6

42.8

40.9-43.9

1.12

0.46

Proximal breadth

6

14.8

13.8-15.7

0.56

0.23

Distal breadth

6

10.7

10.0-11.2

0.38

0.16

New Mexico: One bone from Pleistocene deposits at Dark
Canyon Cave, Eddy County (Howard 1971); two bones rep-
resenting one individual from Una Vida, San Juan County,
dated at A.D. 975-1030; one bone from Picuris Pueblo, Taos
County, dated at A.D. 1300-1350 (the latter two records re-
ported on here).

LTah: One bone from Archaic deposits at the Stansbury II
site, Stansbury Island, The Great Salt Lake, Tooele County
(see Parmalee this vol.).

Washington: Two specimens shot, but not collected, at Spo-
kane Falls by Lt. A.V. Kautz in 1869; also recorded from
Colville and Puget Sound (Jewett et al. 1953).

Wyoming: One skin (USNM No. 13382) collected at Horse-
shoe Creek, Albany County, by Ridgway on 16 September
1859 (Ridgway 1916).

In addition to the above, Schorger (1955) discusses possible
sightings, but no skins, of Ectopistes migratorius in western
Montana, Idaho (Lemi County), and Wyoming (Fort Lara-
mie). The new records from New Mexico are the earliest Ho-
locene records from this state. While all the above records
seem to indicate E. migratorius was an accidental visitor to
the western United States, Phillips (1968) discusses the prob-
ability of locating fossils of accidental species. Concerning the
Pleistocene record of£. migratorius at Rancho La Brea, Phil-
lips (1968:134) states, “The case for considering Ectopistes cas-
ual would be stronger if the abundant and similar (in size, and
in some degree in habits) Columba fasciata were not repre-
sented by exactly the same number and scattering of individ-
uals. Further, other notably gregarious birds (shorebirds,
terns, blackbirds) are also poorly or not at all represented at
Rancho La Brea.” Phillips also points out that while £. mi-
gratorius is absent from the younger tar pits, so are remains
of the condor, Gymnogyps amplus, the presumed direct ances-

Hargrave and Emslie: Passenger Pigeons

259

Figure 2. Locations of Pleistocene and Holocene records of Ectopistes migratorius in the western United States. Line indicates extent of normal
distribution (after Schorger 1955).

tor (Howard 1962) of G. calif ornianus , which is still found not
far from Rancho La Brea.

Based on this argument, we can conclude that Ectopistes
migratorius may have been a common species in southern Cal-
ifornia during the Pleistocene. In fact, the Pleistocene distri-
bution of this species may have extended over much of western
North America. The presence of£. migratorius in Idaho, Ne-

vada, Washington, and Wyoming as late as the 1850’s shows
that the species was still able to survive in isolated pockets in
the west.

The Pleistocene record of Ectopistes migratorius in New
Mexico suggests it was a common species of forests and wood-
lands there as well. Evidence exists that indicates long pluvial
periods in New Mexico and Arizona during the Pleistocene

260

Hargrave and Emslie: Passenger Pigeons

(Wright et al. 1973; Harris 1977; Van Devender et al. 1977;
Porter 1978; Van Devender and Spaulding 1979), and these
periods may have seen the extensive development of forests
and open woodlands. However, it is unlikely that the archae-
ological records represent relict, post-Pleistocene populations
of£. migratorius that resulted from climatic changes and their
effect on vegetation patterns. It is more likely that Holocene
climatic fluctuations, causing minor mesic intervals, allowed
E. migratorius to re-extend its range into the state. Pollen
studies have shown increases in moisture at Picuris Pueblo
from A.D. 1335-1425 (Schoenwetter 1970) and at Chaco Can-
yon beginning about A.D. 1100 (Hall 1977). These periods
may represent brief mesic intervals that allowed E. migrator-
ius to expand into those areas.

CONCLUSIONS

Two new records of Ectopistes migratorius are the earliest
Holocene records known from New Mexico. Pleistocene rec-
ords of this species indicate that it was possibly common in
the state at that time. It is questionable, however, that it was
able to remain in New Mexico as relict populations following
post-Pleistocene climatic changes. Rather, it probably re-ex-
tended its range into certain areas because of temporary cli-
matic fluctuations that caused minor mesic intervals. We ex-
pect there will be other Pleistocene and archaeological records
of£. migratorius in the western United States.

ACKNOWLEDGMENTS

We are very grateful to the following persons for loans of
skeletal material: Herbert Dick, Adams State College, Colo-
rado; Robert Finley, U.S. Fish and Wildlife, Ft. Collins, Col-
orado; D.L. Hamilton, University of Texas at Austin; Paul W.
Parmalee, University of Tennessee at Knoxville; Amadeo M.
Rea, San Diego Museum of Natural History; and Gordon Vi-
vian (now deceased), formerly of the Southwest Archaeological
Center, Tucson, Arizona. Storrs Olson and the late Alexander
Wetmore of the Smithsonian Institution confirmed the iden-
tifications discussed in this report and pointed out certain dis-
tinguishing characters. We also extend our thanks to Thomas
Van Devender, Arthur Harris, and Amadeo Rea for their help-
ful comments and criticisms on earlier drafts of this paper.
Mark Middleton of the Museum of Northern Arizona provided
the photograph. The Max C. Fleischmann Foundation funded
much of the work reported on here.

LITERATURE CITED

Burleigh, T.D. 1971. Birds of Idaho. Caxton Printers, Ltd.
Cadwell, Idaho. 467 pp.

Hall, S.A. 1977. Late Quaternary sedimentation and paleo-
ecologic history of Chaco Canyon, New Mexico. Bull.
Geol. Soc. Amer. 88:1593-1618.

Harris, A.H. 1977. Wisconsin Age environments in the
northern Chihuahuan desert: evidence from the higher
vertebrates. Trans, of the Symposium on Biol. Resources
of the Chihuahuan Desert Region: 23-52. Sul Ross State
University, Alpine, Texas.

Howard, H. 1937. A Pleistocene record of the Passenger
Pigeon in California. Condor 39:12-14.

. 1962. A comparison of prehistoric avian assemblages

from individual pits at Rancho La Brea, California. Nat.
Hist. Mus. Los Angeles Co., Contrib. Sci. No. 58:1-24.

. 1971. Quaternary avian remains from Dark Canyon

Cave, New Mexico. Condor 73:237-240.

Jewett, S.G., W.P. Taylor, W.T. Shaw, and J.W. Al-
drich. 1953. Birds of Washington State. Univ. of Wash-
ington Press, Seattle. 767 pp.

Linsdale, J.M. 1951. A list of the birds of Nevada. Condor
53:228-249.

Phillips, A.R. 1968. The instability of the distribution of
land birds in the southwest. Pp. 129-162 in Collected
Papers in Honor of Lyndon Lane Hargrave (A.H. Schroe-
der, Ed.). Papers Arch. Soc. New Mexico No. 1. Univ.
New Mexico Press, Albuquerque.

Porter, L.S.W. 1978. Pleistocene pluvial climates as indi-
cated by present day climatic parameters of Cryptotis par-
va and Microtus mexicanus . J. Mammal. 59(2):330— 338.

Ridgway, R. 1916. The Birds of North and Middle America.
Part VII. Bull. United States Natl. Mus. No. 50:1-543.

Schoenwetter, J. 1970. Archaeological pollen studies of
the Colorado Plateau. Amer. Antiquity 35(l):35-48.

Schorger, A.W. 1955. The Passenger Pigeon. Univ. of Wis-
consin Press, Madison.

Van Devender, T.R., and W.G. Spaulding. 1979. De-
velopment of Vegetation and Climate in the Southwestern
United States. Science 204:701-710.

Van Devender, T.R., M. Phillips, III, and J.I. Mead.
1977. Late Pleistocene reptiles and small mammals from
the lower Grand Canyon of Arizona. Southwestern Nat-
uralist 22(l):49-66.

Wright, H.E., Jr., AM Bent, B.S. Hansen, and L.J.
Maher, Jr. 1973. Present and past vegetation of the
Chuska Mountains, northwestern New Mexico. Bull.
Geol. Soc. Amer. 84:1155-1180.

alp

W§

,

tee, M' !|) MV" I11'- I
111

jjtRtf ziliwQ. 1 '-'

Wife

Sill

i :::!■

SititP 2f& >

||§L " _ s w

■

•{ .•! .iiM:11 ! ?\l

■ : '.: ii-jj !'l

Kmji®!

. :

iiiifete mm

lifc-.-i: :|fctv'- ■..:' :■ .irjJMjrtJii .■■!■; ,; :<;:

' 1 !- .iii;j; = : .' Si.i ::..,:i,

jBliMiliariiiii'fii.iiliii i"iiii i

PMtKj y*

"y - ’ »» ! m

V ij nM $ ‘ | ll’^l ,j‘f

P|apw#»a

ill

■mmmmpsiii ip

■BIB

I: Oiim ItHlIH! yiti :'■ ’ .'1

■HHl
mmmm I

^''r!; f; -Ml

r.ylijiBilSil^ll'lfS!;!]:;

Ifilliflpl ''

a, — * . , i in

m

pMA i i M-

fesspi&jrtVi |J* wWHhf •<■

UWSWM

■tam

mi ™ art* Sw If?' p re ft'l :;:i •■ ;.v:r ;-

SMfMt mm Isiim'Mftii&v •:« w'iMi'i

RtiffliHSliSSl

Eii

;::V

liiiiiiiiiiiajiPf!!

mmm

JMLaMiSHHi ■ »

Z

CD
X)

>

XI

Sff 4 I TUT -c /

m ''•t «« xgiiTisgx m m

co V S co — co — co

ivy an libraries Smithsonian institution NoiiniiiSNi nvinoshihns S3iavyan libraries smithsonu

W i -##

> N^ii^gx 5 ^ 2 > '4$*' s

co * z c o k- z co 2 co V _

ITUTION NOIlfllllSNI NVINOSH1IWS S3ldVdan LIBRARIES SMITHSONIAN INSTITUTION NOlifUllSNI NVINOSHil :

, ^ ^ W 2 A

^ w 111 00 /SSS8S\ !f . ,u> m

iTPXv-i\ ~ -i .v’S^xK. r. h ~ -/a/ A' h AATfX-a r /JATn&s

O " xc'ousgz cj __ O

IVyan^LIBRARIES2 SMITHSONIAN-1 INSTITUTION NOlinillSNI ^NVINOSHIIIMS S3 I y Vd a 11 ”*11 B RAR I ESZ SMITHSON! >
2 r- z n , z

m /<fU*5x 2 ra m ^ 2 = O

m ^ * ' m m rn

co £ co £ co £ co _

ITUTION NOlifUllSNI NVIN0SH1IWS S3ldVd9n LIBRARIES SMITHSONIAN INSTITUTION NOlifUllSNI NVINOSHil I

co

t s.

z W •' z 5 > 2

B RAR I ES^SMITHSONIAN INSTITUTION NOlifUllSNI _ NVINOSHIIIAIS^ S3 I d Vd 8 11 LIB RAR I ES^SMITHSONI

CO ~ co 7 \ co — CO

ITUTION NOlinillSNI NV1N0SH1IWS S3ldVd8n LIBRARIES SMITHSONIAN INSTITUTION NQIlfUllSNS NVINOSHil

Z n sjfr- | g y x/5 - /&^ts a Ji^S

I- u. — /‘'HlfAA H- / yU' ^

i- >

co

CO

co

m ’■c \c<oii.s^z m X'* pc^z r; m “• ^ugsg^ m

co \ _ co £ co £ co

d V d 8 11 LIBRARIES SMITHSONIAN INSTITUTION NOlifUllSNI NVINOSH1IIMS S3iaVd8ll LIBRARIES SMITHSON!
2 <£ z .v, co z _ co z

CO

8 ft, Jf s lS|x s isor s WL.J#) s I ilk' i

I | ^ 1 l v^/ 1 W I v i ^

CO * Z CO *■ 2 CO 2 CO * z

ITUTION NOlifUllSNI NVINOSHIIINS S3iaVd9n LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHil
co — co = co X co —

^ o

JVda ll^LI BRAR I ES^SMITHSONIAN^INSTITUTION^NOliniliSNI^NVINOSHlItMS S3ldVaan LIBRARIES SMITHSONI

r- Z •” 2 r- Z r~ Z

^STlr,oX

/"X 03

X
>

m ^ m ^ rn - ^ m

ITUTION^NOlinillSNI _ NVINOSHil WS S3 I MVd 9 ll“LI BRAR S ES SMITHSONIAN^NSTlTUTlON^NOliniilSNI NVINOSHil
co 2 W ^ 2 ■, w z co z

“■ %%; o

2 > 2 2 >

HVd 8 II2 LI B RAR 1 ES^SMITHSONIAN INSTITUTION °° NOlinillSNI NVINOSHIIWS^SS I d Vd 8 II^Ll B R AR I ES^SMITHSONI

CO “ CO -7, X CO “ CO

u -#> u>

TITUTION NOIlfUllSNI NVINOSHillNS S3iaVH8l1 LIBRARIES SMITHSONIAN INSTITUTION NOIlfUllSNI NVINOSH

co

CO

iyvaan libraries Smithsonian institution NoiiruusNi nvinoshiiws S3iavaan libraries smithsoi

*=—//> ~ ' f/) — (/)

-

w

ITITUTION NOIlfUllSNI NVINOSH1HNS S 3 I a V a 8 IT LIBRARIES SMITHSONIAN INSTITUTION _NOIinillSNI_NVINOSH.

I 2 I 2 2 l /tm* • ^

, >

m -• vw ^ pi) 'll m ^ \.wy m

_ co £ co £ co 5 co

avaan libraries Smithsonian institution NounuiSNi nvinoshiiiais saiavaan libraries smithso

Z , CO 2 CO Z co 2 CO

5 - =* z 5 § /#%\ 3 vm- o %% s

® co -4il co co v mfi co

> i > _

iTITUTlON tON0liniliSNI NVIN0SH1I S 3 I H V a 3 I l^L I B R A R I E S^SMITHSONIAN INSTITUTION ^ NOIlfUllSN I _ NVINOSH

CO = \ CO ~ CO — CO

lavaan li brar i es_smith$onian_snstitution NounuiSNi nvinoshiiws saiavaan libraries smithsc

.2 ,m 2 z£^x “ jSSu 2 z£5^X m °

CO

* vi ^

m ^ -"jo's/ m m

co ± co £ to \ f= co

5TITUTI0N NOIlfUllSNI NVINOSHitlNS SBiaVHatl LIBRARIES SMITHSONIAN INSTITUTION NOIlfUllSNI NVINOSf
4/5 5 ^ ? -v-- ^ z ,v co z

< S ... < ,vv\\ 2 /^^rx < .tf,, S .< z

X c/'ic CJ asl j- . C4*5 C_> m&U VC: JL ispc ' rrn?\ o Vv. X

co 'i/'/p co Sit S§5| co co ’flpfev. N- co Pg^ «*§S| ^ S to

0 X o X \ O fefc Jy/ x %4\ o

1 'W | I | \ i ^

lavaan libraries Smithsonian institution Noiiniusm^NviNosHiiiNs^sa lavaan 2 li brar 1 es°°smithso

—» co co — to ~ co

JTiTUTION^NOlinillSNI^NVINOSHlIIAIS S3 I avaan LIBRARIES SMITHSONIAN INSTITUTION NOIlfUllSNI NVIN0SH

° 'as5SSv 2 V p 5 I I m

? 2 (is %A 5 S SI: S fe# M I lif^M g

33

co "' _ to £ to £ to

lavaan libraries Smithsonian institution NounuiSNi nvinoshihajs saiavaan libraries smithso

z. co z co z co

<XCI[0,

S - E 1 H„i

2

CO

07 Z \ Ul y. -J.

a « ®p"®i = %% “ /aP’^01 3 /I

z

CO

z

co

H

zTqqvX

EjJ

<r

CO

HI

3