UN

HARVARD UNIVERSITY

Library of the

Museum of

Comparative Zoology

MUS. CO

OCCASIONAL PAPERS

juim 3

of the

MUSEUM OF NATURAL HISTORY
The University of Kansas
Lawrence, Kansas

NUMBER 100, PAGES 1-22 JUNE 16, 1982

PHYLOGENETIC RELATIONSHIPS OF THE

COLUBRID SNAKES OF THE GENUS ADELPHICOS

IN THE HIGHLANDS OF MIDDLE AMERICA

BY

Jonathan A. Campbell and Linda S. Ford1

Snakes of the genus Adelphicos are widely distributed in south-
ern Mexico and northern Central America. Adelphicos quadrivir-
gatus occurs on the Atlantic versant and coastal plain from the
foothills of central Veracruz, Mexico, through Guatemala and Be-
lice, exclusive of the outer Yucatan Peninsula; and on the Pacific
versant from central Oaxaca, Mexico, into central Guatemala. This
species ranges from about sea level to 1400 m in a variety of habitats
including lowland rainforest and quasirainforest, lower montane dry
forest, and lower montane wet forest ( for chorography of vegetation
complexes see Stuart, 1966:692). Adelphicos (fuadrivirgatus is dis-
tinctive from all highland populations of Adelphicos by having chin
shields that are greatly expanded toward the lip and in the anterior
position of the first ventral that is generally at the level of the
posteriormost labials ( Smith, 1942 ) .

Disjunct populations of Adelphicos occur in the Middle Ameri-
can highlands at elevations of 1200-2200 m on the Sierra Madre de
Chiapas in extreme southeastern Oaxaca, the Meseta Central of
Chiapas, the Sierra de los Cuchumatanes, and the Sierra de las
Minas and neighboring mountains in Guatemala (Fig. 1). The
various allopatric populations of highland Adelphicos are isolated
in humid pine-oak or cloud forests and are differentiated from each
other in characters of squamation, color, pattern, osteology, and
body proportions.

1 Division of Herpetology, Museum of Natural History, The University of
Kansas, Lawrence, Kansas 66045.

OCCASIONAL PAPERS MUSEUM OF NATURAL HISTORY

Fig. 1. Distribution of highland Adelphicos in southern Mexico and Guate-
mala. Symbols representing the following: inverted triangles, A. latifasciatus;
squares, A. nigrilatus; circles, A. vcracpacis; triangle, A. daryi; hexagon, un-
allocated specimen.

Some of these populations have been recognized previously as
subspecies of A. veraepacis. However, we believe these populations
have diverged sufficiently from each other to merit recognition as
species. We provide diagnoses, descriptions, and a key for all
known members of the highland group. Our phylogenetic analysis
was determined in the manner suggested by Wiley (1981) using
other burrowing xenodontine colubrids (especially Geophis) as the
outgroup. Although highland Adelphicos are not common in col-
lections, we are fortunate in having observed all four species in the
field. We examined a total of 129 snakes: 4 A. latifasciatus, 58 A.
nigrilatus, 59 A. veraepacis, and 8 of the new species described
herein. These species constitute a monophyletic lineage referred
to as the veraepacis group, endemic to the Middle American high-
lands.

We measured cephalic scales to the nearest 0.1 mm using an
ocular micrometer. Measurements were taken along or perpendicu-
larly to the long axis of the scale. Body lengths were taken to the
nearest mm using a meter stick. In comparing the relative eye sizes
of the various species we took the horizontal distance across the eye
from the posterior edge of the loreal to the level between the postoc-
ulars. This distance was divided into the snout length, the distance
from the anterior edge of the eye to the front face of the rostral.

MIDDLE AMERICAN ADELPIIICOS 3

The first plate bordered on both sides by the first row of dorsals
was considered to be the first ventral (Dowling, 1951). We use
the term cephalic index to denote head width/head length X 100.
Three of the highland populations have been named. AcJcJphicos
veraepacis Stuart was based on four specimens from Alta Verapaz,
Guatemala; A. nigrilatus Smith was proposed for specimens from
the Meseta Central of Chiapas, Mexico; and A. latifasciatus Lynch
and Smith was based on a single specimen from the Sierra Madre
north of Zanatepec, Oaxaca, Mexico. The population from the
mountains southeast of Guatemala City is represented by a series
of eight specimens and is unnamed. We propose that this unique
snake be known as

AdcJphicos danji, new species
Figs. 2, 4—6

Holotype. — KU 187260, an adult female, from San Jorge Muxbal,
5.5 km W San Jose Pinula, Department of Guatemala, Guatemala,
elevation 1844 m, collected by J. A. Campbell on 11 July 1980,
original number JAC 5305.

Paratypes.— UTA R-5743, 5897, 6223, 7129, KU 187259, 190886,
female topotypes taken 30 May 1976, July 1976, November 1976,
12 September 1977, June 1980, and August 1980, respectively, at
elevations of 1829 to 2134 m by J. A. Campbell, L. S. Ford, and P.
Castillo; KU 187258, male topotype, taken September 1979 at 2134
m by P. Castillo.

Diagnosis. — The new species is a member of the veraepacis
group that can be distinguished from all other members of the group
by its large size, females attain lengths of at least 574 mm and
males probably exceed 400 mm; broad frontal plate, at least as
broad as long; dorsal coloration dark brown suffused with black;
relatively small eye, contained 2.7 — 3.2 times in the snout length;
low number of dentary teeth, 8 — 9 ( mode 9 ) .

It usually may be distinguished from A. veraepacis and A. lati-
fasciatus by its low number of ventrals, 120 in male, 128 — 132 in
females, opposed to 120 — 133 and 128 in males, and 132 — 142 and
133 — 138 in females, respectively. The male possesses 28 sub-
caudals, whereas males of A. veraepacis have 29 — 41 and A. lati-
fasciatus have 46 — 51. The number of subcaudals, 19 — 22, dis-
tinguishes female A. daryi from A. veraepacis (24 — 31) and A.
latifasciatus (37 — 41). The modal number of palatine teeth is 7
in A. daryi and 8 and 9 in A. nigrilatus and A. latifasciatus, respec-
tively.

Description of holotype. — An adult female, 487 mm in total
length; tail length 48 mm (9.9% of total); head length 16.3 mm
from front face of rostral to posterior end of mandible; head width

OCCASIONAL PAPERS MUSEUM OF NATURAL HISTORY

*S

i

Fig. 2. Dorsal (upper), lateral (middle), and ventral (lower) aspects of
the head of Adelphicos daryi (holotype, KU 187260).

MIDDLE AMERICAN ADELPHICOS 5

11.2 nun at broadest point at the rictus; head moderately distinct
from neck; snout rounded in dorsal view; eye small, snout 3.0 times
as long as horizontal distance across eye; pupil round; rostral about
1.2 times broader than high; internasals 1.4 times wider than long,
contacting anterior and posterior nasals laterally; prefrontals large,
slightly longer than wide, contacting posterior nasal and loreal
laterally; median prefrontal suture 0.6 times as long as frontal;
frontal as broad as long; parietals about 1.6 — 1.8 times longer than
wide, median suture 1.1 times frontal length; nostrils located in
anterior nasal; loreal about equal to combined length of nasals;
preocular absent; supraocular large, 1.1 times longer than loreal;
two postoculars; temporals 1 — 1, separating supralabials 6 and 7
from parietal; supralabials 7/7, 1 contacting nasals, 2 contacting
postnasal and loreal, 3 contacting loreal, 3 and 4 entering orbit,
5 and 6 contact anterior temporal, 6 largest and contacting posterior
temporal; mental 3.8 times broader than long, separated from chin
shields by first infralabials which contact each other along midline;
chin shields well developed, about twice as long as wide and not
extending to border of lip; posterior chin shields not differentiated
from gulars; anterior chin shields separated from first ventral by
four gulars; infralabials 7/7, 1 — 4 contacting anterior chin shields,
fourth largest; unreduced dorsal scales in 15 smooth rows through-
out length of body; dorsal scales in 6 rows at the level of the tenth
subcaudal; no apical pits apparent; ventrals 133; anal divided; sub-
caudals 21, paired; anal glands paired, extending 2 subcaudals.

Coloration in preservative (alcohol after formalin): dorsal
ground color brown strongly suffused with black; black lateral
stripe on upper half of scale row 1, all of 2, and lower three-fourths
of row 3; paravertebral stripes on upper part of fifth and lower part
of sixth row; middorsal black stripe on vertebral row; ventrals pale
yellow with dark anterior and outer borders; subcaudals yellow
with black outside borders and darkly pigmented along midventral
sutures forming stripe; top of head dark brown extending to upper
part of supralabials; gular area yellow with little mottling on mental,
first three pairs of infralabials, and anterior portion of chin shields.
These colors, taken from the freshly preserved specimen, generally
agree with notes from life, except for the ventrals which were
bright yellow.

Variation. — The following data were taken from the holotype
and seven paratypes. The total body length in females is 395 — 574
mm with the tail comprising 9.6— 10.8% (x = 10.1%) of the total
length. Four of the six females exceed 500 mm. The number of
ventrals is 128—133 (x = 130.7) and subcaudals, 19—22 (x =
20.6). The single male is 339 mm in total length with the tail
comprising 15.0% of the total length. It has 120 ventrals and 28
subcaudals.

6 OCCASIONAL PAPERS MUSEUM OF NATURAL HISTORY

The head length is 3.3—3.9 mm (x = 3.6) of the total length
and the cephalic index is 61 — 69 (x = 66.0). The distance across
the eye is contained from 2.7 — 3.2 (x = 3.1) times in the snont
length. The frontal is 1.0 — 1.1 times broader than long. The supra-
labials are 7/7 in all but one specimen (UTA R-6223) which has 6
on one side. The infralabials are 7/7. The number of teeth is 7 — 10
on the maxillary, 7 — 8 on the palatine, 13 — 17 on the pterygoid, and
8 — 9 on the dentary.

The ground color of A. daryi is brown strongly suffused with
black. The lateral stripes extend along scale row 2 and the upper
and lower parts of scale rows 1 and 3, respectively. The unbroken
paravertebral stripes extend along the upper part of row 5 and
lower part of 6. The vertebral stripe is obscured by dark pigment
in one specimen, only barely discernible in two, and clearly defined
in one. The ventrals are yellow with black-edged outer and anterior
margins; there is no darkening midventrally. The subcaudals are
light yellow with dark brown or black outer borders and a distinct
black midstripe on the undersurface of the tail. The gular area is
immaculate, but the area near the sutures of the anteriormost three
infralabials, mental, and chin shields is mottled with brown or
black.

Distribution. — Known only from the humid pine-oak forest
fringing the Las Nubes block in the southeastern Guatemalan high-
lands at 1829 to 2134 m ( Fig. 1 ) .

Etymology. — The specific name is a patronym for the late Mario
Dary Rivera, outstanding Guatemalan biologist and conservationist.

Specimens examined. — (8) The type series constitutes the only
known specimens of this species.

Adelphicos veraepacis
Figs. 4—6

1941 Adelphicos veraepacis Stuart, Occ. Pap. Mus. Zool. Univ.
Mich., 452:5. Type locality: Finca Samac, 7 km W Coban,
1500 m, Alta Verapaz, Guatemala. Holotype: UMMZ 89073,
an adult male, coll. L. C. Stuart, 21 April 1938.

1942 Adelphicos veraepacis veraepacis Smith, Proc. Rochester Acad.
Sci., 8:180.

Diagnosis. — A moderately large Adelphicos, females to 461 mm
and males to 372 mm total length, possessing a unique combination
of characteristics: females may be distinguished from all A. daryi
and most A. nigrilatus by having a higher number of ventrals (132 —
142); males possess 29 — 41 subcaudals and females have 24 — 31,
distinguishing them from A. daryi and A. latifasciatus which possess
fewer and more subcaudals, respectively. The dark brown or gray
dorsal coloration can be confused only with A. daryi, but the scales

MIDDLE AMERICAN ADELPI1ICOS

Fig. 3. Upper, Adelphicos latifasciatus — Cerro Baul, Oaxaca, Mexico (JAC
5539); lower, A. nigrilatus — 6.9 mi SE Teopisca, Chiapas, Mexico (UTA
R-2452 ) .

of the first scale row are not light posteroventrally or only slightly so,
in contrast to the boldly defined pale spots in A. daryi (Fig. 5).
The dorsolateral stripes generally are not continuous and involve
no more than the edges of two adjacent scale rows. The anterior

8 OCCASIONAL PAPERS MUSEUM OF NATURAL HISTORY

edges of the ventrals are darkly pigmented like A. daryi, but there
is a heavy concentration of dark pigment midventrally. This species
may be distinguished from A. daryi by possessing a relatively
larger eye that is contained 1.3 — 2.6 times in the snout length.

Variation. — In a sample of 29 males, one exceeds 350 mm (356
mm) and 14 are between 300 and 350 mm. Females are larger,
but the largest female, 524 mm in total length, is far greater in
length than other specimens. In a sample of 24 specimens, nine
are between 300 and 350 mm, eight lie between 350 and 400 mm,
and two exceed 400 mm. Males have 120 — 133 (x = 126.0) ven-
trals; relative tail lengths of adults are 14.5 — 19.4% (x = 16.5) and
10.0 — 13.4% (.v = 11.9) of the total for males and females respec-
tively. The frontal is 0.67—0.94 (x = 0.84) times as broad as long,
and the average cephalic index is 61. The eye size is contained
1.4 — 2.6 (x = 2.1) times in the snout length.

Specimens from the Montanas de Cuilco of Guatemala differ in
minor respects of color and pattern from snakes collected in Alta
and Baja Verapaz. Although these differences distinguish snakes
from the two regions from each other in most instances, these
snakes are so similar in other respects we feel they do not warrant
specific recognition.

Snakes from the highlands of Alta and Baja Verapaz are gener-
ally dark brown. Specimens may be so dark as to partially obscure
the stripes. The lateral stripes usually extend along scale row 2
and the upper and lower parts of scale rows 1 and 3, respectively.
In about half of the specimens the lateral stripe extends from the
outer edge of the ventrals to the lower half of scale row 3 with
only a small amount of or no light coloration on scale row 1, a
condition not noted in any specimen from elsewhere. In no snakes
from this region does the lateral stripe extend to scale row 4. The
paravertebral stripes are invariably broken and located on the upper
part of scale row 5 and lower part of scale row 6. The vertebral
stripe is broken in about half of the specimens. The yellow ventrals
are heavily pigmented along the midline producing a series of more
or less triangularly shaped blotches with their apices produced
posteriorly. The subcaudals are mostly dark with no well-defined
midstripe. The gular area and chin shields are heavily mottled.
The upper one-fourth to one-third of the supralabials is black and
dark pigment may extend down the intrasupralabial sutures to the
margin of the lip.

Specimens from the Montanas de Cuilco are generally medium
to dark brown and are frequently paler in over-all appearance than
the snakes from Alta and Baja Verapaz. The lateral stripes are
usually located on scale row 2 and the adjacent parts of scale rows
1 and 3 respectively, but in eight of 37 specimens the lateral stripe
extends up to the lower part of scale row 4, a condition not present

MIDDLE AMERICAN ADELPIIICOS

9

.

.

-Ai ^

*

Fig. 4. Upper, Adelphicos daryi — San Jorge Muxbal, Guatemala, Guate-
mala ( KU 187260, holotype ) ; lower, A. veraepacis — W slope Cerro Verde,
BajaVerapaz, Guatemala ( KU 190891).

in specimens from other parts of the range. The paravertebral
stripes are usually continuous (broken in only one specimen) and
variably located on parts of scale rows 5 and 6 (11 specimens),
6 (13), 6 and 7 (7), or 5, 6, and 7(2). The vertebral stripe is un-
broken and discernible in all specimens except several that appear
to have been darkened by preservative. The ventrals are yellow
with dark brown or black outer edges and with little to moderate
amounts of pigment along their anterior borders. There is no ten-
dency for darkening along the ventral midline. The subcaudals are

10 OCCASIONAL PAPERS MUSEUM OF NATURAL HISTORY

pale with black pigment along their inside borders forming a mid-
ventral stripe on the tail. The gular area and chin shields are
usually immaculate, but a small amount of mottling is occasionally
present. Black pigment may barely encroach on the upper part of
the supralabials.

Etymology. — The specific name is in reference to the depart-
ment of Alta Verapaz, Guatemala, from which the type specimens
originated.

Distrilmtion. — The northern mountains of Guatemala: the Mon-
tahas de Cuilco, the Sierra de las Minas and neighboring mountains,
and probably the Sierra de los Cuchumatanes, from 1200 to 2200
m elevation in cloud and pine-oak forests (Fig. 1). Stuart (1943b)
reported a specimen without precise locality data that may have
originated from either the Sierra de los Cuchumatanes or the Guate-
malan Plateau, but he favored the former.

Specimens examined. — (59) GUATEMALA: Alta Verapaz:
FincaChichen [FMNH 109S61 (paratype), UMMZ 89074-75 (para-
types)]; Baja Verapaz: Cerro Quisis (KU 190890, UTA R-7082);
Cerro Verde ( KU 190891, 190894-95, UTA R-6542, 6551, 6571, 6584,
7081, 7083); El Bagasal (KU 187320); near La Union Barrios
(KU 190889, 190892-93, UTA R-7084); 7.7 km SSE Purulha (UTA
R-6479); 1.6 km SE Purulha (UTA R-8841); Huehuetenango:
Paraiso [UMMZ 127233(17), 127234-37, 127238(4), 127239(2),
127240(3), 127241(2), 127266-70]; No Specific Locality: (UMMZ
90239).

Adelphicos nigrilatus
Figs. 3, 5—6

1942 Adelphicos veraepacis nigrilatus Smith, Proc. Rochester Acad.
Sci., 8:182. Type locality: San Cristobal de las Casas, Chiapas,
Mexico. Holotype: USNM 137219 (formerly EHT-HMS No.
15335), coll. H. Devlin Thomas.

Diagnosis. — The most variable member of the veraepacis group;
most easily distinguished from A. daryi and A. veraepacis by its
bright orange, red, or pale brown ground color. The vertebral scale
row is pale in coloration when the middorsal stripe is absent,
thereby distinguishing this species from A. daryi and A. veraepacis:
when present, it involves only the vertebral scale row, thereby
distinguishing this species from A. latifasciatus. In most specimens
the venter is unpigmented, distinguishing A. nigrilatus from A. daryi
and A. veraepacis, whereas in a few specimens the possession of
heavy pigmentation midventrally distinguishes A. nigrilatus from
A. latifasciatus, and it lacks the dark anterior edging characteristic
of A. daryi and A. veraepacis. Adelphicos nigrilatus may be distin-
guished from A. latifasciatus by having fewer subcaudals (21 — 26

MIDDLE AMERICAN ADELPIIICOS

11

Fig. 5. Patterns of the highland species of Adelphicos: a) A. htifasciatus
from Cerro Baul, Oaxaea, Mexico ( UTA R-8794); b) A. nigrilatus from 6.8
mi ESE San Cristobal de las Casas, Chiapas, Mexico (UTA R-8837); c) A.
veraepacis from Cerro Quisis, near La Union Barrios, Baja Verapaz, Guatemala
(UTA R-7082); and d) A. dartji from San Jorge Muxbal, Guatemala, Guate-
mala (KU 187260, holotype).

and 26 — 36 in males and females, respectively) and relatively
shorter tail in males and females, 13.5—18.9% (16.2%) and 10.3—
12.9% (11.6%) of the total length, respectively. Female A. nigrilatus
generally have fewer ventrals than do those of A. veraepacis and
A. latifasciatus.

Variation. — Females are larger than males; in a sample of 25
females, seven were between 250 and 300 mm, five between 300
and 350 mm, one between 350-400 mm, and one was 451 mm in
total length. In a sample of 32 males, eight were between 250-300
mm, the largest being 293 mm. There are 113 — 128 (x = 118.1)
ventrals in males and 123—135 (x = 127.7) in females. The frontal

12 OCCASIONAL PAPERS MUSEUM OF NATURAL HISTORY

width is from 0.67 to 0.94 (x : : 0.82) of the frontal length; the
average cephalic index is 58. The eye is contained 1.4 — 2.6 (x =
2.0) times in the length of the snout.

The ground color of this species is generally bright orange in
life, fading to a pale beige or tan in preservative. Black lateral
stripes frequently involve the upper portion of scale row 2, all of
3, and the lower portion of scale row 4 (33 of 53 specimens), but
also may be located on scale row 3 (7), upper 3 and lower 4 (6),
upper 1, 2, lower 3 (3), upper 1, 2, 3, lower 4 (2), upper 3, 4,
lower 5 ( 1 ) . The paravertebral stripes may be absent ( 16 or 53
specimens), but usually are present on scale row 6 (18), upper 6
and lower 7 (10), upper 5 and lower 6 (7), 5 (1), or lower 7(1).
The vertebral stripe is absent in 11 specimens, present but broken
in three, and unbroken in 39.

Etymology. — The name nigrilatus is derived from the Latin
nigri meaning black, and lotus meaning side, and refers to the dis-
tinctive lateral stripes on the body.

Distribution. — The Meseta Central of Chiapas, Mexico, from
the vicinity of San Cristobal de las Casas southeastward almost to
the Guatemalan border, from 2200 to 2500 m elevation in cloud
and pine-oak forests ( Fig. 1 ) .

Specimens examined.— (5S) MEXICO: Chiapas: San Cristobal
de las Casas [USNM 137219 (holotype), CAS 20617 (paratvpe),
FMNH 100285-91 (paratypes), MCZ 56017 (paratvpe), AMNH
66961, UF 42073]; 4 mi SE San Cristobal de las Casas (AMNH
99981), 6.8 mi ESE San Cristobal de las Casas (UTA R-8837), 7.3
mi E San Cristobal de las Casas (AMNH 102917), 8.2 mi E San
Cristobal de las Casas [UMMZ 119761(6), 121536(2)], 3 mi W
San Cristobal de las Casas. (AMNH 93317), 15 mi W San Cristobal
de las Casas (UMMZ 94595), 9 mi NW San Cristobal de las Casas
(MVZ 109478), 15.5 mi E San Cristobal de las Casas (UTA R-
2514), 9 km W San Cristobal de las Casas (CAS 145203), 4.0 km
W San Cristobal de las Casas ( KU 37594), 4.3 km W San Cristobal
de las Casas (MVZ 144524), 10 km SE San Cristobal de las Casas
(KU 129230), 6.5 mi SE San Cristobal de las Casas (LACM 58124-
27, TCWC 30410-12, UMMZ 94593-94), 6.8 mi SE San Cristobal
de la Casas (LACM 58128-29, 58149, TCWC 30414-18), 7.4 mi SE
San Cristobal de las Casas (UMMZ 126171), 13 mi E Las Rosas
(TCWC 12801); 6.9 mi SW Teopisca (UTA R-2452); 11.2 mi S
Teopisca (LACM 58899); 24.1 mi SE Teopisca (TCWC 30413);
1.0 mi E Tulanca on Mex Hwy 190 (UTA R-8839-40), 1.1 mi E
Tulanca on Mex Hwy 190 (UTA R-SS3S); Puente de Chamic
(UCM 19039); IS km E Mex Hwy 190 on road to Lago Montebello
(MVZ 112403).

MIDDLE AMERICAN ADELPHICOS

13

Fig. 6. A) From left to right, dorsal views of the parietals of Adelphicos
latifasciatus (UTA R-6213), A. nigrilatus ( UTA R-2514), A. veraepacis
(UTA R-7083), and A. danji (UTA R-5897, paratype); B) Clockwise from
upper left, lateral views of the prefrontals of the same specimens listed above;
C) Lateral views of the mandibles of A. daryi, UTA R-5897 (upper), and
A. veraepacis, UTA R-7083 (lower).

Adelphicos latifasciatus
Figs. 3, 5—6

1966 Adelphicos veraepacis latifasciatus Lynch and Smith, Trans.
Kansas Acad. Sci., 69:66. Type locality: Sierra Madre north
of Zanatepec, Oaxaca, Mexico. Holotype: UIMNH 56147,
coll. Thomas MacDougall, 6—12 Sept. 1963.

Diagnosis. — The species may be distinguished from all other
members of the veraepacis group by the high number of subcaudals
in males (46 — 51) and females (37 — 41), its relatively long tail
comprising 20.4—22.3% and 14.6—15.6% of the total length in males
and females respectively, and its broad middorsal stripe that in-
cludes the vertebra] and adjacent halves of the paravertebral scale
rows. The modal number of teeth tends to be more (11 maxillary,
11 dentary, 9 palatal, and 18 pterygoid) than in the other species.
It can be distinguished from A. daryi and A. veraepacis by its pale

14 OCCASIONAL PAPERS MUSEUM OF NATURAL HISTORY

to reddish brown ground color. It differs from A. daryi, A. verae-
pacis, and most A. nigrilatus by lacking paravertebral stripes and
having an unpigmented venter. In A. latifasciatus the lateral stripe
involves scale row 3 and adjacent halves of scale rows 2 and 4,
whereas in A. daryi and A. oeraepacis the stripe includes scale row
2 and adjacent portions of scale rows 1 and 3.

Variation. — Total lengths in two males are 310 and 333 mm, in
three females the total length is 323-437 mm; ventrals are 12.5 — 12S
(.v = 126.5) in males, 133—138 (x = 135.7) in females; the frontal
is 0.S2 — 1.00 (x = 0.S7) times as broad as long. The average
cephalic index is 60 and the eye is contained 2.2 — 3.2 (x = 2.6)
times in the length of the snout.

In four specimens of A. latifasciatus from Cerro Baul, Oaxaca,
the ground color is reddish to pale brown, the lateral stripes involve
scale row 3 and adjacent parts of rows 2 and 4. Paravertebral stripes
are absent, and the vertebral stripe includes the vertebral and half
of the paravertebral scale rows. The top of the head is dark brown,
extending to the upper part of the supralabials. The gular area and
ventrals arc immaculate except for the pale mottling on the outer
portion of the ventrals. The subcaudals arc pale with a midventral
stripe along their median sutures. Except for their pale dorsal
ground color, these specimens agree in aspects of pattern and
color with the description given for the holotype (Lynch and
Smith, 1966). The terms "dark" and "pale" are relative and we
suspect from viewing the photograph of the holotype (Lynch and
Smith, 1966:67) that even the ground color is in agreement.

Etymology. — The name latifasciatus is derived from the Latin
lati, meaning broad or wide, and fasciatus, meaning a band, and
refers to the distinctive broad dorsal stripe.

Distribution. — The cloud forest on the western portion of the
Sierra Madre de Chiapas in Oaxaca, Mexico, at elevations of 1500
to 1900 m.

Specimens examined. — (4) MEXICO: Oaxaca: Cerro Baul
(UTA R-6213, 8793-94, J AC 5539).

OSTEOLOCICAL DIFFERENCES AMONG HIGHLAND

ADELPIIICOS

Many osteological differences exist among the highland species
of Adelphicos. We have examined skeletal material of the following:
A. latifasciatus (UTA R-6213, JAC 5539), A. nigrilatus (UTA
R-2514, 8840), A. veraepacis (UTA R-7081-83, KU 187321), and
A. daryi (UTA R-5743, 5897).

Premaxillae. — The leading horizontal edge of the premaxilla is
flared in A. daryi, but not in the other species. The configuration
of the premaxilla and nasals causes distinct differences in head

MIDDLE AMERICAN ADELPII1COS

15

shape: in lateral profiles the snouts of A. latifasciatus and A. nigri-
latus are acuminate, A. veraepacis is subtruncate, and A. daryi is
truncate.

Prefrontals. — The extent to which the prefrontal is in contact

Fig. 7. A theory of the relationships of snakes of the veraepacis group of
Adelphicos. Numbers refer to the following characters: 1) single sulcus;
2) broad lateral stripes; 3) increase in number of hemipenial spines; 4) en-
larged chin shields; 5) position of first ventral; 6) subcaudals, a. decrease in
number, b. extreme reduction; 7 ) a. venter with some pigmentation, b. anterior
edges of ventrals dark; 8) decrease in number of teeth; 9) broad vertebral
stripe; 10) loss of paravertebral stripe; 11) dark ground color; 12) lower arm
of prefrontal well developed, a. lower arm less than Yz length of upper, b. lower
arm % to equal upper; 13) posterior median ridge on parietal and supraoccipi-
tal; 14) dorsolateral edges of parietal well defined; 15) large size; 16) profile,
a. subtruncate, b. truncate; 17) greatly expanded hypapophyses in anterior
trunk vertebrae; 18) stout teeth; 19) broad frontal; 20) flared premaxilla.
See text for details.

16 OCCASIONAL PAPERS MUSEUM OF NATURAL HISTORY

with the maxilla varies from A. latifasciatus and A. nigrilatus, in
which the anteriorly projecting lower arm of the prefrontal is poorly
developed, to A. daryi in which the lower ami is almost equal to
the length of the upper arm (Fig. 6). Adelphicos vcraepacis has a
lower prefrontal arm that is intermediate in length between that
of A. daryi and the other species.

Frontols. — The frontals are broader than long in all species, but
in A. nigrilatus they are relatively longer and more tapering pos-
teriorly than in the other species.

Parietals. — Perhaps no character is more distinctive in these
snakes than the condition of the parietal in adult snakes (Fig. 6).
In A. latifasciatus and A. nigrilatus the parietal is broadest at about
its midlength, whereas in A. veraepacis and A. daryi it is broadest
anteriorly. It is broader in A. daryi than any other species, with
well-defined, sharp, dorsolateral edges that are close together pos-
teriorly. The parietal is slightly narrower in A. veraepacis, and the
dorsolateral ridges may coalesce posteriori)7. In both A. latifascia-
tus and A. nigrilatus the dorsolateral ridges are poorly defined and
the parietal is flatter. In A. daryi and A. veraepacis the dorsolateral
parietal ridges are distinctly raised posteriorly forming a median
crest that extends through the supraoccipital, whereas in A. lati-
fasciatus and A. nigrilatus no such crests exists.

Maxillae. — A comparison of the maxillae reveals A. daryi has a
relatively stout maxilla, whereas in the other species it is not as
stout nor as firmly articulated with the prefrontals and ectoptery-
goids.

Mandibles. — Several distinguishing characters concerning the
mandibles exist for these species. The mandibles are more massive-
in the western species. This trend is particularly reflected in the
stout splenial of A. daryi. The position of the mental foramen in
reference to its distance from the anterior tip of the compound is
less in A. nigrilatus and A. veraepacis.

Teeth. — The shape of the teeth is diagnostic for A. daryi which
has short, stout teeth (Fig. 6). In the other species the teeth are
longer, more slender, and more strongly recurved. Also, there are
differences in the number of teeth. Variations in the number of
maxillary teeth with the modal number in parentheses for A. lati-
fasciatus, A. nigrilatus, A. veraepacis, and A. daryi are 11(11),
9—10(9), 8—10(9), and 7—10(9), respectively. The number of
dentary teeth varies from 11—12(11), 10—11(10), 10—11(10), and
8 — 9(9); and the number of palatal teeth ranges from 9(9), 10 — 11
(10), 7(7), and 7 — 8(7) in the above species, respectively. There
are 18—20(18), 14—20(16), 13—16(16), and 13—17(16) ptery-
goid teeth in these species, respectively.

Vertebrae. — The hypapophyses of the anterior trunk vertebrae
are greatly expanded relative to the long axis of the body in A.

MIDDLE AMERICAN ADELPHICOS 17

daryi but are not so expanded in the other species. The interhypa-
pophyseal spaces of the anterior trunk vertebrae are about equal
to or less than the hypapophyses in A. latifasciatus, A. nigrilatus, and
A. veraepacis, whereas in A. daryi the spaces between the hypa-
pophyses arc only about half of the hypapophyses. If the sub-
caudals are an accurate indicator of the number of caudal vertebrae,
the number in A. latifasciatus is far greater than in A. daryi, with
A. nigrilatus and A. veraepacis possessing an intermediate number.

Adelphicos daryi has a comparatively inflexible skull that is more
sturdily constructed than those of the other species of highland
Adelphicos. The maxillary in particular is more firmly attached and
has greater articulating surfaces with the prefrontal and ectoptery-
goid. The significance of these morphological traits might well be
elucidated through a knowledge of the habits of these snakes. The
skeletal characteristics and over-all size suggest that the diet of
A. daryi may vary from that of other Adelphicos, which is composed
largely of earthworms.

The following key will serve to distinguish the species of
Adelphicos.

A KEY TO THE SPECIES OF ADELPHICOS

la. Third infralabial absent or greatly reduced in size; chin
shields greatly expanded toward lip _ A. epiadrivirgatus

lb. Third infralabial not so reduced; chin shields not expanded
toward lip 2

2a. Subcaudals more than 42 in males and 32 in females; dorsal
stripe involving paravertebral scale row; venter unpig-
mented A. latifasciatus

2b. Subcaudals less than 42 in males and 32 in females; dorsal
stripe, if present, involving vertebral scale row only; venter
pigmented or not 3

3a. Venter unpigmented, or darkened only midventrally, anterior
edges of ventrals not conspicuously darkened; ground color
bright orange, red, or tan A. nigrilatus

3b. Anterior edges of ventrals conspicuously darkened; ground
color brown or gray 4

4a. Venter not darkened midventrally; ventrals 120 in single
known male, 12S — 132 in females; subcaudals 28 in male,
19 — 22 in females; frontal at least as broad as long A. daryi

4b. Venter darkened midventrally; ventrals 120 — 133 in males;
132 — 142 in females; subcaudals 29 — 41 in males, 24 — 31 in
females; frontal generally longer than broad .. A. veraepacis

DISCUSSION
The decision of whether or not to recognize allopatric popu-

18 OCCASIONAL PAPERS MUSEUM OF NATURAL HISTORY

lations as species is often questionable. We agree with Wiley
( 1981 ) that populations diagnosable without recourse to geographic
data are indicative of independent evolution from their common
ancestor. Further, we are in accord with the concept that an evo-
lutionary species is a single lineage of ancestor-dependent popula-
tions that maintains its identify from other such lineages and that
has its own evolutionary tendencies and historical fate (Wiley,
1978; modified from Simpson, 1961).

Highland populations of Adelphicos exhibit distinct phenetic
differences so that discrimination between them is facile. Lynch
and Smith (1966:69) were understandably conservative in their
assessment of the status of A. v. latifasciatus because they had only
the holotype on which to base their views. It was perhaps the
uniqueness of the specimen that prompted them to speculate "the
taxonomic rank — although . . . not the taxonomic validity — of lati-
fasciatus remains in doubt." They were further hampered by the
notion that A. latifasciatus and A. wgrilatus occurred in physio-
graphic continuity. The highlands of southeastern Oaxaca are sepa-
rated from the range of A. nigrilatus not only by the formidable
barrier presented by the Rio Grijalva entrenchment, but also by
about 120 km of low and unsuitable habitat. We think it unlikely
that the two will be found sympatrically.

On the basis of the salamander fauna, Stuart (1943a) divided
the highlands of Guatemala into five biotic areas. These were sub-
sequently more appropriately designated as "faunal areas" (Stuart,
1950). Adclpliicos veraepacis occurs in three of these: the Cuchu-
matan, the Quecchian, and the Sierran. Later, Stuart (1954b) rec-
ognized the Esquipulan Biotic Area, to which A. daryi is apparently
confined. Adelphicos daryi is isolated from A. veraepacis by the
xeric lowlands of the Rio Motagua Valley — the Zacapan Faunal
Area of Stuart (1954a) — and the seasonally dry uplands of the
Guatemalan Plateau that is covered largely by pine-oak forest — the
Chimaltenangan Faunal Area of Stuart. The importance of the
Guatemalan Plateau as a barrier to dispersal of mesophiles has been
emphasized by Stuart (1951). The type locality of A. daryi is sepa-
rated from the southwestern highlands and Plateau of Guatemala
by the graben in which lies Guatemala City, the Las Vacas Valley,
which scarcely exceeds 1500 m at the ridge separating the head-
waters of the Rio Las Vacas and Rio Michatoya.

In Guatemala the most northerly mountain ranges block most
of the precipitation. Adelphicos veraepacis occurs in localized
areas of high rainfall. To the south A. daryi occurs in a localized
area of humid pine-oak forest. As noted by Stuart (1954b), most
of the southeastern highlands of Guatemala are subhumid; however
in a few localized areas of high elevation the low temperatures
greatly increase the relative humidity and heavy dews occur even

MIDDLE AMERICAN ADELPHICOS 19

in the dry season resulting in much more mesic conditions than
those at lower elevations. These "temperate-cold upland forests"
(Stanley and Steyermark, 1945; Steyermark, 1950) are characterized
by abundant epiphytes and a deep humus layer. Environmental
conditions within this forest are similar to cloud forest (Stuart,
1954b).

With the possible exception of a few high ridges on the Las
Nubes block about 10 km to the east of the type locality and on
which A. daryi probably occurs, the type locality of A. daryi is far
removed from other mesic forests. Inasmuch as all specimens of the
type series of A. daryi were secured within close proximity of water
and in heavy cover, it seems that this species is a mesophile pre-
ferring extremely humid conditions.

The region around San Jorge Muxbal was almost totally cleared
of native vegetation during the period from 1974 to 19S0; A. daryi
may soon be eliminated from the area.

Highland species of Adelphicos occupy distinct physiographic
regions drained by separate rivers or distinct portions of a drainage
system. Adelphicos latifasciatus occupies a range drained by the
Rio Coatzacoalcos and Rio Grijalva; A. nigrilatus by the Rio
Grijalva and Rio Usumacinta; A. veraepacis by the Rio Polochic and
Rio Usumacinta; and A. daryi by tributaries of the Rio Pinula
which drains into Lake Amatitlan. Adelphicos daryi is distinctive
by being the only member of the veraepacis group inhabiting a
Pacific drainage.

The level of endemism in the southeastern Guatemalan highlands
is not high. It seems as though many species have not been isolated
for sufficient time and/ or have not been subjected to sufficiently
strong selective pressures to reach a specific level of divergence.
However, perhaps the most important contributing factor is the
extremely limited distribution of isolated, high peaks with humid
forests in the southeast. One salamander, Pseudoeurycea expectata,
is endemic, and A. daryi is the first reptile reported whose range
is confined to the southeastern highlands, although further investi-
gations of disjunct populations occurring in the southeastern and
southwestern highlands may reveal more endemics.

The dispersal of inhabitants of mesic highland forests in southern
Mexico and northern Central America was accounted for by alti-
tudinal climatic fluctuations in the Pleistocene (Duellman, 1960).
Stuart (1950) suggested that a combination of world cooling and
the extensive uplift in Nuclear Central America caused the herpeto-
fauna of the region to diverge into highland and lowland groups.
The distributions (Fig. 1), characteristics, and hypothesized rela-
tionships (Fig. 7) of the highland species of Adelphicos suggest
that their ancestor probably originated in the northwestern moun-
tains of Nuclear Middle America subsequent to the Pliocene orog-

20 OCCASIONAL PAPERS MUSEUM OF NATURAL HISTORY

env that split off highland and lowland stocks. According to our
concept of their relationships, A. latifasciatus is the most primitive
species and progressively more derived species are found to the
east of its range. It follows that dispersal and subsequent speciation
has proceeded from extreme southeastern Oaxaca eastward. It
seems probable that the ancestral populations of A. daryi gained
access into the southeastern Guatemalan highlands via the Guate-
malan Plateau during periods of Pleistocene climatic depression
when more mesic conditions prevailed. On the basis of suggestive
physiographic features, Stuart (1954b) hypothesized that the south-
eastern highlands were considerably more humid in the past than
at present, a conclusion that is borne out by zoogeographic patterns
of the herpetofauna.

Stuart (1951) did not report Adelphicos from the southwestern
Guatemalan highlands, and the genus is absent from most of the
Guatemalan Plateau. However, a specimen ( UMMZ 127837) from
the highlands of Chimaltenengo establishes the presence of the
genus in at least one restricted area, and at once creates a problem
and suggests an explanation. The Chimaltenengan specimen lends
evidence that the region was considerably more mesic in the past
allowing for dispersal of mesophiles and that the ancestor of A.
danji could have entered the Las Nubes region via the Guatemalan
Plateau. However, A. daryi most closely resembles A. veracpacis
and the Chimaltenengan specimen appears to have its affinities
with A. nigrilatus. Although it is a female possessing 135 ventrals,
the upper limit for A. ivgrilatus, this specimen agrees well in other
characteristics with A. nigrilatus. Most notably, it has a light orange
ground color, an immaculate venter, and lateral stripes on the
upper part of scale row 2, all of 3, and the lower part of row 4.
On the basis of this single specimen and because of its seemingly
puzzling distribution, we are hesitant to assign it to A. nigrilatus.
The status of this isolated highland population hopefully will be
clarified bv the collection of additional material.

J

The highlands of the Meseta Central of Chiapas, Sierra de los
Cuchumatanes, and the Chuaeus-Minas-Mieo range have continued
to increase in elevation up to the present time producing severe
rain-shadow conditions in the interior valleys (Stuart, 1954a) and
having a drying effect on the Guatemalan Plateau in general. It
has been suggested that the Las Nubes region reached a high eleva-
tion at about the same time as the southern volcanic highlands, the
southwestern highlands, and the Sierra de los Cuchumatanes (Stu-
art, 1954b). Today, some populations of highland Adelphicos arc
isolated by portions of Stuart's (1954) subhumid corridor. The
range of A. daryi is cut off to the north by the dry Rio Motagua
and Rio Negro valleys, which are covered with "desert-chaparral"
(Steyermark, 1950). Adelphicos latifasciatus is isolated by the

MIDDLE AMERICAN ADELPIIICOS 21

xeric Middle Grijalva Valley to the cast-northeast that supports a
vegetation of savanna, dry forest, and brushland ( Miranda, 1952;
Stuart, 1954a).

The Quecchian and Cuchumatan populations of A. vcracpacis
are weakly differentiated from one another. These two populations
are isolated from each other by the deeply entrenched Rio Negro
gorge, but the capture of the upper course of the Rio Polochic by
the Rio Negro has been verified (Sapper, 1937), and may have
been the basis for the vicariance event that separated them. The
weak level of divergence between the two populations suggests that
their separation was recent relative to that between A. veraepacis
and A. daryi,

ACKNOWLEDGMENTS

We thank the officials of the Instituto Nacional Forestal
(INAFOR) in Guatemala and the Direction General de la Fauna
Silvestre in Mexico who kindly issued collecting permits. For allow-
ing us to examine material in their care we are indebted to C. W.
Myers (AMNH), R. C. Drewes and A. E. Leviton (CAS), C. J.
McCoy (CM), H. Marx (FMNH), W. E. Duellman (KU), R. L.
Rezy and J. W. Wright (LACM), E. E. Williams (MCZ), H. W.
Greene and D. B. Wake (MVZ), J. R. Dixon (TCWC), S. -K. Wu
(UCM). P. Meylan (UF), A. G. Kluge and R. A. Nussbaum
(UMMZ), W. R. Heyer and F. I. Irish (USNM), and William F. Py-
burn (UTA). We are grateful to William E. Duellman, Darrel Frost,
and Larry David Wilson for their thoughtful comments on the man-
uscript and to Rafael Joglar for translating the Spanish summary.
Harry W. Greene kindly provided the photograph of A. nigrilatus.
Finally, we express our debt of gratitude to Licenciado Mario Dary
Rivera, the late Rector of the University of San Carlos, who was as-
sassinated on December 15, 1981. He aided us in innumerable ways
and always made our visits to Guatemala most memorable and pleas-
ant. His wisdom, foresight, and sense of humor will be sadly missed.

RESUMEN

Una nueva especie de Adelphicos se describe de las tierras altas
que flanquean el sureste de la Ciudad de Guatemala en el departa-
mento de Guatemala. Esta especie posee una escamacion, colora-
tion, y osteologia distintiva y es notable por su gran tamano. Un
analisis de la variation morfologica de las poblaciones disyuntas
de tierras altas de Adelphicos en el sur de Mexico y Guatemala
revelan diferencias consistentes de suficiente peso que justifican el
reconicimiento de estas poblaciones como especies distintas. Estas
especies juntamente con la especie que aqui se describe, consti-
tuyen un linaje monofiletico conocido como el grupo veraepacis,

22 OCCASIONAL PAPERS MUSEUM OF NATURAL HISTORY

endemico a las tierra altas de el sur de Mexico y el norte de Centro
America.

LITERATURE CITED

Dowlixg, H. G. 1951. A proposed standard system of counting ventrals in

snakes. Brit. J. Herpetol. 1:97-99.
Duellman, W. E. 1960. A distributional study of the amphibians of the

Isthmus of Tehuantepec, Mexico. Univ. Kansas Publ. Mus. Nat. Hist.

13:19-72.
Lynch, J. D. and H. M. Smith. 1966. New or unusual amphibians and

reptiles from Oaxaca, Mexico, II. Trans. Kansas Acad. Sci., 69:58-75.
Miranda, F. 1952. La vegetacion de Chiapas. 2 vols. Tuxtla Gutierrez, Dept.

Prensa y Turismo.
Sapper, K. 1937. Mittelamerika. In: Handbuch der Regionalen Geologic O.

Wilckens, ed., Band 8, Heft 29. Heidelberg, Germany.
Simpsox, G. G. 1961. Principles of animal taxonomy. Columbia University

Press, New York.
Smith, H. M. 1942. A review of the snake genus Adelphicos. Proc. Rochester

Acad. Sci., 8:175-195.
Stanley, P. C. and J. A. Steyermark. 1945. The vegetation of Guatemala.

In Plants and plant scientists in Latin America. Waltham: Chronica

Botanica Co. Pp. 275-278.
Steyermark, J. A. 1950. Flora of Guatemala. Ecology 31:368-72.
Stuart, L. C. 1941. Some new snakes from Guatemala. Occ. Pap. Mus. Zool.

Univ. Mich., 452:1-7.
. 1943a. Taxonomic and geographic comments on Guatemalan sala-
manders of the genus Oedipus. Misc. Publ. Mus. Zool. Univ. Mich.,

56:1-33.
. 1943b. Comments on the herpetofauna of the Sierra de los Cuchu-

matanes of Guatemala. Occ. Pap. Mus. Zool. Univ. Mich., 471:1-28.
. 1950. A geographic study of the herpetofauna of Alta Yerapaz,

Guatemala. Contr. Lab. Vert. Zool., 45:1-77.
. 1951. The herpetofauna of the Guatemalan Plateau, with special

reference to its distribution in the southwestern highlands. Contr. Lab.

Vert. Zool., 49:1-71.
. 1954a. A description of a subhumid corridor across northern Central

America, with comments on its herpetofauna] indicators. Contr. Lab.

Vert. Zool., 65:1-26.
. 1954b. Herpetofauna of the southeastern highlands of Guatemala.

Contr. Lab. Vert. Zool., 68:1-65.
. 1966. The environment of the Central American cold-blooded verte-
brate fauna. Copeia 1966:684-699.
Wiley, E. O. 1978. The evolutionary species concept reconsidered. Syst.

Zool. 27:17-27.
Wiley, E. O. 1981. Phylogenetics, the theory and practice of phylogenetic

systematics. John Wiley and Sons: New York.

University of Kansas Publications

MUSEUM OF NATURAL HISTORY

The University of Kansas Publications, Museum of Natural
History, beginning with volume 1 in 1946, was discontinued with
volume 20 in 1971 . Shorter research papers formerly published in the
above series are now published as Occasional Papers, Museum of
Natural History. The Miscellaneous Publications, Museum of Natural
History, began with number 1 in 1946. Longer research papers are
published in that series. Monographs of the Museum of Natural
History were initiated in 1970. All manuscripts are subject to critical
review by intra- and extramural specialists; final acceptance is at the
discretion of the publications committee.

Institutional libraries interested in exchanging publications may
obtain the Occasional Papers and Miscellaneous Publications by
addressing the Exchange Librarian, The University of Kansas Li-
brary, Lawrence, Kansas 66045. Individuals may purchase separate
numbers of all series. Prices may be obtained upon request addressed
to Publications Secretary, Museum of Natural History, The Univer-
sity of Kansas, Lawrence, Kansas 66045.

Editor: E. O. Wiley
Managing Editor: Joseph T. Collins

printed by

university of kansas printing service

lawrence, kansas

j/0 i \k

3 2044 093 361 699

DATE DUE

ARfcX 3 2QQ4

DEMCO, INC. 38-2931