SANi

6(^44

HARVARD UNIVERSITY

Library of the

Museum of

Comparative Zoology

-.-//•

Systematics of the ntclanacephaia Group
of the Colubrid Snake Genus TuntiUa

LARRY DAVID WILSON

Miami-Dade Community College

and

CESAR E. MENA

Miami, Florida

SAN DIEGO

SOCIETY OF NATURAL HISTORY

Memoir 11

1980

Systeinatics of the melanacephala Group
of the Colubrid Snake Genus Tantillta.

LARRY DAVID WILSON

Miami-Dade Community College

and

CESAR E. MENA

Miami, Florida

SAN DIEGO

SOCIETY OF NATURAL HISTORY

Memoir 11

7 May 1980

CONTENTS

Abstract 5

Introduction 5

Color Pattern Terminology 5

Delineation of the melanocephala Group 6

Species Discrimination in the melanocephala Group 6

The Tantilla melanocephala Population System 6

Color Patterns 6

Scale Counts 12

Relative Tail Length 17

The Status of the Names Associated with the

Tantilla melanocephala Population System 18

Species Accounts 21

Key to the Members of the Tantilla melanocephala Group 30

Distributional Patterns 31

Relationships and Evolution 34

Acknowledgments 38

Literature Cited 39

Figures 44

Systematics of the melanocephala Group
of the Colubrid Snake Genus Tantilla

Larry David Wilson and Cesar E. Mena

Abstract. The Tantilla melanocephala group is herein
recognized as comprising 6 species. These are Tantilla andi-
nista, sp. nov., T. capistrata, T. equatoriana, sp. nov.. T. insu-
lamontana, sp. nov., T. lempira, sp. nov., and T. melano-
cephala. The nominal species Tantilla armillata, T. fraseri, T.
longifrontalis, T. mexicana, and T. ruficeps are synonymized
with T melanocephala.

Tantilla melanocephala is composed of 6 pattern groupings
(A-F), 2 of which (C and D| are sympatric. The status of the 1 1
names associated with the T. melanocephala population
system is discussed.

A key for the identification of the 6 species of the T.
melanocephala group and the 6 pattern groupings of T
melanocephala is provided.

The distribution of the Tantilla melanocephala group
encompasses most of Central America and much of South
America. The nominate species occupies most of the range of
the group and the other 5 species are scattered along the
periphery of the northern and western portions of the range.
A wide range of physiographic types is inhabited by members
of the group ranging from coastal lowlands to the high moun-
tains of the Andes. The altitudinal range is from sea level to
3080 m. A broad range of phytogeographic types is inhabited
as well, ranging from pineland and tropical lowland rain

forest to montane rain forest and from coastal deserts to
semiarid intermontane basins.

Study of the relationships and evolution of the members of
the Tantilla melanocephala group indicates that climatic fluc-
tuations in the Pleistocene and Holocene epochs have acted
alternately to fragment and rejoin segments of the group so
as to produce a picture of varying degrees of relationships of
the included taxa one to another. Data derived from study of
evolutionary trends suggest that the pattern E group of T.
melanocephala, now occurring in the upper reaches of the
Amazon Basin, is closest to the evolutionary stock of the
group. Tantilla equatoriana and T. insulamontana appear to
be early offshoots of the basal stock isolated by the uplift of
the Andes in the upper PUocene and lower Pleistocene. Tan-
tilla andinista may have been isolated in an Andean inter-
montane basin at around the same time. The pattern A group
of T. melanocephala — T. lempira stock then may have moved
into Central America during an arid interglacial phase. The
pattern B and F groups appear to have branched off next,
becoming isolated from other members of the group and from
one another in different forest refuges in northern Colombia.
The pattern C-D stock next moved throughout the Amazon
Basin during a wet glacial phase and gave rise to T.
capistrata, which became isolated in the coastal desert and
the region of the Porculla Pass in Peru.

Introduction

The melanocephala group of the colubrid snake
genus Tantilla has been particularly void of any
illuminative scrutiny for most of its history. This
history has been characterized largely by the
addition of new species to the group within the
period from 1758 to 1896. The fundamental ques-
tions concerning the taxonomy of this group of
Tantilla have remained unanswered during the
present century and what work has been done on
the group has dealt with only small segments of
the overall problem (e.g.. Smith, 1942; Peters,
1960).

This study is concerned with the entire scope of
the melanocephala group (as defined herein-see
below) and is one of a series of papers dealing with
the Neotropical species of Tantilla. This study
began with a consideration of the status of Tan-
tilla armillata Cope in connection with the work
by John R. Meyer and the senior author on the
Honduran snake fauna. The study quickly
expanded to include all species of Tantilla with a
dark middorsal stripe. At an early stage of this
project we segregated a group of 4 closely related

species of Mexican Tantilla with a dark middorsal
stripe from others in Central and South America
and considered the former (calamarina) group in
the Meyer and Wilson work mentioned above.
The present paper thus is restricted to the consid-
eration of those species we consider to comprise
the melanocephala group.

Color Pattern Terminology
Due to the relatively complex nature of the
color pattern of the head and neck of members of
the Tantilla melanocephala group, it is necessary
to provide the following definitions:

1. Dark head cap— the portion of the head pat-
tern occupying most of the dorsal portion of
the head;

2. Pale nuchal band— the pale band, which may
be divided medially or medially and laterally,
located on the posterior portion of the pari-
etals and a variable number of dorsal scales
posterior to the parietals;

3. Dark nape band— the dark band, of variable
length (absent in one species), following the
pale nuchal band;

4. Pale neck band— the narrow pale band follow-
ing (if present) the dark nape band.

Delineation of the melanocephala
Group

The term "melanocephala group" has been
used only 3 times in the literature, so far as we are
aware. The term was first used by Smith (1942),
who included T. armillata, T. deppei, and T. mex-
icana. Mertens (1952) used the phrase and indi-
cated that T. armillata is a member. Finally,
Stuart (1963) included T. mexicana and T. armil-
lata. In all cases, obviously, the implication is
that T. melanocephala is also a member.

Other scattered pieces of information from the
literature indicate tacitly that 6 nominal species
comprise the melanocephala group: Tantilla
armillata Cope, 1876a (type locality: "middle
Costa Rica"); T. fraseri Giinther, 1895 (Quito,
Ecuador and "W. Ecuador"); T. longifrontalis
(Boulenger), 18966 (Call, Colombia); T. melano-
cephala (Linnaeus), 1758 ("America"); T. mexi-
cana (Gunther), 1862 ("Mexico"); T. ruficeps
(Cope), 1894 ("Costa Rica"). Tantilla melano-
cephala has been considered to have 2 subspecies,
T. m. melanocephala and T. m. capistrata (the
latter was described as a distinct species by Cope,
18766, from a specimen from the valley of
Jequetepeque, Peru).

Most of these nominal species agree with one
another in possessing a dark middorsal stripe,
generally confined to the middorsal scale row
(absent in some specimens of T. capistrata and T.
melanocephala), a dark head cap separated from a
dark nape band by a pair of variously-sized
nuchal spots or a complete or medially divided
nuchal band (dark nape band absent in one
species). Some specimens from Amazonian South
America and northwestern Peru lack a middorsal
stripe. In such cases, however, the head pattern is
the same as in striped specimens from the same
or surrounding areas. There are also present a
pale preocular and postocular spots separated by
a dark subocular blotch. All members have 7
supralabials and 2 postoculars. Most members
have a pale lateral stripe (except T. capistrata and
some T. melanocephala).

Study of material of this group has demon-
strated that all the nominal taxa of this group
listed above (except T. capistrata) are actually
part of one wide-ranging variable species. Recent
collections, however, have demonstrated the
presence of additional new species, which are des-
cribed below.

Species Discrimination in the
melanocephala Group

In this study, we divided the available material
into several groups on the basis of color pattern.
Eleven distinct color patterns were thus discrim-
inated (see diagnoses below). Some of these pat-
terns could be associated with that described for
certain nominal species (T. armillata, T. fraseri, T.
longifrontalis. and T. ruficeps) or subspecies (T.
capistrata) and others appeared to belong to
undescribed taxa. In the absence of type material,
the pattern characteristics of T. melanocephala
(sensu stricto) could not be determined. This
problem proved unimportant as all of the above
taxa (except T. capistrata) are part of a single
variable and widely distributed species for which
the name T. melanocephala is used (see reasoning
below). An additional currently recognized
nominal taxon (T. mexicana) proved to be based
on a specimen not exhibiting a unique color pat-
tern and having an incorrect type locality.

The next step of the discrimination analysis
involved examination of the material of the
various color pattern types for evidence of sym-
patry or intergradation. No evidence of sympatry
was found but evidence of intergradation was and
this evidence is used in defining species limits.

Using this approach, the pattern that emerged
was one involving a wide-ranging species (T.
melanocephala) occurring from Guatemala to
Argentina within which 6 distinct color pattern
assemblages (2 of which are sympatric with one
another) could be identified. Most of these color
pattern types have formed the basis in the past
for the recognition of species-level taxa within the
melanocephala group. One currently-recognized
name (T. longifrontalis) proved to be based on
material from a population intergradient between
pattern types within T. melanocephala.

Additional satellite species not connected to
the wide-ranging T. melanocephala by intergrada-
tion were found to occur in lowland and highland
locales in Ecuador, in lowland coastal habitats in
northwestern Peru, and in the pine uplands of
south-central Honduras.

The Tantilla melanocephala
Population System
Color pattern has proved to be the key to
unraveling relationships within the T melano-
cephala population system. And, indeed, charac-
teristics of pattern have been used in the past to
diagnose the species-level taxa that are herein

considered to be part of T. melanocephala. On the
basis of pattern, the T. melanocephala population
system can be seen to be composed of 5 units that
show geographical integrity. For purposes of
analysis one of these units is considered to house
2 pattern morphs, a striped and nonstriped form.
The ranges of these population units are des-
cribed in the discussion of color pattern types
below. Intergradation between populations, as
defined by intermediacy between color pattern
types, as well as in certain scutellational
characters, is also discussed below. Discussion of
variation in scutellation follows the discussion of
color pattern types.

Color Pattern.— There are 6 separate identifi-
able patterns within the species T. melanocephala
which show varying degrees of geographical
integrity and most of which have formed the
basis for the recognition of taxa. These 6 patterns
are described as follows (species name, if any,
applied to the pattern types indicated in
parentheses):

(1) pattern A C'armillata")— this pattern (Figs.
1, 6 A), which is found in specimens from princi-
pally along the Pacific versant (Atlantic versant
in Honduras, Nicaragua, and Costa Rica) from
Guatemala to northern Costa Rica (Fig. 15), is
characterized by: (a) a dark brown head cap
(89.4% of specimens), with usually a small
amount of pale pigment on the anterior or antero-
medial portion of the internasals and pale spots
near the junction of the supraocular, frontal, and
parietal; (b) a pair of small, pale nuchal spots,
usually located on the posterior portion of the
parietals (82.6% of specimens), but sometimes
extending also onto a small portion of the
posterior temporal and/or adjacent postparietal
scale (17.4%); (c) a dark brown nape band 3'/2-6
middorsal scales long (^=4.9); (d) tan (56.3% of
specimens), pale brown (33.3%), or brown (10.4%)
ground color, with narrow dark brown middorsal
stripe confined to middorsal scale row; (e) dark
subocular blotch completely separating (41.9% of
specimens) preocular and postocular pale spots or
not (58.9%); (f) lateral extension of dark head cap
usually present, separating the postocular pale
spot from the pale pigment on the lateral gulars
(opposite the pale nuchal spots) completely
(55.3% of specimens) or not (44.7%); (g) pale neck
band following the dark nape band '/2-2 scales in
length (x' = l.l), in turn bounded posteriorly by a
series of dark brown spots (91.7% of specimens);
(h) pale lateral band usually poorly indicated,
occupying the upper half of row 3 and the lower

half of 4, but frequently absent (31.9% of speci-
mens)—specimens from Honduras show a high
incidence of a lack of a pale lateral stripe (83.3%),
as opposed to the other countries (14.3%), if pre-
sent, not bounded by dark pigment.

(2) pattern B ("ruficeps")— this pattern (Figs. 2,
6B), which is found in specimens from principally
along the Atlantic versant (Pacific versant in
Costa Rica) in Nicaragua (exact locality and ver-
sant unknown), Costa Rica, and extreme western
Panama (Fig. 15), is characterized by: (a) a tan
(11.1% of specimens), pale brown (77.7%), or
brown (11.2%) head cap, with no pale pigment on
snout; (b) an inconspicuous pair of pale nuchal
spots (on posterior portion of parietals and a por-
tion of the posttemporal and adjacent post-
parietal scales), if present (not present in 55.5% of
specimens); (c) a brown to dark brown nape band
3-6 middorsal scales long (T=4.6), sometimes
absent (11.1% of specimens); (d) tan (22.2% of
specimens), pale brown (11.1%), or brown (66.7%)
ground color, with well-developed dark brown
middorsal stripe confined to middorsal scale row;
(e) dark subocular blotch, if present, poorly
developed, consisting of dark pigment along
suture lines between supralabials 3, 4, and 5; (f)
lateral extension of dark head cap usually absent
(66.7% of specimens), but sometimes present
(33.3%) and completely separating the postocular
pale spot from the pale pigment on the lateral
gulars (opposite the pale nuchal spots, if the lat-
ter are present); (g) pale neck band following the
dark nape band absent; (h) pale lateral band
always present, well-developed and contrasting
well with surrounding color, occupying from as
little as the upper half of row 3 and the lower half
of row 4 (22.2% of specimens) to as much as the
upper half of row 3, all of row 4, and the lower half
of row 5 (44.5%)— 33.3% of specimens have the
stripe occupying the upper half of row 3 and all of
row 4, and bounded above and below by dark
brown pigment.

(3) pattern C (stripeless pattern): this pattern,
which is found in southeastern Colombia,
Venezuela, Bolivia, Brazil, British West Indies,
and Argentina (Fig. 14), is characterized by: (a) a
cream (2.1% of specimens), pale brown (16.7%),
brown (47.9%), or dark brown (33.3%) head cap.
with usually pale pigment on the internasals and
frequently also on the prefrontals; (b) nuchal band
complete (2.0% of specimens), divided medially
only (22.4%), or divided both medially and lateral-
ly (75.6%), when the latter the nuchal spots are
present only on the posterior portion of the

parietal (19.4% of specimens), on the posterior
portion of the parietals and adjacent portion of
the posterior temporal (3.2%), on the parietals,
posterior temporal and a portion of 1 adjacent
postparietal scale (58.1%), on the parietals,
posterior temporal and portions of 2 adjacent
postparietal scales (12.9%), or on the parietal,
posterior temporal and adjacent portions of 3
adjacent postparietal scales (6.4%); (c) an usually
dark brown nape band 3-7 middorsal scales long
(^ = 3.9); (d) tan (72.9% of specimens), pale brown
(12.5%), or brown (14.6%) ground color, with a
middorsal dark stripe; (e) dark subocular blotch
touching lip (46.0% of specimens) or not (54.0%);
(f) lateral extension of dark head cap usually pre-
sent and touching lip (80.4% of specimens), but
sometimes not (19.6%); (g) pale neck band follow-
ing dark nape band usually present (82.0% of
specimens), measuring Vi-VA middorsal scales in
length (Y — .IS), not bounded posteriorly by dark
pigment; (h) pale lateral band usually absent
(96.0% of specimens), but present in a few (4.0%),
when present, not bounded by dark pigment.

(4) pattern D (eastern South America): this pat-
tern (Figs. 3, 6C), which is found in southeastern
Colombia, Venezuela, Bolivia, Brazil, British
West Indies, Surinam, Guyana, Paraguay,
Uruguay, and Argentina (Fig. 14), is character-
ized by: (a) a pale brown (16.1% of specimens),
brown (67.8%), or dark brown (16.1%) head cap,
with usually pale pigment on the internasals and
frequently also on the prefrontals; (b) nuchal band
divided medially only (19.2% of specimens) or
divided both medially and laterally (80.8%), when
the latter, the nuchal spots are present only on
the posterior portion of the parietal (13.5% of
specimens), on the posterior portion of the pari-
etals and adjacent portion of the posterior tem-
poral (2.7%), on the parietal, posterior temporal
and a portion of 1 adjacent postparietal scale
(70.3%), on the parietals, posterior temporal and
portions of 2 adjacent postparietal scales (8.1%),
or on the parietal, posterior temporal and por-
tions of 3 adjacent postparietal scales (5.4%); (c)
an usually dark brown nape band 2-5 middorsal
scales long (r = 3.6); (d) tan (50.0% of specimens),
pale brown (24.1%), or brown (25.9%) ground col-
or, with a usually diffuse and oftentimes brown
dark middorsal stripe confined to the middorsal
scale row; (e) dark subocular blotch touching lip
(40.0% of specimens) or not (60.0%); (f) lateral
extension of dark head cap usually present and
touching lip (76.4% of specimens), but sometimes
not (23.6%); (g) pale neck band following dark

nape band usually present (62.5% of specimens),
measuring '2-l'2 middorsal scales in length
(T = .90), not bounded posteriorly by dark pig-
ment; (h) pale lateral band usually absent (63.0%
of specimens), but often present (37.0%), in which
case the stripe usually (75.0% ) occupies the upper
half of row 3 and all of row 4.

(5) pattern E (Amazonian Ecuador and Peru):
this pattern (Figs. 4, 6D), which is found in the
Amazonian drainage of Ecuador and Peru (Fig.
14), is characterized by: (a) a pale brown (1.4% of
specimens), brown (95.8%), or dark brown (2.8%)
head cap, with usually pale pigment on the inter-
nasals and prefrontals; (b) nuchal band divided
medially only (21.3% of specimens) or divided
both medially and laterally (78.7%), when the lat-
ter the nuchal spots are present on the posterior
portion of the parietals and an adjacent portion of
the posterior temporal (1.8% of specimens), on
the parietal, posterior temporal and a portion of 1
adjacent postparietal scale (49.1%), on the
parietal, posterior temporal, and portions of 2
adjacent postparietal scales (41.8%), or on the
parietal, posterior temporal, and adjacent por-
tions of 3 adjacent postparietal scales (7.3%); (c)
an usually dark brown nape band 3-5 middorsal
scales long (Y=3.7); (d) tan (2.5% of specimens),
pale brown (23.5%), or brown (74.0%) ground col-
or, with a well-developed, complete dark middor-
sal stripe confined to the middorsal scale row; (e)
dark subocular blotch touching lip (88.9% of
specimens) or not (11.1%); (f) lateral extension of
dark head cap usually present and touching lip
(97.6% of specimens), but sometimes not (2.4%);
(g) pale neck band following dark nape band
usually present (93. 8%. of specimens), measuring
'^2-2 middorsal scales in length (]r = 1.0), not
bounded posteriorly by dark pigment; (h) pale
lateral band usually present (96.3% of
specimens), in which case the stripe occupies only
the upper portion of row 3 (17.4% of specimens),
the upper half of row 3 to the middle of row 4
(8.1%), the upper half of row 3 and all of row 4
(58.1%), or the upper half of row 3 through the
lower half of row 5 (16.4%), but sometimes absent
(3.7%), bounded by dark pigment.

(6) pattern F {"fraseri"): this pattern (Figs. 5,
6E), which is distributed along the western slopes
of the Andes in Ecuador (Fig. 14), is characterized
by: (a) a brown head cap, with usually no pale pig-
ment on snout (60.0% of specimens), if present,
however, usually occupying all of the internasals
and the anterior one-third to almost all of the
internasals; (b) nuchal band divided medially and

Table 1. Primary distinctions between the color pattern types in Tantilla melanocephala.

Characters

Color pattern types

A

B

C

D

E

F

1.

Ground color of

Brown to

dorsum of head

Dark brown

Pale brown

dark brown

Brown

Brown

Brown

2.

Pale pigment
on internasals

and prefrontals

No

No

Yes

Yes

Yes

No

3.

Pale nuchal

Restricted to

Not present

Divided

Divided

Divided

Divided

spots

parietal scales

or present

medially and

medially and

medially or

medially and

and divided

laterally

laterally

medially and

laterally

medially and

laterally

laterally

4.

Color of dark

Brown to

nape band

Dark brown

dark brown

Dark brown

Dark brown

Dark brown

Brown

5.

Mean length of

dark nape band

4.9

4.6

3.9

3.6

3.7

4.1

6.

Lateral
extension of
dark head cap
present and

touching lip

Yes

No

Yes

Yes

Yes

Yes

7.

Pale neck band

present

Yes

No

Yes

Yes

Yes

Yes

8.

Mean length of

pale neck band

1.1

...

.78

.90

1.0

.78

9.

Pale neck band
bounded
posteriorly by

dark pigment

Yes

No

No

No

No

No

10.

Ground color of

Tan to pale

Tan to

dorsum of body

brown

Brown

Tan

brown

Brown

Brown

11.

Middorsal dark

Yes, diffuse

stripe present

Yes

Yes

No

and broken

Yes

Yes

12.

Pale lateral
band

Present or

absent, but if
present,
poorly
indicated

Present

Absent

Absent

Present

Present

13.

Area covered

Upper half of

Upper half

• •■

...

Upper half

Upper half

by pale

row 3 and

of row 3

of row 3 to

of row 3 to

lateral band

lower half of 4

through all
of row 4 or
lower half of
row 5

upper half of
row 4
through
lower half of

middle of
row 4
through
lower half of

14.

Pale lateral
stripe bounded
by dark
pigment

No

Yes

No

No

row 5

Yes

row 5

Yes

laterally, the nuchal spots covering the posterior
portion of the parietal and a portion of the adja-
cent posterior temporal (6.3% of specimens), the
parietal, posterior temporal, and a portion of 1
adjacent postparietal scale (62.5%), the parietal,
posterior temporal, and a portion of 2 adjacent
postparietals (12.5%), or the parietal, posterior

temporal, and portions of 3 adjacent postpar-
ietals (18.7%); (c) an usually brown nape band 3-5
middorsal scales long {Y — 4.\); (d) pale brown
(17.6% of specimens) or brown (82.4%) ground
color, with a well-developed, complete dark mid-
dorsal stripe confined to the middorsal scale row;
(e) dark subocular blotch touching lip (80.0% of

10

specimens) or not (20.0%); (f) lateral extension of
dark head cap usually present and touching lip
(70.6% of specimens), but sometimes not (29.4%);
(g) pale neck band following dark nape band
usually present (75.0% of specimens), sometimes
confined to one or two scales on either side of the
dark middorsal stripe, when complete, measuring
V2-I middorsal scales in length (x" = .78), but
sometimes absent (25.0%), not bounded posteri-
orly by dark pigment; (h) pale lateral band usually
present (94.1% of specimens), in which case the
stripe occupies only the upper portion of row 3
(6.1% of specimens), the upper half of row 3
through the middle of row 4 (31.3%), the upper
half of row 3 and all of row 4 (31.3%), or the upper
half of row 3 through the lower half of row 5
(31.3%), but sometimes absent (5.9%), bounded
by dark pigment.

The primary distinctions between the color pat-
tern types in Tantilla melanocephala are summar-
ized in Table 1. The characteristics hsted in that
table are the mean or modal characteristics, not
the range of characteristics exhibited in each pat-
tern type.

Two things are evident from a perusal of the
above pattern diagnoses. One, that most of
Panama and all of Colombia have not been
included in any of the range statements for the
various pattern types. This is because we con-
sider this area to be one of integradation between
patterns A and B to the north and patterns C, D,
E, and F to the south and east. This intergrada-
tion is discussed in more detail below. Secondly,
patterns C and D are, in large part, sympatric;
actually, they grade into one another and are
separated purely for the sake of convenience.

When we first began to study the melano-
cephala group, we were perplexed by the material
from Panama, which resembled the nominal
species T. armillata in pattern, but the nominal
species T. ruficeps in ventral and subcaudal
counts. As the study progressed, however, it
became clear that the Panamanian material was
intermediate and, in fact, intergradient between
nominal T. armillata (pattern A) and nominal T.
ruficeps (pattern B). A comparison between the 3
groups of specimens is made in Table 2. The mean
or modal condition of each character is listed in
this table. As can be seen from the table, Panama-
nian specimens agree more closely in pattern with
the members of pattern A than with those of pat-
tern B. Panamanian specimens agree with
members of pattern A in the nature and extent of
the pale neck band, the presence of dark pigment

posterior to the pale neck band, the dorsal ground
color, and the lack of dark pigment on the venter.
The resemblance between the two in pattern and
their distinction from that of the members of pat-
tern B is immediately apparent. Panamanian
specimens resemble the members of pattern B in
the area covered by the pale nuchal spots (if pre-
sent in the latter), the mean length of the dark
nape band, and the presence of dark pigment on
the lateral dorsum (more extensive in the latter).
Panamanian specimens exhibit an intermediate
head color.

The picture in Colombia was initially even more
confused, but again, as the study progressed, it
became apparent that most of Colombia is an
area of intergradation between members of pat-
tern types C and D, E, and F. Also, the same type
of pattern seen in Panamanian specimens is also
seen in specimens from parts of Colombia.

To elucidate pattern variation in Colombia, we
divided the material into several geographic
groupings: (a) Pacific coastal region; (b) Atlantic
coastal region; (c) Cordillera Occidental; (d) Cor-
dillera Central; (e) Cordillera Oriental; (f) Amazo-
nian region. Data on pattern characteristics for
these groupings were then compared to those for
members of pattern groups C, D, E and F. The
summarization follows:

(1) Pacific coastal region.— The few specimens
available from this region all resemble the
Panamanian specimens in pattern. The Pacific
coastal specimens primarily have a brown head
cap with the nuchal spots covering the posterior
portion of the parietals and a portion of the
posterior temporal and one adjacent postparietal
scale, an average dark nape band length of 4.0
scales, pale pigment on all of the internasals and
the anterior one-fourth to one-half of the prefron-
tals, an average pale neck band length of 1.3
scales, pale lateral band covering from the upper
half of row 3 through all of row 4 to the upper half
of row 3 through the lower half of row 5, dark pig-
ment on lower dorsum occupying the upper half
of row 2 and the lower half of row 3, a tan to pale
brown ground color, and no dark pigment on the
venter.

(2) Atlantic coastal region.— The 3 specimens
from this area are all unalike. One specimen
(UMMZ 54957) from Depto. Magdalena
resembles specimens from adjacent Panama. A
second (FMNH 165193) from Depto. Bolivar
strongly resembles specimens from Central
America (Guatemala to Costa Rica-pattern A),
especially in head pattern (small parietal spots,

11

Table 2. Comparison of color pattern characteristics among pattern A, pattern B, and Panamanian specimens.

Characters

Pattern A

Panamanian Specimens

Pattern B

Ground color of

Dark brown

Brown

Pale brown

dorsum of head

Area covered by

Posterior portion

Posterior portion of

Absent or poorly

pale nuchal band

of parietals

parietal and adjacent

indicated on

portions of posterior

posterior portions

temporal and one

of parietal,

postparietal

posterior temporal,
and adjacent
postparietal

Length of dark

nape band

4.9

4.6

4.6

Color of dark

nape band

Dark brown

Dark brown

Dark brown

Amount of pale

Anterior '3 of

Anterior ^ a-^ 4 of

Absent

pigment on snout

internasals

internasals and '4-' i of
prefrontals

Area covered by

1 scale length

1 scale length

1 or 2 scales on

pale neck band

either side of
middorsal line

Dark pigment

posterior to pale

neck band

Present

Present

Absent

Area covered by

Absent or barely

Upper half of row 3 and

WeU-

pale lateral stripe

indicated on

all of row 4

distinguished,

upper half of row

ranges from

3 and lower half

upper half of row

of row 4

3 through lower
half of row 4 to
upper half of row 3,
all of row 4, and
lower half of row 5

Dark pigment on

lateral dorsum

Absent

Present

Present

Ground color of

Tan to pale

Tan to pale

dorsum of body

brown

brown

Brown

Dark pigment on

venter

Absent

Absent

Present

ground color of dorsum of head and dark nape
band the same, well-defined pale neck band, and
band of dark pigment posterior to pale neck band)
and lack of pale lateral stripe. The dark middorsal
stripe is barely discernible, however. The locality
data for this specimen could be incorrect and the
specimen actually could be from Central America,
but the nature of the middorsal stripe and other
scutellational features discussed below casts that
decision in doubt. This particular area needs fur-
ther investigation, as does the whole of the Atlan-
tic coastal region of Colombia. A third specimen
(NMW 20367) from Depto. Atlantico resembles
that of pattern C specimens in all respects, which
is to be expected considering the proximity of its
locality to Venezuela, one of the countries in
which pattern C specimens are found.

(3) Cordillera Occidental.— The 3 specimens
from this area are in relatively poor shape and
their affinities are difficult to judge. A perusal of
their pattern characteristics allies them most
closely with pattern E specimens. This area is
also in need of further study.

(4) Cordillera Central.— Eighteen specimens are
available from this area. Ten specimens show
resemblance to specimens from Panama, 3 to pat-
tern F specimens, 2 to pattern E specimens, and 3
are intermediate between specimens from
Panam^ and pattern F specimens. Most speci-
mens from this area are from Depto. Antioquia,
which appears to be an area of integradation
between Panamanian specimens (which are them-
selves intergradient) and pattern F specimens.
Two specimens from deptos. Caldas and Cauca

12

resemble pattern E specimens.

(5) Cordillera Oriental — Specimens from this
area resemble either Panamanian specimens or
pattern E specimens. The few specimens avail-
able (5) do not give a clear picture of variation in
color pattern and additional material is to be
desired.

(6) Amazonian region. — Specimens from this
area resemble either specimens from Panama
(UTACV 3476 from Depto. Meta), pattern C or D
specimens (MCZ 53236 and UTACV 3535 from
Depto. Vaupes and UTACV 3717 from Depto.
Meta), or pattern E specimens (MCZ 49003 from
Depto. Amazonas and FMNH 37448 from Depto.
Putamayo). These relationsips make geographic
sense as deptos. Putamaya and Amazonas are
close to the Amazonian lowlands of Ecuador and
Peru, where pattern E specimens are found and
deptos. Meta and Vaupes are close to Venezuela
and Brazil from which area pattern C and D speci-
mens are found. This area will, no doubt, prove to
be one of intergradation between Panamanian
specimens, pattern E specimens, and pattern C
and D specimens, when more material becomes
available.

On the basis of color pattern, the entire country
of Colombia is one in which there is a meeting
and, in many cases, a merging of pattern ele-
ments from the north (pattern of Panamanian
specimens), east (patterns C and D), southeast
(pattern E) and south (pattern F). It is primarily
upon this evidence that we consider members of
the pattern types discussed above (patterns A-F)
to constitute a single species. Pattern A and B
specimens in Central America intergrade with
one another in Panam^. The Panamanian inter-
gradient pattern extends into Colombia where it
meets or merges with the patterns found in South
America (patterns C-F).

In summary, 6 pattern types that occupy 5
distinct geographic areas are present within the
Tantilla melanocephala population system. The
status of the various names associated with these
pattern types is discussed below. Intergradation
in color pattern has been shown to occur in
Panama and Colombia between patterns A and B
to the north in Central America and patterns
D-D, E and F to the south and east in the
remainder of the range of the population system
in South America.

Ventrals.— Ventral numbers in the Tantilla
melanocephala population system range from
125-174 in males and 134-177 in females. The
highest counts for males and females are for spec-

imens from Honduras; the lowest are for speci-
mens from the department of Junin in Peru. The
mean values follow the same pattern, except that
the lowest mean for females is for specimens from
the Cordillera Oriental of Colombia (139.0),
followed closely by the mean (141.6) for females
from the eastern slopes of the Andes in Peru
(within which area lies the department of Junin).

As indicated in Table 3, where the sample size
is large enough, females have more ventrals than
males (Y = 7.5).

Ventral numbers are highest in Central
America from Guatemala to northwestern and
central Costa Rica. In this area, ventrals range in
males from 155 to 174 and in females from 158 to
177. The numbers appear to be relatively low in
Guatemala and El Salvador and they increase in
Honduras. They drop again in western Nicaragua
and become elevated again in northwestern and
central Costa Rica.

The ventral numbers drop dramatically on the
Altantic slope and southwestern Pacific slope of
Costa Rica and this trend continues on into
Panama. Compared to numbers in northwestern
and central Costa Rica, there is over a 20-scale
drop in mean ventral number in the area in Costa
Rica mentioned above. Ventral numbers are
similar in Panama to those in Costa Rica. Ventral
range in these lower Central American specimens
is from 139 to 149 in males and 143 to 155 in
females. There is no range overlap between these
specimens and those from upper Central America
(northwestern Costa Rica north).

The numbers increase again in Colombia
(except in the Cordillera Oriental). The range for
Colombia (exclusive of the Cordillera Oriental) is
from 136 to 165 (r = 150.9) for males and 146 to
169 (1F = 158.4) for females. The specimens from
the Cordillera Oriental in Colombia have incon-
gruously low numbers of ventrals. The range for
males is 127-148 (T = 136.0) and for females
135-143 (T = 139.0). The means are = 15 scales
higher in other parts of Colombia than in the Cor-
dillera Oriental. This is interpreted as evidencing
influence from the Amazonian region of Ecuador
where ventral numbers are low as well (see below).

In Venezuela, ventral numbers are relatively
low in the Cordillera Merida (the continuation of
the Cordillera Oriental of Colombia into Vene-
zuela). In other areas, however, numbers are
higher, ranging from 142-162 [Y = 153.7) in males
and 147-167 (Y = 159.2) in females, as compared
to a count of 144 in one male and a range of
142-150 (T = 145.8) in females.

13

Table 3. Variation in the numbers of ventrals and subcaudals in Tantilla melanocephala (range, x , N).

Geographic area

Ventrals

Subcaudals

a-cr

99

crcr

99

Southern Guatemala

160-173(166.5)2

...

44(44.0)2

Central El Salvador'

166-172(168.3)3

170

46-54(50.7)3

45

Honduras

170-174(172.0)2

168-177(171.1)12

54

42-60(47.8)10

Western Nicaragua

155-166(161.0)6

158-173(165.9)9

50-56(53.0)5

44-51(46.6)7

Eastern!?) Nicaragua'

148

59

Pacific slope of northwestern and

163-167(164.9)7

169-174(170.9)7

51-59(55.8)6

48-52(50.0)4

central Costa Rica

Atlantic slope and southwestern

142-148(144.3)4

144-147(145.5)2

64-77(69.7)3

66-75(70.5)2

Pacific slope of Costa Rica

Western Panama

144

143-155(149.0)2

82

63-68(65.5)2

Central and eastern Panam^

139-149(143.5)12

143-154(152.0)10

69-79(75.1)10

66-71(68.3)8

Pacific coastal region of Colombia

159

157-169(164.0)3

...

70-74(72.0)2

Atlantic coastal region of Colombia

150-162(157.7)3

...

55-72(63.5)2

Cordillera Occidental of Colombia

147

160

61

Cordillera Central of Colombia

136-158(150.9)9

146-168(157.3)8

62-92(74.5)6

52-72(64.6)7

Cordillera Oriental of Colombia

127-148(136.0)3

135-143(139,0)2

58-80(71.3)3

55

Amazonian region of Colombia

141-165(150.0)5

152-157(154.5)2

56-69(62.5)4

46

Cordillera de Merida of Venezuela

144

142-150(145.8)4

63

54-56(55.0)2

Caribbean coastal plain of Venezuela

159

154-162(158.0)5

65

53-56(55.0)4

Cordillera de la Costa of Venezuela

142-162(152.6)5

147-167(160.0)7

61-76(66.4)5

53-68(57.8)6

Guiana highlands of Venezuela

149-157(153.7)3

...

53-58(55.5)2

Monaguesa formation of Venezuela

152

Trinidad

137-152(138.8)7

147-159(150.7)13

59-67(63.0)5

47-60(54.5)11

Tobago

157-160(159.0)3

168

73-79(76.0)3

68

Guyana

133-144(136.8)9

140-158(147.3)4

50-66(57.6)7

66

Surinam

139-140(139.5)2

61

Brazil

136-154(143.4)7

141-155(147.9)7

56-63(59.8)6

55-56(55.5)2

Uruguay

155

...

Paraguay

144

...

Argentina

158

167

68

60

Bolivia

126-151(140.0)6

158-165(161.5)2

49-62(55.2)5

49-55(52.0)2

Coastal plain of Ecuador

141-142(141.5)2

50-65(57.5)2

Western slopes of Andes in Ecuador

136-147(142.3)9

144-154(149.6)7

51-72(61.3)7

58-65(62.0)4

Inter-Andean basins of Ecuador

156-158(157.0)2

55-58(56.5)2

(Pacific drainage)

Inter- Andean basins of Ecuador

136

142

59

(Atlantic drainage)

Oriente of Ecuador

134-145(138.0)17

139-150(144.5)8

54-58(56.4)10

51-57(53.3)4

Andean Cordillera in Peru

134-136(135.0)2

143

47

...

East Andean slopes in Peru

125-143(133.9)11

134-153(141.6)12

46-53(49.5)8

41-63(49.1)7

Tropical lowlands of Peru

136-159(144.0)21

142-152(148.4)7

41-62(50.9)16

44-50(46.5)6

'Data on 3 of 4 specimens from this area are from Mertens (1952a).
^Locality uncertain, but specimen thought to be from eastern Nicaragua.

14

Variation in ventrals in the British West Indies
presents an interesting picture. Trinidad is 18 to
19 kilometres from the nearest point on the main-
land of Venezuela in the north and = 13 kilo-
metres in the south (across the delta of the Orin-
oco River) and Tobago is 28 to 29 kilometres from
the nearest point on Trinidad (near Toco). The dif-
ferences in ventral counts between specimens
from Trinidad and adjacent Venezuela, compared
to specimens from Trinidad and Tobago, are
striking. Mean ventral numbers in Trinidad are
lower by approximately 8 to 15 scales than
specimens from adjacent Venezuela. On Tobago,
ventral numbers again increase markedly over
those in Trinidad and are even higher than those
from mainland Venezuela. In fact, only specimens
from upper Central America have higher ventrals
than those of specimens from Tobago. The mean
for ventrals for Trinidad males is 20.2 scales
lower than that for Tobagan males. Comparable
figures for females are not available, but the
count for the single Tobagan female is 9 scales
above the highest count for Trinidad females.

Ventral numbers decrease in the Guianas
(material available only from Guyana and
Surinam) as compared to adjacent areas in
Venezuela, as well as in Brazil, Uruguay and
Paraguay, but increase again in Argentina (the
pattern in the latter 4 countries is poorly
understood due to the paucity of material).

Variation in ventrals in males from Bolivia is
similar to that in other parts of eastern South
America. Females from Bolivia (only 2 specimens
are available), however, have much higher ventral
counts than those of specimens from eastern
South America and are closer to those from the
Cordillera de la Costa of Venezuela. In addition,
the range in Bohvian males (126-151) is the
widest of any of that in any of the areas within
the range.

Ventral numbers decrease in Ecuador and Peru,
compared to those in Colombia (exclusive of the
Cordillera Oriental). Within this area, however,
there are additional patterns worthy of notice.
Specimens from the Pacific coastal plain, western
Andean slopes, and the inter-Andean basins of
Ecuador (Pacific drainage) have more ventrals
than those from the Atlantic drainage inter-
Andean basins and the Amazonian lowlands of
Ecuador (means of 142.2 and 149.6 in males and
females, respectively, in the former and 137.9 and
144.2 in the latter). The numbers of ventrals for
specimens from the Andean Cordillera and the
eastern Andean slopes in Peru are similar to

those from the Amazon drainage of Ecuador, but
those from the Amazonian lowlands of Peru are
higher.

Subcaudals. — Suhcaudal numbers in the Tan-
tilla melanocephala population system range
from 41-92 in males and 41-75 in females. The
highest count for males is that for a specimen
from the Cordillera Central of Colombia and for
females a specimen from the southwestern Pacific
slope of Costa Rica. The lowest count for males is
for a specimen from the tropical lowlands of Peru
and for females a specimen from the eastern
slopes of the Andes in Peru. The mean values do
not show the same pattern, probably primarily
due to atypically high counts being represented
in small samples.

As can be seen in Table 3, where sample sizes
are large enough, males have more subcaudals
than females {Ic =4.3).

Subcaudal numbers are the lowest anywhere
within the population system (with the exception
of the eastern Andean slopes and tropical
lowlands of Peru, where the numbers are about
the same as they are in Central America) in Cen-
tral America from Guatemala to northwestern
and central Costa Rica. In this area, subcaudals
range in males from 46-59 and in females from
42-60. The subcaudal numbers appear to increase
gradually within this area from north to south.

The subcaudal numbers increase markedly on
the Atlantic slope and southwestern Pacific slope
of Costa Rica and this trend continues on into
Panama. Compared to numbers in northwestern
and central Costa Rica, there is a 14-20 scale
increase in mean subcaudal number in the area of
Costa Rica mentioned above. Subcaudal numbers
in Panama are similar to those in the area of
Costa Rica under discussion. Subcaudals range in
these lower Central America specimens from
64-82 in males and 63-75 in females and there is
no overlap of these ranges with those for
specimens from upper Central America.

Subcaudal numbers vary widely in Colombia.
Mean numbers for specimens from the Pacific
coastal region, the Cordillera Central, and the
Cordillera Oriental are similar to those of speci-
mens from Panama but the ranges are much
wider (58-92 in males and 52-74 in females, as
compared to 69-82 and 63-71). Numbers for speci-
mens from the Atlantic coastal region and the
Amazonian region are lower than those for speci-
mens from other areas of Colombia and those
from Panama.

Subcaudal numbers are relatively low in

15

Table 4. Variation in contact of prefrontal and second supralabial in Tantilla melanocephala.

Contact

Geographic area

N

Guatemala

2

El Salvador

4

Honduras

13

Nicaragua

16

Costa Rica

20

Panama

30

Colombia

39

Venezuela

36

Trinidad

20

Tobago

4

Guayana

13

Surinam

2

Brazil

14

Argentina

2

Uruguay

1

Paraguay

1

Bolivia

8

Peru

59

Ecuador

53

None

One side

Both sides

...

2

...

4

1

12

...

16

1

16

...

30

34

3

24

1

17

3

5

4

2

1

3

2

3

19

5

32

5
9
2
1
8
2
10

1
3

37
16

Venezuela and Trinidad, although not as low as
they are in upper Central America. The ranges for
all of Venezuela and Trinidad are 59-76 for males
and 47-68 for females.

As with ventrals, variation in subcaudals in the
British West Indies is interesting. The mean for
subcaudals for Trinidad males is 13 scales lower
than that for Tobagan males. Although a compar-
able figure is not available for females, the count
for the single Tobagan female is 8 scales above
the highest count for Trinidad females.

Subcaudal numbers in the Guianas, Brazil and
Argentina are similar to those in Venezuela, as
best as can be demonstrated with the small
samples from these countries.

Specimens from Bolivia have lower numbers of
subcaudals than other specimens from eastern
South America in both males and females.

Subcaudal numbers decrease in Ecuador, as
compared to those for specimens from Colombia.
Within Ecuador subcaudal numbers are higher
for specimens west of the Continental Divide
than for those east of it (means of 60.5 and 60.2 in
males and females, respectively, in the former
and 56.6 and 53.3 in the latter).

Subcaudal numbers continue to decrease in
Peru, as compared with eastern Ecuador and
average 5-6 scales lower.

Anal plate.— M\ specimens examined have a
divided anal plate (as is characteristic of the
genus Tantilla) except 2 specimens, one from
Panama and another from Colombia.

Dorsal scale rows. — All specimens examined
have 15 dorsal scale rows throughout (which is
characteristic of Tantilla) except for 2 specimens.
One specimen from Bolivia has 13 scale rows on
the neck and another from Peru has 17 scale rows
at midbody and at the vent.

Supralabials.—AU specimens have 7 supra-
labials, with the 3rd and 4th entering the orbit,
with a very few exceptions. Five specimens have
6-7 supralabials, one has 5-7 (supralabials 5, 6 and
7 fused on left side), and one has 7-8. All speci-
mens showing the variances are from South
America, albeit from widely divergent localities.

In fralabials.— The majority of the specimens
examined (92.8%) have 6 infralabials, with 4 in
contact with the anterior chin shields and the 4th
the largest. The range in infralabial number, how-
ever, is 4-7 and as few as 3 or as many as 5 infra-
labials may be in contact with the anterior chin
shields.

Contact of prefrontal and second supra-
labial.— This character has frequently been used
as a distinguishing feature between species of
Tantilla and the melanocephala group is no excep-

16

Table 5. Variation in medial contact of first pair of infralabials in Tantilla melanocephala.

Geographic area

N

Medial contact

Absent

Present

Guatemala

El Salvador

Honduras

Nicaragua

Costa Rica

Panama

Colombia

Venezuela

Trinidad

Tobago

Guayana

Surinam

Brazil

Argentina

Uruguay

Paraguay

Bolivia

Peru

Ecuador

1

4

13

16

20

30

40

35

20

4

13

2

14

2

1

1

8

59

53

4
2
2
1
8

2

28
14

1

4

9

14

18

29

32

35

20

4

11

2

12

2

1

1

6

31

39

tion (see Peters and Orejas-Miranda, 1970: 293).
The pattern of variation in this character (Table
4) does not support its use as a species or sub-
species distinguishing characteristic. Most speci-
mens from Central America, Colombia, Venezuela
and the British West Indies (85.9%) have the
prefrontal and second supralabial in contact on
both sides of the head, separating the postnasal
and preocular scales. Conversely, a large percen-
tage of specimens from Guyana, Surinam, Brazil,
Paraguay, BoHvia, and Peru (62.9%) have no con-
tact between the prefrontal and second supra-
labial. The pattern in Uruguay and Argentina is
not clear but available material agrees with the
condition in the northern segment of the range.
Specimens from Ecuador are, to some extent,
intermediate between those to the north and
those to the south. They agree with specimens to
the north in having the highest percentage of
specimens showing contact between the prefron-
tal and second supralabial on both sides,
although the percentage is considerably lower
than in adjacent Colombia (87.2%), for example.
Ecuadoran specimens showing no contact consti-
tute 30.2% of the sample, a figure intermediate
between that for Peru (62.7%) and that for Colom-
bia (12.8%).

Medial contact of first pair of infralabials. — In
most portions of the T. melanocephala population
system the infralabials of the first pair are in
medial contact (Table b) in the majority of speci-
mens (89.7%), but in Peru and Ecuador this per-
centage drops to 62.5 and a corresponding 37.5%
of the specimens have the first pair of infralabials
separated. In fact, in Peru the percentages are
almost equal to one another (52.5% with medial
contact, 47.5'^r without) and specimens are inter-
mediate between those from Peru and those from
the remainder of the range. Variation in this char-
acter also does not support its use as a species or
subspecies distinguishing characteristic.

Preoculars. — AW specimens examined, except 4,
have a single preocular on both sides of the head.
Three specimens, one each from Nicaragua,
Tobago, and Peru, have no preocular on one side
of the head and one specimen from Costa Rica has
2 on both sides.

Postoculars.— All specimens examined, except
5, have 2 postoculars on both sides of the head.
Two specimens, one from Peru, the other from
Venezuela, have but a single postocular on one
side. A specimen from Guyana has the lower post-
ocular fused with the 4th supralabial; another
from Peru has it fused with the 5th supralabial.

17

Table 6. Variation in relative tail length in Tantilla melanocephala.

Geographic area

Relative tail length

orO"

9 9

Southern Guatemala

Central El Salvador

Honduras

Western Nicaragua

Eastern!?) Nicaragua

Pacific slope of northwestern and central Costa Rica

Atlantic slope and southwestern Pacific slope of Costa Rica

Western Panama

Central and eastern Panama

Pacific coastal region of Colombia

Atlantic coastal region of Colombia

Cordillera Occidental of Colombia

Cordillera Central of Colombia

Cordillera Oriental of Colombia

Amazonian region of Colombia

Cordillera de M6rida de Venezuela

Caribbean coastal plain of Venezuela

Cordillera de la Costa of Venezuela

Guiana highlands of Venezuela

Trinidad

Tobago

Guyana

Surinam

Brazil

Argentina

Bolivia

Coastal plain of Ecuador

Western slopes of Andes in Ecuador

Inter-Andean basins of Ecuador (Pacific drainage)

Inter-Andean basins of Ecuador (Atlantic drainage)

Oriente of Ecuador

Andean Cordillera in Peru

East Andean slopes in Peru

Tropical lowlands of Peru

0.179-0.182(0.181)2

0.210

0.205

0.181-0.202(0.190)8

0.180-0.229(0.210)6

0.168-0.202(0.185)6

0.237

0.189-0.225(0.208)6

0.158-0.209(0.187)4

0.230-0.309(0.275)4

0.274-0.294(0.284)2

0.299

0.252-0.259(0.256)2

0.272-0.306(0.292)9

0.245-0.270(0.259)7

...

0.245-0.266(0.256)2

0.216-0.251(0.234)2

0.247

0.253-0.324(0.281)6

0.213-0.252(0.234)7

0.269-0.337(0.301)3

0.252

0.202-0.257(0.238)3

0.195-0.220(0.208)2

0.256

0.223-0.231(0.227)2

0.265

0.216-0.233(0.226)4

0.238-0.292(0.263)4

0.215-0.248(0.234)6

0.200-0.218(0.209)2

0.219-0.259(0.242)5

0.189-0.236(0.207)11

0.222-0.275(0.256)3

0.241

0.235-0.254(0.248)7

0.252

0.246

0.223-0.263(0.242)6

0.224-0.232(0.228)2

0.259

0.231

0.240-0.257(0.246)5

0.200-0.218(0.209)2

0.227-0.275(0.251)2

0.193-0.359(0.262)7

0.246-0.282(0.263)4

0.223

0.273

0.232

0.229-0.275(0.257)9

0.215-0.249(0.233)4

0.228

0.190-0.247(0.224)8

0.186-0.274(0.218)7

0.192-0.258(0.224)15

0.181-0.215(0.201)6

One specimen from Peru has the left upper
postocular fused with the supraocular.

Temporals. — AW specimens have 1 + 1 tem-
porals, although many specimens (exact number
not determined) have the posterior temporal
divided vertically in two.

Relative tail length. — Relative tail length in the
Tantilla melanocephala population system ranges
from 0.180-0.359 in males and 0.158-0.294 in
females (Table 6). The highest value in males is
for a specimen from the western slopes of the
Andes in Ecuador and for females a specimen
from southwestern Costa Rica. The lowest count
for males is for a specimen from western

Nicaragua, and for females a specimen from
western Nicaragua. The mean values do not show
the same pattern, due primarily to the occurrence
of some extreme values in populations otherwise
characterized by less extreme values. The highest
mean value for males is for specimens from the
Cordillera Oriental in Colombia and for females
for specimens from the Atlantic slope and south-
western Pacific slope of Costa Rica. The lowest
mean values for males are for specimens from
northwestern and central Costa Rica, and for
females for specimens from southern Guatemala.
As may be seen from Table 6, where sample
sizes are large enough, males have a greater

18

relative tail length than females, with one excep-
tion. The exception involves specimens from the
western slopes of the Andes in Ecuador where the
mean values are virtually identical. The average
percentage value difference between males and
females is 0.028.

Relative tail length values are lowest in
western Central America from southern Guate-
mala to northwestern and central Costa Rica in
both sexes. There is a sharp increase in specimens
from the Atlantic slope of Nicaragua and Costa
Rica, southwestern Costa Rica and all of Panama.
The mean values are: northern males, 0.209;
southern males, 0.288; northern females, 0.187;
southern females, 0.261. This trend of high
relative tail length values continues, with some
variation, into the Pacific coastal areas of Colom-
bia, the Andes of Colombia, and the western
slopes of the Ecuadoran Andes.

Values tend to be lower on the Amazonian side
of South America than in the Andes and on the
Pacific slopes. The mean values are, respectively:
males, 0.242, 0.266; females, 0.219, 0.263. There
is variation, however, in relative tail length
values in Amazonian South America, with values
being relatively high in Venezuela, the Guianas,
Brazil, Argentina, and Amazonian Ecuador and
relatively low in Bolivia and eastern Peru.
Relative tail length values decrease in Trinidad,
relative to those in Venezuela, but increase again
in Tobago.

This pattern of variation follows, in broad out-
lines, that in subcaudal numbers.

In summary, as has been demonstrated with
many wide-ranging species of Neotropical rep-
tiles and amphibians, the patterns of variation of
the various characteristics are discordant with
respect to one another. It is for this reason and
because we feel that recognition of infraspecific
taxa does not adequately reflect the complexity
of variation in biological species, that no sub-
species are recognized.

Concordance in variation in color pattern and in
ventrals and subcaudals in Central America, how-
ever, may lead some workers to adopt a some-
what different approach to interpretation of the
above-discussed variational picture. As stated
above, we consider specimens from Panama to be
intermediate between those making up pattern
types A and B. The distinctions between speci-
mens exhibiting patterns A and B, however, are
made even more trenchant with consideration of
ventral and subcaudal differences. Members of
pattern A have 155-174 (jf = 165.0) ventrals in

males and 158-177 (x =169.2) in females. Sub-
caudals vary from 46-59 (x" = 53.9) in males and
from 42-60 (Y=46.9) in females. Members of pat-
tern B have 142-148 (Y = 144.2) ventrals in males
and 143-155 (x' = 147.4) in females. Subcaudals
vary from 64-82 (Y = 72.8) in males and from
63-75 (T = 68.5) in females. There is no overlap
between any of the ranges and the distinctions
could easily point to the existence of 2 species. It
is only with consideration of specimens from cen-
tral and eastern Panama that we are able to
justify considering the specimens exhibiting pat-
terns A and B as part of the Tantilla melano-
cephala population system and it is on this point
that other workers might differ in their interpre-
tation. As mentioned above, specimens from
Panama most closely resemble pattern A speci-
mens in color pattern and pattern B specimens in
ventral and subcaudal numbers. The color pat-
tern situation has been summarized in Table 2
and demonstrates that the pattern of Panama-
nian specimens is intermediate between those of
pattern A specimens and pattern B specimens,
but is closer to that of pattern A specimens.
Panamanian specimens have 139-149 (T = 143.5)
ventrals and 69-79 (x" = 75.1) subcaudals in males;
corresponding ranges and means for females are
143-154 (152.0) and 66-71 (68.3). They most close-
ly resemble, therefore, pattern B specimens and
are not, for the most part, intermediate between
the numbers exhibited by pattern A and B speci-
mens. Some workers might prefer to recognize
pattern A and B specimens as distinct from each
other at the species level and each in turn distinct
from the Panamanian specimens, which would
then represent the population at the northern
edge of the range of the South American species
(comprised of our pattern types C-F). In that case,
pattern A and B specimens would take the names
Tantilla armillata Cope and Tantilla ruficeps
(Cope), respectively. Our stand, of course, is
based on the intermediacy of the pattern of the
Panamanian specimens. Also, the closer resemb-
lance of the pattern of the Panamanian specimens
to that of pattern A specimens and the resemb-
lance of ventral and subcaudal numbers to those
of pattern B specimens lends further support to
this argument.

The Status of the Names Associated

with the Tantilla melanocephala

Population System

Eleven names have been associated in various

ways with the Tantilla melanocephala population

19

system in the past. These names are: (1) Coluber
melanocephalus Linnaeus, 1758; (2) Elapomor-
phus mexicanus Giinther, 1862; (3) Tantilla
armiUatum Cope, 1876a; (4) Homalocranion
melanocephalum var. fusciim Bocourt, 1883; (5)
Tantilla pallida Cope, 18876; (6) Pogonaspis rufi-
ceps Cope, 1894; (7) Homalocranium melano-
cephalum var. fraseri Gunther, 1895; (8) Homalo-
cranium melanocephalum var. pernambucense
Gunther, 1895; (9) Homalocranium longifrontale
Boulenger, 18966; (10) Homalocranium hoffman-
ni Werner, 1909; (11) Elapomorphus nuchalis
Barbour, 1914. Another name, Tantilla capistrata
Cope, 18766, is associated with a taxon that has
been recognized as a subspecies (T melano-
cephala capistrata) by some workers but is herein
raised to species level (see species accounts
below). We have been able to examine the holo-
type or a portion of the syntype series upon which
5 of the above names are based. Types are lost or
were not designated for 2 of the above-listed
names. The holotype upon which one name is
based was not available to us, but notes on it
have been provided, and it was not necessary to
examine 3 others.

The status of these 11 names is discussed
below.

(1) Coluber melanocephalus Linnaeus, 1758
(type locality: "America"). This is the oldest
name in the group and is the name we used for the
above-described population system, in spite of
the fact that no holotype was designated and
there is no specific type locality. Andersson
(1899), however, apparently identified the holo-
type of Coluber melanocephalus from among 2
specimens in the Royal Museum in Stockholm
bearing that name. One of these 2 specimens is a
Sibynophis subpunctatus, which is superficially
similar to Tantilla melanocepkala. The other spec-
imen is a T. melanocephala, according to
Andersson (1899). Unfortunately, Linnaeus
(1758) described the Tantilla in the text, but
figured the Sibynophis. Andersson (1899) chose
to apply the name Coluber melanocephalus to the
specimen described in the text, a decision with
which we are in accordance. This specimen has
142 ventrals and 62 subcaudals and cannot be
pinned down to locality on this basis, except to
say that it came from South America.

(2) Elapomorphus mexicanus Gunther, 1862
(type locality: "?Mexico"). This name is based
upon the holotype (BMNH 1946.1.8.58). which
purportedly came from "Mexico." According to
Hobart M. Smith (in litt.), no other specimens of

this group have been forthcoming from Mexico.
The data accompanying the holotype, which we
have examined, reads "?Mexico," although other
authors (Gunther, 1862, 1895; Smith, 1942;
Stuart, 1963; Peters and Orejas-Miranda, 1970)
have given the type locahty as "Mexico," with-
out the interrogative prefix. Smith (1942) stated
"Boulenger (Cat. Snakes, vol. 3, 1896, p. 216) con-
cludes that the type (labelled Salle's collection)
was not collected by Salle and is not from Mexico.
There are facts, however, which indicate that it
may actually have come from Mexico. Although
the ventral count is matched by some South
American specimens, it is lower than any Central
American specimen related to T melanocephala
(i.e., T. armillata). Another specimen from
Dueiias, Guatemala, referred by Boulenger (Cat.
Snakes, vol. 3, p. 221) to T. fusca and by Gunther
to T. armillata very closely matches the charac-
ters of T. mexicana, to which we have referred it.
There is little doubt that these are distinct from
T. armillata, and it is scarcely likely that they can
be termed the same as the geographically distant
T melanocephala."

As mentioned above, we have examined the
holotype of Elapomorphus mexicanus and, even
though the specimen is faded, it is possible to
determine its affinities clearly. In all features of
color pattern, this specimen agrees with pattern
F specimens from the western slopes of the Andes
in Ecuador. Thus Elapomorphus mexicanus is
technically a synonym of Tantilla fraseri. should
the latter be resurrected in the future. No
members of the Tantilla melanocephala group are
known, therefore, from Mexico.

We have also examined the specimen (BMNH
75.5.1.2) from Duenas. Guatemala, mentioned by
Smith (1942) as very closely matching the charac-
ters of the holotype of T mexicana, and find it
does not resemble that specimen at all but rather
is a typical member of the pattern A group of
Tantilla melanocephala (Tantilla armillata
auctorum).

(3) Tantilla armillata Cope, 1876a (type locality:
"middle Costa Rica"). Cope (1876a) based this
name on a single specimen (ANSP 3363), which
now has the locahty "San Jose" attached to it.
This specimen is typical of pattern A specimens
in all features. Pattern A specimens have been
called Tantilla armillata by most previous
workers (see synonymy of T. melanocephala
below).

(4) Homalocranion melanocephalum var. fus-
cum Bocourt, 1883 (type locahty: "Guatemala").

20

Stuart (1963) used the name T. mexicana hesi-
tantly for material of the T. melanocephala group
from Guatemala. Apparently, he had available 3
specimens (no more are now known). One was the
Duenas specimen mentioned above in the discus-
sion of the name Elapomorphus mexicanus. A
second was a specimen (UMMZ 106758) from
Finca La Gloria, Depto. Santa Rosa, which also
has the A pattern. The third specimen (MNHN
6059) is the holotype of Homalocranion melano-
cephalum var. fuscum Bocourt, 1883, purported-
ly from "Guatemala." This latter specimen did
not come from Guatemala. Rather it resembles
pattern E specimens from the Amazonian low-
lands of Ecuador and Peru. Should a future
worker desire to recognize pattern E specimens
as a subspecies, the name fusca would then be
available.

Boulengers's (1896a) usage of the name fusca
also includes a peculiar Tantilla jani-Uke speci-
men from Nicaragua (Wilson and Meyer, 1971).

(5) Tantilla pallida Cope, 18876 (type locality:
Chupada [ = Chapada do Guimaraes, 56°15'W,
15°34'S] Est. Mato Grosso. Brazil). We have
examined the holotype of Tantilla pallida (ANSP
11215) and find it a typical representative of pat-
tern C type. The head is brown, the nuchal collar
is divided medially, the dark nape band is dark
brown and is followed by a pale neck band, the
dorsum is tan and there is neither a middorsal
dark stripe nor a pale lateral band. There are 148
ventrals and 58+ subcaudals (tip of tail missing).
In these features it agrees with other specimens
from Brazil. Should future workers wish to recog-
nize the C-D pattern specimens as a distinct sub-
species, the above name would be available.

The type locality of this specimen has been
given simply as "Mato Grosso, Brazil" (Peters
and Orejas-Miranda, 1970), but Malnate (1971)
listed it as "Brazil, Matto Grosso, Chapada."
Cope (18876) indicated in his introduction that all
the specimens mentioned in that paper came from
or in the vicinity of Chupada [ = Chapada], which
is a small town located about 48 km northeast of
Cuiaba in the state of Mato Grosso.

(6) Pogonaspis ruficeps Cope, 1894 (type locali-
ty: Jimenez, Costa Rica). Although we had pre-
viously considered the holotype of Pogonaspis
ruficeps to have been in the collections of the
museum of the Universidad de Costa Rica and
not presently locatable Douglas C. Robinson
recently informed us that he had located the type
in the American Museum of Natural History.
This male specimen (AMNH 17272), kindly sup-

plied by Charles W. Myers, is in relatively poor
condition and the pattern is somewhat faded but
is clearly of the pattern B type. In addition, the
ventrals number 144 (our count), which is a
typical count for pattern B males.

(7) Homalocranium melanocephalum var.
fraseri Gunther, 1895 (type locahties: Quito,
Ecuador and "W. Ecuador"). We have examined
1 of 4 syntypes (BMNH 1946.1.9.84) upon which
this name is based and find it to be in complete
agreement with specimens exhibiting pattern F.
This name is thus available for that population,
should some future worker wish to use it.

(8) Homalocranium melanocephalum var. per-
nambucense Gunther, 1895 (type locality: Per-
nambuco, Brazil). We have not examined the syn-
type series upon which this name is based and
A.F. Stimson stated (in litt.) that the British
Museum (Natural History) has only a single spec-
imen from the original series and that the location
of the remainder is not known. Giinther's (1895)
description indicated that the name applies to
members with the type C pattern as he stated
that there is no black dorsal stripe. The ventral
counts of 149-154 given fit within the range for
other Brazilian material, but the subcaudal
counts of 30-32 are obviously for specimens with
incomplete tails.

(9) Homalocranium longifrontale Boulenger,
18966 (type locality: Cali, Colombia). We have not
examined the holotype (BMNH 1946.1.8.84), but
A.F. Stimson has kindly sent us some detailed
notes on pattern and scutellation. The color pat-
tern description agrees with that of intergradient
specimens from Panama and Colombia.

(10) Homalocranium hoffmanni Werner, 1909
(type locality: "Guatemala"). We have been
informed by Werner Ladiges (in litt.) of the
Zoologisches Institut und Zoologisches Museum
of the Universitat Hamburg that the holotype of
Homalocranium hoffmanni was destroyed during
World War II. The description given by Werner
(1909), however, is sufficiently detailed to indi-
cate to us that the holotype did not come from
Guatemala. Werner (1909) described the color
pattern as follows (translation ours): "dorsum
dark gray; a yellow middorsally broken neck band
on the posterior half of the parietals and adjacent
lateral scales. Venter whitish, only the throat
gray (posterior chin shields whitish). Anal flecked
with gray." There is no mention of a dark middor-
sal stripe and the pale nuchal collar is broken mid-
dorsally but not laterally. These 2 characters are
not those of specimens from Guatemala (of the

21

pattern A type), which have 2 small nuchal spots
confined to the parietals and a middorsal dark
stripe. The above-cited characteristics are those
of pattern C specimens, however. The ventral
count of 155 (the tail is incomplete) does not allow
us to narrow the type locahty; all we can say is
that the holotype of Homalocranium hoffmanni
came from somewhere in eastern South America.
(11) Elapomorphus nuchalis Barbour, 1914
(type locality: Villa Bella on the Amazon River,
above Santarem, Brazil). We have not examined
the holotype (MCZ 1164) upon which this name is
based, but Barbour's (1914) description is suffi-
ciently detailed to indicate to us that the speci-
men exhibits a pattern intermediate between that
of C and D. He mentioned that there are "three
very indistinct darker longitudinal lines which
are just visible on the nape region which disap-
pear along the body and reappear upon the
posterior extremity of the body and on the tail."
Also, he pointed out that the nuchal collar is
divided medially only. The ventral count of 147
(the tail is incomplete) falls within the range for
other Brazihan material.

Species Accounts
Tantilla andinista, sp. nov.

Holotype.-KH 135209.

Type locality. — 5 km E Alausi, Prov. Chimbor-
azo, Ecuador, elevation 2600-2750 m, collected 9
June 1970 by T.H. Fritts.

Diagnosis.— A member of the Tantilla melano-
cephala group with a pattern consisting of a large
amount of pale markings on the head on a dark
brown background, a tan dorsal ground color, a
narrow dark middorsal stripe (confined to middle
of the middorsal row), and a narrow pale lateral
stripe (confined to the upper half of row 3). The
head pattern consists of an extensive pale mark-
ing on the snout, continuing posteriorly over the
supraocular and anterolateral corner of the
parietal downward onto the postoculars, anterior
portion of the anterior temporal and supralabials
4, 5, and 6. The dark head cap is followed by a
wide middorsally divided pale nuchal band, then
a dark nape band, and finally by a pale neck band
(Figs. 7, 12A).

Description of the holotype. — An adult female
with 15 dorsal scale rows, smooth throughout,
157 ventrals, divided anal plate, 50 subcaudals,
total length 285 mm, tail length 56 mm, relative
tail length 19.6%.

Nasal scale completely divided, the posterior
section separated from the single preocular on

both sides of the head; 2 postoculars, subequal in
size; one anterior temporal, posterior temporal
divided into 2 small scales, approximately the
size of a dorsal scale; supralabials 7-7, with the
3rd and 4th entering the orbit, last one largest;
infralabials 6-7, four in contact with the anterior
chin shields on left and 5 on right, 4th largest on
left and 5th on right, first pair in broad medial
contact; anterior chin shields larger than
posterior pair.

The middorsal stripe is dark brown, narrow,
and confined to the middle of the middorsal scale
row. Anteriorly this stripe contacts the dark nape
band and continues to the end of the tail. The dor-
sal ground color is tan (pale brown in life, accord-
ing to collector's field notes). The pale lateral
stripe is confined to the upper half of row 3 and is
bounded below by brown pigment on the lower
half of row 3 and the upper edge of row 2. There is
a brown spot at the end of each of the scales in
row 1. The venter is immaculate cream (white
with greenish tint in life).

The head is dark brown above (black in life)
with cream-colored markings. Most of the snout
to the level of the posterior quarter of the prefron-
tals is cream, except for dark brown pigment on
the lower half of the rostral and around the
anterior end of the naris. The pale snout is con-
nected to a pale spot on the supraocular, which is,
in turn, connected (but not on left) to a pale spot
on the anterolateral corner of the parietal. This
latter spot is in contact with the postocular pale
spot occupying both postoculars, the anterior
one-third of the anterior temporal, the posterior
half of supralabial 4, all of 5, and the anterior half
of 6. A dark subocular spot is present on the
posterior half of supralabial 2, all of 3, and the
anterior half of 4. The dark head cap is bounded
posteriorly by a cream-colored nuchal band begin-
ning on the posterior portion of the parietals and
extending about one dorsal scale posterior to the
parietals. This pale nuchal band is completely
divided medially by a connection about one dorsal
scale wide between the head cap and the dark
nape band. The dark nape band is three dorsal
scales long and is followed by a well-defined
cream-colored neck band about one dorsal scale in
length. There is a dark blotch on the mental and a
circular dark blotch on the anterior portion of
supralabial 4 (5 on right).

Range.— Known only from the type locality in
the Alausi Basin of Ecuador (Fig. 13).

Etymology. — The name andinista is from the
Spanish, meaning "wanderer among the Andes"

22

in allusion to the montane distribution of this
species.

Habitat notes.— The holotpe was found under a
rock during the day.

Tantilla capistrata Cope
Tantilla capistrata Cope, 1876b.- 181.
Homalocranium melanocephalum: Boulenger,

1896a.- 215 (part).
Tantilla melanocephala: Dunn, 1923: 186; Parker,

1938: 445.
Tantilla melanocephala capistrata: Schmidt and

Walker, 19436: 318; Peters, 1960: 539; Peters

and Orejas-Miranda, 1970: 295.

i/o/o^ype. -ANSP 11581.

Type locality.— VaWey of Jequetepeque, Depto.
La Libertad, Peru.

Diagnosis. — A member of the Tantilla melano-
cephala group with a head pattern consisting of a
pale brown to dark brown head cap separated
from a dark brown nape band by a well-defined
complete or medially divided pale nuchal band.
There is an extensive pale marking on the snout
covering all the internasals and the anterior one-
half to three-quarters of the prefrontals. The dor-
sal pattern consists of a narrow, diffuse dark
brown middorsal stripe confined to the middorsal
scale row (if present) on a tan ground color. There
is no pale lateral stripe (Figs. 8, 12B).

Description. — Tantilla capistrata has a tan dor-
sum without a dark middorsal stripe or only a
slight indication of one. There is no pale lateral
band. The venter is immaculate cream.

The head is dark brown (juveniles) to brown or
pale brown (adults) above. The snout is pale
cream up to the level of the posterior one-fourth
to two-thirds of the prefrontals. A pale to dark
brown head cap follows and covers the top of the
head except for the posterior tips of the parietals.
The preocular pale spot is in contact with the pale
snout markings dorsally and with the postocular
pale spot posteriorly underneath the subocular
dark spot, which does not reach the lip. The post-
ocular pale spot is separated from the pale nuchal
band by an extension of the dark head cap.

The pale cream nuchal band is complete (4 spec-
imens) or divided medially (4 specimens) and is
bounded posteriorly by a dark brown nape band
2'2-5 dorsal scales long. This band is in turn
bounded posteriorly by a pale neck band (absent
in one specimen) 1-1 ' ! dorsal scales long. There is
very little dark pigment on the infralabials, con-
sisting of dark smudges on some of the first 4
infralabials, if present.

Variation in scutellation may be summarized as
follows: prefrontals in contact with second supra-
labial on both sides in all specimens; first pair of
infralabials usually in contact medially (87.5%),
but may be separated (12.5%); ventrals in males
130-150 (r = 142.0), in females 135-156
(Y = 147.5); subcaudals in males 53-61 (^=57.0),
in females 46-71 (^^ = 56.3).

Total length in snakes of all ages ranges from
134-360 mm (one specimen measures 394 mm
overall with an incomplete tail; judged by relative
tail length of other females, this snake may have
had a total length of 428-462 mm), and tail length
ranges from 31-87 mm (the large specimen men-
tioned above may have had a tail length of from
84-119 mm). Relative tail length ranges from
19.5-25.4%.

Geographic variation.— The range of Tantilla
capistrata is broken into 2 disjunct sections, one
in the northwestern portion of coastal Peru and
the other in the semiarid valleys of the Maranon.
Chinchipe, and Chamaya rivers. Although the
basic details of pattern are similar, there are dif-
ferences between the two geographic groupings,
nonetheless. All specimens from the valley of the
upper Maranon (Atlantic drainage— hereinafter
referred to as the eastern segment) have no evi-
dence of a dark middorsal stripe. All specimens
from northwestern coastal Peru (Pacific drain-
age—hereinafter referred to as the western seg-
ment) have a dark middorsal stripe, although it is
very narrow and diffuse (even reduced to one spot
per scale), except one (ANSP 11581, the
holotype). The dark nape band of members of the
eastern segment ranges from 2' 2-4 scales long
(^" = 3.0), that of members of the western segment
3-5 (x"=3.7). Both ventral and subcaudai
numbers are higher in the eastern segment,
except for subcaudals of females (Table 7). The
mean for subcaudals of females from the western
segment is as high as it is because the subcaudai
count of the holotype (ANSP 11581) is 71. This is
the highest count for the species (males
included— the highest male count is 61). The ven-
tral count of 156 is also the highest for the
species. Thus the total of ventrals and subcaudals
for the holotype is also the highest for the species
(227, as opposed to a high count of 210 for males
and 190 for other females). The holotype is also
the only specimen from the western segment
without a vestige of a middorsal stripe. The
reason for the differences between the holotype
and other specimens of the species remains unex-
plained and will until additional material from

23

Table 7. Comparison of ventral and subcaudal numbers in the eastern and western segments of the range of
Tantilla capistrata.

Character

Eastern segment

Western segment

Ventrals

CTO"

9 9

Subcaudals

cr cr
9 9

147-150(148.5)2
155

60-61(60.5)2

130-141(135.5)2
135-156(145.0)3

53-54(53.5)2
46-71(56.3)3

critical areas accumulates.

Range.— From near sea level to moderate eleva-
tions of northwestern coastal Peru and the
valleys of the upper Rio Maranon, Rio Chinchipe,
and Rio Chamaya (Fig. 14).

Ecological distribution.— No direct information
on ecological distribution is available but judged
by the available locality records, Tantilla capis-
trata occurs in at least 2 apparently disjunct pop-
ulations, one occupying the arid coastal desert
regions of northwestern Peru and the other the
semiarid inter-Andean basin formed by the
valleys of the upper Rio Maranon, Rio Chinchipe
and Rio Chamaya. According to Dixon and
Wright (1975), the "xerophyllous vegetation of
these valleys is not continuous with that of the
coastal area."

Locality records. — Peru: Depto. Cajamarca, 3
km NE Chamaya (LSUMZ 19557); 5 km S Jaen,
900 m (LSUMZ 19558); Perico (MCZ 17391);
Depto. La Libertad, ChicUn (FMNH 33749,
34286-87); valley of [Rio] Jequetepeque (ANSP
11581); Depto. Piura, Sullana (NMB 15631).

Tantilla equatoriana, sp. nov.

Holotype.-USNM 198530.

Type locality.— San Lorenzo, Prov.
Esmeraldas, Ecuador, collected in May 1951 by
M. Olalla.

Diagnosis.— A member of the Tantilla melano-
cephala group with a head pattern consisting of a
dark head cap grading into a dark nape band and
punctuated by pale spots located near the
posterior end of the parietals (and extending onto
the posterior temporal and 1 or 2 adjacent
postparietal scales), a preocular pale spot that
continues posterodorsally onto the preocular
scale to touch the eye, and a postocular spot that
is in contact with pale pigment on the lateral
gulars (opposite the parietal spots), not separated
by a dark lateral extension of the head cap. No
pale neck band is present posterior to the dark

nape band. The dorsal pattern is bold with dark
brown stripes on a tan ground color located on the
middle of row 1, from the upper half of row 2 to
the lower half of row 3, upper edge of row 4 to
lower middle portion of row 5, middle of row 6 and
the middle of the middorsal scale row. The sub-
caudals are relatively high, ranging from 77 to 79
(Figs. 9, 12C).

Description of holotype.— An adult (?) male
with 15 dorsal scale rows, smooth throughout,
147 ventrals, divided anal plate, 79 subcaudals,
total length 253 mm, tail length 72 mm, relative
tail length 28.5%.

Nasal scale completely divided, the posterior
section separated from the single postocular on
both sides of the head; two postoculars, subequal
in size; one anterior and one posterior temporal,
both elongate; supralabials 7-7, with the 3rd and
4th entering the orbit, last one largest; infra-
labials 6-6, four in contact with the anterior chin
shields, 4th largest, first pair in broad medial con-
tact; anterior chin shields larger than posterior
pair.

The middorsal stripe is dark brown, well-
defined, covering the middle of the middorsal
scale row and expanding at intervals to cover the
medial apices of the scales of the paravertebral
row. Anteriorly this stripe contacts the dark nape
band and continues to the end of the tail. The dor-
sal ground color is tan. The lateral stripe pattern
consists of: a narrow dark brown stripe on the
middle of row 1; a dark brown stripe extending
from the upper half of row 2 to the lower half of
row 3; a pale cream-colored stripe from the upper
half of row 3 through most of row 4; a narrow, dif-
fuse dark brown stripe from the upper edge of row
4 through the middle of row 5; a narrow dark
brown stripe through the middle of row 6. The
venter is immaculate cream.

The head is brown above with pale markings on
the upper portion of the rostral and anterior half
of the internasals and a pair of pale spots located

24

on the posterior portion of the parietals and a por-
tion of the adjacent posterior temporal and 2
postparietal scales. The pale preocular spot occu-
pies most of supralabials 1 and 2 and extends
posterodorsally onto the preocular and touches
the eye. The preocular and postocular pale spots
are separated by a dark subocular blotch occupy-
ing most of supralabials 3 and 4 (and a thin strip
along the posterior edge of supralabial 2), except
for the anteroventral corner of each. This dark
pigment is also present along the lower edge of
supralabials 5 through 7. The postocular pale
spot is present on the lower portion of the
anterior temporal and the upper portion of supra-
labials 5 through 7 (separated from the brown
head cap by a dark brown stripe coursing through
the middle of the anterior temporal and onto the
dorsal portion of the 7th supralabial) and is in
contact with the pale pigment on the lateral
gulars (opposite the parietal spots) and is not
separated by a lateral extension of the dark head
cap.

The dark head cap grades into a dark brown
nape band which extends posteriorly to the
fourth dorsal scale posterior to the parietals.
There is no pale band posterior to the dark nape
band. The mental is largely dark brown and there
is a dark brown blotch along the posterior suture
of each infralabial.

Paratype. — USNM 198429, same locality as
holotype.

Variation in the paratype.— The dorsal pattern
of the male paratype is essentially like that of the
holotype. The only pattern differences are minor;
the dorsum of the head is dark brown and the pale
parietal spots are slightly elongated anteriorly.

Scutellation is the same for the paratype as for
the holotype, except that the paratype has 142
ventrals and 77 subcaudals. Total length is 218
mm, tail length 59 mm, and the relative tail
length is 27.1%.

Range. — Known only from the type locality
(Fig. 13).

Etymology.— The name equatoriana is from the
Latin meaning "equator," referring to its distri-
bution and its country of origin.

Ecological distribution.— hov/land rain forests
of northwestern coastal Ecuador.

Tantilla insulamontana, sp. nov.

Holotype.-K\] 152207.

Type locality. — Rio Minas, 15.1 km W Santa
Isabel, Prov. Azuay, Ecuador, elevation = 1250
m, collected 12 June 1971 by A.C. Echternacht.

Diagnosis. — A member of the Tantilla melano-
cephala group with a head pattern consisting of a
brown head cap with 2 horn-shaped anterior
extensions, nuchal spots confluent with dorsal
ground color, and no dark nape band. The dorsal
pattern consists of numerous dark stripes on a
more-or-less uniform tan ground color. There are
diffuse stripes through the middle of scale row 1,
on the upper half of row 2 and lower half of row 3,
through the middle of row 4, on the upper half of
row 5 and the lower half of row 6. A relatively
dark, well-defined stripe courses down the middle
of the middorsal row and is connected to dark
markings along the medial edges of the scales of
the paravertebral rows, giving the whole stripe a
hatched appearance (Figs. 10, 12D).

Description of holotype. — An adult (?) male
with 15 dorsal scale rows, smooth throughout,
144 ventrals, divided anal plate, 65 subcaudals,
total length 249 mm, tail length 67 mm, relative
tail length 26.9%.

Nasal scale completely divided, the posterior
section separated from the single postocular on
both sides of the head; 2 postoculars, subequal in
size; one anterior and one posterior temporal,
both elongate; supralabials 7-7, with the 3rd and
4th entering the orbit, last one largest; infra-
labials 6-6, four in contact with the anterior chin
shields, 4th largest, first pair broadly separated
by contact of mental and anterior chin shields;
anterior chin shields larger than posterior pair.

The middorsal stripe is dark brown, occupies
the middle of the middorsal scale row and is con-
nected to dark markings along the medial edges
of the scales of the paravertebral rows. Anteriorly
this stripe contacts the dark head cap at the
posterior end of the medial parietal suture and
continues onto the end of the tail. The dorsal
ground color is tan. The lateral stripe pattern con-
sists of: a diffuse brown stripe through the middle
of scale row 1 accentuated by a dark brown spot
at the posterior end of each scale; a diffuse brown
stripe on the upper half of row 2 and the lower
half of row 2 with a dark dot at the posterior end
of the scales of each of those 2 rows; a broad, dif-
fuse brown stripe through the middle of row 4,
each scale of which is rimmed anteriorly by
darker pigment; a diffuse brown stripe on the
upper half of row 5 and the lower half of row 6,
each scale of both rows outlined anteriorly by
dark pigment. The venter is cream with a single
brown spot at the lateral edges of each ventral.

The head is brown above. From the anterior
edge of the head cap extend 2 horn-shaped or

25

comma-shaped dark brown marks onto the mid-
dle of the internasals. They border a postero-
medial extension of the pale pigment on the snout
that extends back to the juncture between the
prefrontals and frontal. A pair of pale parietal
spots are present on the posterior portions of the
parietals and are confluent with the dorsal
ground color on either side of the middorsal
stripe. There is no dark nape band. The pale pre-
ocular spot on supralabials 2 and 3 is confluent
with the pale pigment on the snout and also with
the pale postocular spot on the posterior portion
of supralabial 4, all of 5, most of supralabial 6, the
lower portion of the upper postocular, all of the
lower postocular, and the anterior third of the
anterior temporal. The subocular dark spot is
limited to a narrow rim of dark pigment around
the bottom of the eye. A lateral dark extension of
the head cap on the posterior edge of supralabial
6 and the anterodorsal two-thirds of supralabial 7
completely separates the postocular pale spot
from the pale pigment of the lateral gulars oppo-
site the parietal spots.

There are spots of brown pigment on the men-
tal, all infralabials except the last, and the
anterior chin shields.

Paratypes.—TYa-ee, all from Prov. Azuay,
Ecuador: Llunguilla [ = Yunguilla] (CAS
94090-91) [not mapped]; 10 km SW Giron, 2100 m
(USNM 198430).

Variation in the paratypes.—The dorsal pattern
of all paratypes, which are all females, is essen-
tially the same as that of the holotype, even
granting that one specimen is slightly faded (CAS
94091) and another was preserved just prior to
shedding (USNM 198430). One specimen (CAS
94090) shows evidence of a complete pale nuchal
collar extending posteroventrally from the parie-
tals. It terminates at the dark stripe on rows 2
and 3.

Variation in scutellation in the female para-
types may be summarized as follows: ventrals
152-157 (T = 154.3)-3 specimens; subcaudals
59-61 (T=60.0)-2 specimens. All other aspects of
scutellation agree with that of the holotype. Total
length of snakes of all ages ranges from 310-404
mm, and tail length ranges from 70-95 mm. Rela-
tive tail length ranges from 22.5-23.5%.

i?an^e. — Intermediate elevations of the Hoya
de Jubones (Pacific drainage) in southern
Ecuador (Fig. 13).

Etymology.— The name insulamontana is from
the Latin, meaning "island in the mountains,"
referring to the distribution of this species in an

isolated intermontane basin.

Habitat notes.— The holotype was collected as
it moved along an embankment adjacent to a
road beside a bamboo thicket along a river at
13:30 hours on an overcast day.

Tantilla lempira, sp. nov.

Holotype.-LSVMZ 26093.

Type locality.— 41 km NW Tegucigalpa, Depto.
Francisco Morazan, Honduras, collected 3 June
1968 by Ernest A. Liner.

Diagnosis.— A member of the Tantilla melano-
cephala group with a head pattern consisting of a
brown head cap with a pair of small, indistinct
nuchal spots largely confined to the scales post-
erior to the parietals. There is no dark lateral
extension of the head cap between the postocular
pale spot and the pale pigment on the lateral
gulars. The dorsal ground color is pale brownish-
gray and there is a dark ventral border below the
pale lateral stripe. The prefrontal and second
supralabial are separated from one another by
contact of the postnasal and preocular. The ven-
trals and subcaudals are relatively low (Figs. 11,
12E).

Description of holotype. — An adult 9 with 15
dorsal scale rows, smooth throughout, 153 ven-
trals, divided anal plate, 36 subcaudals, total
length 258 mm, tail length 40 mm, relative tail
length 15.5%

Nasal scale completely divided, the posterior
section in contact with the single preocular on
both sides of the head; 2 postoculars, subequal in
size; one anterior and one posterior temporal
(posterior temporal on right divided vertically
into 2 scales), both elongate; supralabials 7-7,
with the 3rd and 4th entering the orbit, last one
largest; infralabials 6-6, four in contact with the
anterior chin shields, 4th largest, first pair in nar-
row medial contact; anterior chin shields larger
than posterior pair.

The middorsal stripe is dark brown, narrow,
disjunct, composed of a series of dots, one each at
the posterior end of each scale of the middorsal
row. Anteriorly this stripe contacts the dark nape
band and fades out just posterior to the vent. The
dorsal ground color is pale brownish-gray. The
pale lateral stripe is present on the upper half of
row 3 and the lower half of row 4 and is barely
discernible. The pale lateral stripe is bounded
below by a dark Une on the lower half of row 3.
The venter is immaculate cream.

The head is brown above. A barely visible pale
spot is present on the upper portion of the rostral

26

and medial portions of the internasals. The pale
nuchal spot on the right side of the head (none is
visible on the left) is small and barely discernible
and is confined to the point near the juncture of
the parietal, posterior temporal, and adjacent
postparietal scales. The preocular pale spot is pre-
sent on supralabials 1 and 2 and is separated from
the postocular pale spot by a subocular dark spot
on supralabials 3 and 4. The postocular pale spot
is in contact with the pale pigment on the lateral
gulars (opposite the pale nuchal spots), not separ-
ated by a dark lateral extension of the dark head
cap. The dark head cap grades into the dark
brown nape band, which extends 4 scales post-
erior to the parietals and is bounded posteriorly
by a pale neck band. This pale neck band is one
and one-quarter scales long and is divided mid-
dorsally by the dark middorsal stripe.

Paratypes.—Two, both from Honduras: 41 km
NW Tegucigalpa, Depto. Francisco Morazan
(LSUMZ 33737); Montserrat, Depto. El Paraiso
(MCZ 49961).

Variation in the paraty pes.— The dorsal pattern
of both paratypes is essentially like that of the
holotype. The middorsal dark stripes of both
paratypes are better defined than that of the holo-
type. The dark nape band extends 3 scales post-
erior to the parietals in both specimens.

Variation in scutellation in the paratypes, both
females, may be summarized as follows: ventrals
148-150 (x =149.0); subcaudals 36-44 (T = 40.0).
All other aspects of scutellation agree with that
of the holotype. Total length ranges from 206-254
mm, and tail length ranges from 38-43 mm.
Relative tail length ranges from 16.9-18.4%.

Range.— Moderate elevations of the Pacific ver-
sant of Honduras (Fig. 15).

Etymology.— The name lempira is that of an
Indian chief who fought against the Spanish inva-
sion of Honduras. It is also the name of a depart-
ment in Honduras and of the Honduran monetary
unit.

Ecological distribution.— All specimens came
from within the Subtropical Moist Forest forma-
tion (Pinus oocarpa forest).

Tantilla melanocephala (Linnaeus)
Coluber melanocephalus Linnaeus, 1758: 218,

1766: 378; Gmelin, 1789: 1095; Shaw, 1802:

490; Daudin, 1803: 367.
Matrix melanocephalus: Merrem, 1820: 110.
Elaps melanocephalus: Wagler, 1824: 8.
Duberria melanocephala: Fitzinger, 1826: 55.
Lycodon melanocephalum: Boie, 1827: 526.

Cloelia melanocephala: Wagler, 1830: 187; Fitz-
inger, 1843: 23.

Calamaria melanocephala: Schlegel, 1837: 38.

Homalocranion melanocephalum: Dumeril,
Bibron and Dumeril, 1854: 855; Jan and
Sordelli, 1866: livr. 15; Bocourt, 1883: 588.

Tantilla melanocephala: Cope, 1861: 74, 1866
126, 1868: 102, 1871: 205, 1876a: 142, 18766
181, 1887a: 83, 1900: 1111; Stejneger, 1902
157; Griffen, 1916: 209; Ruthven, 1922: 68
Dunn, 1923: 186; Amaral, 1929a: 107, 19296
46, 1929c: 221, 1931; 93; Dunn and Bailey,
1939: 19; Smith, 1941: 115; Schmidt and
Walker, 1943a: 293; Beebe, 1946: 43; Hoge,
1952: 221; Brongersma, 1956: 184; Roze, 1966:
224.

Elapomorphus mexicanus Giinther, 1862: 57
(holotype, BMNH 1946.1.8.58; type locahty:
"?Mexico," in error).

Tantilla armillata Cope, 1876a: 143 (holotype,
ANSP 3363; type locahty: "middle Costa
Rica"), 1887a, 1900: 1111; Dunn and Bailey,
1939: 19; Taylor, 1951: 152, 1954: 763; Mertens,
1952a: 74, 19526: 93; Rand, 1957: 531; Smith,
1958: 224; Smith and Grant, 1958: 212; Hardy
and Cole, 1967: 194; Myers and Rand, 1969: 4;
Clark, 1970: 130; Peters and Orejas-Miranda,
1970: 294.

Homalocranium melanocephalum var. fuscum
Bocourt, 1883: 589 (holotype, MNHN 6059;
type locahty: "Guatemala," in error).

Tantilla pallida Cope, 18876: 56 (holotype, ANSP
11215; type locality: Chupada [=Chapada dos
Guimaraes], Est. Mato Grosso, Brazil), 1900:
1111.

Pogonaspis ruficeps Cope, 1894: 204 (holotype,
AMNH 17272; type locahty: Jimenez, Costa
Rica).

Homalocranium armillatum: Giinther, 1895:
149.

Homalocranium melanocephalum var. fraseri
Giinther, 1895: 148 (syntypes, BMNH
1946.1.9.43-44 and BMNH 1946.1.8.77,
1946.1.8.80, 1946.1.9.84; type locahty of first 2
specimens: Quito, Ecuador; of remaining 3: "W.
Ecuador").

Homalocranium melanocephalum [var. melano-
cephalum]: Giinther, 1895: 147.

Homalocranium melanocephalum var. pernam-
bucense Giinther, 1895: 148 (single available
syntype, BMNH 80.11.25.5, location of remain-
ing syntypes unknown fide A.F. Stimson; type
locahty: Pernambuco, Brazil).

27

Homalocranium mexicanum: Giinther, 1895:

153.
Homalocranium melanocephalum: Boulenger,

1896a: 215; Andersson, 1899: 10; Schmidt and

Inger, 1951: 463.
Homalocranium fuscum: Boulenger, 1896a: 220

(part).
Homalocranium longifrontale Boulenger, 18966:

17 (holotype, BMNH 1946.1.8.84; type locality:

Cali, Colombia), 1896a: 218.
Homalocranium ruficeps: Boulenger, 1896a: 223.
Homalocranium hoffmanni: Werner, 1909: 239

(number of holotype unknown; type locality:

"Guatemala," in error); Amaral, 19296: 46.
Elapomorphus nuchalis Barbour, 1914: 199

(holotype, MCZ 1164; type locaHty: Villa Bella,

on the Amazon River above Santarem, Brazil);

Amaral, 19296: 46.
Tantilla longifrontale: Ruthven, 1922: 68; Dunn,

1944: 29; Beebe, 1946: 42.
Tantilla longifrontalis: Amaral, 1929c: 220, 1931:

93; Dunn, 1944: 208; Peters, 1960: 539; Roze,

1966: 222; Peters and Orejas-Miranda, 1970:

295.
Tantilla fusca: Amaral, 1929c: 220 (part).
Tantilla ruficeps: Amaral, 1929c: 221; Taylor,

1951: 156; Smith, 1958: 224; Smith and Grant,

1958: 212; Peters and Orejas-Miranda, 1970:

296.
Tantilla melanocephalus: Shreve, 1947a: 315,

19476: 533.
Tantilla mexicana: Smith, 1942: 37; Hartweg,

1944: 4; Smith and Taylor, 1945: 139; Stuart,

1963: 118; Peters and Orejas-Miranda, 1970:

296.
Tantilla melanocephalum: Wehekind, 1955: 12.
Tantilla fraseri: Peters, 1960: 539; Peters and

Orejas-Miranda, 1970: 294.

Tantilla melanocephala melanocephala: Peters,
1960: 539; Underwood, 1962: 169; Verteuil,
1968: 103; Peters and Orejas-Miranda, 1970:
295; Mertens, 1972: 16; Cordeiro and Hoge,
1973: 283; Achaval and Vaz-Ferreira, 1974: 107;
Dixon and Soini, 1975: 5, 1977: 72.

Holotype.— Not designed; stated by Andersson
(1899) to be in the Royal Museum in Stockholm.
Type locality.— "America."

Diagnosis. — A member of the melanocephala
group with a variable color pattern that may be
divided into 6 discrete types (see section on the T.
melanocephala population system above). All
populations of Tantilla melanocephala differ from
T. andinista in not having the extreme amount of

pale markings on the head of the latter, but rather
having the pale markings restricted to the snout
and nuchal collar. All populations of T. melano-
cephala, except for those containing pattern C
specimens, differ from T capistrata in having the
nuchal collar divided medially and laterally,
sometimes reduced to 2 parietal spots and a dark
middorsal stripe present. Pattern C specimens
differ from T. capistrata in usually (75.6% of spec-
imens) having the nuchal collar divided both
medially and laterally, whereas in T capistrata
the nuchal collar is always divided medially only
or complete. In addition, smaller specimens of T.
capistrata have a virtually black head and dark
nape band, separated distinctly from the cream-
collared snout markings and nuchal collar. Tan-
tilla equatoriana differs from all T. melano-
cephala in having the pale preocular spot entering
the orbit and in details of pattern (especially the
presence of a dark stripe on dorsal scale row 6).
The subcaudal counts of T. equatoriana are also
higher than those of any T. melanocephala except
for pattern B specimens and the intergradient
specimens from Panama and Colombia. All T.
melanocephala differ from T. insulamontana in
having the pale nuchal collar or spots not con-
fluent with the dorsal ground color, but rather
separated by a dark nape band, and in having no
dark stripe located on the adjacent halves of dor-
sal scale rows 5 and 6. Tantilla insulamontana
appears to the eye to be a Tantilla with a multi-
lineate dorsal pattern on a pale background, with-
out a dark nape band and with the pale nuchal
spots confluent with the dorsal ground color, a
pattern very different from that of any T. melano-
cephala. Tantilla lempira differs from T. melano-
cephala in having the pale nuchal spots largely
confined to scales posterior to the parietals. It dif-
fers from most T. melanocephala (except pattern
B specimens and occasional specimens in other
pattern groups) in not having a dark lateral
extension of the head cap separating the
postocular pale spot and the pale pigment on the
lateral gulars (opposite the pale nuchal spots).
From pattern A specimens, the range of which is
virtually parapatric with that of T. lempira, T.
lempira differs in the following suite of characters
(characters of T. lempira indicated in
parentheses): nuchal spots distinct and occurring
usually only on the parietals (nuchal spots indis-
tinct and largely confined to scales posterior to
the parietals); dorsal ground color tan to brown
(pale brownish-gray); dark ventral border of pale
lateral stripe not present (present); dark lateral

28

extension of head cap separating postocular pale
spot and pale pigment on lateral gulars (opposite
pale nuchal spots) present (absent); preocular and
postnasal in contact separating the prefrontal
and second supralabial (not in contact); ventrals
158-177 in females (148-153); subcaudals 42-60 in
females (36-44) (Figs. 1-6).

Description. — Tantilla melanocephala is a
widely-distributed snake and, as such, shows a
wide range of variation in pattern and scutella-
tional characteristics. What follows is a summari-
zation of those characteristics: (1) head cap cream
to dark brown, usually pale brown to dark brown;
(2) pale nuchal collar complete, divided medially,
divided medially and laterally or reduced to two
small spots centered on the parietals (usually not
complete); (3) dorsal ground color tan to brown;
(4) dark subocular blotch completely separating
pre- and postocular pale spots or not; (5) lateral
extension of dark head cap present or not; (6)
brown to dark brown dark nape band present or
not; (7) pale neck band following dark nape band
present or not, if present, occupying '/2-2 dorsal
scale lengths, sometimes reduced to small spots
on either side of the dark middorsal stripe; (8) pale
lateral band present or not, if present, usually
occupying from as little as the upper portion of
row 3 to as much as the upper portion of row 3, all
of row 4, and the lower portion of row 5, bounded
above and/or below with dark pigment or not.

Variation in scutellation may be summarized as
follows: prefrontal in contact with second supra-
labial on one or both sides or not (in contact on
both sides in 66.2% of specimens); first pair of
infralabials in contact medially or not (in contact
in 80.7% of specimens); ventrals in males
125-174, in females 134-177; subcaudals in males
41-92, in females 41-75.

Total length in snakes of all ages ranges from
95-486 mm and tail length ranges from 15-140
mm. Relative tail length ranges from 15.8-35.9%.

Range. —From near sea level to intermediate
elevations along both versants from Guatemala
throughout the length of Central America into
South America as far south as southern Peru,
Bohvia, northern Argentina, and Uruguay; also
on the islands of Trinidad and Tobago (Figs. 14,
15).

Ecological distribution.— See section on distri-
butional patterns.

Locality records.— We have arranged the speci-
mens examined according to the pattern types
discussed above.

Pattern A.— GUATEMALA: Depto. Sacatepe-

quez: Duenas (BMNH 75.5.1.2); Depto. Santa
Rosa: Finca La Gloria (UMMZ 106758). EL SAL-
VADOR: Depto. San Salvador: San Salvador
(FMNH 64964). HONDURAS: Depto. Francisco
Moraz^n: El Zamorano (AMNH 70180, 70195;
MCZ 48701-02, 49882-85. 49920-21); 11 km N
Sabana Grande (LSUMZ 26092); Tegucigalpa
(BYU 18262); Depto. Olancho: Pataste (MSU
4724); Depto. Valle: San Lorenzo (AMNH 70228).
NICARAGUA: Depto. Chontales: Juigalpa (KU
174315); Depto. Managua: Managua (UMMZ
89478); Las Robles (KU 174314); Depto.
Matagalpa: 2 mi N Matagalpa (UMMZ 116517);
0.5 mi W Matagalpa (UMMZ 116515-2 speci-
mens); 1 mi W Matagalpa (UMMZ 116514); 1.5
mi W Matagalpa (UMMZ 116516—2 specimens);
Depto. Rivas: 12 km N Costa Rican border
between Pan American Hwy. and Lake
Nicaragua (KU 140077); Finca Amayo, 13 km S,
14 km E Rivas (KU 101922); within 3 km
Moyogalpa, Isla de Ometepe (KU 86258). COSTA
RICA: Prov. Guanacaste: 0.5 mi N Colonia Car-
mona (CRE 8221); 6 km NW Las Canas (CRE
2905); Finca Taboga, = 20 km SE Las Canas (KU
102528-29; LACM 64595); 1 mi N, 2 mi E Santa
Rosa (TCWC 17169); 1 km W Tilaran (UCR 1927);
4 km E Los Angeles de Tilaran (KU 35516-17); 0.1
mi SE Finca Jimenez (CRE 3087); 8.4 mi from
Tilaran to Las Cafias (CRE 6280); Prov. Heredia:
La Ribera, Belen(UCR 1038); Prov. San Jose: San
Jose (ANSP 3363).

Pattern B.— NICARAGUA: no other data
(USNM 15205). COSTA RICA: Prov. Cartago: 1
mi E Puente Rio Reventazon (UCR 2144); Prov.
Limon: Rio Jimenez, Pococi (UCR 3370); Prov.
Puntarenas: Finca Las Cruces, 6 km N San Vito
de Java (RWM 6756); [Rincon, Peninsula de] Osa,
near airport (LACM 125563); 4 km S San Vito
(LACM 125562); Prov. San Jose, 10 mi NW San
Isidro El General (KU 31960). PANAMA: Prov.
Chiriqui: Finca Palosanto, = 6 km WNW El
Volcan (KU75758); Prov. Code: El Valle de
Anton (KU 112495).

Intermediate specimens. — PANAMA: Prov.
Darien: 0.5 mi S Camp Chucunaque (FMNH
170141); Prov. Panama: Rio Chagres (MCZ
37152); Rio Chagres, Camp San Juan de Pequeni
(MCZ 38243); Agua Clara, Chagres (ANSP
22267); Arraijan, Chorrera (ANSP 22277); El
Aquacate, N of Capira (UF/FSM 29869-70); La
Cumbres (KU 75757); Isla Taboga (MCZ 9999);
Panama (MCZ 28048, 37114); Prov. Veragua:
Macaraquito Camp (ANSP 22464-65); Zona del
Canal: Ancon (MCZ 23870-72); Barro Colorado

29

Island (ANSP 24581; KU 75756, 117020);
Cardenas (UF/FSM 29868); near Fort Clayton
Reservation (UIMNH 41733); Fort Randolph
(MCZ 18849); Fort Sherman (MCZ 26764);
Howard Air Force Base (KU 117021); Isla
Gigante (ANSP 24750); Summit (FMNH 83554).
COLUMBIA: Depto. Antioquia: 4 km NE
Bellavista, above Rio Force (FMNH 63756);
Jerico, W of Medellin (AMNH 35243); Medellin
(AMNH 35470, 35472, 35474-79, 35483; BMNH
97.11.12.14-15); Rio Negrito, 15 km E Sonson
(FMNH 63757); Santa Rosa de Osas (AMNH
35781); Valdivia (FMNH 69659); Depto. Bolivar:
Cartegena (FMNH 165193); Depto. Choco:
Camino de Yupe (LACM 72749); Condoto
(BMNH 1910.7.11.32); Depto. Cundinmarca: La
Mesa, near Bogota (MCZ 46418); Sasaima
(UMMZ 78264); near Bogota (ZSM E4/1973);
Depto. Magdalena: Santa Marta Mts. (UMMZ
54957); Depto. Meta: = 15 km W Vista Hermosa
(UTACV 3476); Depto. Nariiio: Sandona (FMNH
54884); Tumaco (AMNH 20400); Depto. San-
tander: 42 km NW Bucaramanga, near Rio
Negro, Hac. "BerUn" (ZSM 96/1934-1-2); Depto.
Valle: Cah (AMNH 4480); La Concepcion (MCZ
6552).

Patterns C and D.-ARGENTINA: Prov.
Buenos Aires: Buenos Aires (NMW 13813); Prov.
Salta: Embarcacion (LACM 73988). BOLIVIA:
Depto. La Paz: Tumupasa (=Tumapasa] (AMNH
22504); Depto. Santa Cruz: Buenavista (CM 2823,
2826, 2908; FMNH 35642-43; ZSM E2/1973); La
Perdiz (CM 2851) [not mapped). BRAZIL: Est.
Amazonas: Rio Manjuru [ = Rio Mamuru]
(AMNH 101970-71); Rio Purus (IRSNB 14918);
Est. Goias: Santa Isabel, Ilha de Bananal, Rio
Araguaia (AMNH 87965); Est. Mato Grosso:
Chupada | = Chapada dos Guimaraes] (ANSP
11215); near Sao Cruz, road to Villa de Mato
Grosso (CM 337); Tapirape, confluence of
Araguaia and Tapirape rivers (AMNH 87964);
Est. Para: Belem (IRSNB 14915; KU 140178-79);
3 km E Belem (KU 128261); Santarem (AMNH
2841); Est. Rio Grande do Norte: Extremoz (CAS-
SU 6971); Est. Sao Paulo: Sao Paulo (MCZ
20805-06). BRITISH WEST INDIES: Tobago:
no other data (MCZ 6337); Bacolet Guest House
(AMNH 81447); 4 mi N Roxborough (RT 1221);
Scarborough- Bacolet rd. burial ground (MCZ
56266); Trinidad: no other data (NMB 8993-94);
Diego Martin (UF/FSM 16513); Guayaguayare
(NMB 11363); Mt. Catherine (AMNH 64543);
Piarco (FMNH 40445); Pointe-a- Pierre (MCZ
49068; UF/FSM 16768); Port-of-Spain (MCZ

8987, 87341, 126232); St. Ann's Ward, 7.5 mi N
San Juan (RT 1217); San Raphael | = San Rafael]
(FMNH 49970); Simla (AMNH 73092-93); Toco
(AMNH 94888; MCZ 10138); Tucker Valley
(AMNH 64473-74, 101341-42). COLOMBIA:
Depto. Atl^ntico: mouth of Rio Magdalena
(NMW 20367); Depto. Meta: Lomalinda (UTACV
3717); Depto. Vaupes: Timbo (UTACV 3535);
Gino-Goye, Rio Apaporis (MCZ 53236).
GUYANA: N Acarahy Mountains, W New River
(KU 69836) [not mapped]; Dawa (ASFS 1002) [not
mapped]; Georgetown (AMNH 36139); Kalacoon
(AMNH 8099); Kartabo (AMNH 15147,
98216-17); Lama Creek, Demerara River (AMNH
36109) [not mapped[; Oko Mountains (FMNH
26667-68) [not mapped[; Paruima Mission
(UMMZ 85292) [not mapped]; Placers de
Carsevenne (MNHN 99-81) [not mapped];
between Wichabai and Isheartun (AMNH 60828).
PARAGUAY: Depto. Cordillera: San Bernardino
(ZSM 50/1925). SURINAM: no other data (MCZ
5394); Suriname Dist.: Powakka (CM 44323).
URUGUAY: Depto. Tacuarembo: Banda (NMB
2118). VENEZUELA: Est. Aragua: Maracay
(ZSM 136/1930); Est. Bolivar: Arabupu
[ = Arabopo] (UMMZ 85294-95); Guri (MBUCV
III-1825); Mt. Roraima, Paulo side (UMMZ
85293); Rio Uraricapara [ = Rio Uraicapa]
(MBUCV III-8151); Est. Carabobo: Belen
(MBUCV III-3142); Borburata (MBUCV
III-8148); Dist. Federal: Caracas (NMB 11921);
El Mamo (MBUCV III-3141); La Mariposa
(MBUCV III-3842); Naiguata (CM 22774); Est.
Merida: Merida (FMNH 4073; MCZ 17096,
17102; UMMZ 46620, 55901); Est. Falcon: El
Mene (NMB 9236-37, 9982, 12862); Pauji (MCZ
48762-65, 49033; NMB 13859); Riecito (NMB
13914); Est. Sucre: 24 km (by Puerto La Cruz
Road) W Cumana (KU 117056); Peninsula de
Paria, Hac. Yacua (MCZ 43854).

Pattern E.— COLOMBIA: Depto. Amazonas:
Leticia (MCZ 49003): Depto. Caldas: Laguneta
(ANSP 25438); Depto. Cauca: El Tambo (ANSP
25428); Depto. Cundinmarca: near Bogota (ZSM
E4/1973); Depto. Putamayo: Putamayo, Rio
Putamayo (FMNH 37448). ECUADOR: Prov.
Chimborazo: Riobamba (AMNH 23266, 23283);
Prov. Morona: Gualaquiza (AMNH 25190);
Miazal (USNM 198737); Prov. Napo: Concepcion
(USNM 198734); Lago Agrio (KU 126049-51);
Limoncocha (LACM 73334; UIMNH 61257);
headwaters of Rio Arajuno, tributary Rio Napo
(USNM 198735); Santa Cecilia (KU 109848; MCZ
96667-68); Sarayacu (AMNH 28795); Prov. Pas-

30

taza: Alto Rio Curaray (USNM 198717, 198722)
[not mapped]; Canelos (AMNH 49060-61; NMW
20366:3-4; USNM 198721); Mera (KU 98624,
121342); Puyo (USNM 198715); Rio Corrientes
(USNM 198720); Rio Pindo, near village of Rio
Tigre (USNM 198716) [not mapped]; Rio Villano
(USNM 198718-19); Prov. Santiago: Rio Upano
(AMNH 28813). PERU: Depto. Amazonas: upper
Rio Nieva (AMNH 55887); Depto. Ayacucho:
Sivia, on Rio Apurimac (FMNH 39645); Depto,
Cuzco: Marcapata (FMNH 59173-74, 62128)
Depto. Huanuco: [Rio] Pachitea (AMNH 52683)
= 35 km NE Tingo Maria (LSUMZ 25441)
Depto. Junin: La Merced (ZSM 138/1930-1-4)
Depto. Loreto: Centro Union (TCWC 38209,
39123, 40532, 41436, 44105); Iquitos (AMNH
52496, 52753, 53122, 53164, 54891, 56121); Lago
Mirano, lower Rio Napo (AMNH 53310); Mishana
(TCWC 38206-08, 44106); Monte Carmelo (Re-
quena) (AMNH 55928); Nanay (AMNH 52496)
Pampa Hermosa (AMNH 53562, 55440, 55759
55768, 55859, 55894, 55966); Panya, Rio Cusha
batay (AMNH 52353); Pevas (CAS-SU 12497)
Rio Curanja [ = Rio Curanga], Balta (LSUMZ
14595); Rio Samiria and Parinari Caiion (AMNH
57265); mouth of Rio Santiago (AMNH 52687,
53251); Roaboya (AMNH 52476, 52480, 53093);
Tipishca (Contamana) (AMNH 52172); Yarino-
cocha, Rio Ucayali (FMNH 56125); Depto. Pasco:
Huancabamba (NMW 20363:1-3); Depto. San
Martin: Rio Ponasa (AMNH 52820); Tocache
(AMNH 52391).

Pattern F.-ECUADOR: Prov. Esmeraldas:
Quininde [ = Rosa Zarate] (NMW 20366:2); Prov.
Guayas: Guayaquil (ANSP 3353); Prov. Loja:
Loja(USNM 198736); Prov. Pichincha: Chiriboga
(USNM 198733); Meme (USNM 198729); below
Pacto (USNM 198731); Quito (ANSP 24777); Rio
Baba, 19 km S, 5 km E Santo Domingo de los Col-
orados (UIMNH 92236-37); Rio Baba, 24 km S
Santo Domingo de los Colorados (UIMNH
92238); Rio Toachi, 5 km N, 1 km E Santo Dom-
ingo de los Colorados (UIMNH 92239) [not
mapped]; 4 km E Rio Baba "bridge," 24 km S
Santo Domingo de los Colorados (UIMNH
92240); Santo Domingo de los Colorados (NMW
20366:1; MCZ 147302; USNM 198723-25,
198728); below Sigchos (USNM 198730).

Key to the Members of the Tantilla
melanocephala Group
1. Nuchal spots confluent with dorsal body
ground color; no dark nape band; dorsal
ground color pale with diffuse dark stripes

on middle of row 1 , upper half of 2 and lower
half of 3, middle of 4, upper half of 5 and
lower half of 6; middorsal dark stripe well-
defined, present on middorsal scale row, con-
nected to dark markings along medial edge
of paravertebral scales . . T.insulamontana
Color pattern not as above 2

2. Dorsal ground color pale with bold dark

stripes on middle of row 1, upper half of row
2 and lower half of row 3, upper edge of row
4 and lower middle portion of row 5, middle
of row 6, and the middle of the middorsal
scale row; preocular spot continues poster-
iorly onto preocular scale to touch eye

T. equatoriana

Color pattern not as above 3

3. Dorsum of head with much pale pigment, con-

sisting of extensive pale markings on snout
continuing posteriorly over supraocular and
anterolateral corner of parietal and down-
ward to connect with the postocular pale
spot; dorsal pattern consisting of dark mid-
dorsal stripe confined to middorsal row and
pale lateral stripe confined to upper half of
row 3 on a pale ground color . . T. andinista
Color pattern not as above 4

4. No dark middorsal stripe 5

Dark middorsal stripe present 6

5. Pale nuchal band complete or divided medi-

ally; pale neck band following dark nape

band T. capistrata

Pale nuchal band usually divided both
medially and laterally; no pale pigment

following dark nape band

T. melanocephala (pattern C)

6. Dark pigment bounding pale neck band

posteriorly 7

No dark pigment bounding pale neck band (if
present) posteriorly 8

7. Nuchal spots distinct and usually restricted

to parietal scales; no dark ventral border of
pale lateral stripe (if present); lateral exten-
sion of dark head cap separating postocular
pale spot and pale pigment on lateral gulars

(opposite pale nuchal spots)

T. melanocephala (pattern A)

Nuchal spots indistinct and largely confined
to scales posterior to parietals; dark ventral

31

border of pale lateral stripe present; no
lateral extension of dark head cap separ-
ating postocular pale spot and pale pigment
on lateral gulars (opposite pale nuchal spots)
T. lempira

8. Middorsal stripe poorly developed, diffuse,

often broken; pale lateral band usually
absent, but, if present, not bounded by dark

pigment T. melanocephala (pattern D)

Middorsal stripe well-developed, complete;
pale lateral band usually present and
bounded above and/or below by dark
pigment 9

9. Dark nape band well-developed, well-distin-

guished from ground color of head and body;
pale nape band usually present and not

reduced in extent

T. melanocephala (pattern E)

Dark nape band usually poorly developed,
barely distinguishable from ground color of
head and/or body; pale neck band posterior
to dark nape band usually reduced to small
spots on either side of middorsal dark stripe
or absent 10

10. Lateral extension of dark head cap separating
postocular pale spot and pale pigment of
lateral gulars (opposite pale nuchal spots)
present; color of dorsum of head essentially
the same as that of the dorsum of body ....

T. melanocephala (pattern F)

Lateral extension of dark head cap separating
postocular pale spot and pale pigment of
lateral gular (opposite pale nuchal spots)
absent; color of dorsum of head paler than

that of dorsum of body

T. melanocephala (pattern B)

Distributional Patterns
Geographic patterns.— The most striking
aspect of the geographic distribution of the mem-
bers of the Tantilla melanocephala group is the
wide range of the nominate species and the
peripheral or virtually peripheral ranges of the
remainder of the species. As herein interpreted,
Tantilla melanocephala has the widest range of
any species in the genus, occurring from Guate-
mala south to southern Peru, Bolivia, northern
Argentina, and Uruguay. The other 5 species are
scattered along the edges of the range of the cen-
tral species, apparently in habitats that are
marginal for or outside the habitat span of T.

melanocephala.

Tantilla lempira sits almost at the northern
extremity of the range of the group and occupies
the upland pine forests along the Pacific versant
of Honduras. Because much of the interior of
Honduras is covered by such forest, perhaps T.
lempira has a much larger range in the country
than is now known.

Two species, T andinista and T. insulamon-
tana, occur in relatively isolated high intermon-
tane basins in the Andean chain in Ecuador. Tan-
tilla andinista is known only from the Alausi
Basin and T. insulamontana from the Jubones
Basin. Both of these basins or hoyas are drained
by rivers that flow into the Pacific and the Pacific
slope of Ecuador is sparsely populated by mem-
bers of the Tantilla melanocephala population
system (only members of pattern type F).

Tantilla equatoriana occurs in extreme north-
western Ecuador. This area is covered by rain
forest (Savage, 1960) and it is not immediately
apparent why T. melanocephala does not occur in
this area, so far as is known, because it occurs
across the border along the Pacific coastal plain
of Colombia. Perhaps more collecting in north-
western Ecuador will demonstrate the existence
of T. melanocephala there.

Tantilla capsitrata occupies the hostile coastal
deserts and the semiarid inter-Andean basin
formed by the valleys of the upper Rio Maranon,
Rio Chinchipe and Rio Chamaya in northwestern
Peru. Again the area appears to be inhospitable
to the more mesic-adapted T. melanocephala.

As mentioned above, T. melanocephala is wide-
ranging and a centralized member of the group.
In Central America it occupies both subhumid
(pattern A) and humid (pattern B) environments.
In northern Central America (Guatemala to
Nicaragua) the species occurs primarily along the
Pacific versant, whereas in lower Central
America (Costa Rica and Panama) it is widely
distributed on both versants. In South America it
appears to occur in any habitat in the portion of
the continent east of the Andes as far south as
the humid pampa that is sufficiently humid and
low in elevation. It also occurs along the lower
western flank of the Andes in northwestern
Ecuador and in the lowland and riverine fingers
that interdigitate with the 3 cordilleras in
Colombia.

Physiographic patterns.— As might be pre-
dicted by its extensive geographic distribution,
Tantilla melanocephala inhabits a wide range of
physiographic types. In Central America, pattern

32

A individuals principally occupy the Pacific
coastal plain but also range into the lower eleva-
tions of the Pacific slopes of the central mountain
ranges (range in elevation sea level to 1450 m).
Although principally occurring on the Pacific ver-
sant, pattern A individuals are known to occur on
the eastern side of Lake Nicaragua on the Carib-
bean side.

Pattern B individuals occur in the Atlantic low-
lands in Nicaragua (?) and Costa Rica. They also
occur along the Pacific lowlands and the foothills
of southwestern Costa Rica and western Panama.
They have been collected at elevations ranging
from 30 to 762 m.

The intermediate specimens in Panama occupy
the lowlands in the area of the Canal Zone and the
Pacific lowlands of the eastern portions of the
country at elevations not exceeding 600 m.

In Colombia the intermediate specimens occur
principally in the lowland region (Atlantic and
Pacific) and in the river valleys which divide the 3
Cordilleras (Cordillera Occidental, Cordillera Cen-
tral, and Cordillera Oriental) from one another.
The Atlantic coastal specimens are from eleva-
tions near sea level, with the exception of a speci-
men from the Santa Marta Mountains, which
came from an unknown locality. Ruthven (1922)
mentioned specimens, which we have not seen,
from Cerro San Lorenzo and Palomina at eleva-
tions, respectively, of 1220 and 1525 metres. He
also mentioned specimens from Fundacion and
Valencia, which are on either side of the main
mass of the Santa Marta Mountains but gave no
elevation for either. The specimens from the
Pacific coastal region are from the Serrania de
Baudo at an elevation of 420-625 m, the valley of
the Rio San Juan near the southern end of the
Serrania de Baudo, and from sea level along the
southwestern Pacific lowlands. The latter speci-
men shows no resemblance to the 2 known speci-
mens of Tantilla equatoriana, although the
localities whence they came are but approximate-
ly 70 airline kilometres apart and covered with
presumably the same type of vegetation. Speci-
mens from the Colombian Andes have come pri-
marily from the rift valleys between the 3 Cordill-
eras. These 3 mountain ranges merge at their
southern ends but separate at the northern end
providing a relatively low elevation dispersal
route for Atlantic coastal animals into the Colom-
bian Andes. Tantilla melanocephala has been
recorded from this area at elevations as low as
135-180 m at the northern end of the Cordilleras
to 2801 m in the Medellin Valley.

Tantilla melanocephala also occurs along the
narrow Caribbean coastal plain in Venezuela from
the area east of the Segovia Highlands to that
north of the mountains of the northern range of
the Central Highlands northeast of the valley of
Caracas at elevations not exceeding 152 m. It also
occurs in the Merida Basin along the Rio Chama
at an elevation of 1640 m in the Cordillera de
Merida. In the Central Highlands it occurs in
both the northern and southern ranges and in the
valleys that separate them (valley of the Rio Tuy,
valley of Caracas, and the Valencia Basin) at
elevations ranging from 305-1000 m. The species
also inhabits the Northeastern Highlands from
the mountains southwest of the Golfo de Cariaco
to the Paria Peninsula.

Tantilla melanocephala is also found on the
islands of Trinidad and Tobago, geologically part
of South America. The species is widespread on
Trinidad, occurring in most (if not, in fact, all) of
the physiographic divisions of the island. It has
been recorded from the area of the Northern
Range at elevations from near sea level to 457 m.
It occurs in the northern lowland belt along the
Caroni and Talparo rivers (one of 2 locahties from
this region is at = 14 m). In the southern lowland
belt specimens have been found along the western
coast at or near sea level. Specimens from the
region of the southern highlands are from the
southeastern coast at or near sea level.

Back on the mainland, Tantilla melanocephala
has been recorded from the eastern portion of the
Orinoco llanos. The species has also been found in
the eastern and extreme western portions of the
Guiana highlands in, respectively, southeastern
Venezuela (at or near 305 m) and southern
Guyana, on the one hand, and in southeastern
Colombia, on the other. Its lack of occurrence in
the central portions of the Guiana highlands is
due, no doubt, to insufficient collecting. North-
east of the Guiana highlands is a coastal plain
where T melanocephala also occurs in Guyana
and Surinam (no records are available for French
Guiana but the snake undoubtedly occurs there).

Occurrence of T. melanocephala in the alluvial
plain of the Rio Amazonas is also documented.
Near the mouth of the river it has been found at
elevations of 12 m and farther upstream at 20 m
near Santarem. In the upper drainage of the
Amazon Basin the distribution is better known.
It occurs in the Amazon lowlands of southern Col-
ombia, eastern Ecuador, and eastern Peru from
elevations as low as 84 m in southeastern Colom-
bia, into the eastern flanks of the Andes in

33

Ecuador up to 1020 m and along the valleys and
tributaries of the Ucayali, Huallanga and Madre
de Dios rivers in the eastern Peruvian Andes up
to 3080 m. The species also occurs in the southern
extension of the Amazon Basin along the Rio
Grande in eastern Bolivia (823 m) and along the
base of the front range of the Bolivian Andes
(457-610 m).

In the vast region of the Brazilian highlands T.
melanocephala has been sparsely but so widely
collected that we can say that it probably occurs
in much of the area. It has been recorded from the
coastal plain of northeastern Brazil, Ilha de
Bananal, Matto Grosso Plateau, the Serra do
Mar, Parana Plateau, and the Uruguayan
highlands.

Tantilla melanocephala also is found in the
Gran Chaco and in the Humid Pampa, the latter
being the southernmost physiographic province
in which the species is known to occur.

Finally, although most of the distribution of T.
melanocephala is east of the Andean chain, the
species also occurs west of the Andes in the
Guayas lowland (at sea level) and the coastal hilly
belt (600-1863 m) in Ecuador. It also is found in
the intermontane basins of Quito (2851 m), Rio-
bamba (2700 m), and Loja (2200 m).

The other species of the melanocephala group
occur in areas outside of the range of Tantilla
melanocephala. Tantilla lempira is distributed in
the eastern portion of the southern cordillera of
the serrania region of Honduras {sensu Meyer,
1969). Tantilla equatoriana has been found only
in the coastal lowlands north of the coastal hilly
belt in Ecuador. Tantilla andinista is known from
the Alausi intermontane basin at an elevation of
2600-2750 m. Tantilla insulamontana occurs in
the Jubones intermontane basin at known eleva-
tions of from about 1250-2100 m. Tantilla capis-
trata is distributed along the northern coastal
plain of Peru and in the valleys of the Maranon,
Chinchipe, and Chamaya rivers in north-central
Peru.

Phytogeographic patterns. — Little direct
evidence of the types of vegetation inhabited by
members of the Tantilla melanocephala group is
available. What follows is an effort to reconstruct
a phytogeographic picture for the group from
information provided by a wide variety of
sources. It must be acknowledged at this point
that it is difficult to reconcile the various systems
that have been used to describe the vegetational
patterns of different areas within the range of the
melanocephala group. We have tried to apply the

Holdridge (1964) system of vegetatonal forma-
tions as broadly as possible but have found it dif-
ficult to do so in some cases due to lack of suffi-
cient information and have reverted to other, less
precise sorts of descriptions.

In Central America, individuals belonging to
the pattern A grouping are principally distri-
buted in the Tropical Dry Forest formation (in El
Salvador, Honduras, Nicaragua, and Costa Rica)
but they also extend marginally into the Sub-
tropical Moist Forest formation in Guatemala,
Nicaragua, and Costa Rica, the Subtropical Wet
Forest formation in Guatemala, and the Tropical
Moist Forest formation in Costa Rica.

Pattern B individuals are widely distributed in
wet forest types, occurring in Tropical Wet
Forest, Subtropical Wet Forest and Subtropical
Rain Forest formations in Costa Rica and in the
Subtropical Wet Forest formation in Panama.

The intermediate specimens in Panama came
principally from the Tropical Moist Forest forma-
tion but also from the Tropical Dry Forest
formation.

Tantilla melanocephala is widely distributed in
South America and, as such, inhabits many of the
vegetational types found on that continent east
of the Andean chain and south to the Pampas. In
Colombia the intermediate individuals have been
found in 3 areas: (1) the Atlantic coastal plain: (2)
the Pacific coastal area; (3) the rift valleys
between and the lower slopes of the Colombian
Andes. The Atlantic lowlands of Colombia are in
Holdridge's Tropical Thorn Forest formation or
thorn scrub. Along the Pacific coast of Colombia
the snake occurs in the Tropical Wet Forest for-
mation or tropical lowland rain forest. In the rift
valleys between the 3 cordilleras of the Colom-
bian Andes, Tantilla melanocephala occurs at
lower elevations in the Tropical Moist Forest for-
mation and at higher elevations in formations
presumably ranging from Subtropical Moist
Forest through Lower Montane Moist Forest and
into Montane Wet Forest.

Moving eastward from western Colombia, T.
melanocephala ranges along the Venezuelan
coastal plain into the strip of thorn scrub (Tropi-
cal Thorn Forest formation) that is an extension
of that along the Caribbean coastal plain of Col-
ombia. It also occurs in the Lower Montane Moist
Forest formation in the Merida Basin in the Cor-
dillera de Merida. In the Central Highlands of
Venezuela, the species ranges into the Subtropi-
cal Dry Forest formation.

On Trinidad the eastern coast receives from

34

2540-3810 mm of rainfall. The rainfall decreases
toward the west to between 1270 and 1524 mm
along the west coast (James, 1959). The eastern
portion of Trinidad would fall within Holdridge's
(1967) Tropical Moist Forest formation (quasi-
rain forest) and the western coast within the
Tropical Dry Forest formation. Tantilla melano-
cephala occurs virtually throughout the island.

Tantilla melanocephala has also been recorded
in the eastern portion of the Orinoco llanos, an
area of grassland with "semideciduous trees
growing singly or in groups or patches" (Uzzell,
1973). In the Holdridge system the area would
fall in his Tropical Arid Forest formation, a
slightly misleading application. In the Guiana
Highlands the species occurs in the Tropical
Moist Forest and Tropical Wet Forest
formations.

Along the coastal plain of the Guianas T.
melanocephala occurs in the Tropical Moist
Forest formation. In the lower Amazon alluvial
plain the species is distributed in the Tropical
Moist Forest and Tropical Dry Forest forma-
tions. In the upper drainage it occurs principally
in the Tropical Moist Forest formation but also in
the Subtropical Wet Forest formation of the
eastern flanks of the Ecuadorian and Peruvian
Andes.

In the Brazilian highlands Tantilla melano-
cephala has been found in the caatinga, the
thorny deciduous woodland of northeastern
Brazil, and the woodland savanna of central
Brazil. Along the coast of southeastern Brazil it
occurs in the Tropical Moist Forest formation, in
tropical lowland rain forest or in luxuriant
semideciduous forest (James, 1959).

In the Uruguayan highlands T. melanocephala
occurs in tall-grass prairies and in Paraguay in
tropical semideciduous forest. In Argentina the
species has been found in the vast grass-covered
plain of the Humid Pampa and in the deciduous
scrub forest interspersed with patches of grassy
savanna known as the Gran Chaco.

The ecological distribution of T melanocephala
in Bolivia is not clear but it appears that the
species may occur in the arm of selva that
extends south along the front ranges of the Boli-
vian Andes or in the palm savanna to the east of
there.

On the Pacific side of Ecuador Tantilla melano-
cephala is distributed in the deciduous broadleaf
forests in the Guayas lowlands, in the Subtropi-
cal Moist Forest formation or premontane rain

forest of the coastal hilly belt, and in the Montane
Moist Forest formation or montane grasslands of
the Quito and Riobamba basins and the semiarid
vegetation (Montane Dry Forest formation) of
the Loja basin.

Tantilla lempira occurs in the Subtropical
Moist Forest formation (Pinus oocarpa forest) in
south-central Honduras. Tantilla equatoriana
occurs in the Tropical Moist Forest formation of
coastal northwestern Ecuador. Tantilla andinista
is found in the montane grasslands of the Alausi
basin. Tantilla insulamontana is known only from
the Jubones basin which supports semidesertic
vegetation (Acosta-Solis, 1966). Arthur C.
Echternacht, who collected the holotype
informed me (personal communication) that "the
valley of the Rio Jubones is increasingly drier as
one moves east from the coast, and was quite dry
at the Rio Minas." Tantilla capsitrata occurs in
the hostile barren coastal desert of northwestern
Peru and the semiarid inter-Andean valleys of the
Maraiion, Chinchipe and Chamaya rivers. The
extremely arid coastal desert results from a rain-
shadow effect produced by a combination of the
effects of the Andes to the east and the cold-water
Humboldt Current offshore (Dixon and Wright,
1975). The inter-Andean valleys of north-central
Peru lie in a ". . .rain shadow formed by the
eastern front ranges of the Andes" (Dixon and
Wright, 1975). Dixon and Wright (1975) noted 3
plant formations characteristic of the coastal
desert: (1) desert— an area with no or only isolated
plants; (2) lomas— isolated, elevated hills fre-
quently shrouded with fog and supporting
ephemeral herbs; (3) riparian communities— trees,
shrubs, and herbs growing adjacent to streams
having their origin high in the Andes. Although
we have no direct evidence about which of these
plant formations T. capistrata inhabits it seems
more likely that they would occur in the riparian
areas. Dixon and Wright (1975) described the
nonriparian vegetation of the inter-Andean
valleys as consisting of "thorn scrub with mes-
quite (Prosopis), paloverde (Cercidium), and col-
umnar cactus (Cereus and Cephalocereus)
dominating."

Relationships and Evolution
Study of interspecific relationships within the
melanocephala group of the genus Tantilla has
been hampered by the small sample size of most
of the species herein recognized, which precluded
examination of internal characters. At any rate.

35

internal characters have not been subjected to
analysis in the genus as a whole. This is a subject
the senior author hopes to take up in future
papers. Thus, we have been limited in attempting
to assess relationships to consideration of exter-
nal characteristics. It is the characteristics of the
pattern that appear to be most useful in this
regard, inasmuch as the characters of scutellation
are either relatively invariable or too variable.

Relationships.— The pattern character states
common to all members of the melanocephala
group are as follows: (1) dark head cap present; (2)
pale marking present on nuchal area; (3) preocular
pale blotch present; (4) postocular pale blotch pre-
sent. Some of the remainder of the pattern char-
acter states are so distributed that small clusters
of pattern groupings are apparent. The pattern
grouping clusters and the significant features
shared in common are as follows:

A. Tantilla melanocephala (pattern A)— T. lem-
pira: little, if any, distinction in intensity of color
between dark head cap and dark nape band; pale
pigment on snout restricted to dorsal apex of
rostral and anteromedial portion of internasals;
nuchal spots small, well-separated from dorsal
extension of pale gular color; pale neck band pre-
sent; dorsal ground color tan to pale brownish-
gray; middorsal dark stripe narrow, confined to
middle of middorsal scale row; pale lateral stripe
confined to upper portion of dorsal scale row 3
and lower portion of 4.

B. Tantilla equatoriana—T. insulamontana:
nuchal spots relatively large, present on posterior
portion of parietals and postparietal scales; mid-
dorsal stripe extends onto paravertebral scales;
dark stripe through middle of dorsal scale row 1;
dark stripe on upper half of scale row 2 and lower
half of 3; pale lateral stripe on upper portion of
scale row 3 and lower portion of 4; dorsal scale
rows 4, 5, and 6 with dark striping; pale dorso-
lateral field extending from upper half of scale
row 6 to lateral apex of scales of middorsal row.

C. Tantilla melanocephala (pattern C and
D)—T. melanocephala (pattern E): dark head cap
paler in color than dark nape band; pale pigment
present on internasals and prefrontals; pale
nuchal band medially divided; pale neck band pre-
sent; pale lateral stripe, if present, relatively
broad, usually occupying the upper half of scale
row 3 and all of 4.

The remainder of the pattern groupings are
such that the number of distinctive pattern
features outweigh those held in common with any

other taxon or pattern group. These pattern
groupings and their distinctive suites of pattern
character states are as follows:

A. Tantilla melanocephala (pattern B).— dark
head cap much paler in coloration than dark nape
band; no pale pigment on snout; pale nuchal spots
often not present; dark nape band sometimes not
distinguishable from dorsal ground color; pale
neck band usually absent; dark subocular blotch,
if present, poorly developed, consisting of dark
pigment along suture lines between supralabials
3, 4, and 5; lateral extension of dark head cap
usually absent; pale lateral stripe relatively
broad, usually occupying the upper half of scale
row 3 and all of 4 or the upper half of 3, all of 4,
and the lower half of 5.

B. Tantilla melanocephala (pattern F).— usually
no distinction between the color of the dark head
cap and dark nape band, which in turn is usually
not distinguishable from the dorsal ground color;
no pale pigment on snout; pale nuchal spots
muted, occupying the posterior portion of the
parietals and postparietal scales; pale neck band
either absent or confined to one or two scales on
either side of the dark middorsal stripe.

C. Tantilla andinista.— extensive pale markings
on the internasals, prefrontals, supraoculars, and
parietals; pale nuchal band broad and medially
divided; pale lateral stripe narrow, confined to
upper half of row 3.

D. Tantilla capistrata,— dark head cap same
color as dark nape band, which both are much
darker than the dorsal ground color; snout to
level of anterior half to three-quarters of prefron-
tals pale in color; nuchal band well-defined, med-
ially divided or complete; dorsum pale in color,
with no dark middorsal stripe or only sUght indi-
cation of one; no pale lateral stripe.

These groupings seem to point to relatively
close genetic relationships between those pattern
groupings showing a high number of shared
character states, on the one hand, and more dis-
tant relationships of those having a high number
of distinctive character states with any other pat-
tern grouping, on the other.

Further analysis reveals that the pattern B, E,
and F specimens of T melanocephala resemble
one another in having a relatively dark dorsal
ground color, a well-defined middorsal dark
stripe, a well-defined, contrasting, relatively
broad lateral pale stripe, usually occupying at
least the lower half of scale row 3 and all of 4. In
addition, T. capistrata resembles patterns C and

36

D specimens of T. melanocephala in having an
indistinct middorsal dark stripe (if any), a
relatively pale dorsal ground color, no pale lateral
stripe and a well-defined medially divided or com-
plete pale nuchal band. A close evolutionary rela-
tionship between the members of the latter group
is equivocal given the differences in habitat and
the degree of isolation from one another (but see
below). Tantilla andinista remains a distinctive
isolate.

Evolutionary trends. — A worker attempting to
construct a phylogeny based on an analysis of
evolutionary trends determined by changes in
pattern character states is treading on thin ice
indeed. Nonetheless, we have made an effort to do
so, risking a fall into the cold water of sceptical
criticism. Following the arrangement used by
Savage (1975), we have ranked below each of the
important character states discussed as primitive
(P), intermediate derived (I), or advanced (A).
Where the character state trends are multidirec-
tonal they are listed as i, ii and iii:

than the dark nape band; no or few pale markings
on the snout; pale nuchal band medially divided;
subocular blotch well-developed; lateral extension
of the dark head cap present; dark nape band pre-
sent and darker than ground color; pale neck
band well-developed.

The above description fits perfectly members of
the pattern E group of Tantilla melanocephala.

Evolutionary model.— The data derived from a
study of evolutionary trends in color and pattern
indicate that the pattern E group, located in the
upper reaches of the Amazon basin, is closest to
the postulated basal evolutionary stock of the
melanocephala group. The pattern C-D group,
occupying the bulk of the South American range
of the species, appears to be the most recent off-
shoot of the basal stock, differing from the pat-
tern E group in having a paler ground color with
less well-developed (or nonexistant) middorsal
dark stripe and pale lateral stripe. Another level
of evolutionary change is represented by pattern
F and pattern B specimens. Each of these pattern

P

I

A

middorsal stripe

i) well-developed

reduced

absent

ii) well-developed

broadened

pale lateral stripe

well-developed

reduced

absent

dorsolateral field

uniform

striped

dark head cap

i) uniform

pale

andinista

snout

pattern

ii) slightly paler

much paler
than dark
nape band

iii) slightly paler

same color as
dark nape
band

pale nuchal band

i) medially

small

absent

divided

spots

ii) medially

complete

divided

subocular blotch

well-developed

reduced

lateral extension of

dark head cap

present

absent

pale neck band

well-developed

reduced

absent

dark nape band

i) present
ii) distinct from

absent

ground color

not distinct

This disposition of character states suggests
that the basic evolutionary stock of the Tantilla
melanocephala group possessed the following
features: middorsal stripe well-developed; pale
lateral stripe well-developed; dorsolateral field
uniform in color; dark head cap sUghtly paler

groups presently has a relictual distribution, pat-
tern B in the humid lowlands of Nicaragua, Costa
Rica and western Panama and pattern F along
the western flanks of the Ecuadorian Andes. A
third level of differentiation is represented by the
pattern A group and T. lempira, both occurring at

37

the northern extremity of the range of the
melanocephala group. The principal distinction of
this group is the reduction of the pale nuchal
band to 2 spots near the posterior end of the
parietals and the lack of distinction between the
ground color of the head and the color of the dark
nape band.

The relationships of T. capistrata appear to lie
most closely with the pattern C-D grouping of T.
melanocephala. They agree in having a poorly
developed dark middorsal stripe (or none at all),
in having no pale lateral stripe (usually) and in
having a medially divided or complete (in T.
capistrata) pale nuchal band. Tantilla capistrata
remains a distinctive isolate, however, in the
deserts of northwestern coastal Peru and the arid
intermontane valleys of the area around the Por-
culla Pass in north-central Peru. A hypothesis to
account for the present isolation of these 2 groups
and their separation by populations of the pat-
tern E group of T. melanocephala is offered in the
succeeding section.

Tantilla andinista is a distinctive isolate in the
Alausi intermontane basin in central Ecuador. It
appears to have no close relatives within the
group which means that it may be a remnant of a
coastal Ecuadoran population no longer extant or
as yet undiscovered.

The last spectrum of evolutionary differentia-
tion is represented by the species T. equatoriana
and T. insulamontana, both occurring as isolates
on the Pacific versant of Ecuador. The distribu-
tion of neither is well understood but presently
they are well-separated geographically and altitu-
dinally from one another. They resemble one
another in having a multilineate pattern and a
broad dark middorsal stripe. They differ from one
another principally in that T. insulamontana
lacks a dark nape band. These 2 species appear to
be early offshoots from the basal evolutionary
stock of the group and to have diverged from one
another relatively early in the evolutionary
history of the group.

Evolutionary history.— There is ever-increasing
information to indicate that the significant
events in the evolutionary history of many recent
groups of South American vertebrates has taken
place within relatively recent times in the
Pliocene, Pleistocene and Holocene epochs. In
addition, this same evidence points to a succes-
sion of several alternating arid and humid
cUmatic phases during which there occurred alter-
nate contractions and expansions of closed

canopy forest and open savanna or thorn forest
formations (van der Hammen, 1961; Haffer, 1969;
Vanzolini and Williams, 1970; Vuilleumier, 1971;
Duellman, 1972; Heyer, 1973; Duellman and
Crump, 1974).

The conclusions arising from this evidence are
in marked contrast to the traditional view of long-
term stability of the South American rain forest
(Darlington, 1957; Schwabe, 1969), on the one
hand, and the antiquarian view of the biogeo-
graphic history of the Central American herpeto-
fauna (Duellman, 1958; Savage, 1966, 1975;
Wilson and Meyer, 1971), on the other. These
disparities were pointed out by Myers (1974).

The evidence concerning the evolutionary and
distributional relationships of the members of the
Tantilla melanocephala lends support to a
hypothesis of the occurrence of marked arid-
humid climatic fluctuations in the Quaternary.

As pointed out by Savage (1975), "the test of
any phylogenetic model is to compare the evolu-
tionary divergences and relations of the studied
group against the known environmental history
of the geographic region where the group occurs."

We submit the following evolutionary model to
account for the present-day evolutionary relation-
ships:

1. Movement of the basal evolutionary stock of
the Tantilla melanocephala group into South
America following the reemergence of the isth-
mian link connecting Middle and South America
at the beginning of the Pliocene (Savage, 1966).

2. Early evolution of the basal stock in the
lowland forest in the upper Amazon region (Haf-
fer, 1969).

3. Isolation of the T. equatoriana— insulamon-
tana stock along the west coast of South America
by the uplift of the Andes in the upper Pliocene
and lower Pleistocene (Haffer, 1969).

4. Isolation, perhaps, of the T. andinista stock
in an intermontane basin during this same period
of Andean orogeny. Another alternative is that
the T. andinista stock moved into the presently
semiarid Alausi Basin from the Pacific coastal
lowlands during an interglacial arid period subse-
quent to the Andean orogeny and became isolated
subsequently during the next glacial pluvial
period.

5. Movement of the pattern A—T. lempira
stock into Central America during an interglacial
arid phase.

6. Isolation of the pattern F stock and pattern
B stock in the Choco and Nechi forest refuges of

38

Haffer (1969), respectively, perhaps, during the
same interglacial period that allowed for the
northward movement of the pattern A-T. lempira
stock.

7. Movement of the pattern C-D stock through
the Amazon Basin presumably during a pluvial
glacial phase. Perhaps, during this period the pat-
tern B group was able to expand its range into
humid environments along the Atlantic coast of
lower Central America and the pattern F group
into similar environments along the western
slopes of the Ecuadoran Andes.

8. Movement of the T. capistrata stock into the
arid Pacific lowlands of northwestern Peru and
the area of the Porculla Pass during an arid phase
and its subsequent isolation during the establish-
ment of the present patterns of vegetational dis-
tribution. Also at this time the patterns A and B
group came into secondary contact and these in
turn with the South American pattern groups to
produce the current picture of distribution in
northwestern South America.

The above explanation of the evolutionary
history of the melanocephala group corresponds
well with that presented by Duellman and Crump
(1974) for the Hyla parviceps group even though
that group's distribution is cis-Andean exclusive-
ly. It also corresponds well with data presented
on Quaternary climatic changes in South Amer-
ica (van der Hammen, 1961; Haffer, 1969, 1970;
Vuilleumier, 1971). However, it is uncertain
which of the postulated major dry- humid alterna-
tion sequences during the Pleistocene or the more
minor, less severe ones during the Holocene are
correlated with the above-detailed evolutionary
events.

Acknowledgments
We thank the following curators who were kind
enough to loan us specimens (museum abbrevia-
tions indicated parenthetically): Walter Auffen-
berg, Florida State Museum, University of
Florida (UF/FSM); Rollin H. Baker, Museum,
Michigan State University (MSU); G.F. de Witte,
Institut Royal des Sciences Naturelles de Belgi-
que, Bruxelles (IRSNB); James R. Dixon, Texas
Cooperative Wildlife Collection, Texas A&M
University (TCWC); William E. Duellman,
Museum of Natural History, University of Kan-
sas (KU); J. Kustos Eiselt, Naturhistorisches
Museum, Wein, Austria (NMW); Alice G.C.
Grandison and A.F. Stimson, British Museum
(Natural History) (BMNH); Ulrich Gruber,

Zoologische Sammlung des Bayerischen Staates,
Munich (ZSBS); Jean Guibe, Museum National
d'Historie Naturelle, Paris (MNHN); Alan E.
Leviton, California Academy of Sciences (CAS);
C.J. McCoy, Carnegie Museum (CM); Roy W.
McDiarmid, University of South Florida, Tampa,
private collection (RWM); Edmond V. Malnate,
Academy of Natural Sciences, Philadelphia
(ANSP); Hymen Marx, Field Museum of Natural
History (FMNH); William F. Pyburn, Vertebrate
Collection, University of Texas at Arlington
(UTAVC); J. Racenis, Museo de Biologia, Univer-
sidad Central de Venezuela (MBUCV); Ohvier
Rieppel, Naturhistorisches Museum Basel, Swit-
zerland (NMB); Douglas C. Robinson, Museo de
Biologia, Universidad de Costa Rica (UCR);
Douglas A. Rossman, Museum of Zoology, Loui-
siana State University (LSUMZ); Jay M. Savage,
University of Southern CaUfornia (CRE); Albert
Schwartz, Miami-Dade Community College,
private collection (ASFS); Dorothy M. Smith,
Museum of Natural History, University of Illi-
nois (UIMNH); Wilmer W. Tanner, Museum of
Natural History, Brigham Young University
(BYU); Richard Thomas, University of Puerto
Rico, private collection (RT); Donald W. Tinkle,
Museum of Zoology, University of Michigan
(UMMZ); Ernest E. Williams, Museum of Com-
parative Zoology (MCZ); John W. Wright, Los
Angeles County Museum (LACM); George R.
Zug, National Museum of Natural History
(USNM); Richard G. Zweifel and Charles W.
Myers, American Museum of Natural History
(AMNH).

We are particularly indebted to Alice G.C.
Grandison and A.F. Stimson, Jean Guibe, and
Edmond V. Malnate for the loan of type material
and to Werner Ladiges for information on the
holotype of Homalocranium hoffmanni.

We would also like to extend our gratitude to
A.F. Stimson for descriptive information on the
holotype of Homalocranium longifrontale,
George R. Zug for the gift of the late James A.
Peters' notes on Ecuadoran Tantilla, and Norman
J. Scott for data on Costa Rican Tantilla and for
depositing material from his personal collection
in the Los Angeles County Museum. We are
grateful to Charles W. Myers for making avail-
able material from Panama and Colombia in
advance of his study on Tantilla from those coun-
tries and to Alan H. Savitzky for information on
color and habitat for the holotypes of Tantilla
andinista and T. insulamontana. Arthur C.

39

Echternacht also supplied additional information
on the holotype of T. insulamontana and other
matters and Douglas C. Robinson provided infor-
mation on the whereabouts of the holotype of
Tantilla ruficeps; to both we are very grateful.
Finally, we thank William E. Duellman, Carl S.
Lieb, Charles W. Myers, William F. Pyburn,
Douglas A. Rossman, and Philip A. Silverstone
for information regarding obscure localities.

Literature Cited
Achaval, F., and R. Vaz-Ferreira

1974. La presencia de Tantilla melano-
cephala melanocephala (L.) en la fauna
Uruguaya. Neotropica 20(62): 107-112.
Acosta-Solis, M.

1966. Las divisiones fitogeograficas y las for-
maciones geobotanicas del Ecuador.
Revista de la Academia Colombiana de
Ciencias Exactos Fisicas y Naturales
12(48): 401-447.
Amaral, A. do.

1929a Contribuigao ao conhecimento dos
ophidios do Brasil. IV. List remissiva
dos ophidios do Brasil. Memorias do
Institute Butantan (Sao Paulo) 4: i-iv,
71-125.
19296. Estudos sobre ophidios neotropicos.

XVII. Valor systematico de varias for-
mas de ophidios neotropicos.
Memorias do Institute Butantan (Sao
Paulo) 4: 3-68.

1929c. Estudos sobre ophidios neotropicos.

XVIII. Lista remissiva dos ophidios da
regiao neotropica. Memorias do Insti-
tute Butantan (Sao Paulo) 4: i-viii,
129-271.

1931. Studies of Neotropical Ophidia. XXVI.
Ophidia of Colombia. Bulletin of the
Antivenin Institute of America 4 (4):
89-94.
Andersson, L.G.

1899. Catalogue of Linnean type-specimens
of snakes in the Royal Museum in
Stockholm. Bihang Till K Svenska
vetenskapsakademien, (4)24(6): 1-35.
Barbour, T.

1914. A new snake from northern Brazil. Pro-
ceedings of the Biological Society of
Washington 27: 199-200.
Beebe, W.

1946. Field notes on the snakes of Kartabo,
British Guiana, and Caripito, Vene-

zuela. Zoologica 31: 11-52.
Bocourt, M.

1883. Mission scientifique au Mexique et
dans I'Amerique Centrale - Recherches
zoologiques. Part 3, Sect. 1. Etudes sur
les reptiles. Imprimerie Imp6riale,
Paris. Livr. 9: 529-592.
Boie, F.

1827. Bemerkungen iiber Merrem's Versuch
eines Systems der Amphibien. 1-te
Lieferung: Ophidier. I sis (von Oken)
20(3): 508-566.
Boulenger, G.A.

1896a. Catalogue of the snakes in the British
Museum (Natural History). Volume
III. Taylor and Francis, London. xiv+
727 pages.

18966. Descriptions of new reptiles and
batrachians from Colombia. Annals
and Magazine of Natural History,
series 6, 17(97): 16-21.
Brongersma, L.D.

1956. On some reptiles and amphibians from
Trinidad and Tobago. British West
Indies Proceedings of the Akademie
van Wetenshappen, Amsterdam 59C:
165-188.

Clark, D.R., Jr.

1970. Loss of the left oviduct in the colubrid
snake genus Tantilla. Herpetologica
26(1): 130-133.

Cope, E.D.

1861. [No title]. Proceedings of the Academy
of Natural Sciences of Philadelphia 13:
73-75.

1866. Fourth contribution to the herpetology
of tropical America. Proceedings of the
Academy of Natural Sciences of Phil-
adelphia 18: 123-132.

1868. An examination of the ReptiHa and
Batrachia obtained by the Orton Expe-
dition to Ecuador and the upper
Amazon, with notes on other species.
Proceedings of the Academy of
Natural Sciences of Philadelphia 20:
96-140.

1871. Ninth contribution to the herpetology
of tropical America. Proceedings of the
Academy of Natural Sciences of Phila-
delphia 23: 200-224.

1876a. On the Batrachia and Reptilia of Costa
Rica. Journal of the Academy of
Natural Sciences of Philaelphia 8(4):

40

93-154.

18766. Report on the reptiles brought by Pro-
fessor James Orton from the middle
and upper Amazon, and western Peru.
Journal of the Academy of Natural
Sciences of Philadelphia 8(2): 159-183.

1887a Catalogue of batrachians and reptiles
of Central America and Mexico. Bulle-
tin of the United States National
Museum 32: 1-98.

18876. Synopsis of the Batrachia and Reptilia
obtained by H.H. Smith, in the pro-
vince of Mato Grosso, Brazil. Proceed-
ings of the American Philosophical
Society 24: 44-60.

1894. Third addition to a knowledge of the
Batrachia and Reptilia of Costa Rica.
Proceedings of the Academy of
Natural Sciences of Philadelphia 46:
194-206.

1900. The crocodilians, lizards, and snakes of
North America. United States Nation-
al Museum Annual Report, 1898:
153-1270.
Cordeiro, C.L., and A.R. Hoge.

1973. Contribuicao ao conhecimento das
serpentes do Estado de Pernambuco.
Memorias do Institute Butantan (Sao
Paulo) 37: 261-290.
DarUngton, P.J., Jr.

1957. Zoogeography: The geographic
distribution of animals. John Wiley
and Sons, Inc., New York, USA. xi +
675 pages.
Daudin, P.M.

1801- Histoire naturelle, generate et partic-

1803. uliere des reptiles; ouvrage faisant
suite a I'histoire naturelle generale et
particuliere, composee par Leclerc de
Buffon, et redigee par C.S. Sonnini,
membre de plusiers societes savantes.
F. Dufart, Paris, 8 volumes.
Dixon, J.R., and P. Soini

1975. The reptiles of the upper Amazon
basin, Iquitos region, Peru. I. Lizards
and amphisbaenians. Milwaukee
Public Museum Contributions in
Biology and Geology 4: 1-58.

1977. The reptiles of the upper Amazon
basin, Iquitos region, Peru. II.
Crocodilians, turtles and snakes.
Milwaukee Public Museum Contribu-
tions in Biology and Geology 12: 1-91.

Dixon, J.R., and John W. Wright

1975. A review of the iguanid genus
Tropidurus in Peru. Natural History
Museum of Los Angeles County Con-
tributions in Science 271: 1-39.
Duellman, W.E.

1958. A monographic study of the colubrid
snake genus Leptodeira. Bulletin of the
American Museum of Natural History
114(1): 1-152.
1972. South American frogs of the Hyla
rostrata group (Amphibia, Anura,
HyUdae). Zoologische MededeUngen
47: 177-192.

, and M.L. Crump

1974. Speciation in frogs of the Hyla parui-
ceps group in the upper Amazon basin.
Occasional Papers of the Museum of
Natural History, University of Kansas
23: 1-40.
Dumeril, A.M.C., G. Bibron, and A.H.A.
Dumeril.

1854. Erpetologie generale ou histoire
naturelle complete des reptiles. Paris.
Volume 7.
Dunn, E.R.

1923. Some snakes from northwestern Peru.
Proceedings of the Biological Society
of Washington 36: 185-188.
1944. Los generos de anfibios y reptiles de
Colombia, III. Tercera parte: Reptiles;
Orden de las serpientes. Caldasia 3:
155-244.

, and J.R. Bailey

1939. Snakes from the uplands of the Canal
Zone and of Darien. Bulletin of the
Museum of Comparative Zoology
86(1): 1-22.
Fitzinger, L.J. F.J.

1826. Neue Classification der Reptilien nach
ihren naturlichen Verwandtschaften
nebst einer Werwandtschafts-Tafel
und einem Verzeichnisse der Reptilien-
Sammlung des k.k. zoologischen
Museums zu Wein. Huebner, Wein. viii
+ 66 pages.
1843. Systema reptilium. Fasciculus primus.
Amblyglossae. Vindobonae, Brau-
muller und Beidel. 106 + x pages.
Gmelin, J.F.

1789. Caroli a Linne, . . . Systema Naturae
per regna tria natural, secundum
classes, ordines, genera, species, cum

41

characleribus differentiis, synonymis,
locis. Edition 13. Volume 1(3): 1038-
1516. Leipzig.

Griffen, L.E.

1916. A catalog of the Ophidia from South
America at present (June, 1916) con-
tained in the Carnegie Museum with
descriptions of some new species.
Memoirs of the Carnegie Museum 7(3):
163-228.

Gunther, A.

1862. On new species of snakes in the collec-
tion of the British Museum. Annals
and Magazine of Natural History,
series 3, 9: 52-59.
1895. Biologia Centrali-Americana. Reptilia
and Batrachia. Porter, London. Parts
19-25: 145-196.

Haffer, J.

1969. Speciation in Amazonian forest birds.
Science 165(3889): 131-137.

1970. Geologic-cUmatic history and zoogeo-
graphic significance of the Uraba
region in northwestern Colombia.
Caldasia 10: 603-636.

Hardy, L.M., and C.J. Cole.

1967. The colubrid snake Tantilla armillata
Cope in Nicaragua. Journal of the
Arizona Academy of Sciences 4(3):
194-196.
Hartweg, N.

1944. Remarks on some Mexican snakes of
the genus Tantilla. Occasional Papers
of the Museum of Zoology, University
of Michigan 486: 1-9.
Heyer, W.R.

1973. Systematics of the marmoratus group
of the frog genus Leptodactylus
(Amphibia, Leptodactylidae). Natural
History Museum of Los Angeles Coun-
ty Contributions in Science 251: 1-50.
Hoge, A.R."

1952. Notas erpetologicas. la. Contribuicao
ao conhecimento dos ofidios do Brasil
Central. Memorias do Instituto Butan-
tan (Sao Paulo) 24(2): 179-214.
Holdridge, L.R.

1964. Life zone ecology. Tropical Science
Center, San Jose, Costa Rica.
James, P.E.

1959. Latin America. Odyssey Press, Inc.
New York, New York, USA. xviii +
942 pages.

Jan, G., and F. SordelU

1866. Iconographie generale des ophidiens.
Atlas. Milan and Paris. Volume 1, livr.
15.
Linnaeus, C.

1758. Systema naturae per regna tria
naturae, secundum classes, ordines,
genera, species cum characteribus, dif-
ferentiis, synonymis, locis. Tenth edi-
tion. Volume 1. L. Salvius, Stockholm,
iv -f 826 pages.
1766. Systema naturae per regna tria
naturae, secundum classes, ordines,
genera, species cum characteribus, dif-
ferentiis, synonymis, locis. Twelfth edi-
tion, Volum 1. L. Salvius, Stockholm.
532 pages.
Merrem, B.

1820. Versuch eines Systems der Amphibien.
Tentamen systematis amphibiorum.
Marburg, xv + 191 pages.
Mertens, R.

1952a. Die Amphibien und Reptilien von El
Salvador, auf Grund der Reisen von R.
Mertens und A. Zilch. Abhandlungen
der Senckenbergischen Naturfor-
schenden Gesellschaft 487: 1-120.
19526. Neues uber die Reptilienfauna von El
Salvador. Zoologischer Anzeiger
148 (3/4): 87-94.
1972. Herpetofauna tobagana. Stuttgarter
Beitrage zur Naturkunde 252: 1-22.
Meyer, Jr.

1969. A biogeographic study of the amphi-
bians and reptiles of Honduras. Doc-
toral thesis. University of Southern
California, Los Angeles, California,
USA. 589 pages.
Myers, C.W.

1974. The systematics of Rhadinaea (Col-
ubridae), a genus of New World snakes.
Bulletin of the American Museum of
Natural History 153(1): 1-262.

, and A.S. Rand.

1969. CheckUst of amphibians and reptiles of
Barro Colorado Island, Panama, with
comments on faunal change and sampl-
ing. Smithsonian Contributions to
Zoology 10: Ml.
Parker, H.W.

1938. The vertical distribution of some rep-
tiles and amphibians in southern
Ecuador. Annals and Magazine of

42

Natural History, series 11, 2: 438-450.
Peters, J. A.

1960. The snakes of Ecuador. A checklist and
key. Bulletin of the Museum of Com-
parative Zoology 122(9): 491-541.
, and B.R. Orejas-Miranda.

1970. Catalogue of the Neotropical
Squamata: Part I. Snakes. Smithso-
nian Institution Press, Washington,
District of Columbia, USA. ill + 346
pages.
Rand, A.S.

1957. Notes on amphibians and reptiles from
El Salvador. Fieldiana Zoology 34(42):
505-534.
Roze, J. A.

1966. La taxonomia y zoogeografia de los
ofidios de Venezuela. Servicio de
Distribucion de Publicaciones,
Biblioteca, Universidad Central de
Venezuela, Caracas. 362 pages.
Ruthven, A.G.

1922. The amphibians and reptiles of the
Sierra Nevada de Santa Marta, Colom-
bia. Miscellaneous Publications
Museum of Zoology, University of
Michigan 8: 1-69.
Savage, J.M.

1966. The origins and history of the Central
American herpetofauna. Copeia
1966(4): 719-766.

1975. Systematics and distribution of the
Mexican and Central American stream
frogs related to Eleutherodactylus
rugulosus. Copeia 1975(2): 254-306.
Schlegel, H.

1837. Essai sur la physionomie des serpens.
Kips and Van Stuckum, La Hague.
Partie generale, 28, 251, 314 pages;
partie descriptive, 606 pages.
Schmidt, K.P., and R.F. Inger

1951. Amphibians and reptiles of the
Hopkins-Branner Expedition to Brazil.
Fieldiana Zoology 31(42): 439-465.
, and W.F. Walker, Jr.

1943a. Peruvian snakes from the University
of Arequipa. Field Museum of Natural
History Publications, Zoology series
24(26): 279-296.

19436. Snakes of the Peruvian coastal region.
Field Museum of Natural History Pub-
lications, Zoology series 24(27):
297-324.

Schwabe, G.H.

1969. Towards an ecological characterisation
of the South American continent. In,
Fittkau, E.J., J. lilies, H. Klinge, G.H.
Schwabe, and H. SioU. Biogeography
and ecology in South America. Volume
1. Dr. W. Junk N.V. PubHshers, The
Hague, xvi + 445 pages.

Shaw, G.

1802. General zoology or systematic natural
history. Thomas Davidson, London.
Volume 3, part 2: i-vi, 313-615.

Shreve, Benjamin

1947a. On Colombian reptiles and amphibians
collected by Dr. R.E. Schultes. Cal-
dasia 4(19): 311-316.
19476. On Venezuelan reptiles and amphi-
bians collected by Dr. H.G. Kugler.
Bulletin of the Museum of Compara-
tive Zoology 99(5): 519-537.

Smith, H.M.

1941. A new genus of Central American
snakes related to Tantilla. Journal of
the Washington Academy of Sciences
31(3): 115-117.

1942. A resume of Mexican snakes of the
genus Tantilla. Zoologica 27: 33-42.

1958. Handlist of the snakes of Panama.
Herpetologica 14: 222-224.
, and C. Grant

1958. New and noteworthy snakes from
Panama. Herpetologica 14: 207-215.
Smith, H.M., and E.H. Taylor

1945. An annotated checklist and key to the
snakes of Mexico. Bulletin of the
United States National Museum 187:
i-iv, 1-239.
Stejneger, L.

1902. The reptiles of the Huachuca Moun-
tains, Arizona. Proceedings of the
United States National Museum 25:
149-158.
Stuart, L.C.

1963. A checklist of the herpetofauna of
Guatemala. Miscellaneous Publica-
tions Museum of Zoology, University
of Michigan 122: 1-150.
Taylor, E.H.

1951. A brief review of the snakes of Costa
Rica. University of Kansas Science
Bulletin 34(1): 3-188.

1954. Further studies on the serpents of
Costa Rica. University of Kansas

43

Science Bulletin 36(11): 673-801.

Underwood, F.

1962. Reptiles of the Eastern Caribbean.
Caribbean Affairs, New Series 1: 1-192.

Uzzell, T.

1973. A revision of the lizards of the genus
Prionodactylus, with a new genus for
P. leucostictus and notes on the genus
Euspondylus (Sauria, Teiidae). Postilla
159: 1-67.

van der Hammen, T.

1961. The Quaternary climatic changes of
northern South America. Annals of the
New York Academy of Sciences 95:
676-683.

Vanzolini, P.E., and E.E. Williams

1970. South American anoles: The
geographic differentiation and evolu-
tion of the Anolis chrysolepis species
group (Sauria, Iguanidae). Arquivos de
Zoologia 19(3-4): 1-298.

Verteuil, J. P. de

1968. Notes on the snakes and lizards of
Tobago. In Alford. The island of
Tobago. Seventh edition. Hampstead,
London.

Vuilleumier, B.S.

1971. Pleistocene changes in the fauna and
flora of South America. Science

173(3999): 771-780.
Wagler, J.G.

1824. Serpentium Brasiliensium species
novae, ou histoire naturelle des especes
nouvelles de serpens. . . Jean de Spix,
Monaco, viii + 75 + (1) pages.
1830. Naturliches System der Amphibien,
mit voranghender classification der
Saugethiere und Vogel. Ein Beitrag
zur vergleichender Zoologie. Munich,
vi + 354 pages.
Wehekind, L.

1955. Notes on the foods of the Trinidad
snakes. British Journal of Herpetology
2(1): 9-13.
Werner, F.

1909. Ueber neue oder seltene Reptilien des
naturhistorischen Museums in Ham-
burg. I. Schlangen. Mitteilungen der
Naturhistorisches Museum in Ham-
burg 26(2): 205-247.
Wilson, L.D., and J.R. Meyer

1971. A revision of the taeniata group of the
colubrid snake genus Tantilla.
Herpetologica 27(1): 11-40.
Division of Intercurricular Studies, Miami-
Dade Community College, South Campus, Miami,
Florida 331 76 and 1561 S. W. 12th Avenue, Miami,
Florida 33129.

44

Fig- 1- Lateral, dorsal, and ventral views of head of a member of the pattern
A group of Tantilla melanocephala (KU 102529, Finca Taboga, =20
km SE Las Canas, Prov. Guanacaste, Costa Rica). All figures are
drawn on a generalized scale pattern.

45

Fig. 2. Lateral, dorsal, and ventral views of head of a member of the pattern
B group of Tantilla melanocephala (KU 31960. 16 km NW San Isidro
el General, Prov. San Jose, Costa Rica).

46

Fig. 3. Lateral, dorsal, and ventral views of head of a member of the pattern
D group of Tantilla melanocephala (MBUCV III-1825, Guri, Est.
Bolivar, Venezuela).

47

Fig. 4. Lateral, dorsal, and ventral views of head of a member of the pattern
E group of Tantilla melanocephala (TCWC 38209. Centro Union,
Depto. Loreto, Peru).

46

ERRATUM

The drawings, but not the captions, of Figs. 3 and 4
were inadvertently exchanged. That is, the drawing on
page 46 is explained by the caption on page 47, and
the drawing on page 47 is explained by the caption on
page 46.

Systematics of the melanocephala group of the colubrid
snake genus Tantil la.

Wilson, Larry David, and Cesar E. Mena.

San Diego Society of Natural History Memoir 11. 1980.

Fig. 3. Lateral, dorsal, and ventral views of head of a member of the pattern
D group of Tantilla melanocephala (MBUCV III-1825, Guri, Est.
Bolivar, Venezuela).

47

Fig. 4. Lateral, dorsal, and ventral views of head of a member of the pattern
E group of Tantilla melanocephala (TCWC 38209, Centro Union,
Depto. Lore to, Peru).

48

Fig. 5. Lateral, dorsal, and ventral views of head of a member of the pattern
F group of Tantilla melanocephala (NMW 20366:1, Santo Domingo
de los Colorados, Prov. Pichincha, Ecuador).

49

Fig. 6. Dorsal color patterns of members of five pattern groups of Tantilla
melanocephala: (A) pattern A (KU 102529); (B) pattern B (KU 31960);
(C) pattern D (MBUCV 111-1825); (D) pattern E (TCWC 38209); (E)
pattern F (NMW 20366:1). All locality data same as that for head
figures.

50

Fig. 7. Lateral, dorsal, and ventral views of head of Tantilla andinista (KU
135209, 5 km E Alausi, Prov. Chimborazo, Ecuador).

51

Fig. 8. Lateral, dorsal, and ventral views of head of Tantilla capistrata
(FMNH 34287, Chiclin, Depto. La Libertad, Peru).

52

Fig. 9. Lateral, dorsal, and ventral views of head of Tantilla equatoriana
(USNM 198530, San Lorenzo, Prov. Esmeraldas, Ecuador).

53

Fig. 10. Lateral, dorsal, and ventral views of head of Tantilla insulamontana
(KU 152207, Rio Minas, 15.1 km W Santa Isabel, Prov. Azuay,
Ecuador).

54

Fig. 11. Lateral, dorsal, and ventral views of head of Tantilla lempira
(LSUMZ 26093, 41 km NW Tegucigalpa, Depto. Francisco Moraz^n,
Honduras).

55

Fig. 12. Dorsal color patterns of 5 members of the Tantilla melanocephala
group: (A) T. andinista (KU 135209); (B) T. capistrata (FMNH
34287); (C) T. equatoriana (USNM 198530); (D) T. insulamontana
(KU 152207); (E) T. lempira (LSUMZ 26093). All locality data same
as that for head figures.

56

▲ T. equatoriana
• T. andinista
■ T. insulamontana

80"

78"

Fig. 13. Distribution of 3 members of the Tantilla melanocephala group.

57

Fig. 14. Distribution of 4 pattern groups of Tantilla melanocephala (patterns
C and D— triangles; pattern E— black squares; pattern F— open
squares), T. melanocephala intermediates (half-open circles), and T.
capistrata (circles with stars).

58

90'

85"

80^

kilometers

90^

85*

80^

Fig. 15. Distribution of 2 pattern groups of Tantilla melanocephala (pattern
A— black circles; pattern B— open circles), T. melanocephala inter-
mediates (half-open circles), and T. lempira (triangles).

Date Due

ACME

BOOKPlfiDir^G CO., INC.

Nnv; 28 1984

100 CAMBRIDGE J^iRECT
Ct*,RLESTOWN, MASS.