S-KJfl-S 
iVlUS. COM p. ZO^-'L. 
LIBRARY 
JUN 2 1 10RR 
HAKVmimJ 
UNIVERSITY 
TRANSACTIONS 
OF THE 
SAN DIEGO SOCIETY OF NATURAL HISTORY 
Volume 14, No. 11, pp. 137-156 
BIOGEOGRAPHY AND DISTRIBUTION OF 
THE REPTILES AND AMPHIBIANS ON ISLANDS 
IN THE GULF OF CALIFORNIA, MEXICO 
BY 
MICHAEL SOULE 
Stanford University, Stanford, California 
(Present address: University of Malawi, Limbe, Malawi) 
AND 
ALLAN J. SLOAN 
Assistant Curator of Reptiles and Amphibians 
San Die°o Natural History Museum 
SAN DIEGO, CALIFORNIA 
Printed for the Society 
June 10, 1966 
J. EL MUERTO 
2. ENCANTADA GRANDE 
3. MEJIA 
4. ANGEL DE LA GUARDA 
5. SMITH 
6. TIBURON 
7. PARTIDA NORTE 
8. RAZA 
9. SALSIPUEDES 
10. SAN LORENZO 
LORENZO 
ESTEBAN 
PEDRO Ma'rTIR 
PEDRO NOLASCO 
5. TORTUGA 
16, SAN MARCOS 
II. SAN 
-j 12. SAN 
I 3. SAN 
14- SAN 
Fig. 1. Islands in the Gulf of California referred to in this report. 
JUN 2 1 1966 
C, . ^ Z HARVARD 
UNIVERSITY 
BIOGEOGRAPHY AND DISTRIBUTION OF 
THE REPTILES AND AMPHIBIANS ON ISLANDS 
IN THE GULF OF CALIFORNIA, MEXICO 
BY 
MICHAEL SOULE and ALLAN J. SLOAN 
Introduction 
The first comprehensive analysis of the herpetofauna of the islands in the Gulf of Cali- 
fornia was that by Schmidt (1922); in his report of the results of the 1911 "Albatross" 
expedition, 54 records of island reptiles are given. In 1921 most of the islands were visited by 
the California Academy of Sciences' "Silvergate" expedition under the leadership of Joseph 
Slevin. Van Denburgh's (1922) tabulation of 127 records from the Gulf islands includes the 
material collected by the "Silvergate" personnel. CliflF (1954a) studied the Gulf island herpeto- 
fauna and published on the snakes (1954b). At that time the total known reptile fauna 
amounted to some 200 insular populations. Since then about 50 new records have accumulated. 
Responsible for the acquisition of much of this new material was the Belvedere expedition to 
the Gulf in the spring of 1962. Lindsay (1962) provided an account of the Belvedere expedi- 
tion in which he mentioned many of the new records which are formally presented here. 
Many persons and institutions have made important contributions to this project. Charles 
E. Shaw, of the Zoological Society of San Diego, and the senior author were responsible for 
the collection of reptiles obtained on the Belvedere expedition, although all the biologists 
contributed. Permits to collect reptiles in the Republic of Mexico were granted by the late Luis 
Macias Arellano, and Rodolfo Hernandez Corzo, Direccion General de Caza, Departamento 
de Conservacion de la Fauna Silvestre. Field work has been supported by grants from the 
Belvedere Scientific Fund and the National Science Foundation (G-14426, GB-2317), and 
donations from Richard M. Adcock, Richard F. Dwyer, and Roy E. Marquardt. 
Basis for Taxonomtc Decisions. — Many of the Gulf populations are known by two or 
more names. The majority of these taxonomic uncertainties stem from differences of opinion 
as to whether a population deserves specific or subspecific recognition. Where decisions as to 
specific or subspecific rank of insular populations have been necessary, we have in general fol- 
lowed the suggestion ot Mayr (1942:121): "If we examine the 'good' species of a certain 
locality we find that the reproductive gap is associated with a certain degree of morphological 
difference. If we find a new group of individuals at a different locality, we use the scale of 
diflferences between the species of the familiar area to help us in determining whether the new 
form is a different species or not. These scales of differences are empirically reached and differ 
in every family and genus." 
Comments on Distribution Table. — The distribution of reptiles and amphibians on islands 
in the Gulf of California is shown in table 1. The table includes 15 new records reported in 
this paper (see below). The islands are arranged approximately from north (left side of table) 
to south. Several very small islands close to the mainland on either side of the Gulf, as those 
in Bahias de los Angeles, Concepcion, and Kino, and near Mazatlan, are not listed. Several 
very small satellite islands from which only one or two lizards are known are also omitted. 
References for the table are: Banks and Farmer, 1963; Banta and Leviton, 1963; Cliff, 1954a, 
1954b, 1958; Crippen, 1962; Dickerson, 1919; Dixon, 1964; Etheridge, 1961; Figg-Hoblyn and 
Banta, 1957; Klauber, 1956, 1963; Lowe and Norris, 1955; Savage and Cliff, 1954; Schmidt, 
1922; Shaw, 1945; Smith and Taylor, 1945, 1950; Soule, 1961 and in preparation; Van Den- 
burgh, 1922; Zweifel and Norris, 1955. The location of the islands is shown in figure 1. 
New Records of Snakes 
This section contains data on a number of snakes which appear to represent new Gulf 
island records. Except as noted, the specimens referred to are in the collection of the San 
Diego Society of Natural History (SDSNH). 
140 
San Diego Society of Natural History 
[Vol. 14 
J 
U 
CS 
c 
-^ 
O 
•0 ^ 
o B 
IS u 
U G 
•-t-i -rt 
O C 
3 
4) 
'C 
.S CO 
-13 
C 
< 
-T3 
C 
.2 
15 
n 
<u 
__ C 
6 c 
rt rt 
-T3 .^ 
S 2 
_ o 
3 
Q 
u 
'B 
u 
c 
m 
.Si 
a, 
»> 
3 
-a 
c 
cd 
s 
II 
u 
1966} 
SouLE AND Sloan: Herpetofauna of Gulf Islands 
141 
142 
San Diego Society of Natural History 
[Vol. 14 
OA1BJJ33 £9[ isL Z 
Ku^ipg ^•> 69 — 
BpiwBd-oauBS niuidsg 55 9^6 9 
03SIDUEJJ uBg 9-3 013 Z 
3sof UBS t'61 ff9 L 
U 
u 
(/) u 
w 
u 
u 
u 
u 
SBlUIUy SB^ 
oSaiQ uBg 
zru3 BauBg 
bui|bjb;]) BauBg 
8JEJJaSUOJ/\J 
9'n 
t'"6I 
ozz 
£33 
U 
u 
ajuBzuBQ 6> 901 
SOPEUOJ03 ^8 £83 
OSUOppiI UEg 9-3 ^n 
(jng) sauj BjuES ^- > 6 
SOJJBI^ UBS ^"I^ \Ll 
U 
u u 
u 
u 
u 
u u 
u 
3 
.2 
c 
o 
U 
< 
BSniJOj^ £-9 
ODSB|0|^ ojpaj UBS Zi 
JUJBi^ ojpad UBS 8' I 
uB(ja:sg ues ^'^ 
jns ozusjo'^j UBS ^^^ 
60 f 
93f 
03f 
U 
3»0N OZU3J01 UBS S'8 003 
sapandisps 8"1 t'll 
BZBy I'l [f 
ajaof^ EpijjBd 1-3 331 
"o-inqij. 31 31 81 31 
31 
U 
U 
Qi 
u 
u 
u 
ijjiuis 
upjEHQ B[ ap jaSuy 
Biraj^ 
apUBjr) BpBJUBDUg 
oajanj^ |g 
3 £9 
^•£ 
60 
^9t' 
91£I 
393 
£33 
361 
a: 
u 
u 
V 
2 
3 
ST 
s 
V 
3 
g 
X 
c 
ra 
■c 
a 
CO 
in 
o 
■-CJ 
2 
bo 
bo bo ,bo ^^ bo 
bo 
bo 
O o 
o o 
o o 
5 
o 
o 
s s 
C 
O >o ►o ►J 
bo 
"3 
w bo 
.5 -^ ~c 
S S 
:^ :^ '^ 
O O uj 
"-^ tj Q 
>- ? != 
u 
2 3 3 
T3 "3 T3 
~C ~C -J5 
Vj U fj 
c c c 
?s ^ ?s 
-c ~c -c 
■* -c -c 
>>> <J (J 
000 
.2 c 
bo bo 
o 
■Q. «». -a. 
^ ^ ?N 
■ci. 
o 
A. 
5 
a: i: 1: ^ ^ 
1966] 
SouLE AND Sloan: Herpetofauna of Gulf Islands 
143 
3 i: 
5-c Si 
Ji §H 
a in C 
C b O 
-* o '-» 
cy_§ ° 
■^ < o 
« •< 
c o- q7 
to 
O O o 
•™ gi 
(J ^ ^ — 
tX M O 
ig > — ja 
o •« 3 
'-« -D O 
_ o a. 
e _2 
i^ o E 
as 
3 
u u 
O X nj 
J2 u 
<! O ■^ 
u 
o 
jaco « 
C _ 0) 
■-^ *T-i £^ 
-15 c-H 
o 
n, « 
a 
c 
ti O 
< g Q 
-^ 3 
^ H 
CQ 
O 
o « 
U- JO u 
c 
nj 
u 
o c 
o 
U i-i U 
o 
u 
o « 
o — : Q- 
'•? 
to 
.•^ 
^3.° 
- u o 
-d o 
144 San Diego Society of Natural History [Vol. 14 
Leptotyphlops humilis slevini Klauber. The single blind snake collected during the 
Belevedere expedition constitutes an addition to the herpetofauna of Isla Carmen. A female 
(SDSNH 44386) was found near Bahia Marquer by Charles E. Shaw and George E. Lind- 
say, April 4, 1962. The total length is 221 mm., the tail length, 8 mm. There are 14 scale 
rows at midbody, 14 subcaudals, 240 scales between the rostral and tail spine, and 12 scale 
rows around the tail. The first four median dorsal scales are nearly equal in size. The seven 
dorsal scale rows are medium brown. 
This specimen is assigned to the subspecies slev-.ni until additional specimens are available 
to determine the variation in this island population. Most characters are within the range of 
variation ascribed to this subspecies by Klauber (1940a: 133). The number of dorsal scales is 
lower, and is very close to the mean number for dugesi. This specimen is darker than typical 
slevini, although not nearly as dark as coastal himiilis. 
The only previous report of worm snakes on Gulf islands is that by Banks and Farmer 
(1963) for Isla Cerralvo. 
Lichanura trivirgata gracia Klauber. An adult female (SDSNH 51999) was collected 
by Howard E. Everette at Puerto Refugio, on the north end of Isla Angel de la Guarda, 
March 19, 1963. It was found about two hours after sundown. 
This specimen has a total length of 724 mm., and a tail length of 90 mm. There are 235 
ventrals, 49 subcaudals, and 41 scale rows at midbody. The three longitudinal stripes are 
orange-brown, and adhere closely to the edges of the limiting scales. Unlike typical specimens 
of this subspecies, which have drab ground color, the dorsal ground color is light to medium 
gray, and the stripes tend to be outlined in black. This black outline is particularly prominent 
on the lower edge of each lateral stripe, and is composed of a concentration of minute black 
spots on the inner edges of adjacent scales. This specimen is similar in coloration to one from 
six miles south of Socorro, in northern Baja California (considered a possible intergrade by 
Klauber, 1931a:312), and to recently collected material from the peninsula opposite the island. 
The ovary of this snake contains eggs of three distinct size classes. Five eggs average 
6.8 X 7.2 mm., six average 2.5 X 3.2 mm., and eight others are less than 1 mm. in greatest 
diameter. 
The desert rosy boa was expected to occur on Isla Angel de la Guarda since a dried 
example was collected on nearby Isla Mejia by Joseph R. Slevin in 1921 (Van Denburgh, 
1922:633). 
Lichanura trivirgata trivirgata Cope. Two specimens of this subspecies were collected 
during the Belvedere expedition, and represent the first insular records of this subspecies. 
An adult male (SDSNH 52898) was secured in Arroyo Sausal, on the southwest side 
of Isla Tiburon, March 20, 1962, by Charles F. Harbison. This snake is 678 mm. long, in- 
cluding the 84 mm. tail. There are 222 ventrals, 45 subcaudals, and 37 scale rows at midbody. 
The second specimen is also an adult male (SDSNH 44389) , collected in Arroyo de los 
Chivos, on the northwest side of Isla San Marcos, March 29, 1962, by Richard C. Banks. The 
total length is 627 mm., the tail length, 79 mm. There are 217 ventrals, 41 subcaudals, and 
39 scale rows at midbody. 
Both specimens have the typical coloration and contrasting pattern (dark brown stripes 
on a cream colored background) of southern Baja California material; however, the San 
Marcos snake has exceptionally wide stripes. 
Lampropeltis getulus calijorniae (Blainville) . The California king snake is tentatively 
added to the faunas of three islands in the Gulf of California, Islas Angel de la Guarda, 
Salsipuedes, and San Lorenzo Norte, on the basis of one shed skin from each. The first 
(SDSNH 19989) was collected in Arroyo Estaton, on the west side of Isla Angel de la 
Guarda, by Chris Parrish, June 6, 1961. This shed skin is nearly complete, with only a few 
of the head scales and the tip of the tail missing. It has 248 ventrals, 61+ subcaudals, and 
23 scale rows at midbody. The skin is 1480 mm. long. 
The second shed skin (SDSNH 45150) was found by Soule on Isla San Lorenzo Norte, 
March 24, 1962, and the third (SDSNH 45003) by Richard C. Banks on Isla Salsipuedes 
1966] SouLE AND Sloan: Herpetofauna of Gulf Islands 145 
October 23, 1964. Both are incomplete, and the identifications are questionable. However, the 
characters of smooth scales in 23 rows, paired scale pits, undivided anal, divided subcaudals, 
and an alternating dark and light pattern eliminated other kinds of snakes considered possible 
by us. No snakes have previously been reported from these two islands. 
Two living specimens have recently been collected on Isla Monserrate. An immature male 
(SDSNH 44631) was found in sand dunes on the northeast side of the island by Sloan on 
the evening of April 28, 1964. The total length is 482 mm., and the tail length is 67 mm. It 
has 235 ventrals, 54 subcaudals, 25 scale rows at midbody, 7-7 supralabials, and 10-11 infrala- 
bials. There are 38 light rings on the body and eight on the tail. An immature female 
(SDSNH 44632) was collected at the north end of the island by Richard C. Banks on 
June 24, 1964. This snake has a total length of 452 mm , and tail length of 56 mm. It has 
243 ventrals, 49 subcaudals, 25 scale rows at midbody, 7-7 supralabials, and 11-10 infralabials. 
There are 37 light rings on the body and eight on the tail. Both snakes have light rings 1-2 
scales long which are separated by 4-6 dark scales. These specimens are very similar to those 
found on the desert slopes of San Diego County, California. 
Phyllorhynchus deciirtatus decurtatus (Cope). An immature male (SDSNH 44682) 
was collected in Arroyo de Aquada, on the northeast side of Isla San Jose, by George E. 
Lindsay on June 26, 1964. The total length of this snake is 230 mm., and the tail length is 
30 mm. There are 153 ventrals, 33 subcaudals, and 19 scale rows at midbody. There are 23 
dark blotches on the body and five on the tail. The blotches near the head are rectangular, 
those near the anus are square. Most of them are rounded diamonds, 3-4 scales (end to end) 
long, occupy 6-7 scale rows, and are separated by three scales (end to end). 
Except for the low number of ventrals and the shape of the blotches, this specimen agrees 
with the description of P. d. decurtatus (as redefined by Klauber, 1940b: 206), but it keys out 
to P. d. norrisi (Savage and Cliff, 1954:74). This snake is abnormal in having the first in- 
fralabial on the right side split transversely, and the frontal incompletely separated from the 
single scale occupying the prefontal area. This is the first insular record for this subspecies. 
Phyllorhynchus decurtatus perkmsi Klauber. An immature female (SDSNH 52000) 
was collected on a sand dune at the northeast end of Isla Angel de la Guarda by Sloan, 
March 20, 1963. It was found about one hour after sundown among the exposed roots of 
A triplex poly car pa. This is the first island record for this subspecies. 
The total length of this snake is 263 mm., the tail length, 24 mm. It has 183 ventrals, 
29 subcaudals, and 19 scale rows at midbody. There are 43 dark blotches on the body and 
four on the tail. The blotches near midbody are two scales (end to end) long and are 
separated by 2-3 scales (end to end) . The length of the interspaces is less than in typical 
specimens of this subspecies (Klauber, 1935), and in this respect this snake is similar to a 
specimen of P. d. decurtatus from central Baja California reported by Murray (1955:45). The 
high ventral and blotch counts, however, indicate a close relationship to populations in north- 
eastern Baja California and southeastern California rather than to those in southern Baja 
California. 
Pituophis melanoleucus affinis Hallowell. An adult female (SDSNH 44387) was col- 
lected by Soule at Bahia Aqua Dulce, on the north end of Isla Tiburon, March 18, 1962. 
This snake is 1555 mm. long, including the 175 mm. tail. There are 240 ventrals, 61 sub- 
caudals, and 33 scale rows at midbody. There are 46 dark body blotches, and 11 tail blotches. 
The supralabials are 9-9 (the fifth enters the orbit on each side), and the infralabials are 
14-14; there are two pre- and four postoculars on each side. 
In meristic characters, and in general appearance, this specimen is readily assigned to the 
subspecies affinis, which occurs on the adjacent mainland of Sonora, Mexico. Some peculiarities, 
however, seem worthy of note. The anterior nasal on each side is divided transversely, and 
there are creases on several head scales including the frontal, rostral, some infralabials and 
genials. A single shortened, irregularly shaped scale occupies the parietal region. There are 
three loreals on each side; the lower one is very small and does not contact the nasal or preocular. 
There are two series of scales in the prefrontal area, an anterior row of four followed pos- 
teriorly by a row of three. 
146 San Diego Society of Natural History [Vol. 14 
This appears to be the first definite Gulf record of a gopher snake; it is surprising that 
such a large diurnal snake could have escaped detection for so many years, but perhaps it is 
a recent immigrant. 
Rhinocheilns lecontei lecontei Baird and Girard. An adult male was collected by Richard 
C. Banks and Soule near the former Ruffo Ranch, on the southwest side of Isla Cerralvo, 
October 29, 1961. EXie to misidentification, this specimen was reported as Lampropeltis getuliis 
conjuncta (Banks and Farmer, 1963). The genus Rhinocheilus has not previously been re- 
ported from any Gulf island. Cerralvo is approximately 540 miles southwest of Cape Colnett, 
the southernmost record for the genus in Baja California considered reliable by Klauber 
(1941:300). 
The total length of this snake is 1100 mm., including the 146 mm. tail. There are 210 
ventrals, 53 subcaudals (the first 35 are entire), and 23 scale rows at midbody. There are 
25 dark blotches on the body, and nine on the tail. 
This specimen (Calif. Acad. Sci. 98095) is reported through the courtesy of A. E. Leviton. 
Tantilla planiceps (Blainville) . One of the more interesting discoveries made during 
the Belvedere expedition was the presence of a black-headed snake on Isla Carmen. The single 
female (SDSNH 44388) was collected near Bahia Marquer by Charles E. Shaw and George 
E. Lindsay, April 4, 1962. It is 284 mm. long including the 66 mm. tail. There are 179 ven- 
trals, 65 subcaudals, and 15 scale rows at midbody. In preservative, the specimen is uniform 
brown posterior to the head, all scale rows being pigmented. The black head cap extends down 
approximately to the angle of the mouth. The light nuchal band is one scale wide, and is 
bordered behind by four small dark brown spots. 
Crotalus atrox Baird and Girard. The discovery of two specimens on Isla Santa Cruz is 
a new insular record for this species. They were collected on the west side of this island by 
Reid Moran and Chris Parrish on April 18, 1962. 
An adult male (SDSNH 44350) has a total length of 909 mm., a tail length of 70 mm., 
and a head length of 41 mm. It has 183 ventrals, 22 subcaudals, 25 scale rows at midbody, 
16-15 supralabials, 16-16 infralabials, 39 body blotches, and four tail blotches. An adult female 
(SDSNH 44351) has a total length of 737 mm., a tail length of 48 mm., and a head length 
of 34 mm. It has 181 ventrals, 17 subcaudals, 25 scale rows at midbody, 15-15 supralabials, 
18-18 infralabials, 31 body blotches and four tail blotches. Both have 12 scales bordering the 
proximal rattle segment, and 15 scale rows at the middle of the tail. 
The male was heavily infested with the parasitic linguatulid Porocephalus crotali. These 
were present nearly the entire length of the respiratory tract, with one actually occluding the 
glottis. The majority, including females heavily laden with eggs, were found in the anterior 
portion of the lung, which was distended and nearly filled the body cavity. Only a few para- 
sites were found in the posterior portion of the lung; these included an attached individual 
opposite ventral 104. 
Cliff (1954b: 80) reported the first specimen of Crotalus atrox from Isla San Pedro 
Martir. Members of the Belvedere expedition collected six additional specimens on the east 
side of this island on March 21, 1962. Measurements and scale counts are summarized in 
table 2. Cliff mentioned the possibility of this being a dwarf population on the basis of the 
single individual. It is true that these additional specimens also have relatively shorter heads 
than mainland specimens, but in the absence of other differences the recognition of an insular 
subspecies is not warranted. 
Crotalus catalinensis Cliff. Two specimens of this recently described species (Qiff, 
1954b: 80) were collected on Isla Santa Catalma by members of the Belvedere expedition. 
During 1964, 10 additional specimens were secured by various individuals. Data now being 
accumulated on this interesting rattlesnake will be published separately. 
Crotalus enyo enyo (Cope). An immature male (SDSNH 44355) was collected at 
Los Ostiones, on the northwest side of Isla San Jose by Richard C. Banks and William K. 
Emerson, April 12, 1962. This subspecies had been known previously from Islas San Fran- 
cisco, Carmen, and Espiritu Santo (including Partida portion) in the Gulf of California 
(Klauber, 1956:122). 
1966} 
SouLE AND Sloan: Herpetofauna of Gulf Islands 
147 
Table 2 
Measurements (in mm.) and scale counts for Crotalus atrox from Isia San Pedro Martir 
SDSNH Catalog 
The total length of this snake is 659 mm.; the tail length is 69 mm. There are 174 ven- 
trals, 27 subcaudals, 25 scale rows at midbody, and 13 scale rows at the middle of the tail. 
There are 13-13 supralabials, and 14-14 infralabials. The proximal rattle segment is black, and 
is bordered by 10 scales. There are 36 body blotches and six tail blotches. 
This specimen seems to fall within the range of variation of the subspecies enyo as de- 
fined by Klauber (193 lb: 363) in meristic characters and coloration, but it lacks preciliary 
scales and the lower loreal is slightly larger than the upper on both sides. The presence of 
of this species was expected on Isla San Jose since it has been known to occur on nearby Isla 
San Francisco for many years. 
Crotalus t7iitchelli mitchelli (Cope). An adult male (SDSNH 45002) was collected 
at the north end of Isla Monserrate by Richard C. Banks, June 24, 1964. The total length 
of this snake is 810 mm., and the tail is 60 mm.; the head is 29 mm. long (contained in the 
total length 27.9 times). There are 173 ventrals, 24 subcaudals, and 23 scale rows at midbody. 
There are 14-14 supralabials (the last conspicuously larger than the one preceding), and 15-15 
infralabials. The body has 37 blotches, and the tail, five. Speckled rattlesnakes have not pre- 
viously been reported from this island. 
Klauber (1956:38, 122) listed C. ?». 772itchelli on Isla Santa Cruz. This appears to be 
an error, and Klauber (personal communication) advises us to delete this island from the 
range. 
The Environment 
Climatic and Vegetational Patterns. — The Gulf of California is a roughly rectangular 
body of water about 1400 km. long and from about 100 km. to 200 km. wide. It lies 
between the peninsula of Baja California to the west and the mainland Mexican states of 
Sonora and Sinaloa to the east. Although geographically the Gulf is in the Temperate Zone, 
its marine biota is predominantly Panamic (eastern tropical Pacific), reflecting the isolation 
of the Gulf from the cool water and air masses of the northern Pacific. The marine inlet to 
the Gulf is at 23° North Latitude, 50 km. south of the Tropic of Cancer. 
The almost complete encirclement by land imposes on the Gulf and its islands a climate 
like that of an inland body of water, marked by extreme seasonal variation in air and water 
temperatures. This is especially true at the northern end of the Gulf, as illustrated by the 
annual variation of the surface temperatures of its waters. At Puerto Periasco in the north, the 
annual range is 16° C (14.9° C in January to 31.2° C in August); at Cabo San Lucas in 
the south, on the Pacific side, the annual range is about 9° C (19.8 C in January to 29.1° C 
in August) (Roden and Groves, 1959). Similarly, air temperature extremes are greater in the 
northern Gulf coast area than in the southern (Roden, 1958). 
148 
San Diego Society of Natural History 
[Vol. 14 
CENTRAL GULF COAST '^ 
DESERT 
\;^ LESS THAN 507< 
PRECIPITATION IN SUMMER 
MORE THAN 70% 
PRECIPITATION IN SUMMER 
I - - I , . . . I .- 
' i~ 
' . . I 
Fig. 2. Baja California and the Gulf of California. Precipitation belts are indicated by 
isoprecipitation lines. Numbers refer to inches of rainfall per year. (Adapted from Aschmann. 
1959, and Shreve and Wiggins, 1964.) 
1966] SouLE AND Sloan: Herpetofauna of Gulf Islands 149 
The Gulf islands and the surrounding continental areas are extremely arid. Except for 
part of the Cape Region and the southernmost 400 km. of the mainland coast, the 
Gulf coast and adjacent areas lie entirely within the boundaries of the Sonoran Region of the 
North American Desert (Shreve and Wiggins, 1964). All but the southernmost parts of 
the Gulf probably receive less than seven inches of rain annually (fig. 2), although occasional 
hurricanes, known locally as "chubascos", may bring more than this amount within a few 
hours. At such times the typically dry arroyos become torrents. 
The data pertaining to the rainfall pattern over the peninsula have been summarized by 
Aschmann (1959). He found that in the northern peninsular area to Bahia de los Angeles 
(29' N. Lat.), most of the precipitation is from Pacific cyclonic storms, and occurs in the 
winter. The actual amount of precipitation from these storms is slight, however, because this 
part of the Gulf coast is in the rain shadow of the lofty Sierra San Pedro Martir and Srerra 
Juarez. Aschmann pointed out that most of the precipitation from the vicinity of Santa 
Rosalia (27~ 20' N. Lat.) southward is from convectional showers and chubascos and occurs 
in the summer (fig. 2). The central part of the peninsula lies between the northern winter 
storm belt and the southern summer storm belt. Few records are kept in this area, but it seems 
that some parts may receive less than three or four inches annually, and even this meager 
amount is not dependable. In fact, droughts lasting a year or longer have been reported. 
With the exception of the Islas Las Encantadas group in the north, Shreve and Wiggins 
(1964) considered all of the Gulf islands studied here to be in the Central Gulf Coast sub- 
division of the Sonoran Desert (fig. 2). In emphasizing the vegetational uniformity of this 
region, Shreve states: "Throughout the two parts of this area [the central regions of both 
Gulf coasts] the vegetation is almost identical in physiognomy, and a large number of its com- 
ponent species are the same" (Shreve and Wiggins, 1964: 53). As in desert areas in general, 
vegetation is tallest and thickest in arroyos and at the bottoms of canyons. It is generally true 
that the larger the water course, the lusher the vegetation along its banks. Hillsides on some 
of the northern islands are completely barren of seed plants. 
Shreve designated this desert the Biirsera-Jatropha region because of the relative abundance 
of these thick-trunked, deciduous trees. Other common genera of shrubs and trees of the area 
are Olneya, Cercidium, Fouquieria, Prosopis, Larrea, Viscainoa, Maytenus, Stegnosperma, 
Encelia, Lyciujn, Euphorbia, Atriplex, Franseria, and Fagon'ia. In some areas cacti are very 
abundant and may be the dominant plants locally. The most common of the larger forms of 
cacti are Pachycereus pringlei, Machaerocereus gianmosus, Lophocereus schottii; Lemaireocereus 
thurberi, cylindrical forms of Opiintia, and species of Ferocactus. 
Granting the overall vegetational uniformity of the Central Gulf Coast and the adjacent 
islands, there are, nevertheless, two vegetational gradients associated with the seasonal differ- 
ences in the amount and pattern of precipitation. First, the amount of rainfall increases from 
the central Gulf area to the southern Gulf area. This produces a denser and taller vegetational 
aspect, especially in the large arroyos, southward from about Isla San Marcos to Isla Cerralvo. 
A possible faunistic concomitant of this north-south precipitation gradient is the restriction of 
amphibians to the southernmost islands, Islas Espiritu Santo-Partida Sur and Cerralvo 
(table 1 ) . Second, the farther south one goes in the Gulf, the more the precipitation tends to 
be restricted to the summer months. The result is a higher proportion of plants in the south 
whose germination and flowering depend on this periodicity of the tropical weather pattern. 
Geological History. — Biogeographic analysis is facilitated if the area under study has 
had a relatively quiet and unspectacular geological history. Such is not the case with the Gulf 
and the adjacent peninsular region. Recent geological and oceanographic research has shown 
that the shoreline of the Gulf basin has long been a dynamic physiographic feature. Apparently 
there has been extensive tectonic activity in the region, at least since the mid-Cretaceous orogeny 
(Durham and Allison, 1960) . For the present purpose, however, it is unnecessary to trace the 
geological developments occurring before the late Cenozoic since the biogeographic and evolu- 
tionary patterns found in the islands today are probably interpretable, in large part, as the 
result of Pleistocene events. 
150 San Diego Society of Natural History [Vol. 14 
The most far-reaching Pleistocene events were the cyclic glaciations. Changes in the eustatic 
sea level associated with the advance and retreat of Pleistocene continental ice sheets must have 
repeatedly modified the shoreline of the Gulf, alternately creating and drowning islands. It is 
generally agreed by geologists that the water trapped in ice sheets during the last glaciation 
accounted for a drop in sea level of approximately 1 10 meters. These ice sheets began to recede 
about 17,000 years ago, the sea level having risen to its present level about 6,000 years ago 
(Fairbridge, 1960; Godwin, Suggete and Willis, 1958). This last eustatic rise inundated large 
parts of the coast and created many of the offshore islands in the Gulf. Eustatic fluctuations 
have probably not amounted to more than a foot or so for the last 4,000 years (Russell, 1963). 
If crustal movements are ignored, it is reasonable to assume that those islands separated from 
the mainland by less than 110 meters depth (fig. 1) are less than 17,000 years old. 
This category of probable recent or young shallow-water islands includes Islas Tiburon, 
San Marcos, Coronados, San Jose, San Francisco, and Espiritu Santo, as well as many small 
coastal and satellite islands. 
The apparent tectonic instability of the region around the islands of Carmen, Monserrate 
and Danzante precludes accurate dating of these islands at this time. This instability is strik- 
ingly shown by Pleistocene wave-cut terraces at elevations of 250 to 300 feet in the vicinity 
of Santa Rosalia and on Islas Carmen and Monserrate (Anderson, 1950; Wilson and Rocha, 
1955). Edward C. Wilson (personal communication) found evidence for a recent uplift 
amounting to about 250 feet on Salsipuedes. Tectonic activity is further evidenced by faulting 
and warping of Pleistocene sediments along the western margin of the Gulf (Anderson, 1950) 
accompanying the elevation of this area (Durham and Allison, 1960). All in all, eustatic 
changes can serve only as a very rough approximation of island age. 
Some recent oceanographic investigations suggest that the great amount of volcanism, 
tectonic mobility, and overall geological diversity of the Gulf area are probably related to its 
geographical coincidence with the crest of the East Pacific Rise. The crest disappears at the 
southern end of the Gulf, but the exceptionally high flow of heat through the crust, as well as 
the frequency of shallow earthquakes in the Gulf area, both characteristic of the crest, attest 
to its extension into the Gulf (Menard, 1960). In fact, Menard's description (1960:1745) 
of the topography near the crest could easily be a description of the submarine topography 
of the Gulf: ". . . it appears to consist of volcanoes, low domes, and troughs with adjacent 
tilted fault blocks trending at various angles to the crest of the rise." Shepard (1950) con- 
cluded that the coastal and submarine scarps, as well as the extensive system of basins in the 
Gulf, are the result of faulting, some of it rather recently. 
Endemicity and its Biogeographical Correlates 
Johnston (1924) found that less than one per cent of the plants on the islands in the 
Gulf were endemic, and Gentry (1949) remarked, "The most striking aspect of the island 
floras is the apparent lack of divergent evolution." The terrestrial vertebrates, however, present 
quite a different picture. Nearly all of the mammals have been considered taxonomically 
distinct from their closest mainland relatives. In general, the reptiles also show a high degree 
of differentiation, although the percentage of endemics (about 20 per cent) is much lower 
than that for mammals. 
Any comparison of endemicity between plants and animals must be subjective. Never- 
theless, the striking difference in the frequency of endemics seems worthy of comment. Two 
reasons are suggested for this difference. First, dispersal in plants is generally more efficient 
over short stretches of water; accordingly, the relatively high rates of gene flow from mainland 
populations of plants to island populations would impede differentiation. Peter H. Raven 
suggested the second reason in conversation. Edaphic and climatic fartors are probably para- 
mount in the evolution of most plant species; since the islands and adjacent mainland coasts 
are apparently very similar in soil type and climate, little evolutionary adjustment would 
be expected. 
1966] SouLE AND Sloan: Herpetofauna of Gulf Islands 151 
It is probable that the difference in evolutionary divergence between the mammals and 
the reptiles, as indicated by taxonomic designation, is explicable partly as a difference in the 
characters used by the taxonomists dealing with the respective groups. Mammalogists rely 
heavily upon coat color differences in making taxonomic decisions about races; in rodents, 
especially, coat color is usually correlated with the substrate color (Dice and Blossom, 1937). 
Substrate-matching is common in reptiles as well, but herpetologists often ignore color because 
of the metachromatic capacities of many reptiles, and because reptile colors fade or are altered 
by preservation. 
Populations of reptiles often seem to react to insular environments in predictable ways. 
Among the most frequently observed of these evolutionary tendencies is gigantism and 
"melanism" in lizards, and dwarfism in snakes (Klauber, 1963; Mertens, 1934). Insular gigan- 
tism, here meaning a larger mean body size than typically found in mainland populations, 
occurs in a number of forms on the deep-water, distant islands. Two species of chuckwallas, 
Sauromalus variiis on Isla San Esteban and S. hispidus on most of the islands in the Isla 
Angel de le Guarda-San Lorenzo chain are bigger than the mainland forms. The Cerralvo 
Island whiptail lizard, Cnemidophorm ceralbensis, seems to be a gigantic insular derivative of 
the orange-throated whiptail lizard, C. hyperythrus. One of us (Soule, 1966) studied 
gigantism in the side-blotched lizards {Uta stansbiiriana, sensu lato) of the Gulf islands. The 
largest utas occur on Isla San Pedro Martir and on the Islas San Lorenzos; Uta is the only 
sceloporine lizard on those islands. It was concluded that the presence of other sceloporine 
lizards somehow inhibits an evolutionary increase in body size, and that the level of inhibition 
is inversely proportional to the square root of the number of sympatric sceloporine species. The 
large speckled rattlesnake of Isla Angel de la Guarda, Crotalus mitchelli angelensis, is an 
exception to the typical tendency for dwarfism in insular rattlesnakes in the Gulf (Klauber, 
1963). 
Insular melanism, or more generally, a dark albedo relative to the color of continental 
forms, is characteristic of some of the same populations mentioned in the discussion of gigan- 
tism. The utas on Islas San Pedro Martir, San Pedro Nolasco, San Lorenzos, and Salsipuedes 
are noticeably darker than those on shallow-water islands or on the mainland. Similarly, the 
whiptail lizards on Islas San Lorenzos, San Pedro Martir, and San Pedro Nolasco are rela- 
tively dark. The chuckwalla, Sauromalus hispidus, that is endemic to the Salsipuedes archi- 
pelago^, is uniformly dark brown or black when adult. The significance of insular "melanism" 
is not definitely known, although the high rate of heat transference provided by a dark albedo 
would seem to be advantageous to ectotherms (see Kramer, 1949; Mertens, 1952). 
Species-Area Relationship 
Population biologists have recently paid increasing attention to the challenging question 
of species diversity. Quantitative models capable of predicting the number and the relative 
abundance of species in a given area (with varying degrees of success) have been proposed 
by a number of authors (MacArthur, 1957; MacArthur and Wilson, 1963; Preston, 1962; 
Hairston, 1959; Hamilton et al., 1963). Some of these models are more or less a matter of 
curve- fitting, while others rest on a combination of curve-fitting and theoretical considerations. 
It is the purpose of this section to describe the species-area phenomenon as it pertains to the 
vertebrate fauna of the Gulf islands. This analysis must be considered preliminary since it is 
only semi-quantitative and since new records of snakes are still being reported rather frequently. 
The positive relationship between the numbers of species and the areas of islands is shown 
for terrestrial vertebrates (excluding birds and bats) in figure 3. A notable feature of this 
figure is the relative paucity of species on the deepwater, distant (from the mainland) islands. 
^The occurrence of this and other deep-water island forms of chuckwallas on shallow-water islands adjacent 
to the peninsula, e.g., on Islas Smith, Afio Nuevo, Cabeza de Caballo, and Coronados, is very likely a 
reflection of the gustatory significance of these animals to the Indians that once occupied the region (see 
Aschmann, 1959). 
152 
San Diego Society of Natural History 
[Vol. 14 
to 
UJ 
o 
UJ 
Q. 
</) 
O 
cc 
UJ 
CO 
400 600 1000 1500 
ISLAND AREA IN SQUARE KILOMETERS 
Fig. 3. Species-area relationship of terrestrial vertebrates (excluding bats and birds) on 
the major islands in the Gulf of California. 
The line drawn through the upper points is a conservative "saturation curve" (see text). See fig. 2 
for island-number key. 
The curve drawn in this figure between the highest points (the greatest species abundance for 
their area) is a conservative estimate of the relationship between maximum species abundance 
and island area for terrestrial vertebrates in the Gulf. 
MacArthur and Wilson (1963) have treated such data in an original way to demonstrate 
the "distance effect." Their method is applied to our data in figure 4; the abscissa represents 
the distance from the land mass that is the likeliest source of immigrants. The ordinate 
represents the percentage "saturation"; that is, the ratio of the number of species known on an 
island to an estimate of the maximum number of species possible (the curve in fig. 3). The 
obvious negative association is consistent with the findings of MacArthur and Wilson (1963) 
that the "equilibrium" number of species on an island is a function of its distance from the 
source of immigrants. 
Recently, Preston (1962) has emphasized a most important point in distinguishing between 
the biotas of islands and of mainland areas. The biota of any mainland area is essentially a 
sample of the biota of a much larger region. Many of the species are represented by only a 
few individuals or by migrants and would not be expected to survive long if they were not 
periodically replenished from outside. On islands, replenishment is obviously less frequent than 
it is on most continental regions of equal size. As such, islands more closely approach the 
"correct" number of species for their areas. The view of MacArthur and Wilson is in general 
agreement with that of Preston, but emphasizes (1) the dynamic nature of the equilibrium — 
the actual equilibrium number for any island would be a biogeographical constant, but the 
species composition is constantly changing — and (2) the effect of distance from the source (s) 
1966] 
SouLE AND Sloan: Herpetofauna of Gulf Islands 
153 
DISTANCE FROM SOURCE IN KILOMETERS 
Fig. 4. Percentage saturation of terrestrial vertebrate faunas (excluding bats and birds) 
on major islands in the Gulf of California. 
For Islas Partida Norte, San Esteban, San Diego, and San Francisco, the distance from the source 
is estimated from the nearest large island. For all other islands, the distance is measured from the mainland. 
See fig. 2 for island-number key. 
of immigrants on the equilibrium number. Hamilton and Rubinoff (1963) demonstrate a 
distance effect relative to endemicity of Darwin's Finches on the Galapagos Islands. 
Most quantitative studies of insular biogeography have concerned oceanic islands. During 
the colonization of such islands the equilibrium number of species is approached from below 
by the gradual accumulation of successful colonizers. In contrast, continental islands are 
formed with their biotas intact. At the time of isolation the biota of a continental island is 
a sample of the mainland biota and includes more species than would be predicted for a 
biotically equilibrated island of comparable size and distance from the source (s) of immigrants. 
For continental islands, the approach to the equilibrium number of species would be from 
above, i.e., extinction will proceed at a greater rate than will colonization. Consequently, the 
biotas of recently isolated continental islands are "supersaturated." It is problematical how 
long a biota will remain supersaturated. This depends on the organisms being considered, the 
nature and constancy of the physical environment, the size and ecological diversity of the 
island, and other factors. 
With respect to the Gulf islands, the greater species abundance on the shallow-water 
islands could be attributed either to supersaturation or to the effect of distance on the equili- 
brium number. If further analysis confirms the lit of these data to the latter model, it would 
mean that equilibrium has been reached in the few thousands of years since these islands 
were formed. 
The effects of species interactions on distribution and species abundance of reptiles on 
the Gulf islands is discussed elsewhere (Soule, 1966). 
154 San Diego Society of Natural History [Vol. 14 
Summary 
The distribution table lists some 250 records of insular occurrences of reptiles and am- 
phibians of 33 genera (2 anurans, 14 lizards, 16 snakes, and 1 tortoise), from 31 islands in 
the Gulf of California. Of these genera, only the lizard genus Sat or is endemic to the Gulf. 
Most of the 60 species, including the two amphibians and the tortoise, are considered mainland 
forms. Twelve of the species are sufficiently divergent from the mainland species to be con- 
sidered endemic to the Gulf. Most of the endemic subspecies are isolated populations of 
widely distributed mainland species. 
Extensions of the known range to additional islands in the Gulf of California or first 
reports of occurrence on any Gulf island are given for 12 species of snakes. 
Insular gigantism and melanism are discussed, and examples from the Gulf herpetofauna 
are given. The relationship of the number of terrestrial vertebrates to the size of the island, 
and to the distance from the probable source of immigrants, is discussed and illustrated. A 
positive correlation is demonstrated for the former, and a negative correlation for the latter. 
Literature Cited 
Anderson, C. A. 
1950. 1940 E. W. Scripps cruise to the Gulf of California. Part 1. Geology of islands 
and neighboring land areas. Geol. Soc. Amer. Mem. 43. 53 pp. 
ASCHMANN, H. 
1959. The central desert of Baja California: Demography and ecology. Ibero- Americana : 
42. Univ. Calif. Press, Berkeley. 282 pp. 
Banks, R. C, and W. M. Farmer 
1963. Observations on reptiles of Cerralvo Island, Baja California, Mexico. Herpeto- 
logica 18:246-250. 
Banta, B. H., and a. E. Leviton 
1963. Remarks on the colubrid genus Chilomeniscus (Serpentes: Colubridae). Proc. 
Calif. Acad. Sci., 4th Ser., 31:309-327. 
Cliff, F. S. 
1954a. Variation and evolution of the reptiles inhabiting the islands in the Gulf of 
California, Mexico. Unpublished Ph. D. dissertation. Stanford Univ. 
1954b. Snakes of the islands in the Gulf of California, Mexico. Trans. San Diego Soc. 
Nat. Hist. 12:67-98. 
1958. A new species of Sauromalns from Mexico. Copeia 1958:259-261. 
Crippen, R. G. 
1962. Two new insular records of amphibians from Baja California. Herpetologica 
18:137. 
Dice, L. R., and P. M. Blossom 
1937. Studies of mammalian ecology in southwestern North America with special 
attention to the colors of desert mammals. Carnegie Inst. Wash. Publ. 485. 
129 pp. 
Dickerson, M. C. 
1919. Diagnoses of twenty-three new species and a new genus of lizards from Lower 
California. Bull. Amer. Mus. Nat. Hist. 41:461-477. 
DixoN, J. R. 
1964. The systematics and distribution of lizards of the genus PhyllodactyUis in North 
and Central America. New Mexico State Univ. Research Center, Sci. Bull. 64-1. 
139 pp. 
Durham, J. W., and E. C. Allison 
1960. The geologic history of Baja California and its marine faunas. Syst. Zool. 
9:47-91. 
Etheridge, R. 
1961. Additions to the herpetological fauna of Isla Cerralvo in the Gulf of California, 
Mexico. Herpetologica 17:57-60. 
1966] SouLE AND Sloan: Herpetofauna of Gulf Islands 155 
Fairbridge, R. W. 
1960. The changing level of the sea. Sci. Amer. 202(5) : 70-79. 
FiGG-HoBLYN, J. P., AND B. H. BaNTA 
1957. Lampropeltis getuliis coiijiincta (Cope) on Cerralvo Island, Gulf of California, 
Mexico. Herpetologica 13:192. 
Gentry, H. S. 
1949. Land plants collected by the Velero III, Allan Hancock Pacific Expedition 
1937-1941. Allan Hancock Pac. Exped. 13:5-245. 
Godwin, H., R. P. Suggate, and E. H. Willis 
1958. Radiocarbon dating of the eustatic rise in ocean-level. Nature 181:1518-1519. 
Hairston, N. G. 
1959. Species abundance and community organization. Ecology 40:404-416. 
Hamilton, T. H., and I. Rubinoff 
1963. Isolation, endemism, and multiplication of species in the Darwin finches. Evolu- 
tion 17:388-403. 
Hamilton, T. H., I. Rubinoff, R. H. Barth, Jr., and G. L. Bush 
1963. Species abundance: natural regulation of insular variation. Science 142:1575-1577. 
Johnston, I. M. 
1924. Expedition of the California Academy of Sciences to the Gulf of California in 
1921. The botany (The vascular plants). Proc. Calif. Acad. Sci., 4th ser., 
12:951-1218. 
Klauber, L. M. 
1931a. A new subsp>ecies of the California boa, with notes on the genus Lichanura. 
Trans. San Diego Soc. Nat. Hist. 6:305-318. 
1931b. Crotalus tigris and Crotaliis enyo, two little known rattlesnakes of the Southwest. 
Trans. San Diego Soc. Nat. Hist. 6:353-370. 
1935. Phyllorhynchns, the leaf-nosed snake. Bull. Zool. Soc. San Diego No. 12. 31 pp. 
1940a. The worm snakes of the genus Leptotyphlops in the United States and northern 
Mexico. Trans. San Diego Soc. Nat. Hist. 9:87-162. 
1940b. Two new subspecies of Phyllorhynchus, the leaf-nosed snake, with notes on the 
genus. Trans. San Diego Soc. Nat. Hist. 9:195-214. 
1941. The long-nosed snakes of the genus Rhinocheilus. Trans. San Diego Soc. Nat. 
Hist. 9:289-332. 
1956. Rattlesnakes: Their Habits, Life Histories, and Influence on Mankind. Univ. 
Calif. Press, Berkeley. 2 vol. xxix + 1476 pp. 
1963. A new insular subspecies of the speckled rattlesnake. Trans. San Diego Soc. Nat. 
Hist. 13:73-80. 
Kramer, G. 
1949. Uber Inselmelanismus bei Eidechsen. Z. Ind. Abst. u. Vererb. 83:157-164. 
Lindsay, G. E. 
1962. TTie Belvedere expedition to the Gulf of California. Trans. San Diego Soc. Nat. 
Hist. 13:1-44. 
Lowe, C. H., Jr., and K. S. Norris 
1955. Analysis of the herpetofauna of Baja California, Mexico. III. New and revived 
reptilian subspecies of Isla de San Esteban, Gulf of California, Sonora, Mexico, 
with notes on other satellite islands of Isla Tiburon. Herpetologica 11:89-96. 
MacArthur, R. H. 
1957. On the relative abundance of bird species. Proc. Nat. Acad. Sci. 43:293-295. 
MacArthur, R. H., and E. O. Wilson 
1963. An equilibrium theory of insular zoogeography. Evolution 17:373-387. 
Mayr, E. 
1942. Systematics and the Origin of Species. Columbia Univ. Press, New York. 334 pp. 
156 San Diego Society of Natural History [Vol. 14 
Menard, H. W. 
1960. The East Pacific Rise. Science 132:1737-1746. 
Mertens, R. 
1934. Die Insel-Reptilien, ihre Ausbreitung, Variation, und Artbildung. Zoologica 
32:1-209. 
1952. Schwarzblaue Insel-Eidechsen und die neueren Ansichten iiber ihr Farbkleid. 
Natur u. Volk- 82:386-395. 
Murray, K. F. 
1955. Herpetological collections from Baja California. Herpetologica 11:33-48. 
Preston, F. W. 
1962. The canonical distribution of commonness and rarity. Parts I, II. Ecology 
43:185-215, 410-432. 
RODEN, G. I. 
1958. Oceanographic and meteorological aspects of the Gulf of California. Pacific Sci. 
12:21-45. 
RoDEN. G. I., AND G. W. Groves 
1959. Recent oceanographic investigations in the Gulf of California. Jour. Mar. Res. 
18:10-35. 
Russell, R. J. 
1963. Recent recession of tropical cliffy coasts. Science 139:9-15. 
Savage. J. M., and F. S. Cliff 
1954. A new snake, Pbyllorhytichus arenicola, from the Gulf of California, Mexico. 
Proc. Biol. Soc. Wash. 67:69-76. 
Schmidt, K. P. 
1922. The amphibians and reptiles of Lower California and the neighboring islands. 
Bull. Amer. Mus. Nat. Hist. 46:607-707, pi. 47-62. 
Shaw, C. E. 
1945. The chuckwallas, genus Sauromalus. Trans. San Diego Soc. Nat. Hist. 
10:269-306. 
Shepard, F. p. 
1950. 1940 E. W. Scripps cruise to the Gulf of California. Part III. Submarine topog- 
graphy of the Gulf of California. Geol. Soc. Amer. Mem. 43. 32 pp. 
Shreve, F.. and I. L. Wiggins 
1964. Vegetation and Flora of the Sonoran Desert. Stanford University Press. 2 
vol. X + 1740 pp. 
Smith, H. M., and E. H. Taylor 
1945. An annotated checklist and key to the snakes of Mexico. U. S. Nat. Mus. Bull. 
187. 239 pp. 
1950. An annotated checklist and key to the reptiles of Mexico exclusive of the snakes. 
U. S. Nat. Mus. Bull. 199. 253 pp. 
SOULE, M. E. 
1961. Eridiphas slcvini (Tanner) on Cerralvo Island, Gulf of California, Mexico. 
Herpetologica 17:61. 
1966. Trends in the insular radiation of a lizard. Amer. Nat. 100:47-64. 
Van Denburgh, John 
1922. The reptiles of western North America. Occ. Pap. Calif. Acad. Sci. no. 10. 2 vol. 
1028 pp. 
Wilson, I. F., and V. S. Rocha 
1955. Geology and mineral deposits of the Boleo copper district, Baja California, 
Mexico. U. S. Geol. Survey Prof. Paper 273. 134 pp. 
ZwEiFEL, R. G., and K S. Norris 
1955. Contribution to the herpetology of Sonora, Mexico: Descriptions of new sub- 
species of snakes {Micruroides euryxanthus and Lampropeltis getulus) and mis- 
cellaneous collecting notes. Amer. Midi. Nat. 54:230-249.