HARVARD UNIVERSITY
LIBRARY
OF THE
Museum of Comparative Zoology
,, ^^e Library
••useu™ of Ccnaratlve Zoology Harvard University
1
i4y.ui
Internet Archive
RoQtnn ?^rpnninn P.ontor
:'> t !
UNIVERSITY OF KANSAS
MUSEUM OF NATURAi-RlSTORY- ^ ^^
LIBRARY
^P3 1 4 Wl
MISCELLANEOUS
PUBLICATION
No. 53
m
Evolutionary Relationships, Osteology, and Zoogeography of Leptodactyloid Frogs
By
John D. Lynch
UNIVERSITY OF KANSAS LAWRENCE 1971
June 30, 1971
138
1967
MISCE
SubfamiliE Phylloston tera ) base and myok lane Univ. Weinmann, J. P. 1966. Emption in Orban'; ology (H Mosby Co
V-:
'- - i,iU;^. CU.ViP. ZCCL
LIBRARY
UNIVERSITY OF KANSAS miscellaneous
MUSEUM OF NATURAL HISTORY JiW-il^Ajgpjj
No. 55
HARVARD^ UNlVERSl-ni
Middle American Lizards of the Genus Ameiva (Teiidae) with Emphasis on Geographic Variation
By
Arthur C. Echternacht
UNIVERSITY OF KANSAS
LAWRENCE 1971 December 14, 1971
Jk..
UNIVERSITY OF KANSAS PUBLICATIONS MUSEUM OF NATURAL HISTORY
The University of Kansas Publications, Museum of Natural History, beginning with volume 1 in 1946, was discontinued with volume 20 in 1971. Shorter research papers formerly published in the above series are now published as Occasional Papers, Museum of Natural History, The Miscellaneous Publications, Museum of Natural History, began with number 1 in 1946. Longer research papers are pub- lished in that series. Monographs of the Museum of Natural History were initiated in 1970.
Institutional libraries interested in exchanging publications may obtain the Occa- sional Papers and Miscellaneous Publications by addressing the Exchange Librarian, University of Kansas Library, Lawrence, Kansas 66044. Individuals may purchase separate numbers of all series. Prices may be obtained upon request addressed to Publications Secretary, Museum of Natural History, University of Kansas, Law- rence, Kansas 66044.
University of Kansas Museum of Natural History
Miscellaneous Publication No. 55 December 14, 1971
Middle American Lizards
of the Genus Ameiva (Teiidae)
with Emphasis on Geographic Variation
By
Arthur C. Echternacht
Department of Biology
Boston University
Boston, Massachusetts 02215
University of Kansas
Lawrence
1971
University of Kansas Publications, Museum of Natural History
Editors: Craig C. Black, William E. Duellman, Philip S. Humphrey
Miscellaneous Publication No. 55
pp. 1-86; 28 figures
Published December 14, 1971
Museum of Natural History
University of Kansas
Lawrence, Kansas 66044
U.S.A.
Printed by
University of Kansas Printing Service
Lawrence, Kansas
CONTENTS
INTRODUCTION 4
ACKNOWLEDGMENTS 5
MATERIALS AND METHODS 6
Characters 6
Statistical Procedures 11
Sexual Dimorphism 1 2
Sources of Specimens Examined 12
SYSTEMATIC ACCOUNT OF THE SPECIES 13
KEY TO THE MIDDLE AMERICAN SPECIES OF AMEIVA 13
Ameiva ameiva ( Linnaeus ) 14
A meiva leptoph njs Cope 20
Ameiva f estiva ( Lichtenstein and Von Martens) 26
Ameiva quadrilineata (Hallowell) 34
Ameiva undulata ( Wiegmann ) 40
Ameiva chaitzami Stuart 59
DISCUSSION ■ 63
INTERSPECIFIC RELATIONSHIPS 63
INTERSPECIFIC PATTERNS IN GEOGRAPHIC VARIATION 64
HISTORY OF AMEIVA IN MIDDLE AMERICA 67
SUMMARY 71
RESUMEN 72
APPENDIX A: SUMMARY OF STATISTICS 75
APPENDIX B: SPECIMENS EXAMINED 76
LITERATURE CITED 83
INTRODUCTION
In their revision of the Mexican species of Ameiva, Smith and Laufe ( 1946 ) prefaced their comments with the remark that interest in these species had "been rising to a crescendo whose peak surely could soon be anticipated." Little could they have foreseen the surge of interest, both taxonomic and ecological, that the entire genus has enjoyed over the past five years. This attention has been precipitated by the realization that Arneiva, by virtue of its wide distribution in the Neotropical Realm and the extremes of variation exhibited by its species, is especially well suited for ecological, biogeograph- ical, and evolutionary studies. Coupled with this has been the discovery that some species of Ameiva act as vectors for disease-causing organisms actually or potentially dangerous to humans (for example, see Schneider, 1965).
There are three natural geographic groupings of Ameiva: those found on the various islands of the Caribbean, those on the continent of South Amer- ica, and those from Middle America (Mexico and Central America). Taxo- nomic studies of the Caribbean Arneiva have been conducted by Baskin and Williams ( 1966 ) , Heatwole and Torres (1967), Schwartz (1965, 1967, 1968), Schwartz and Klinikowski ( 1966 ) , and Schwartz and McCoy (1970). Very little recent work has been done on the Ameiva of South America; the only study is that of Peters (1964) on the Ecuadorian species. No recent compre- hensive taxonomic studies have been made of the Middle American species, although Stuart (1942) and Smith and Laufe ( 1946 ) contributed to our knowl- edge of certain species. The most re- cent revision of the entire genus is that of Barbour and Noble (1915). Their conclusions concerning the Middle American species of Ameiva were based on examination of only 41 specimens, a situation which largely masked the re-
markable degree of geographic varia- tion exhibited by some of the species. It is the purpose of this study to present an analysis of the geographic variation in each of the Middle American species of Ameiva as well as to summarize known information on the biology and ecology of these lizards.
Some persons will look upon the present work as justification for the rec- ognition of subspecies of Middle Amer- ican Ameiva. The controversy over sub- species has been debated extensively and heatedly elsewhere (Wilson and Brown, 1953; Peters, et ah, 1954; Smith and White, 1956; Savage and Heyer, 1967; Smith, 1967; and others), and I will not dwell on it here other than to state my conviction that a thorough de- scription of geographic variation may make unwarranted and often undesir- able the arbitraiy nomenclatural frag- mentation of highly variable species.
There is a growing body of evi- dence indicating the congeneric status of Ameiva and Cnemidophorus. Van- zolini and Valencia ( 1965 ) , Estes (1969), and Gorman (1970) demon- strated the closeness of the relationship and discussed the difficulties encoun- tered in trying to diagnose the two genera on the basis of known charac- ters. Gorman (1970:240), however, was hesitant to make a decisive statement on the matter when he commented: "The similarity of karyotypes between Ameiva and the southernmost Cnemido- phorus species group does not neces- sarily demonstrate close relationship, but there seem to be no characters that clearly separate the two genera." I be- lieve that the argument will not be re- solved until increased knowledge of South American Ameiva, especially the perplexing A. lacertoides, is available. The condition of the base of the tongue (free and heart-shaped in Cnemido- phorus, sheathed in Ameiva) has been considered to be diagnostic (Burt,
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
1931), but A. lacertoides has a Cnemi- dophorus-\ike tongue (Vanzolini and Valencia, 1965; personal observation). I prefer not to comment on the issue, but instead await the completion of relevant investigations by Richard Estes, George C. Gorman, William P. Mac- Lean III, and William Presch.
ACKNOWLEDGMENTS
During the course of this study, I have built up a huge debt of gratitude to many people who have provided as- sistance in one way or another. I have probably taxed the patience of a num- ber of museum curators, curatorial as- sistants, and collectors with requests for the loan of large numbers of specimens and locality or ecological data on spe- cific specimens. I especially acknowl- edge the following: Kraig K. Adler, the late Doris M. Cochran, James R. Dixon, William E. Duellman, Henry S. Fitch, George Foley, Jean Guibe, Giinther Peters, Robert F. Inger, Daniel A. Jan- zen, John M. Legler, Charles H. Lowe, Edmond V. Malnate, Hymen Marx, John R. Meyer, Charles W. Myers, Thomas Olechowski, James A. Peters, Douglas C. Robinson, Richard D. Sage, Jay M. Savage, Norman J. Scott, Hobart M. Smith, Dorothy M. Smith, A. G. Stimson, L. C. Stuart, Stephen G. Til- ley, Donald W. Tinkle, Robert G. Tuck, Jaime Villa, Ernest E. Williams, John W. Wright and Richard G. Zweifel.
I am indebted to Richard F. John- ston, Irving Pfau, James D. Rising, F. James Rohlf, Gerald R. Smith, and especially Dennis M. Power for advice on statistical methods or computer pro- cedures. The computer programs uti- lized in this study were provided by Dennis M. Power.
I have had the benefit of a great deal of constructive criticism and advice from my colleagues at the University of Kansas Museum of Natural History. William E. Duellman has been a con- stant source of encouragement, and his patience with me has, at times, been
remarkable. He, along with Henry S. Fitch and Ronald L. McGregor, have reviewed this manuscript. Donald R. Clark, Ko Ko Gyi, John D. Lynch, Rus- sell J. Hall, Charles W. Myers, E. H. Taylor and Linda Trueb have been will- ing to discuss special problems as they arose, and their help is gratefully acknowledged.
Richard Estes, George C. Gorman and James A. Peters have kindly al- lowed me to read their unpublished manuscripts, and Estes has also been gracious with his time and knowledge of fossil teiids.
E. Raymond Hall, Philip S. Hum- phrey, and William E. Duellman of the University of Kansas Museum of Nat- ural History have allowed me free ac- cess to the collections and facilities' of the museum. The Panamanian collec- tions made by Duellman and Charles W. Myers have been especially useful. Field work was financed in part by Na- tional Science Foundation Summer Fel- lowships for Graduate Teaching As- sistants ( 1965 and 1966 ) and by travel grants and traineeships provided by the Committee fof Systematics and Evo- lutionary Riology from funds granted by the National Science Foundation (Grant No. GR-4446X) to the Univer- sity of Kansas, Dr. William A. Clemens and Dr. George W. Byers, principal in- vestigators (1966-1968). A grant-in-aid from the Society of the Sigma Xi pro- vided funds for illustrations, which were executed by Mrs. Susann Jacobshagen. Computer-time was made available by the University of Kansas and Boston University Computer Centers.
To my wife Sandra, I owe my great- est thanks. She has been a constant source of encouragement and support and has proven to be an able assistant in both field and laboratory. She has cheerfully endured the many incon- veniences experienced by the families of graduate students, and, with the realiza- tion that the debt can never be repaid, I dedicate this work to her.
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
MATERIALS AND METHODS
This study is based on the examina- tion of over 3000 specimens of Ameiva, 2848 of which were examined in detail with respect to morphology, color and pattern. Samples were drawn from throughout the ranges of Ameiva fes- tiva, A. leptophrys, A. quadriUneata, A. tindulata, and A. chaitzami. Onlv Pan- amanian specimens of Ameiva ameiva were examined, because to include specimens from throughout the range of the species would have necessitated examination of material from South America, where inadequate geographic representation of samples would have made any conclusions more than usually speculative. Ameiva ameiva will be in- cluded in a projected study of South American members of the genus which will be completed only after field work necessary for the accumulation of sam- ples representative of the entire South American range of the species is pos- sible. South American samples of Ameiva festiva have been included in the present study because the range of the species is primarily in Central America.
For purposes of statistical analysis, I attempted to amass samples of at least 30 specimens (ideally 15 males and 15 females) from relatively restricted localities. This attempt was only par- tially successful in that, with the excep- tions of Guatemala and parts of Mexico, large series of Ameiva from single lo- calities are not available. Of necessity, samples of ten or more have been used; for characters showing sexual dimor- phism (Table 1), the samples were still smaller. Greater success was achieved in accumulation of samples representa- tive of the ranges of the species, al- though lack of collections from critical areas in Panama has precluded exact statements of variation in A. ameiva and, especially, A. leptophrys.
In each of the species accounts, a short synonymy, statement of distribu- tion, diagnosis, and description are fol-
lowed by an extensive discussion of infraspecific geograpliic variation, eco- logical notes, and general remarks. A discussion of interspecific relationships and the history of the genus Ameiva in Middle America follows the species accounts.
Characters
In the interests of standardization of methods, a definition of each of the characters analyzed in this study fol- lows. Unless stated otherwise, charac- ters are defined in the same manner as by Smith ( 1946 ) . An asterisk ( * ) indi- cates a character analyzed for geo- graphic variation by Power's (1970) modification of Gabriel's ( 1964 ) Sums of Squares Simultaneous Test Proce- dure (STP). Interpretation of the re- sults of STP analysis is summarized in the section on statistical procedures. It will be noted that certain characters were not taken for some species, where- as in other instances a character was taken but not analyzed by the STP method.
Snoiit-Vent Length (SVL). — The distance, to the nearest millimeter, from the tip of the snout to the anterior mar- gin of the vent. Maximum observed values for males and females of each species are given. Collections seem to be biased in favor of larger (male) in- dividuals. Measurements were taken with a flexible plastic rule.
Head Length. — The distance to the nearest 0.1 millimeter from the tip of the snout to the posterior surface of the quadrate bone. The latter point was found by piercing the tympanum with one tip of the dial calipers and hooking the tip behind the bone. That the quad- rate was, in fact, the bone encountered was verified by x-ray photography.
^Supralahials. — The total number of supralabials on both sides of the head. The posterior supralabial is herein de- fined as the last enUu'ged scale in con- tact with both the lip and a subocular scale (Fig. la).
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
'^Infralabiah. — The total number of infralabials on both sides of the head. The posterior infralabial is determined by the point at which the masseter muscle and associated membranes pass from the lower to the upper jaw.
'* Terminal Siiblabials. — This charac- ter was noted for Ameiva festiva only. Stuart (1943:21) utilized the number of terminal sublabials to distinguish A. festiva festiva from A. /. edwarchi. I counted the sublabials present in the position of the two terminal sublabials of a "typical" A. /. festiva, totaling the numbers obtained for both sides of the head. By this method, a "typical" A. /. festiva (sensu Stuart) has four, whereas A. /. echvardsi has six or more.
'* Supraoculars. — The total number of supraoculars on both sides of the head. Supraoculars are numbered from front to back (3a-c, Fig. lb); there are three or four on each side of most specimens of Ameiva from Middle America.
'^Loreals. — The total number of loreals on both sides of the head. If more than one loreal is present on one side of the head, the second is smaller and usually located at the angle formed by the contact between supralabials and suborbitals anterior to the eye (3, Fig. la). Occasionally, there are two loreals about equal in size, one dorsal to the other. If a third scale is present in the area encompassed by the loreal(s), it is counted as a third loreal only if it is one-half the size of the smaller of the two other loreals or larger.
^Degree of Circumorbital Granule Contact With the Frontal (COF).— The degree of contact by the circumorbital granular scales with the frontal was coded for purposes of statistical analysis as follows: 1) circumorbital granules not in contact with frontal-frontopari- etal suture (Fig. lb), 2) circumorbital granules in contact with frontal-fronto- parietal suture, but not extending past it (Fig. Ic), 3) circumorbital granules extending forward past the frontal-
frontoparietal suture (Fig. Id). The coded values obtained for both sides of the head were totaled. Thus, the low- est possible value for a given specimen is two; the highest is six.
"^Circumorbital Pattern (COP). — The position of the anteriormost circum- orbital granule relative to the supra- ocular scales was coded as follows: 1) granules extending to the level of the middle of the first supraocular scale, 2) granules extending to the level of the suture between the first and second supraocular scales, 3) granules reach- ing the level of the middle of the sec- ond supraocular scale (Fig. Id), 4) granules extending to the level of the suture between the second and third supraocular scales (Fig. Ic), 5) gran- ules extending to the level of the mid- dle of the third supraocular scale (Fig. lb), 6) granules extending to the level of the suture between the third and fourth supraocular scales or, if there is no fourth supraocular, granules not reaching the level of the middle of the third supraocular, 7) granules extend- ing to the middle of the fourth supra- ocular scale. Coded values for both sides of the head were totaled for analysis.
'^Extent of the Double Row of Gran- ules Between the Supraocular and Su- perciliary Scales (SO-SC). — This char- acter was coded as follows: 1) single row of granules between the super- ciliary and supraocular series ( Fig. Id ) , 2) double row of granules extending foi-ward to the level of the middle of the third supraocular scale, 3) double row of granules extending forward to the suture between the second and third supraocular scales, 4) double row of granules extending forward to the level of the middle of the second supra- ocular (Fig. Ic), 5) double row of granules completely separating the su- perciliary scales from the second and third supraocular scales (Fig. lb). Coded values for both sides of the head were totaled for analysis.
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
B
Fig. I. Details of head and preanal scutellation of Ameiva. (A) Lateral view of head: 1. — Postnasal; 2. — Anterior loreal; 3. — Posterior loreal; 4. — Suborbital scale in contact with posterior supralabial (small, rectangular scale below 4). (B, C, and D) Dorsal view of head, left supraorbital region: 1. — Superciliary series; 2. — Supraocular-superciliary granules; 3A, 3B, and 3C. — Anterior, mid- dle and posterior supraocular scales, respectively; 4. — Prefrontal scale; 5. — Frontal scale; 6. — Frontoparietal sca'e; 7. — Circumorbital granules. (E) Pre-
anal region: 1.-
-Anterior scale; 3.-
(1st) preanal —Terminal row
scale; 2. — Posterior (last) preanal of ventral scutes.
Gular Scales. — The central gular scales are small and irregular in ar- rangement, large and irregular, or large and oriented longitudinally in the mid- line. Intermediate conditions exist, but most specimens can be assigned without
difRculty. If enlarged midgular scales are present, the smaller scales around them may become gradually smaller to- ward the periphery of the gular region or may become gradually smaller laterally and anteriorly but abruptly
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
9
very small posteriorly. If the latter is true, the gular region is distinctly di- vided into pregular and postgular areas (see Fig. 5c, A. leptophrys) .
MesoptychiaJ Scales. — The meso- pt\'ehial scales are small and irregular, continuous with the posterior gular scales, or they may form a band of abruptly enlarged scales across the throat.
"^Granules Around the Body (GAB). — The number of granules around the body was shown by Zweif el ( 1959 ) to be of diagnostic importance in distin- guishing species of Cnemidophorus, and this character has been used extensively by students of that difficult genus. In most cases, the character has been given as "scales around midbody" (I believe the term "granules" to be more de- scriptive), but rarely is there a state- ment as to what "midbody" refers. This is important inasmuch as the character is intended to be a measure of granule size, and the granules vary in size over the length of the body. In this study, GAB counts were made starting at a point adjacent to the 15th transverse row of enlarged ventral scutes, the first transverse row of ventral scutes is taken as the first row extending entirely across the chest behind the gular fold.
*" Paravertebral Granules (PV). — The number of granules between, and in- cluding, the paravertebral stripes of Ameiva leptophrys, quadriUneata, un- dulata, and chaitzami.
'^Vertebral Stripe (VS). — The num- ber of granules included within the vertebral light stripe of Ameiva festiva.
PV and VS counts were omitted for A. ameiva because the stripes are lack- ing or very faint on most adults of that species.
""PV/GAB Ratio.— The ratio of para- vertebral granules, as defined above, to granules around the body was calcu- lated.
^Granules Occiput to Rump (GOR). — The number of granular scales be- tween the enlarged occipital scales and
the large, keeled dorsal caudal scales was counted for all species. Although there is usually a high correlation be- tween values obtained for this count and those for granules around the body (correlation coefficient=0.75 and 0.54 for males and females of A. leptophrys, respectively), GOR sometimes shows sexual dimorphism and occasionally is an easier and more accurate count than GAB. The latter is especially true in poorly preserved specimens that have been coiled or that have numerous folds in the skin of the flanks. With experi- ence, neither GAB nor GOR counts are difficult or time-consuming to make, and error is not excessive (2-3% based on recounts of the same specimens). The most suitable technique for count- ing dorsal granules seems to be to use a fine insect pin as a pointer and marker while counting granules beneath a wide- field, binocular microscope. Ideally, the specimens should be slightly dry, since fluid between the granules results in glare and makes counting difficult.
''GAB /GOR Ratio.— This ratio was calculated for all species.
Prebrachials. — The number of rows of prebrachials was noted.
Posthrachials. — T h e postbrachials are enlarged or not.
Preantebrachials. — The number of rows of preantebrachial scales was re- corded.
Prefe moral Scales. — Prefemoral scales are defined herein as the large scales on the anterior and ventral surfaces of the upper hind leg as far back as the femoral pore series. Thus defined, pre- femorals include the infrafemoral and prefemoral scales of Smith (1946:28). The number of rows was counted.
Infratibial Scales. — The number of rows of infratibial scales was counted.
The number of rows of preante- brachials, prebrachials, prefemorals and infratibials is variable throughout the length of the series on a single speci- men. In many specimens, there is a greater number of scales medially than
10
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
proximally. This variation is discussed for each species.
'* Femoral Pores. — The number of femoral pores was recorded; counts given in the text and tables are for the total number of pores on both legs.
'^Subdigital Lamellae. — The number of subdigital lamellae on the anterior margin of the fourth (longest) toe of the left hind foot was recorded. The starting point for this count is the tubercule immediately proximal to the enlarged tubercule at the base of the first phalanx. If there is no tubercule in this position, the enlarged basal terminal is itself used as the starting point.
'* Longitudinal Rows of Ventral Scutes. — The number of rows of en- larged ventral scutes was counted be- tween the gular fold and the first pre- anal scale, as defined below. The first row is the anteriormost row extending entirely across the chest posterior to the gular fold.
'^Transverse Rows of Ventral Scutes. — The number of transverse rows of ventral scutes was counted at the level of the 15th longitudinal row of ventrals. The outer row on each side is reduced in size in some specimens.
'^ Total Preanal Scales. — The number of preanal scales from just posterior to the tenninal row of longitudinal ven- tral scutes to the enlarged scale(s) pre- ceding the vent were counted. The first preanal scale is usually single, lies between the proximal femoral pores of each leg, and is distinguished from the last row of ventrals by its unpaired con- dition (1, Fig. le). Smith and Laufe (1946:20) encountered difficulty in de- termining which preanal scale is termi- nal because the preanal scales occa- sionally grade in size into the granular scales immediately anterior to the vent. When in doubt as to which preanal or row of preanals is terminal, I have sim- ply recorded the last row having three or fewer enlarged scales. In most cases, this is obviously the terminal row any-
way. In those lizards with a single terminal preanal, the scales lateral to it are usually very much smaller and create no confusion as to whether or not they are part of the terminal pre- anal series.
'^Termitml Preanal Scales. — The number of scales in the temiinal posi- tion in the preanal series was deter- mined by the method outlined above (2, Fig. le).
''Scales Around the Tail (SAT).— The number of scales around the tail was counted at the level of the 15th verticil of caudal scales; the first verticil was taken as that row of caudal scales which first completely encircles the tail posterior to the vent.
Various aspects of color and color pattern were noted on a sample-by-sam- ple basis. Recurring pattern compo- nents noted were as follows:
Vertebral Stripe. — A narrow stripe originating on the snout or occiput and terminating on the tail unless com- pletely or partially lost in adults.
Middorsal Field or Area. — A broad stripe, containing a vertebral stripe or not, originating on the occiput and terminating on the tail, bordered later- ally by the dorsolateral light stripes, if present, or the dorsolateral dark field.
Dorsolateral Light Stripe. — A nar- row, light-colored stripe originating on, or posterior to, the superciliary rows and terminating on the tail and which may be entirely or partially lost in adults.
Secondary Stripe. — A narrow, dark- colored stripe, the lateral border of which is the dorsolateral light stripe, if present, or the dorsolateral dark field and the medial border is the lighter part of the middorsal field in which it lies.
Dorsolateral Dark Field. — A dark, often black or dark brown, area bounded above by the dorsolateral light stripe, if present, or the middorsal field and below by the ventrolateral light stripe, if present, or the ventrolateral field; originating on the loreal scales,
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
11
terminating on the tail and interrupted by the eye; may be indistinct or lost in adults.
Ventrolateral Light Stripe. — A nar- row, light-colored stripe originating at the posterior margin of the eye or pos- terodorsal margin of the ear and ex- tending along the body to the groin, usually continuing behind the leg and slightly onto the tail.
Ventrolateral Field or Area. — Ven- trolateral one-half or two-thirds of the flanks bordered above by the ventro- lateral light stripe, if present, or the dorsolateral dark field and below by the enlarged ventral scutes; distinguished from the dorsolateral dark field by its generally lighter color; may contain irregularly arranged spots, blotches and/ or bars or regularly spaced vertical bars.
Lateral Bars or Blotches. — Regular blotches in the dorsolateral dark field or bars on the flanks.
Head scutellation is extremely vari- able in Ameiva. This is especially evi- dent in Ameiva leptophrys and, to a lesser extent, in festiva and quadriline- ata. The supraoculars of specimens of quadrilineata from the Bocas del Toro region of Panama are so fragmented as to preclude counting them. There may be three or four supraoculars per side in any of the Middle American species of Ameiva, and specimens of all species are known which have four on one side, three on the other. The number of supraoculars is thought to be sufficiently constant in Cnemidophorus that the number is used to distinguish species groups ( Burt, 1931 ) . The posterior dor- sal head scales of Ameiva leptophrys (Fig. 7) and some festiva are variously fragmented, so that in extreme cases parietal scales are not recognizable, and the frontoparietals and frontal are only slightly less disrupted. It might seem that such variability in the posterior head scales of Ameiva renders them useless for taxonomic purposes, but it seems that the amount of variation is a
useful taxonomic character. For exam- ple, the presence of small, irregular scales separating the inteiparietal from the parietals and frontoparietals can be used to distinguish leptophrys from other Middle American Ameiva.
Statistical Procedures
Statistical summaries of all species examined comprise Appendix A.
Power's ( 1970 ) modification of Gab- riel's (1964; see Gabriel and Sokal, 1969) Sums of Squares Simultaneous Test Procedure ( STP ) provides a test of significance or non-significance of differences among sample means and is employed only after an overall analysis of variance demonstrates significant dif- ferences among the means. Samples are then ranked in decreasing order of their means and sums of squares calculated by sequentially adding means (be- ginning with the largest) until a maxi- mal non-significant subset is delimited. The procedure is then repeated, delet- ing one or more of the larger means in the previously defined subset, until another maximal non-significant subset is described. This procedure is repeated until all samples have been included in at least one such subset. A constant probability level of 0.05 was used in the study, as recommended by Power ( 1970 ) . Calculations were performed by the University of Kansas Computa- tion Center on a GE-625 computer, and at Boston University on an IBM 360-40.
In addition to STP, a number of statistics were calculated for each spe- cies and each sample. Certain of these, along with the STP results, are sum- marized in the tables accompanying each species account. Information per- taining to the total number of femoral pores of Panamanian Ameiva ameiva will serve as an example for the inter- pretation of STP results (see Appendix A). The first column of numbers in- cludes the locality code, keyed to the range map of Panamanian A. ameiva (Fig. 2). The localities are ordered in
12
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
TABLE 1. — Occurrence of sexual dimorphism among Middle American Atneiva as determined by STP analysis. A, ameiva; L, Icptophnjs; Q, quadrilineata; F, festiva; U, undulata; C, chait-
:ami: N, not tested.
Species
Character ~A L Q F U C~ Character
Supralabials — + — + + — PV/GAB
Infralabials ______ GOR
Supraoculars ______ GAB/GOR
COF values +_____ SAT
COP \'alues ______ Total Preanals
SO-SC values N — N N — — Terminal Preanals
Loreals N — N N — — Subdigital Lamellae
Sublabials N N N + N N Trans. Rows of Ventrals
GAB _ _ _ _^ _ _ Long. Rows of Ventrals
PV N — — + + — Femoral Pores
Species
Q
u
N
- - + -
+
+
|
+ |
— |
+ |
— |
+ |
— |
|
— |
— |
+ |
— |
+ |
— |
|
— |
+ |
+ |
+ |
+ |
— |
|
+ |
+ |
+ |
+ |
— |
— |
|
— |
— |
+ |
— |
— |
— |
+ + + + +
terms of tlie decreasing magnitude of their means, given in column two. Fol- lowing the mean values are the standard deviation of the mean, sample size (in parentheses), and observed range. To the right of each block of statistics are the STP results, appearing as vertical lines. There are no STP results given if the overall analysis of variance indi- cated no significant differences among the sample means, and in these in- stances the samples are ordered by their locality numbers irrespective of their mean values. In the example, samples 6 and 5, with means of 35.0 and 34.1, respectively, form the first maximal non- significant subset and are so indicated by the left-most line. Localities 5, 7, 3, 2 and 1 constitute the second subset (middle line), and localities 7, 3, 2, 1 and 4 are the third (right line). The mean for locality 6 is significantly dif- ferent from all other means, except that of locality 5 as indicated by the obser- vation that the mean for locality 6 ap- pears in no subsets containing means other than that of locality 5. Likewise, locality 5 is significantly different from locality 4. Localities 5 and 7 are not statistically different, although it would appear so if only the third subset were considered, but in this subset locality 5 was omitted from the comparison in
order to allow other means into the sub- set at the lower end of the scale. Locali- ties 5 and 7 are from the same statistical population, because they are both in- cluded in the second subset.
Statistical significance must be in- terpreted with caution. One cannot infer gene pool similarities from knowl- edge of a single character; for example, it would be improper to consider liz- ards from central Mexico and Costa Rica as belonging to the same biological population, even if they are shown to belong to the same statistical popula- tion. It is well known that phenotypic expression of the genotype can be modi- fied by environmental pressures and similarities, or differences in a particu- lar character among samples from di- vergent localities simply may be an expression of this pressure on a rela- tively plastic genome.
Sexual Dimorphism
Presence or absence of sexual dimor- phism for individual characters was de- termined by analysis of variance. This was done for each locality sample of Ameiva festiva and undulata, for all males against all females of A. ameiva, chaitzami, and leptophrys, and for At- lantic slope males versus Atlantic slope
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
13
females and Pacific slope males versus Pacific slope females of quadrUineata. If any sample as so defined was sevually dimorphic for a character, the entire If any sample as so defined was sexually dimorphic for that character for pur- poses of statistical analysis. Presence or absence of sexual dimorphism for most of the characters utilized in this study is indicated in Table 1.
Sources of Specimens Examined
A complete list of specimens exam- ined, along with locality data for each, is included in Appendix B. The follow- ing abbreviations have been used in referring to individual specimens and indicate the museum or private collec- tion in which the specimens are housed.
ANSP Academy of Natural Sciences of
Philadelphia
AMNH American Museum of Natural
History
ARP A. R. Philips Collection
(University of Arizona)
|
BMNH |
British Museum (Natural History) |
|
FMNH |
Field Museum of Natural History |
|
JRM |
John R. Meyer (Private |
|
Collection) |
|
|
LACM |
Los Angeles County Museum |
|
MCZ |
Museum of Comparative Zoology |
|
MNHN |
Museum National d'Histoire |
|
Naturelle, Paris |
|
|
RDS |
Richard D. Sage (Private |
|
Collection) |
|
|
TCWC |
Texas Cooperative Wildlife |
|
Collection |
|
|
USNM |
United States National Museum |
|
UCR |
Universidad de Costa Rica |
|
UCLA |
University of California at Los |
|
Angeles (Specimens now |
|
|
housed in the Los Angeles |
|
|
County Museum) |
|
|
UIMNH |
University of Illinois Museum of |
|
Natural History |
|
|
KU |
University of Kansas Museum of |
|
Natural History |
|
|
UMMZ |
University of Michigan Museum |
|
of Zoology |
|
|
USC(CRE) |
University of Southern California |
|
(Costa Rican Expedition) |
|
|
UU |
University of Utah |
|
ZMB |
Zoologisches Museum, Berlin |
SYSTEMATIC ACCOUNT OF THE SPECIES
I have recognized six species of KEY TO THE MIDDLE AMERICAN
Ameiva in Middle America. All of these SPECIES OF AMEIVA
are easily recognizable in the field by , „ i i . t i r
the experienced student, but are easily ^- ^^" ^? *^^^^^^ longitudmal rows of
confused when dealing with preserved ^^^^''''^ /^'"^f^; "^^^optychial scales
specimens. As is often the case, the subequal; adults spotted _. .. amewa
species cannot be distinguished readily Eight longitudinal rows of ventral
on the basis of single characters, and scutes; transverse row of enlarged
identification of preserved specimens mesoptychial scales; principally
can be difficult even if a suite of char- striped 2
acters is used. The following artificial 2. Posterior gular scales smaller than
key should be used in conjunction with anterior gular scales 3 the tables in the analyses of geographic
variation and with the descriptions of Anterior and posterior gular scales
color and pattern variation. No attempt subequal m size 4
has been made to place the key into a 3. Postnasal scales not in contact with
phylogenetic framework. Fortunately, prefrontals; lateral parietal and fron-
the situation is not so bad as that which toparietal scales separated by one or
caused Stuart (1955), in discussing his more rows of small, irregular scales;
key to Guatemalan Anolis, to comment midgular scales greatly enlarged, ir-
that ". . . the worker who knows what regular in arrangement leptophrys
species he has before him should ex- Postnasal scales in contact with pre-
perience few difficulties in its use." frontals; parietal and frontoparietal
14
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
82°
78°
50 100
KILOMETERS
82°
78°
Fig. 2. Map showing locality records for Ameiva ameiva in Panama. To avoid crowding of sym- bols, some localities have been omitted. Samples used in statistical analyses were ( 1 ) Vicinity of Panama City, (2) Vicinity of Colon, (3) Vicinity of Niievo Gorgona, (4) Isla San Jose, (5) Prov. Herrera, (6) Prov. Los Santos, and (7) Vicinity of Cerro Campana.
.scales in contact; midgular scales slightly to moderately enlarged, ir- regular or in a single longitudinal row quadrilineata
Midgular scales much enlarged, ir- regular in arrangement; narrow, light-colored vertebral stripe present except in large adults festiva
Midgular scales much enlarged, in longitudinal arrangement or not, moderately enlarged and irregular in arrangement; no vertebral stripe . 5
Small (maximum observed SVL 85 mm for males, 75 mm for females); paravertebral stripe narrow (mean PV = 37.1); dorsolateral blotches in males fused to dorsolateral light stripe chaitzami
Moderately large (maximum ob- served SVL 129 mm for males. 111 for females ) ; paravertebral
mm
stripe broad (mean PV = 47.4 for males, 46.4 for females); dorsolateral blotches in males, if present, do not
fuse with dorsolateral light stripe
undulata
Ameiva ameiva (Linnaeus)
Lacerta ameiva Linnaeus, Systema naturae, ed. 12, 1:362, 1766.— America.
Cnemidophorus praesignis Baird and Girard, Proc. Acad. Nat. Sci. Philadelphia, 6:129, 1852 [Syntypes: USNM 5519 and UMMZ 3823, Chagres, Panama; C. B. Adams, col- lector].
Ameiva praesignis — Cope, Proc. Acad. Nat. Sci. Philadelphia, 14:67, 1862.
Ameiva praesigna — Bocourt, Mission Scienti- fique au Mexique et dans I'Amerique Cen- trale; fitudes sur les reptiles, livr. 3, p. 265, pi. 20b, fig. 9-9d, 1874.
Ameiva surinamensis — Boulenger, Catalogue of the Lizards in the British Museum (Natu- ral History), 2nd Ed., vol. 2, p. 352, 1885 (part). Giinther, Biologia Central!- Ameri- cana; Reptilia and Batrachia, p. 22, pi. 19, 1885 (part).
Ameiva ameiva praesignis — Barbour and Noble, Bull. Mus. Comp. Zool., 59(6):468, 1915. Dunn, Proc. Acad. Nat. Sci. Phila- delphia, 92:113, 1940.
Distribution in Tanaind. — Grasslands and savannas of the Pacific slopes near Chepo, Panama, to David, Chiriqui, and in open areas across the Canal Zone to Colon and Portobelo, Colon, on the Atlantic side (Fig. 2).
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
io
Diagnosis. — Ameiva ameiva can be distinguished from all other Middle American Ameiva by the following com- bination of characters: Size large, max- imum observed SVL 197 mm for males and 157 mm for females; mesoptychial scales small, irregular; scales around the tail more than 30; ventrals at midbody usually more than eight; supraoculars usually eight; color pattern in adult males emphasizes spots, rather than longitudinal stripes or lateral bars and blotches.
Description. — Maximum observed SVL 197 mm for males, 157 mm for females; 11-16 (12.9) supralabials; 8-12 (10.1) infralabials; 6-10 (8.0) supraocu- lars; COF values 2-6 (2.1), but note geographic variation discussed below; COP values 8-16 (10.6); nostril in pre- nasal-postnasal suture; prefrontals in contact with postnasals; loreals 2; GAB 128-204 (146.0), 204 an exceptional value — the second highest value re- corded was 169; GOR 229-324 (287.8) for males, 225-394 (295.5) for females; GAB/GOR 0.45-0.68 (0.51) for males, 0.41-0.67 (0.49) for females; SAT 34-45 (39.5); total preanal scales 7-16 (10.8); terminal preanal scales 2-3 (2.4) for males, 2-3 (2.2) for females; one row of enlarged prebrachials with irregularly arranged accessory rows on either side; postbrachials slightly to moderately en- larged, irregularly arranged; preante- brachials enlarged, one (occasionally two) rows; prefemoral scales enlarged, in four or more rows; infratibial scales in three rows; femoral pores 25-41 (32.4); subdigital lamellae 26-36 (30.9); longitudinal rows of ventral scutes 28- 33 (31.2); transverse rows of ventral scutes 6-12 (11.0); dorsal caudal scales weakly keeled. For details of scutella- tion see figure 3.
Color and Pattern in Alcohol. — Juve- nile males: Top and upper half of side of head gray-green to brown with black markings immediately anterior to eye; lower one-half of head from eye to post- nasal white; supralabials black above,
Fig. 3. Head and preanal scutellation of Ameiva ameiva from Panama (KU 108258): ( A ) Lateral view of head, ( B ) Dorsal view of head, (C) Ventral view of head. X L3. (D) Preanal region. X 2.
white below to lip; infralabials black below, white above to lip; chin and gular region white or cream with small, scattered flecks and often a transverse row of black flecks between larger me- dian gulars and smaller posterior gulars; middorsal area between occiput and tail black or dark brown with brown or gray-green middorsal stripe, anterior one-fifth of which as broad as paraver- tebral field, gradually narrowing to a width of about 10 granules near base of tail and with irregular margins through- out; cream or gray-white paravertebral stripes from (and including) supercili- ary series well onto tail; dorsolateral dark stripe black with or without a brown median stripe and with many small, white or blue- white spots; lateral light stripe cream, originating at dorso- posterior margin of ear and extending
16
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
to point of insertion of hind leg and onto anterior surface of thigh, beginning again on posterior surface of thigh and extending well onto tail; area between lateral light stripe and ventral scutes black with small, white spots (some in- distinct); venter white or cream medial- ly with lateral scutes and (occasionally) chest marked with black Hecks; dorsal surfaces of fore and hind limbs black with blue, blue-white or pale brown spots and blotches, this color and pat- tern grading to uniform cream or white ventrally with or without black Hecks; pre- and postfemoral light stripes cream. Tail brown with black or blue-gray flecking above, flecks sometimes ar- ranged to give the appearance of whorls; white dorsolateral stripe; sides black with white flecking; ventrolateral light stripe white, blending into white or cream ventral coloring; subcaudal area sometimes with black blotches; short, black postanal stripes extending from each corner of the vent.
Through ontogeny there is a gradual loss of stripes and an increased empha- sis on spotting, so that in adult males the dorsal pattern is one of transverse rows of white or blue-white spots on a black background with indistinct gray- green or pale brown areas between the rows of spots (Fig. 4). Anteriorly, the pattern breaks down altogether, leaving only fine, light brown reticulations on a black background. This pattern carries over onto the dorsal surfaces of the forelimbs and the granules of the side of the head. The hind limbs become black with blue flecks, again sometimes arranged to give the appearance of whorls. Ventrally, the body, tail and hind limbs are blue; the chin, gular region and forelimbs white or cream with black flecks. In adult males the chin, gular region, and the forepart of the chest may be pale orange. Males from eastern Panama retain a bright orange or reddish-tan vertebral stripe as adults.
Females: Juvenile females resemble
juvenile males, but there is strong sex- ual dimorphism in adult color patterns ( Fig. 4 ) . Adult females retain longi- tudinal stripes, and spotting does not become as pronounced as in adult males. Adult females have brownish- white to white dorsolateral and lateral stripes bounded by black. Middorsally, there may be a remnant of the vertebral stripe (this is always true of specimens from eastern Panama) with the remain- ing area mottled gray-green or brown on black with scattered white spots. A short cream stripe originating at the posteroventral corner of the ear and ending at the shoulder may be present. The dorsolateral dark stripe is black with white spots and may contain a dif- fuse, brown secondary stripe. Between the lateral light stripe and the enlarged ventral scutes there is a broad, light brown stripe. Some females have no spotting at all, the areas between the paravertebral and lateral light stripes appearing uniformly green or pale brown. All combinations of plain and spotted patterns in middorsal, dorsolat- eral and ventrolateral fields may be found.
Color and Pattern in Life. — The fol- lowing description of color in life of Atneiva ameiva is quoted (with modi- fications to eliminate abbreviations) from the field notes of C. W. Myers and
Fig. 4. Ontogenetic Change and Variation in Pattern of Ameiva ameiva from Panama. Males: (A) KU 107521, SVL 68 mm, from Guanico, Los Santos Prov.; (B) KU 107542, 107 mm, from Pese, Herrera Prov.; (C) KU 107532, 160 mm, from Guanico, Los Santos Prov. Females: (D) KU 76106, 45 mm, from S Slope Cerro Campana, Panama Prov.; (E) KU 95546, 92 mm, from the Rio Salado, Chiriqui Prov.; (F) KU 107524, 132 mm, from Guanico, Los Santos Prov.; (G) KU 108259, 150 mm, from Juan Mina, Chagres River, Canal Zone. Small males resemble small fe- males as exemplified by KU 76106 (D above). KU 108259 (G above) is included to illustrate the presence of a well defined vertebral stripe in adult lizards from eastern Panama ( see te.xt for discussion).
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
17
18
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
pertains to specimens collected at Guanico Arriba, Los Santos, Panama: (KU 107521-3; juveniles) ". . . Mid- dorsum brown with darker brown marks and bounded by pale yellow dorsolat- eral stripe. Sides rich brown or blackish brown, with pale tan dots, bordered below by a creamy stripe." ( KU 107524- 7; adult females) ". . . Dorsolateral and lateral stripes remain conspicuous, and the dorsum takes on a greenish tinge or not; a variable amount of tan to yellow dorsal spots. Throat light orange, and undersides of arms orange tinged to light dull red; rest of venter white in one and pale blue in the others; rear of the thighs white or pinkish white. Iris light brown." (KU 107528-31; adult males) ". . . Body stripes absent or in- conspicuous and body with many con- spicuous light spots which tend to be greenish yellow above and pale yellow on sides. Venter as in females but there seems to be a tendency for the colors, especially the orange, to be brighter (but the belly of one is bluish white). Iris light brown."
Geographic Variation. — Sample lo- calities are indicated in figure 2. Of the characters analyzed by STP, the follow- ing show no among-sample statistical significance (ranges of sample means in parentheses ) : Total infralabials ( 9.8- 10.2), total supraoculars (8.0-8.1), COF values for females (2.0), GOR (males, 282.5-295.8; females, 288.5-318.2), and GAB/GOR (males, 0.50-0.51; females, 0.47-0.50). As previously stated, A. ameiva from eastern Panama differ from individuals from western Panama in their retention of a distinct vertebral stripe as adults. Specimens collected west of a line between Penonome, Code, and the mouth of the Rio Estan- cia, Code, lack the stripe as adults.
On the basis of STP results, the sample from Isla San Jose in the Bay of Panama (Sample 4) can be distin- guished from other samples in three characters (within sample ranges fol- lowed by sample means in parenthe-
ses): Total supralabials 14-15 (14.4), other samples 11-16 (12.3-12.8); COF values for males 2-6 (4.5), other sam- ples 2 (2.0); COP values 8-12 (9.2), other samples 8-15 (10.0-12.0). Addi- tionally, the Isla San Jose sample has the largest sample means for the fol- lowing characters, although the sample does not differ statistically from main- land samples for these characters: total supraoculars, GAB, GOR for females, GAB/GOR for males, terminal preanals for females, and transverse rows of ven- trals. The sample values were lowest among the localities compared for GAB/ GOR for females, femoral pores, ven- trals at midbody and COP values. With respect to the total number of supra- labials, the significant difference be- tween sample 4 and other samples is probably a function of the few large samples available for analysis. That in- clusion of other samples in the compari- son might alter the picture is suggested by values obtained for this character from small, mainland samples and from small samples from other islands in the Bay of Panama. Values intermediate between that obtained for Isla San Jose and mainland samples used in STP analysis were obtained from Isla Taboga (mean = 13.0, n = 5); Rio Mamoni, Panama (13.6, 10); vicinity of Nata, Code (13.6, 5); Rio Chorcha east of Chiriqui, Chiriqui (13.8, 5); and Isla San Miguel (14.0, 6). A sample of five specimens from Tres Puntas, Herrera, has a mean of 14.6, higher than that of the Isla San Jose sample for total num- ber of supralabials.
Samples 5 and 6 from the Azuero Peninsula can be distinguished statisti- cally as a unit from other samples for COP values, 12 (12.0) and 10-13 (11.9), respectively; other samples 8-15 (9.2- 10.5). These samples possess the great- est (although not statistically signifi- cant) mean values for total number of femoral pores and number of subdigital lamellae.
On the mainland of Panama, distinct
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
19
geographic trends are apparent only for the number of femoral pores, number of termhial preanals of males and COP values. All of these characters show in- creasing values from east to west. Values obtained for small samples not used in STP analysis indicate the same trends. Other characters demonstrate no geographic trends although certain localities differ from others statistically. Ecological Notes. — Ameiva ameiva is characteristic of open areas in Pana- ma. It is common in the savannas of the lower Pacific slopes and in cleared areas across the Canal Zone. Cochran ( 1946 ) reported it in the same habitat on Isla San Jose. It has adapted well to the presence of man and is common in resi- dential parts of Panama City. Heatwole (1966) reported A. ameiva to be absent in the Darien Gap and noted that the Panamanian population of the species is disjunct from South American popula- tions. He postulated that the species was once continuously distributed over the isthmus but that post-Columbian reforestation isolated the Panamanian population. Heatwole further demon- strated that A. ameiva is presently ex- panding its range eastward in Panama by moving into areas recently cleared by man.
Remarks. — It is difficult to explain the presence of two geographically dis- tinct color types (presence or absence of a reddish vertebral stripe in adults) of Ameiva ameiva in Panama. Present environmental conditions are similar east and west of the Penonome-Rio Estancia line. Bennett ( 1968 ) provided a clue to solution of the problem with his suggestion that the eruption of Vol- can El Valle, which may have occurred during the Pleistocene (Terry, 1956), was an event of biogeographic im- portance. The ejecta from El Valle may have covered many square miles and could have divided the Panamanian population of Ameiva ameiva for suf- ficient time for divergence of color pat- tern to occur. Detailed ecological in-
vestigations of A. ameiva might yield infomiation on this point, as well as on the origins of east-west clines for certain other characters, but no such studies have been undertaken. Even statistical studies aimed at correlating geographic variation in characters with environ- mental factors cannot be carried out until samples of adequate size for analy- sis can be accumulated from through- out the range of the species in Panama. Taxonomic recognition of segments of the Panamanian population of Ameiva ameiva would be premature until analy- sis of character variation throughout the range of the species in South America can be completed.
Two specimens of Ameiva suritia- mensis ('=A. ameiva) were listed by Boulenger (1885:353) as collected in Costa Rica by "Mr. Geale." There are no records at the British Museum that these lizards (BMNH 80.6.21.1-2) were collected by Ceale himself, although the museum bought a number of specimens from him between 1866 and 1880 (Mr. A. F. Stimson, pers. comm. ) . The speci- mens are indeed A. ameiva and can be referred to the subspecies praesignis which ranges through northwestern Venezuela, northeastern Colombia and Panama, but it is doubtful that they were collected in Costa Rica. Both specimens possess relatively well-de- fined middorsal stripes, suggesting that they may have come from eastern Panama, if they were collected in Cen- tral America at all. There is presently no suitable habitat for A. ameiva in Costa Rica with the possible exception of the dry Guanacaste region where the species is not now known.
Barbour and Noble ( 1915 ) examined specimens of Ameiva ameiva ostensibly from Acapulco, Mexico. Re-examina- tion of the specimens (MCZ 2728 and 2730) confirms the identification, but the locality information is obviously erroneous.
Yunker and Radovsky ( 1966 ) and Brennan and Yunker ( 1966 ) reported
20
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
10'
8'
50 100
I -I
KILOMETERS
Fig. 5. Map showing locality records of Ameiva Jeptophnjs. To avoid crowding of symbols, some localities have been omitted. Samples used in statistical analyses were ( 1 ) Osa Peninsula, Punta- renas Prov., Costa Rica, (2) Palmar-Golfito Region, Puntarenas Prov., Costa Rica, (3) Barro Colo- rado Island, Canal Zone, Panama, and (4) Eastern Darien Prov., Panama.
the acarinids Dracomjsus belgicae Yunker and Radovsky, Eutrombicula alfreddugesi (Oudemans) and E. goeldi (Oudemans) parasitic on Ameiva hi- frontata from Panama. Examination of the series of Uzards from which these parasites were taken proves them to be A. ameiva.
Ameiva leptophrys Cope
Amiva leptophrijs Cope, Proc. Amer. Phil. Soc, 31:341-342, 1893 [Holotype: Apparently lost; given by Cope as "No. 318." Type Locality: "Buenos Ayres" (= Buenos Aires, Puntarenas Prov., Costa Rica)].
Ameiva leptophrys — Dunn, Proc. Acad. Nat. Sci. Philadelphia, 92:114, 1940. Stuart, Proc. Biol. Soc. Washington, 55:146, 1942. Taylor, Univ. Kansas Sci. Bull., 38(1): 257-260, 1956.
Ameiva nithveni Barbour and Noble, Bull. Mus. Comp. Zool., 59(6) :471-473, 1915 [Holotype: MCZ 10927. Type Locality: near Panama City. Collector: W. W. Brown, Jr.]. Burt and Burt, Trans. Acad. Sci. St. Louis, 28(1 ):53, 1933.
Ameiva festiva — Wettstein, Sitzungsli. Akad. Wiss. Wien, math.-natiuw. Kb, Abt. I, 143. Bd., 1.U.2 Heft. p. 30, 1934 (part).
Distribution. — Forested areas of Pacific slopes from Parrita, Puntarenas
Prov., and San Isidro del General, San Jose Prov., Costa Rica to Jaque, Darien Prov., and the Rio Mono, Darien Prov., within a few kilometers of the Colom- bian border in Panama; across the Canal Zone and east along the Atlantic slopes at least as far as the Rio Sasardi, San Bias Prov., Panama (Fig. 5). The species probably occurs in extreme northwestern Colombia.
Diagnosis. — Ameiva leptophrys can be distinguished from its Middle Amer- ican congeners by the following com- bination of characters: Separation of parietal and frontoparietal scales from the interparietal by one or more rows of small, irregular scales; abrupt reduction in size of small scales immediately pos- terior to the enlarged, median gular scales; large number of femoral pores (mean number for males, 49.1; females, 44.7); a transverse row of abruptly en- larged mesoptychial scales; eight longi- tudinal rows of ventral scutes; color pattern emphasizing longitudinal stripes rather than spotting, but lacking a light vertebral stripe.
Description. — Maximum observed
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
21
Fig. 6. Head and preanal scutellation of Arneiva Icptophrijs: (A) Lateral view of head, KU 96827, (BJ Dorsal view of head, KU 96827, (C) Ventral view of head, KU 96827. X 1.2. (D) Preanal region, KU 96815. X 1.
SVL 133.0 for males, 129.0 for females; supralabials 11-15 (12.5) for males, 12- 15 (12.7) for females; infralabials 10-12 (10.1); supraoculars 6-8 (6.2); COF values 2-6 (3.5); COP values 5-10 (8.1); SO-SC values 2-10 (5.9); loreals 2-6 (3.7); nostril in prenasal-postnasal suture; prefrontal usually not in contact with postnasal (96.3%, N = 136); GAB 127-182 (153.8); PV 28-63 (41.0); PV/ GAB 0.19-0.35 (0.26); GOR 182-285 (235.6); GAB /GOR 0.54-0.74 (0.66); SAT 18-25 (21.0); total preanal Scales 4-10 (6.6) for males, 6-10 (8.1) for fe- males; terminal preanal Scales 1-3 (2.1) for males, 1-3 (1.9) for females; one row of enlarged prebrachials; postbra- chials not enlarged to moderately en- larged, irregularly arranged or in one or two rows; preantebrachials enlarged, in
two rows proximally, one row distally; 3 (usually) or 4 rows of enlarged pre- femoral scales; two rows of enlarged infratibials; femoral pores 42-61 (49.1) for males, 34-53 (44.7) for females; sub- digital lamellae 24-35 (28.1); longi- tudinal rows of ventral scutes 26-31 (28.8); transverse rows of ventral scutes 8-10 (8.1). For details of scutellation see figure 6.
Individual variation in dorsal head scutellation is extreme (Fig. 7). There is no predictable pattern to scutellation posterior to the frontal, and even this scale may be divided. However, usually two frontoparietal and three parietal scales can be recognized. The degree of variation in posterior head scales is unequalled by other species of Middle American Ameiva, although it is ap- proached in the closely related A. fes- tiva.
Fig. 7. Variation of dorsal head scutellation of Ameiva leptophrys. (A) KU 96820, X 2, (B) KU 96826, X 2.2, (C) KU 96814, X 2.3, (D) KU 95577, x 3.8, (E) KU 25641, x 4.1.
22
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
Color and Pattern in Alcohol. — Juve- nile males: Top of head light brown; rostral and anterior part of frontonasal pearl-white with a slight greenish tinge; side of head dark brown above grading to gray-brown below; area between orbit and ear brown with darker brown postorbital bar dorsally; white crescent beneath and slightly behind eye; chin, infralabials, gular region and throat im- maculate. Middorsal stripe broad, slightly irridescent, gray-brown, bor- dered laterally by white paravertebral stripes originating on superciliary scales and extending well onto tail; paraverte- bral stripes (and edges of middorsal stripe) undulating on posterior three- quarters of body; velvet-black dorso- lateral stripe from above shoulder to insertion of hind legs, bordered below by broken white line originating on shoulder and extending to insertion of hind limbs and onto anterior surface of femur; area below this lateral light stripe brown, grading to gray-brown above ventral scutes. Venter immacu- late. Dorsal surfaces of fore and hind limbs dark brown with lighter brown blotches, spots or mottling; no post- femoral light stripe; ventral surfaces of limbs immaculate to gray-white. Tail brown with black Hecks above, brown laterally (below dorsolateral light stripe) blending to white or gray- white ventrally; short postanal white stripe bounded by brown below for about six scales, then joining white ventral pat- tern.
Through ontogeny ( Fig. 8 ) the mar- gins of the middorsal stripe become more undulating and the paravertebral stripes are lost except where the mid- dorsal stripe remains broadest. Gray- brown or dark brown bars or large spots appear laterally overlaying the dorso- lateral and ventrolateral dark fields. In some large adult males the lateral light stripes are lost entirely, and the venter becomes blue-white or coppery from chin to vent and onto the ventral sur- face of the tail.
Females: Juvenile females are simi- lar to juvenile males. Through ontogeny ( Fig. 8 ) the middorsal stripe often be- comes interrupted by narrow, dark brown cross-bars. The adult pattern of females is similar to that of adult males, faint to relatively distinct bars appear- ing in the dorsolateral and ventrolateral dark fields. However, some females lack lateral bars, and the lateral light stripes are retained throughout life.
Color and Pattern in Life. — The dor- sal coloration in life of Ameiva lepto- phry.s does not differ markedly from that seen in preserved specimens. The middorsal stripe is metallic brown and may have a greenish tint. Paravertebral and lateral light stripes are yellowish to yellowish-gray, and lateral bars or blotches, if present, are yellowish-gray to rust colored. Ventral coloration is variable. The entire ventral surface may be white, cream or copper colored in juvenile males and females. Adults of both sexes may retain these colors, but some adult males become pale blue over the entire ventral surface, or they may have greenish-yellow, orange or brick red on chin, gular region and throat with the ventral surfaces of body, limbs and tail blue. Some adult males exhibit a variation of the latter color pattern in which the hind limbs pos- terior to the femoral pores and the ventral surface of the tail is copper- colored. Variability in ventral colora-
FiG. 8. Ontogenetic Change and Variation in Pattern of Anieiva Jeptophnjs. Males: (A) KU 107555, SVL 85 mm from N slope Cerro Cambutal, Los Santos Prov., Panama; (B) KU 76142, 115 mm from Barro Colorado Island, Canal Zone, Panama; (C) KU 95582, 118 mm from 6 km W Palmar Norte, Puntarenas Prov., Costa Rica (included for comparison with KU 76142). Females: (D) KU 95578, 44 mm from 6 km W Palmar Norte, Puntarenas Prov., Costa Rica; (E) KU 76144, 89 mm from Barro Colorado Island, Canal Zone, Panama; (F) KU 76163, 112 mm from El Valle, Code Prov., Panama. Small males are similar to small females as exemplified by KU 95578 ( D above ) .
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
23
24
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
tion might be correlated with reproduc- tive condition in these Hzards, but this aspect of their biology has not been investigated. It \\'ould be instructive to retain specimens in captivity in order to ascertain not only when color changes occur, but whether or not greenish-yel- low and orange pigmentation represent stages leading to brick red coloration.
Geographic Variation. — Any analysis of geographic variation of characters of Ameiva leptophnjs must be considered preliminary, pending the acquisition of additional large samples of specimens from throughout the range of the species. Only four such samples have been available for this study (Fig. 5). There is a noticeable lack of specimens from western Panama, and the follow- ing results should be viewed with both the paucity of large series of specimens and the lack of good geographic repre- sentation in mind.
The following characters exhibit no significant geographic differences among sample means (ranges of sample means in parentheses): Supralabials (males, 12.1-12.9; females, 12.4-13.2), infra- labials (10.0-10.3), SO-SC values (5.2- 6.3) and terminal preanal scales of males (1.9-2.3).
Samples 1 and 2 (Costa Rica) are significantly different from samples 3 and 4 (Panama) with respect to the following characters (locality number followed by sample mean in parenthe- ses): Number of subdigital lamellae (Sample 1, 29.2; 2, 29.4; 3, 26.5; 4, 27.2), paravertebral granules (1, 45.9; 2, 48.9;
3, 38.1; 4, 37.4) and PV/GAB (1, 0.28; 2, 0.30; 3, 0.25; 4, 0.25). Samples 1, 2 and 3 differ significantly from sample 4, but not from each other, in mean COF values (1, 7.4; 2, 7.3; 3, 8.0; 4, 8.7). One character, total number of femoral pores, shows a significant difference be- tween samples 1 and 2 ( 1, 43.6; 2, 48.4 ) . In addition to COF values, mentioned above, sample 3 differs significantly from sample 4 in GAB/GOR (3, 0.68;
4, 0.65), but sample 4 is not significantly
different from samples 1 and 2 (both 0.66). There are a number of charac- ters for which mean values of either sample 1 or 2 differ significantly from that of sample 3, but not from sample 4, and a few (COP, GAB, Longitudinal Rows of Ventrals) for which mean values of sample 1 and/ or sample 2 are not significantly different from sample 3, but do differ from the mean value for sample 4.
East-west clines are apparent for a number of characters. In terms of mean values, there is a decrease from east to west ( infralabials, longitudinal rows of ventrals) or an increase from east to west (COF values, terminal preanals of females, total preanals of females). If samples 1 and 2 are combined (it is legitimate to do so according to STP results), additional characters show the same clinal tendencies. An increase in mean value from east to west is noted for number of supralabials of males, number of supraoculars, GAB, PV and PV/GAB; decreasing mean values from east to west are noted for COP values. A small sample (n =r 10) of leptophnjs from near the Rio Sasardi in San Bias Prov., Panama, was collected by C. W. Myers and arrived too late to be in- cluded in the STP analysis. Mean values calculated for each character for the Rio Sasardi sample are in accord with the east-west trends discussed above for (Rio Sasardi means in paren- these): COF values (2.8), total pre- anal scales of females (8.4), number of supralabials of males (12.2) and COP values (9.1). The Rio Sasardi sample runs counter to east-west trends estab- lished on the basis of mean values of samples 1-4 for number of terminal pre- anals of females (2.0), number of in- fralabials (10.0), longitudinal rows of ventrals (28.7), number of supraoculars (6.3), GAB (160.1) and PV/GAB (0.24). These results once again call attention to the necessity of obtaining additional samples from intermediate localities before attempting a really
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
25
definitive analysis of geographic varia- tion of leptophrys.
With the possible exception of color pattern, none of the characters not ana- lyzed by STP appears to vary geo- graphically. Male leptophrys from cen- tral and eastern Panama tend to have more emphasis on spotting (rather than bars) in the lateral dark fields then do specimens from western Panama and Costa Rica. With the acquisition of more specimens from the hiatus in west and central Panama this difference may also prove to be clinal. In addition, specimens from the Osa Peninsula of Costa Rica ( sample 1 ) appear darker in overall coloration than those collected at other localities, including those from the adjacent non-peninsular mainland of Costa Rica. Since my sample from the Osa Peninsula was collected over a short period of time and by the same collectors, it may be that the darkness of the specimens can be attributed to preservation in warm formalin or to an unusually long period of time in fomia- lin before transfer to ethyl alcohol.
Ecological Notes. — Amewa lepto- phrys is a forest species; at various lo- calities throughout its range it is found sympatrically with all other species of Middle American Ameiva, except chait- zami. Hillman (1969) studied niche specificity of leptophrys, festiva and quadrilineata near Rincon on the Osa Peninsula of Costa Rica. He found that adult leptophrys forage deeper into the forest than its sympatric congeners; the latter search for food in more open areas. Roth leptophrys and festiva seem to utilize the same areas for basking, but there is little overlap with the more heliophilous quadrilineata. Hillman found hatchling leptophrys occupying more open areas than the adults but virtually identical to areas utilized by adult festiva. Presumably the size dif- ference between juvenile leptophrys and adult festiva acts to lessen direct com- petition. In the vicinity of Palmar Norte, Puntarenas Prov., Costa Rica,
festiva is not abundant, and I found leptophrys and quadrilineata more closely associated ecologically than either is to festiva. The two species were seen basking within a few meters of one another, but whereas quadriline- ata utilized clearings, leptophrys basked in the sun-flecked areas beneath the dense, low vegetation. Ameiva lepto- phrys would allow approach to within a meter before bolting when so hidden. Tree-fall clearings in a swamp forest near Palmar Norte were occupied by juvenile quadrilineata and leptophrys, but I never found the young of both species utilizing the same clearing. Dunn (1940a) found leptophrys and festiva sympatric on Rarro Colorado Island but reported that festiva oc- cupied the more shaded parts of the habitat, whereas leptophrys utilized more open areas, the opposite of the situation in Costa Rica. Charles W. Myers (personal communication) found leptophrys and ameiva sympatric at Guanico Arriba, Los Santos Prov., Pana- ma. There leptophrys occupies dry, partly disturbed forest and ameiva is common in open pastures. Recently col- lected material from near Parrita, Pun- tarenas Prov., Costa Rica, confirms the sympatry of leptophrys and undulata hesitantly reported by Dunn (in Stuart, 1942).
Remarks. — Rarbour and Loveridge (1929b: 213) listed two paratypes of Ameiva ruthveni (^=A. leptophri/s): MCZ 10925-10926. Rarbour and Noble (1915:471-472) acknowledged examina- tion of only two specimens in their de- scription of A. ruthveni: MCZ 9931 (re- catalogued MCZ 10927; E. E. Williams, in. litt.), the holotype, and MCZ 9932 (recatalogued MCZ 10926). The third hzard, MCZ 10925 (originally MCZ 9928), is not a paratvpe. One paratype (MCZ 10926) was not located.
Brennan and Yunker (1966:235-236, 258) reported Ameiva undulata from Panama as host for the acarinids Eu- tromhicula alfreddugesi (Oudemans)
26
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
and E. goeldi (Oudemans). Examina- tion of the specimens of Ameiva prove them to be leptophrys. There are no vahd records of A. undiihta from Panama.
Ameiva festiva (Lichtenstein and Von Martens)
Cnemidophorus festivus Lichtenstein and Von Martens, Nomenclator Reptilium et Am- philiiorum Musei Zoologici Berolinensis, p. 13, 1856 [Lectotype, herein designated: ZMB 881a; Paralectotype, formerly syn- type: ZMB 881b. Type Loca]it>-: "Vera- gua" = Veragiias, Panama].
Ai7iewa eutropia Cope, Proc. Acad. Nat. Sci. Philadelphia, 14:62, 1862 [Syntypes: ANSP 9071, USNM 4320(3). Type lo- cality: "Region of the Triiando, New Gre- nada" (=Colombia). Collector: Arthur Schott].
Aiyjeiva edwardsi Bocourt, Ann. Sci. Nat. 17 (17): 1-2, 1873 [Holotype: MNHN 5480. Type Locality: Izabel and Santa Maria de Panzos near Rio Polochic, Guatemala. Col- lector : unknown] .
Ameiva festivus — Bocourt, Mission Scientifique au Mexique et dans TAmerique Centrale; fitudes sur les reptiles, livr. 3, pp. 260-263, pi. 20A, Fig. 10; pi. 20D, Fig. 3; pi. 20, Fig. 2; 1874.
Ameiva festiva — Giinther, Biologia Centrali- Americana; Reptilia and Batrachia, p. 24, 1885. Boulenger, Catalogue of the Lizards in the British Museum (Natural History), vol. 2, p. 347, 1885. Barbour and Noble, Bull. Mus. Comp. Zool., 59(6):473-4, 1915. Burt and Burt, Bull. American Mus. Nat. Hist., 61(5) :309, 1931. Burt and Burt, Trans. Acad. Sci. St. Louis, 28(1): 53, 1933. Wettstein, Sitzungsb. Akad. Wiss. Wien, math.-naturw. Kl., Abt. I, 143. Bd., l.u. 2 Heft., p. 30, 1934.
Ameiva festiva festiva — Dunn, Proc. Acad. Nat. Sci. Philadelphia, 92:114, 1940. Tay- lor, Univ. Kansas Sci. Bull., 38(1):260- 264, 1956. Ameiva festiva edwardsi — Stuart, Occ. Pap. Mus. Zool., Univ. Michigan, No. 471, p. 21, Figs. 6-7, 1943. Smith and Laufe, Univ. Kansas Sci. Bull. 31(2):64-66, 1945. Smith and Taylor, U. S. Nafl. Mus. Bull, 199:174, 1950. Stuart, Misc. Publ. Mus. Zool., Univ. Michigan, No. 122, p. 77, 1963. Ameiva festiva occidentalis Taylor, Univ. Kan- sas Sci. Bull., 38(l):260-264, 1956 [Holo- type: FMNH 120236. Type Locality: 8.0 km E San Isidro del General, San Jose Prov., Costa Rica. Collector: Richard C. Taylor] .
Distribution. — Ameiva festiva is found in the forests of the Caribbean lowlands from Tabasco, Mexico to Depto. Boyaca, Colombia and on the Pacific lowlands from the vicinity of San Isidro del General, San Jose Prov., Costa Rica, at least to the Condoto- Andagoya area in the Intendencia de Choco, Colombia. The species also is known from the vicinity of Tilaran, Guanacaste Prov., Costa Rica, just west of the continental divide (Fig. 9).
Diagnosis. — Ameiva festiva can be distinguished from other Middle Amer- ican Ameiva by the following combina- tion of characters: Central gular scales much enlarged, irregular in arrange- ment; gradual reduction in size of scales radiating outward from enlarged, cen- tral gular scales; many femoral pores (mean for males, 40.8; females, 40.0); a transverse row of abruptly enlarged mesoptychial scales; eight longitudinal rows of ventral scutes; a distinct, light colored vertebral stripe present in all but large adults. No other species of Ameiva from Middle America possesses such a well-defined vertebral stripe, al- though several South American Ameiva do have such a stripe (edracantha, or- cesi, bridgesi, and septemlineata).
Description. — Maximum obsei"ved SVL 114.0 mm. for males, 129 mm. for females; supralabials 12-18 (14.2) for males, 12-17 (13.9) for females; infra- labials 8-15 (10.8); sublabials 4-9 (4.7) for males, 4-7 (4.5) for females; supra- oculars 4-10 (6.5); COF values 2-6 (2.6); COP values 2-14 (9.4); loreals 2; nostril in prenasal-postnasal suture; prefrontals in contact with postnasal (41.1%) or no (54.5%), 4.5 percent in contact on one side but not the other (n = 448); GAB 145-221 (182.5) for males, 143-213 (179.8) for females; VS 5-29 (12.8) for males, 4-27 (11.2) for females; VS/GAB 0.02-0.14 (0.06); COR 209-324 (261.1) for males, 215- 312 (259.5) for females; GAB/GOR 0.57-0.85 (0.70); SAT 16-22 (18.3); total preanal scales 4-10 (6.6) for males,
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
27
Fig. 9. Map showing locality records of Aineiva festiva. To avoid crowding of symbols, some localities haxe been omitted. Samples used in statistical analyses were ( 1 ) Piedras Negras, Depto. El Peten, Guatemala, (2) Vicinity of Chinaja, Depto. Alta Verapaz and Depto. El Peten, Guate- mala, (3) Finca Chama, Depto. Alta Verapaz, Guatemala, (4) North Coast Localities, Honduras, (5) Bonanza, Depto. Zelaya, Nicaragua, (6) Cukra and Kanawa, Depto. Zelaya, Nicaragua, (7) Rio Mico and vicinity, Depto. Zelaya, Nicaragua, (8) Tuli Creek, Depto. Rio San Juan, Nicaragua, (9) Puerto Viejo, Heredia Prov., Costa Rica, (10) Turrialba, Cartago Prov., Costa Rica, (11) Tortuguero, Limon Prov., Costa Rica, (12) Los Diamantes and vicinity, Limon Prov., Costa Rica, (13) Almirante, Bocas del Toro Prov., Panama, (14) Isla Escudo de Veraguas, Panama, (15) Barro Colorado Island, Canal Zone, Panama, (16) Rio Mono, Darien Prov., Panama, (17) Golfito Region, Puntarenas Prov., Costa Rica, (18) Vicinity of Rincon, Osa Peninsula, Puntarenas Prov., Costa Rica, (19) Rio Ure and vicinity, Depto. Cordoba, Colombia, (20) Condoto-Andagoya-Tado
Area, Intendencia de Choco, Colombia.
4-9 (6.4) for females; terminal preanal scales 1-3 (2.0) for males, 1-3 (2.2) for females; one row of enlarged prebra- chials; postbrachials enlarged, in a single short row; preantebrachials en- larged, two rows proximally, one dis- tally; three rows of enlarged prefemoral scales; two rows of enlarged tibial scales; femoral pores 33-55 (40.8) for males, 32-50 (40.0) for females; sub-
digital lamellae 23-34 (28.4); longitu- dinal rows of ventral scutes 25-33 (27.8); transverse rows of ventral scutes 6-8 (8.0).
Individual variation in posterior head scutellation is great, second only to that of Ameiva leptophrijs, but two frontoparietal and three parietal scales are usually recognizable (Fig. 10).
Color and Pattern in Alcohol. —
28
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
Fig. 10. Head and preanal scutellation of Ameiva festiva (ZMB 881a; Lectotype): (A) Lateral view of head, ( B ) Dorsal view of head, (C) Ventral view of head. X 1.5. (D) Preanal region. X 1-6.
There is extensive geographic variation in color and pattern over the range of Ameiva festiva. As a basis for discuss- ing this variation, the following com- plete description of color and pattern of specimens from Puerto Viejo, Here- dia Prov., Costa Rica, is presented.
Juvenile males: Blue-gray vertebral stripe 5-13 granules wide at midbody extending from rostral scale well onto tail, the margins slightly scalloped on the posterior three-fourths of body; top of head lateral to vertebral stripe dark gray; side of head dark gray above be- coming lighter below, anterior to eye; ear encircled by interrupted white line; chin, infralabials, gular region and throat light gray or white; dorsal sur- face of body very dark brown lateral to vertebral stripe; dorsolateral light stripe white, narrow and interrupted, originat- ing on canthus and extending onto tail; dorsolateral dark stripe velvet black,
beginning immediately posterior to eye and extending onto tail, bordered below by white, narrow, interrupted ventro- lateral light stripe originating beneath eye, passing over ear to the groin, thence onto anterior surface of femur at juncture of dorsal granular scales and enlarged ventral scales; vertical bars ex- tending ventrally from ventrolateral light stripe immediately anterior and posterior to forelimbs; ventrolateral field dark brown with irregular blue- gray spots or narrow, regularly spaced blue-gray bars, or both. Ventral sur- faces of body, limbs and tail gray-white. Dorsal surface of forelimbs and femur dark brown to black; shank dark brown with blue-gray spots above.
Through ontogeny (Fig. 11) the vertebral stripe becomes sinuous and less well defined. It may be lost entirely on the head, posterior part of the body and tail. A light-colored rostral scale is retained, as usually is the vertebral stripe on the anterior third of the body. The head becomes uniformly brown above and the pale markings above and below the eye, on the side of the neck and around the ear are lost. The velvet- black dorsolateral dark stripe is re- tained. In larger individuals, the mid- dorsal area immediately lateral to the position occupied by the vertebral stripe becomes regularly marked with dark and light brown blotches, the darker occupying positions where the sinuous or scalloped margin of the ver- tebral stripe curved inwards toward the midline; this gives the back a ladder- like pattern. The dorsal surface of the tail becomes imiformly brown or brown with darker brown blotches which may be paired or nearly so. The dorsolateral and ventrolateral light stripes are re- tained posteriorly only. Most males be- come blue-gray overall ventrally, the lateral ventral scutes and enlarged fem- oral scales marked with black. The in- fralabials may also be flecked with black.
Females: The color and pattern of
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
29
Fig. 11. Ontogenetic Pattern Change of Ameiva f estiva from the Atlantic
Slopes of Costa Rica. Males: (A) AMNH 89210, SVL 43 mm; (B) AMNH
89191, 80 mm; (C) AMNH 89203, 106 mm. Females: (D) AMNH 89212,
47 mm; (E) AMNH 89204, 69 mm; (F) AMNH 89193, 90 mm.
30
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
juvenile females is similar to that of juvenile males. Females differ from males in that the juvenile color and pat- tern is retained with little fading in all but very large adults and adult females do not develop blue colors ventrally.
Color and Pattern in Life. — The fol- lowing notes refer to specimens col- lected by me on 1 August 1965 at Puerto Viejo, Heredia Pro v., Costa Rica. Adult male (KU 95558), SVL 99 mm: Neck and chest reddish-orange; chin and gular region pale blue; belly pale blue medially and pale blue broken by black blotches laterally; middorsal area brown with indications of a faint yellow verte- bral stripe extending from occiput to base of tail; dorsolateral dark field black, bordered above and below by narrow, interrupted yellow stripes; ven- trolateral field olive with narrow yellow bars. Adult female (KU 95559), SVL 79 mm: Venter pale bronze; dorsal pat- tern similar to KU 95558 except verte- bral stripe more pronounced and ex- tending from rostral well onto tail; yel- low on head and body becoming pale blue on tail.
Juvenile Ameiva festiva from near Bluefields, Depto. Zelaya, Nicaragua (KU 101214-101215) possess vertebral stripes which are yellow anteriorly be- coming blue-green on the tail. The dor- solateral and ventrolateral light stripes are yellow to yellowish-orange. An adult female (KU 101213) from the same locality has the same color and pattern as KU 95559 from Puerto Viejo.
As in other species of Ameiva I have examined, gular coloration varies in adult males. Collections are not such that this can be correlated with season. I suspect the color to be indicative of reproductive state, at least in males, but lack evidence to substantiate this hy- pothesis. Field notes taken by W. E. Duellman, C. W. Myers, and myself, as well as color photographs provided by D. H. Janzen disclose that the gular region and chest of adult male festiva may be uniformly blue (if so, the entire
venter is blue), copper, dull brick-red, yellow-orange, golden-yellow or yellow- green (all commonly associated with a blue venter, although individuals hav- ing a copper gular region often have the entire venter that color). There seems to be no geographic pattern associated with these colors or none that I can de- tect from the field notes available to me. The ventral surfaces of females may be uniformly copper or cream (often with a pinkish tint), or may be one of these colors with the gular region and chest gray.
Geographic Variation. — Only the number of transverse rows of ventral scutes and the number of terminal pre- anals of males show no significant dif- ferences among sample means. A num- ber of characters exhibit variation among samples, but show no marked geographic trends. These characters are: GOR, GAB, number of femoral pores, number of longitudinal rows of ventrals, total number of preanal scales and SAT (although SAT values from samples in the Atlantic lowlands may tend to be higher than values from Pacific samples).
A striking aspect of geographic vari- ation among samples of festiva is the apparent "center" of variation for cer- tain characters in the Caribbean low- lands of Costa Rica and Nicaragua. Values for VS, VS/GAB, and number of sublabials of males gradually increase in samples to the north and east of these lowlands. A decrease in number of supralabials and in COF values is ap- parent in the same directions.
A clinal increase in mean values from north (Guatemala) to southeast ( Panama and Colombia ) is expressed for GAB/GOR, but samples 17 (Golfito Region of Costa Rica) and 18 (Osa Peninsula of Costa Rica) have high means, nearer to values noted for Sam- ples 16 (Rio Mono, Darien Prov., Pana- ma) and 20 ( Condoto-Andagoya-Tado Area, Intendencia de Choco, Colombia) to the south and east. The mean value
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
31
for sample 19 ( Rio Ure and vicinity, Depto. Cordoba, Colombia) is lower than would be predicted by the trend. A clinal decrease from north to south- east is expressed in number of infra- labials ( Samples 1 and 19 have lower and higher mean values, respectively, than anticipated by the trend), number of sublabials of females (although high values are noted for samples 14 and 20 ) , COP values (locality 20 has a very low mean and represents a separate subset) and the number of subdigital lamellae. As is evident, these clinal tendencies are never perfectly expressed: one or more samples disrupt the trend. Samples 19 and 20 (in South America) counter in some respects geographic trends estab- lished in the more northerly parts of the range of festiva. These two localities are separated by high mountains and differ from one another in a statistically significant way for a number of charac- ters. Samples 15 (Barro Colorado Is- land) and 16 (Rio Mono, Darien Prov., Panama) both show greater similarities to Sample 20 than to Sample 19. Sam- ples 17 and 18, in southeastern Costa Rica, show similarities to Sample 19 for some characters, to sample 20 for others. It would be instructive to have addi- tional samples from the Pacific slopes of western Panama in order to see whether there is further evidence for the "Panamanian crossover" suggested by Dunn (1940b).
The more northerly samples of fes- tiva (Samples 1-4) differ from others in a number of characters, none of which set these samples apart statistically from all others. The samples from Guatemala (Samples 1-3) and Honduras (Sample 4), taken as a group, differ from adja- cent samples in a statistically significant way for number of supralabials of males and females, VS of males (data from Sample 4 unavailable) and total pre- anal scales of females.
As alluded to earlier, there is marked geographic variation in the color and pattern over the range of festiva. There
are four distinct pattern types with variation to a lesser degree within two of these. The four patterns are dis- cussed separately below relative to the description of specimens from Puerto Viejo, Costa Rica, given above.
Pattern Type A: The color and pat- tern described for festiva from Puerto Viejo is characteristic in major aspects of specimens collected in the Caribbean lowlands of Costa Rica, western Pana- ma, and northward into Mexico. Liz- ards comprising Samples 1-7 and pos- sibly 8 (Nicaragua northward) tend to be browner in overall coloration than those from Costa Rica and Panama (Samples 9-13), although this is a sub- jective observation and not based upon quantitative evidence. The first impres- sion one gets upon seeing living festiva from Costa Rica and Panama is one of a black lizard with a prominent vertebral stripe. The lizards from Honduras and Guatemala have a broader vertebral stripe than others with Pattern Type A; this is especially striking in specimens from Guatemala.
Pattern Type B: A second widely distributed pattern type, largely allo- patric to Pattern Type A, is peculiar to lizards from the Pacific slopes of south- eastern Costa Rica and western Pana- ma, the San Bias Coast of Panama and Colombia east of the Cordillera Central (Samples 15-19). Specimens from this area are characterized by a relatively broader vertebral stripe than that of other pattern types and by having the ventrolateral light stripe either lacking, present or present only posteriorly. The vertebral stripe broadens posteriorly and is broadest in specimens from eastern Panama and Colombia. Lizards from the San Bias Coast of Panama and from east of the Cordillera Central in Colom- bia possess vertebral stripes that may become as broad as the middorsal area at the base of the tail (Fig. 12B). Ameiva festiva from the Osa Peninsula of Costa Rica (Sample 18) are darker in overall coloration than those from the
32
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
Fio. 12. Patterns of Ameiva f estiva from Colombia. (A) FMNH 165748, male, SVL 44 mm, from the Upper Rio Ure, Depto. Cordoba. (B) FMNH 165227, male, 101 mm, from the Upper Rio Ure, Depto. Cordoba. (C) FMNH 43828, female, 81 mm, from Pizzaro, Intendencia de Choco (?).
adjacent mainland. Thi.s is not the re- sult of preservation, for color photo- graphs taken by Daniel H. Janzen be- fore the lizards were preserved show them to be darkly pigmented. Two specimens of f estiva ( KU 80210 and KU 110745) from Almirante, Bocas del Toro Prov., Panama, exhibit a pattern similar to that described here, although others from the same locality show Pattern Type A. Ameiva festiva from Barro Colorado Island and southeastern Pana- ma are somewhat intermediate between Pattern Types A and B, but because the population more closely resembles the latter, I have included them here.
Pattern Type C: Ameiva festiva from the Choco of Colombia (including Sample 20) have a pattern similar to Type A, but the vertebral stripe is ex- tremely straight and well defined even in large specimens ( Fig. 12C ) . The dorsolateral and ventrolateral light stripes are also well defined, may be entire or dotted (rarely dashed), and are not lost in large individuals. Pre- served specimens are, with the excep-
tions of the light stripes and vertebral stripe, very dark.
Pattern Type D: Adult Ameiva fes- tiva from Isla Escudo de Veraguas, off the northwest coast of Panama, are melanistic. That this is so is supported by color photographs of freshly col- lected specimens provided by Charles W. Myers. Very little pattern is ap- parent on specimens I have examined and, although examination of juvenile specimens (unavailable to me) might prove them to belong to Pattern Type A or B, I prefer to assign them to a separate pattern type for the present. Baskin and Williams (1966) summa- rized information on insular populations of melanistic Ameiva in the Caribbean, and Mayr (1963) discussed similar populations of Lacerta in the Mediter- ranean. Such melanistic island popula- tions are usually found on small, diy, barren islands ( Mayr, 1963 ) ; Heatwole and Torres ( 1966 ) attributed the phe- nomenon in Ameiva to local diftcrentia- tion, those species of Ameiva exhibiting melanism not being closely related.
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
33
Mayr ( 1963 ) suggested that the dark pigmentation might function in heat regulation. The sample from Isla Es- cudo de Veraguas docs not fit the pat- tern in temis of habitat as outlined above. According to C. W. Myers (per- sonal communication), one of the few biologists to have visited the island, Isla Escudo de Veraguas ". . . is covered with humid forest, including swamp forest, except for a well developed line of strand vegetation at some points." Specimens of festiva from the Bocas del Toro Archipelago to the west of Isla Escudo de Veraguas are not notably mclanistic and conform to Pattern Type A, although, according to Myers, habi- tats on all of these islands are similar. That the melanism of festiva on Isla Escudo de Veraguas is the result of local differentiation can hardly be denied, but the adaptive significance of such pigmentation remains a mystery.
Ecological Notes. — In the northern part of its range, Ameiva festiva in- habits heavy forest (Stuart, 1943; Smith and Laufe, 1946; Echternacht, 1968) when sympatric with other macroteiids. If other large teiids are not present, the niche of the species is expanded to in- clude more open habitats (Echternacht, 1968). In Mexico, Guatemala, and Honduras, festiva is partly sympatric with undulata and with various species of Cnemidophorus. In Costa Rica and Panama, festiva and either leptophnjs or quadrilineata, or all three species may occur together. In the vicinity of Almirante and on the islands of the Bocas del Toro Archipelago, Bocas del Toro Pro v., Panama, festiva and quad- rilineata are sympatric, as they are on Isla Escudo de Veraguas. Ameiva fes- tiva typically occupies the more shaded parts of the habitat, whereas quadriline- ata occurs in the open (C. W. Myers, personal communication). In southeast- ern Costa Rica (vicinity of San Isidro del General, San Jose Prov.; vicinity of Palmar Sur and on the Oso Peninsula, Puntarenas Prov.) festiva is sympatric
with both quadrilineata and leptophnjs. Ameiva festiva occupies a habitat in- termediate to those of the other two species: leptophnjs occupies more shaded areas and quadrilineata, the open areas. This has been noted by Hillman (1969) and myself, but Dunn ( 1940a ) reported the habitats of lepto- phnjs and festiva reversed on Barro Colorado Island in Panama; there fes- tiva occupies the less shaded areas. Hillman ( 1969 ) found hatchling A. lep- tophnjs sympatric with adult festiva on the Osa Peninsula and concluded that the difference in size between the two minimizes competition. The reproduc- tive biology of festiva has been investi- gated by R. E. Smith (1968a, b) in the only such study of Ameiva that has been undertaken. Smith's results suggest that festiva is reproductively active through- out the year, and that clutch sizes aver- age between 2.22 and 2.29 eggs.
Nothing has been reported on the ecology of A. festiva in South America.
Remarks. — Ameiva festiva was de- scribed by Lichtenstein and von Mar- tens ( 1856 ) on the basis of two speci- mens, which are catalogued under one number (ZMB 881) in the Zoological Museum of Berlin. The description is based on a specimen "Long, a rostro ad caud. bas. 3 1/2", caud. ultra 6"" (88.9 mm). The syntypes presently have snout- vent lengths of 86 and 58 mm; the tail of the larger is broken. I am herein designating the larger syntype as the lectotype; shrinkage resulting from over one hundred years in preservative would account for the discrepancy in snout- vent length between that given in the description and my measurement. The smaller syntype becomes a paralecto- tyi^e.
Dunn (1943) described Ameiva fes- tiva nicefori from Saisima, Depto. Cun- dinamarca, Colombia, but I have pre- sented evidence elsewhere ( Echter- nacht, 1970) that the population de- serves specific recognition. Dunn (1940a) and Villa (1968) allied the population
34
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
Fig. 13. Map showing locality records of Ameiva quadrilineata. To avoid crowding of symbols, some localities have been omitted. Samples used in statistical analyses were ( 1 ) Southeastern Nica- ragua, (2) Tortuguero, Limon Prov., Costa Rica, (3) Limon and vicinity, Limon Prov., Costa Rica, (4) Bocas del Toro Prov., Panama, (5) Isla Escudo de Veraguas, Panama, (6) San Isidro del General, San Jose Prov., Costa Rica, (7) Palmar Sur and vicinity, Puntarenas Prov., Costa Rica, (8) Vicinity of Rincon, Osa Peninsula, Puntarenas Prov., Costa Rica, (9) Progreso, Puerto Ar- muelles and vicinity, Chiriqui Prov., Panama, (10) Boquete, Chiriqui Prov., Panama.
of Ameiva on the Islas del Maiz, Nica- ragua with undulata rather than festiva as originally described (Barbour and Loveridge, 1929a). My investigations support this conclusion, and the popu- lation is discussed with undulata in the present work.
Ameiva quadrilineata (Hallowell)
Cnemidophorus quadrilineatus Hallowell, Proc. Acad. Nat. Sci. Philadelphia, 1860, p. 483, 1860 [Syntypes: USNM 6053A, 6053B. Type Locality: "Nicaragua," restricted to Greytown, Depto. Rio San Juan, by Tay- lor, Univ. Kansas Sci. Bull., 38(1):271, 1956].
Ameiva quadrilineata — Cope, Proc. Acad. Nat. Sci. Philadelphia, 1862, p. 61, 1862. Bo- court, Mission Scientifique au Mexique et dans TAmerique Centrale, fitudes sur les Reptiles, livr. 4, p. 259, pi. 20A, Fig. 9, 1874. Giinther, Biologia Centrali-Ameri- cana, Reptilia and Batrachia, p. 24, 1885. Dunn, Proc. Acad. Nat. Sci. Philadelphia, 92:113, 1940. Taylor, Univ. Kansas Sci. Bull., 38(1):271, 1956.
Ameiva <^ahhiana Cope, J. Acad. Nat. Sci. Philadelphia, ser. 2, 8:33, pi. 28, Fig. 3, 1875 [Syntypes: USNM 32614-16. Type Locality: Old Harbor (= Puerto Viejo), Limon Prov., Costa Rica. Collector: W. M. Gabb]. Giinther, Biologia Centrali-Ameri- cana, Reptilia and Batrachia, p. 23, 1885.
Ameiva undulata — Boulenger, Catalogue of the Lizards in the British Museum (Na-
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
35
Fig. 14. Head scutellation of Ameiva quad- lilineata: (A) Lateral view of head, (B) Dor- sal view of head, (C) Ventral view of head; KU 95593 from 1.5 km E Palmar Norte, Puntarenas Prov., Costa Rica. X 2. (D) Dor- sal view of head; KU 96849 from Cayo Zapatillo Grande, Bocas del Toro Prov., Panama, X 2.7.
tiiral History), 2nd Ed. 1885 (part).
vol. 2, p. 348,
Distribution. — Atlantic coastal low- lands from extreme southeastern Nica- ragua to northwestern Panama; Pacific slopes of Costa Rica from Quepos, Pun- tarenas Prov. (H. S. Fitch, personal communication) to the Azuero Penin- sula, Panama; the Bocas del Toro Archi- pelago, Isla Escudo de Veraguas and Isla de Cebaco of Panama and Isla del Cafia of Costa Rica. Specimens have been collected at elevations of 1100- 1650 m above Paso Ancho and on Cerro Hornito, Chiriqui Prov., Panama (Fig. 13). _
Diagnosis. — A small lizard which can be distinguished from other Middle American Ameiva by the following com- bination of characters: Central gular scales moderately enlarged, irregular,
hJ^^S
Fig. 15. Preanal region of Ameiva quadrili- ncata: (A) KU 95636 from 6 km W Palmar Norte, Puntarenas Prov., Costa Rica, X 2. (B) KU 95593 from 1.5 km E Palmar Norte, Puntarenas Prov., Costa Rica, X 2.
scales of posterior one-half of gular re- gion abruptly smaller than scales of anterior one-half (Figs. 14-15); moder- ate number of femoral pores (mean number for males, 34.2; females, 27.8); a transverse row of abruptly enlarged mesoptychial scales; eight longitudinal rows of ventral scutes; narrow middor- sal field (mean PV: 26.1) which does not become darker laterally; continu- ous, well defined dorsolateral stripe and continuous to interrupted ventrolateral stripe, both usually retained throughout life. The middorsal field of Ameiva un- diilata is broad (mean PV: males, 46.8; females, 45.6) and the ventrolateral light stripe is usually interrupted. Ameiva chaitzami also has a relatively broad middorsal field (mean PV: 38.1) and interrupted ventrolateral light stripes and, in addition, lacks a uni- formly colored middorsal field.
Description. — Maximum observed SVL 88 mm for males, 82 mm for fe- males; supralabials 11-16 (13.7); infra- labials 10-13 (10.1); supraoculars 6-9 (6.7); COF values 2-6 (3.0); COP values 4-12 (9.1); loreals 2-4; nostril in prenasal-postnasal suture; prefrontals in contact with postnasals; GAB 116-177 (140.5); PV 16-36 (26.1); PV/GAB 0.09-0.25 (0.19); GOR 185-258 (218.6); GAB/GOR 0.41-0.78 (0.64) for males, 0.56-0.75 (0.64) for females; SAT 17-22 (19.8) for males, 17-21 (19.5) for fe- males; total preanal scales 4-10 (7.1) for males, 4-9 (6.8) for females; termi- nal preanal scales 1-3 (1.4) for males,
36
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
37
1-3 (1.3) for females; prebrachials en- larged, in a single row; postbrachials moderately enlarged, in one row or ir- regular; preantebrachials in three rows proximally, one or two rows distally; three rows of enlarged prefemoral scales; infratibials enlarged, in two rows; femoral pores 22-42 (34.2) for males, 18-39 (27.8) for females; sub- digital lamellae 25-34 (29.5) for males, 25-34 (28.8) for females; longitudinal rows of ventral scutes 28-33 (29.8); transverse rows of ventral scutes 4-10 (8.0).
Color and Pattern in Alcohol. — Juve- nile males: Top of head light brown to gray-brown; side of head gray with black wash above grading to gray on lower half of postnasals, loreals and supralabials; middorsal field dark brown to dark gray-brown, narrow, bounded laterally by narrow, white dorsolateral light stripes, which originate above the eyes and extend to the base of the tail where they coalesce and form a verte- bral stripe extending well onto the tail; dorsolateral dark fields black; ventro- lateral light stripe narrow, white, orig- inating beneath eye and extending to the groin, thence onto anterior surface of femur where it runs between the dor- sal granular scales and the ventral plates; short, narrow, interrupted white
<1
Fig. 16. Ontogenetic change and variation in pattern of male Ameiva quadrilineata. A-C; Pacific population. D-G; Atlantic population. Females do not show marked ontogenetic changes in color or pattern and resemble juvenile males as to these characters. (A) KU 95623, SVL 31 mm, from the Rio Zapote, 8 km E Palmar Norte, Puntarenas Prov., Costa Rica; (B) KU 95583, 68 mm, from 8 km SE San Isidro del General, San Jose Prov., Costa Rica; (C) KU 93984, 84 mm, from 2-5 km ESE Piedras Blancas, Puntarenas Prov., Costa Rica; (D) KU 96837, 40 mm, from Isla Basti- mentos, Bocas del Toro Prov., Panama; (E) KU 108282, 62 mm, from E end Isla Escudo de Veraguas, Panama; (F) KU 96845, 79 mm, from Cayo Zapatillo, Bocas del Toro Prov., Panama; (G) Ventral view of KU 96845 to show dark pigmentation.
stripe from beneath ear to shoulder; ventrolateral dark field black with brown mottling below; dorsal surfaces of forelimbs mottled brown and black, an interrupted white line extending along the back of the arm from shoulder to wrist; dorsal surface of hind limbs mottled black and brown with white spots on shank and narrow, white post- femoral stripe extending onto tail for ca. 10-12 scale whorls; ventral surfaces immaculate except for faint gray mark- ings in the gular region and a faint, transverse gray line across the gular re- gion at the juncture of the large anterior and small posterior scales.
Through ontogeny (Fig. 16) black mottling appears in the middorsal area, but never enough to disrupt the visual impression of a unicolored field. Gray or blue-gray reticulations appear in the ventrolateral dark field, this pattern ex- tending onto the lateral-most enlarged ventral scutes and the enlarged ventral femoral scales. The dorsolateral dark field remains black above, but brown vertical bars or reticulations appears in the lower half. The white spots on the hind limbs and the short stripe between ear and shoulder disappear, the side of the neck becoming uniformly gray- brown or black with gray reticulations. Ventral surfaces may be pale blue over- all or the chin, gular region and throat may be salmon colored, the remainder of the venter pale blue. There may be black flecks in the midventral region (see discussion of geographic varia- tion ) .
Females: Juvenile females resemble juvenile males. Adult females tend to retain the juvenile pattern except that black flecking appears in the middorsal field and small, brown, longitudinally oriented flecks and blotches may appear in the dorsolateral dark field (Fig. 16).
There is some geographic variation in color and pattern and the foregoing description applies to specimens from the Pacific lowlands of Costa Rica.
Color and Pattern in Life. — ^The fol-
38
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
lowing description is extracted from the field notes of C. W. Myers and refers to a series of Ameiva quadriUneata (KU 96853-96865) collected 13 km N Puerto Armuelles, Chiriqui Prov., Panama, on 22 April 1965. Ameiva quadriUneata from this locality do not differ in color and pattern from individuals found in the Pacific lowlands of Costa Rica, the latter having been used for the descrip- tion of color and pattern in alcohol. The following account (modified from Myers' field notes) sei-ves to illustrate the variability of ventral pigmentation. Dorsolateral and lateral stripes vivid yellow in small individuals and some- what duller in larger ones; dorsolateral stripes nearly disappearing in largest males. Middorsal region light yellow- brown or gray-brown, with or without irregular black spotting, and turning to a whitish-gray or pale brown dorsal stripe on basal one-third of tail. Venter of tail base copper, but all but basal part of tail is uniformly brown or gray above and below (in a juvenile male this tail is light blue-gray). Dorsolat- eral area of body (between yellow stripes) bright rust, variably suffused with black (in some nearly uniformly black with rust spots). Flanks light brown or light gray-brown, heavily spotted with black. Ventral color vari- able, as follows: KU 96853-5—3 juve- niles ( at least two are males ) : Ventral surfaces pale copper. KU 96856-9 — 4 females: Ventral surfaces pale copper, perhaps with a yellow tinge on throat. KU 96860-1—2 males: Underside of head pale blue; rest of venter light blue, except tail base and thighs behind fem- oral pores, which are light copper. KU 96862-3—2 males: Underside of head light coppery orange; rest of venter light blue, except for copper thighs be- hind femoral pores, and light blue and copper-mottled tail base. KU 96864-5 — 2 males: Underside of head red-orange (pale in smaller specimen and deep in larger) with an occasional speck or spot of blue showing through; rest of venter
light blue, except rear of thighs, and tail base (copper).
In addition to the ventral colors given above, both males and females of all sizes may have immaculate or pearly- white ventral surfaces. The chin, gular region and throat may be various shades of yellow.
Geographic Variation. — The range of Ameiva quadriUneata is divided by the high mountains of the Cordillera de Talamanca of Costa Rica and Panama. Although the resulting Atlantic and Pa- cific populations do not show statisti- cally significant differences for any char- acters, it is apparent that divergence has taken place. Mean values for num- ber of femoral pores of females and number of terminal preanals of both sexes are higher and lower, respectively, on the Atlantic slope than on the Pacif- ic. The mean number of terminal pre- anal scales of females from Boquete, Chiriqui Prov., Panama (Sample 10), is statistically different from mean values of all other samples and constitutes a separate subset. Several other charac- ters show similar trends, but with one sample running counter to the trend established by the others. Characters having lower values along the Atlantic slope are GAB, GAB/GOR for both sexes, and total number of preanal scales for both sexes. The sample from Boquete, which is intermediate in geo- graphic position to Puerto Armuelles (Pacific slope) and Almirante (Atlantic slope) in Panama, has intermediate mean values to those of Puerto Armu- elles and Almirante for some characters and more closely resembles samples from the Atlantic slope than those from the Pacific slope for other characters. Boquete lies on the Pacific slope of the Cordillera de Talamanca but a few kilometers from the continental divide in western Panama. Sample 10 has in- termediate values for number of supra- labials, PV, and SAT of males. Mean values for number of supraoculars, GAB, GAB/GOR of males and females,
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
39
and total number of preanals of both sexes are nearer mean values for Atlan- tic samples than for Pacific samples. Occasionally, other samples on the Pacific slope resemble the aggregate of Atlantic slope samples in terms of mean values of one character or another, but none with such noticeable frequency as does Sample 10.
Samples 8, 9 and 10 show a statis- tical relationship to one another for sev- eral characters and constitute a separate subset for COF and COP values. Sam- ples 8 and 9 on a separate subset for GAB, having very high means for this character. For GAB, Sample 10 has the lowest observed mean value and is simi- lar to Atlantic slope samples in this. Sample 6 has the lowest observed mean PV and constitutes a separate subset.
General clinal tendencies are noted for some characters. The total number of preanal scales decreases from north to south on the Atlantic slopes, but in- creases from north to south in samples from the Pacific slopes (except Sample 10, which has a very low mean); COP values decrease from north to south on the Pacific side; COF values increase from north to south on the Pacific slope; the number of subdigital lamellae of females decreases from north to south on the Atlantic slope. Inclusion of more samples in the analysis is necessary in order to substantiate these trends.
Mean values for total number of in- fralabials and for numbers of rows of longitudinal and transverse ventral scutes show no significant differences among the ten samples.
Some geographic variation is appar- ent in the color and pattern of Ameiva quadrilineata. Although specimens from Atlantic and Pacific slope popula- tions do not differ markedly among themselves, there are differences be- tween the two populations. Adult males from the Atlantic side are never as ornate in terms of dorsal and lateral color and pattern as those from the Pacific slopes (Fig. 16). Males from
the Atlantic slopes, especially those from northwestern Panama (including Samples 4 and 5) have all ventral sur- faces variably flecked with black (Fig. 16); the dorsolateral light stripes of specimens from this area may be scal- loped over the posterior third of the body and may become diffuse, blending with middorsal coloration, in large in- dividuals; the lateral light stripe may be interrupted and the dorsal caudal stripe is less evident; the lateral ground color of some adults is black mottled with brown above, gray below. In gen- eral, it seems that the chin, gular region and throat of adult male quadrilineata from the Atlantic slopes becomes yel- low when brightly pigmented, but that such lizards from the Pacific slopes show red or orange coloration. I know of only one exception to this: KU 95600 collected at Quebrada Boruca at the Rio Puente Nuevo, 22 km E Palmar Norte, Puntarenas Pro v., Costa Rica, on 15 August, 1965, was yellow on the ventral surface of the head. This gener- ality needs to be substantiated; I have had no color-in-life notes on Atlantic slope specimens from west of the Bocas del Toro region of Panama.
Ameiva quadrilineata from Isla Es- cudo de Veraguas of Panama and from the Osa Peninsula of Costa Rica do not appear to be any darker in coloration than those from the adjacent mainlands.
Ecological Remarks. — Ameiva quad- rilineata is ecologically the best known member of the genus ( Hirth, 1963, 1965; R. E. Smith, 1968a, b). Both Hirth and Smith worked with populations of quad- rilineata on the Atlantic lowlands, Costa Rica. Hirth (1963) studied all aspects of the ecology of A. quadrilineata at Tortuguero, Limon Prov., Costa Rica. At Tortuguero these lizards were abun- dant on the beaches, which they shared with Basiliscus vittatus. Ameiva quad- rilineata is also characteristic of the beach fauna on the islands of the Bocas del Toro Archipelago and Isla Escudo de Veraguas of Panama (C. W. Myers,
40
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
personal communication) and near the northern (western) Hmit of its range on the Pacific coast of Costa Rica ( H. S. Fitch, personal communication). I have collected it in open areas (clearings in forested areas, along roadsides, etc.) near San Isidro del General and Palmar Norte in southeastern Costa Rica. Dvmn ( 1940a ) reported the species to be com- mon in open areas along large rivers on the Atlantic side of Costa Rica. Hill- man (1969:477), who studied quadrili- neata on the Osa Peninsula of Costa Rica, found them foraging "in low vege- tation adjacent to an exposed surface such as a road." The species is defi- nitely more heliophilic than either fes- tiva or leptophrys, with which it is sympatric in some areas. It is similar in habitat preference to ameiva. The rela- tively small size of adult qiiadrilineata undoubtedly helps to reduce competi- tion between it and its larger, sympatric congeners. Competition may be rela- tively more intense between juvenile qiiadrilineata and juvenile leptophrys or festiva. I have no experience with festiva on this point, but I found juve- niles of the other two species common in tree-fall clearings in a swamp forest near Palmar Norte in Costa Rica. The clearings were always occupied by one of the two, but not both species.
R. E. Smith (1968a, 1968b) studied Ameiva qiiadrilineata and festiva at Pandora, Limon Pro v., Costa Rica. His investigations indicate that both species may be reproductively active through- out the year. The mean clutch size for qiiadrilineata is 2.05-2.07 eggs, depend- ing upon method of counting, and ovi- position occurs between 17 and 21 days following ovulation. Smith also dis- cussed the adaptive significance of fat bodies and provided experimental evi- dence for a relationship between fat body development and gonadal cycles.
Ameiva are usually diurnal, but W. E. Duellman (personal communica- tion) found an adult male quadrilineata active at night on a road in western
Panama. There had been a heavy rain earlier in the day, and possibly the liz- ard was flooded from its burrow.
Henry S. Fitch (personal communi- cation) is studying a population of quadrilineata near Quepos, Puntarenas Pro v., Costa Rica. The nearest locality from which specimens of undulata are available is near Parrita, about 25 km to the northwest, but Fitch has observed both species at Quepos. There quad- rilineata occurs only on the beach, and undulata is inland in plantations or pas- tures where shade and ground cover are abundant. Quepos is the northern-most locality for quadrilineata on the Pacific coast, and the southern-most locality for undulata.
Remarks. — Barbour and Noble ( 1915 ) apparently confused Ameiva quadrilineata and A. undulata pulchra. Both were described by Hallowell ( 1860 ) from "Nicaragua," but quad- rilineata is restricted to the Atlantic coast of Nicaragua and pulchra to the Pacific coast; Barbour and Noble had before them specimens from "Chinan- dega, Nicaragua," which is on the Pacif- ic side. Moreover, the color description given by Barbour and Noble fits that of undulata, not quadrilineata.
Ameiva undulata (Wiegmann)
Cnetnidophorus undulatus Wiegmann, Herpe- tologica Mexicana, pp. 27-28, 1834 [Syn- types: ZMB 867-869, 869 missing. Type Locality: "Mexico," restricted to Tehiian- tepec, Oaxaca, Mexico by Hobart M. Smith, Proc. Biol. Soc. Washington, 53:56, 1940. Collector: Unknown].
Ameiva undulatus — Gray, Catalogue of the lizards in the British Museum ( Natural History), p. 20, 1845.
Ameiva undulata — Cope, Proc. Acad. Nat. Sci. Philadelphia, 14:62-63, 1862. Bocourt, Mission Scientifique an Mexique et dans TAmerique Centrale; fitudes sur les rep- tiles, livr. 3, pp. 254-259, pi. 20A, Fig. 7a- 7e, pi. 20B, Fig. 1. Gunther, Biologia Cen- trali-Americana; Reptilia and Batrachia, pp. 23-24, 1885. Boulenger, Catalogue of the lizards in the British Museum ( Natural History), vol. 2, pp. 347-348, 1885. Hart- weg and Oliver, Occas. Papers Mus. Zool., Univ. Michigan, 359:7-8, 1937. Dunn,
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
41
»
Proc. Acad. Nat. Sci. Philadelphia, 92:114- 115, 1940.
Ameiva pulchra Hallowell, Proc. Acad. Nat. Sci. Philadelphia, p. 483, 1860 [Syntypes: ANSP 9133-9134. Type Locality: "Nica- ragua." Collector: Unknown].
Cnemidophorus amivoides Cope, Proc. Acad. Nat. Sci. Philadelphia, pp. 198-199, 1894 [Holotype: AMNH 16316. Type Locality: La Carpintera, San Jose Prov., Costa Rica. Collector: A. Alfaro].
Ameiva undulaia undulata — Barbour and Noble, Bull. Mus. Comp. Zool, 59(6):474- 475, 1915. Stuart, Proc. Biol. Soc. Wash- ington, 55:145, 1942. Smith and Laufe, Univ. Kansas Sci. Bull., 31(2):62-64, Fig. 2b, 1946. Smith and Taylor, U. S. Natl. Mus. Bull., 199:174, 1950.
Ameiva undulata parva Barbour and Noble, Bull. Mus. Comp. Zool., 59(6) :476-477, 1915 [Holotype: MCZ 5831. Type Locali- ty: "Guatemala," restricted to Mazatenan- go, Depto. Suchitepequez by Smith and Laufe, Univ. Kansas Sci. Bull., 31(2) :51, 1946. Collector: Unknown]. Stuart, Proc. Biol. Soc. Washington, 55:145, 1942. Smith and Laufe, LTniv. Kansas Sci. Bull., 31(2):51-54, Fig. la. Pi. la, 1946. Smith and Taylor, U. S. Natl. Mus. Bull., 199: 173, 1950. Taylor, Univ. Kansas Sci. Bull., 38(1):26S-271, 1956. Stuart, Misc. Publ. Mus. Zool., Univ. Michigan, 122:77, 1963.
Ameiva undulata pulchra — Dunn, Proc. Acad. Nat. Sci. Philadelphia, 92:115, 1940. Stuart, Proc. Biol. Soc. Washington, 55: 146, 1942.
Aineiva undulata haitwegi Smith, Proc. Biol. Soc. Washington, 53:55, 1940 [Holotype: USNM 108600. Type Locality: Across the Rio Usumacinta from Piedras Negras, Depto. El Peten, Guatemala. Collector: H. M. Smith]. Stuart, Proc. Biol. Soc. Washington, 55:145, 1942. Smith and Laufe, Univ. Kansas Sci. Bull., 31(2):35- 36, Pi. lib, 1946. Smith and Taylor, U. S. Natl. Mus. Bull., 199:171-172, 1950. Stuart, Misc. Publ. Mus. Zool., Univ. Mich- igan, 122:77, 1963.
Ameiva undulata stuarti Smith, Proc. Biol. Soc. Washington, 53:55-56, 1940 [Holotype: USNM 108601. Type Locality: Palenque, Chiapas, Mexico. Collector: H. M. Smith]. Stuart, Proc. Biol. Soc. Washington, 55: 146, 1942. Smith and Laufe, Univ. Kansas Sci. Bull., 31(2):50-51, Fig. lb, Pi. lb, 1946. Smith and Tavlor, U. S. Natl. Mus. Bull., 199:173, 1950.
Ameiva undulata amphigramma Smith and Laufe, Trans. Kansas Acad. Sci., 48(3): 338-344, 1945 [Holotype: FMNH 100020. Type Locality: San Andres Tuxtla, Vera- cruz, Mexico. Collector: H. M. Smith]. Smith and Laufe, Univ. Kansas Sci. Bull.,
31(2):43-47, PI. Ic-Id, 1946. Smith and Tavlor, U. S. Natl. Mus. Bull., 199-172, 19.50.
Ameiva undulata gaigeae Smith and Laufe, Univ. Kansas Sci. Bull., 31(2):37-39, Fig. Ic, Pi. lie, 1946 [Holotype: FMNH 100030. Type Locality: Progreso, Yucu- tan, Mexico. Collector: H. M. Smith]. Smith and Taylor, U. S. Natl. Mus. Bull., 199:172, 1950. Stuart, Misc. Publ. Mus. Zool., Univ. Michigan, 122:77, 1963.
Ameiva undulata podarga Smith and Laufe, Univ. Kansas Sci. Bull., 31(2):40-43, Fig. Id, 2a, 1946 [Holotype: FMNH 100050. Type Locality: 7 mi. W Victoria, Tamau- lipas, Mexico. Collectors: H. M. Smith and D. H. Dimkle]. Smith and Taylor, U. S. Natl. Mus. Bull., 199:172, 1950.
Ameiva imdidata dextra Smith and Laufe, Univ. Kansas Sci. Bull., 31(2) :54-,59, Fig. 2c, 1946 [Holotype: FMNH 100080. Type Locality: nr. Rincon, Guerrero, Mexico. Collectors: E. H. Taylor and H. M. Smith]. Smith and Taylor, U. S. Natl. Mus. Bull., 199:173, 1950.
Ameiva undulata sinistra Smith and Laufe, Univ. Kansas Sci. Bull., 31(2):59-62, 1946 [Holotype: FMNH 100054. Type Locali- ty: Manzanillo, Colima, Mexico. Collec- tor: H. M. Smith]. Smith and Taylor, U. S. Natl. Mus. Bull., 199:174, 19.50.
Ameiva festiva miadis Loveridge and Barbour, Bull. Mus. Comp. Zool., 69(7) :141, 1929 [Holotype: MCZ 26970. Type Locality: Isla del Maiz Grande, Depto. Zelaya, Nica- ragua. Collector: James L. Peters].
Ameiva undulata miadis — Dunn, Proc. Acad. Nat. Sci. Philadelphia, 92:115, 1940. Villa, Rev. Biol. Trop., 15(1):119, 1968. Echter- nacht, Breviora, 354:5, 1970.
Distribution. — The most widely clis- tiibuted of the Middle American Amei- va (Fig. 17), Ameiva umJidata occm-s from near Las Varas, Nayarit, Mexico, to the vicinity of Quepos, Puntarenas Prov., Costa Rica, on the Pacific slopes. On the Atlantic slopes, the species ranges from Rancho Santa Ana (12.8 km SE Podilla), Tamaulipas, Mexico, to Puerto Lempira, Cabo Gracias a Dios, Honduras, and on more westerly Atlan- tic slopes of Nicaragua from Corozo, Depto. Nuevo Segovia to Tuli Creek, Depto. Rio San Juan. Four specimens (ANSP 1543S-41) from Huaunta, Depto. Zelaya, Nicaragua, seemingly document the presence of Ameiva undulata on the Caribbean coast of that country (see
42
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
24^
2
8^
5°
2'
9'
£>.
-24^
21
27-.a^^S^° •34, ,' W
0
200 400
KILOMETERS
105^
99^
Fig. 17. Map showing locality records of Ameiva undulata. To avoid crowding of symbols, some localities have been omitted. Samples used in statistical analyses were ( 1 ) Gomez Farias and vicin- ity, Tamanlipas, Mexico, (2) Tamazunchale and vicinity, San Luis Potosi, Mexico, (3) Tierra Colorado, Veracruz, Mexico, (4) Cuautlapam, Veracruz, Mexico, (5) San Andres Tuxtla and vicinity, Veracruz, Mexico, (6) Vicinity of Jesus Carran?a, Veracruz, Mexico, (7) Frontera and
(8) Tres Brazos and Balchacaj, Campeche, Me.xico, (9) Dzilbalchen,
Piste, Yucatan, Mexico, (11) El Cayo and vicinity, Cayo Dist., British
and Tikal, Depto. El Peten, Guatemala, (13) Piedras Negras, Depto.
Canihor, Depto. Alta Verapaz, Guatemala, (15) El Rancho and Finca
vicinity. Tabasco, Mexico, Campeche, Mexico, ( 10 ) Honduras, ( 12 ) Uaxactiin El Peten, Guatemala, (14)
Bucural, Depto. Progreso, Guatemala, (16) Copan, Depto. Colon, Honduras, (17) La Ceiba, Depto. Atlantida, Honduras, (18) Coyoles Central, Depto. Yoro, Honduras, (19) Trujillo, Depto. Colon, Honduras, (20) Catacamas, Depto. Olancho, Honduras, (21 ) Isla del Maiz Grande, Depto. Zelaya, Nicaragua, (22) Manzanillo-Pacific Coast Localities, Colima, Mexico, (23) Colima and vicinity, Colima, Mexico, (24) Acapulco and vicinity, Guerrero, Mexico, (25) Chilpancingo and vicinity, Guerrero, Mexico, (26) Chacalapa and vicinity, Oaxaca, Me.xico, (27) Tres Cruces, Oaxaca, Mexico, (28) Tolocito, Oaxaca, Mexico, (29) Tapanatepec, Oaxaca, Mexico, (30) Vicinity of Cintalapa, Chiapas, Mexico, (31) Las Tazas, Chiapas, Mexico (51 km E Altamirano), (33) Esquintla, Chiapas, Mexico, (34) Sabana de San Quintin, Chiapas, Mexico, (35) Panajachel, Depto. Solola, Guatemala, (36) Finca La Trinidad, Depto. Jutiapa, Guatemala, (37) JuHapa and Finca Mongoy, Depto. Jutiapa, Guatemala, (38) Vicinity of La Libertad, Depto. La Liber- tad, El Salvador, (39) Chinandega, Depto. Chinandega, Nicaragua, (40) Matagalpa, Depto. Malagalpa, Nicaragua, (41) Hacienda La Norma and vicinity, Guanacaste Prov., Costa Rica, (42) Canas, Guanacaste Prov., Costa Rica, (43) Tilaran, Guanacaste Prov., Costa Rica, (44) El Zamo- rano and vicinity, Depto. Francisco-Morazan, Hondmas, (46) Isla Mujeres, Quintana Roo, Mex- ico, (48) Apatzingan, Michoacan, Mexico. Open symbols indicate literature records of specimen-S
not examined for the present study.
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
43
Fig. 18. Head and preanal scutellation of Ameiva nndulata (KU 94112; Sabana San Quintin, Chiapas, Mexico): (A) Lateral view of head, (B) Dorsal view of head, (C) Ventral view of head, X 1.1- (D) Preanal region, X 1.5.
Remarks). The species is known from Isla Mujeres, east of Quintana Roo, Mexico, and from the Islas del Maiz east of Nicaragua. Gene flow between populations on the Atlantic and Pacific slopes occurs across the Isthmus of Te- huantepec, in the lowlands of eastern Guatemala, and in Honduras and Nica- ragua south of the highlands that mark Nuclear Central America. Ameiva tin- chiJata is generally found at elevations below 1500 m, although there are ex- ceptions (e.g., at Panajachel, Depto. Solola, Guatemala).
Diagnosis. — Ameiva umJulata can be distinguished from other Middle Ameri- can Ameiva by the following combina- tion of characters: Gentral gular scales slightly to greatly enlarged, regular or irregular in arrangement; gradual re- duction in size of scales radiating out-
ward from central gular scales; prefron- tal scales in contact with postnasal scales; three parietal scales (four if me- dian parietal divided); a transverse row of abruptly enlarged mesoptychial scales; usually eight rows of ventral scutes; a relatively broad middorsal stripe bounded by narrow light stripes which may be lost in adults; no narrow vertebral stripe; lateral pattern empha- sizes longitudinal stripes or vertical bars in adults, never spots.
Description. — Maximum observed SVL 129 mm for males, 111 mm for females; supralabials 12-18 (14.1) for males, 12-18 (14.1) for females; infra- labials 7-14 (10.2); supraoculars 6-10 (6.4); COF values 2-6 (2.2); GOP values 2-14 (10.6); SO-SG values 2-10 (3.8); loreals 2-8 (4.0); nostril in pre-
nasal-postnasal suture; contact with postnasal;
prefrontal in median gular
Fig. 19. Head and preanal scntellation of Ameiva unchdata (KU 26963; 25 km SE Jesus Carranza, Veracruz, Mexico): (A) Lateral view of head, (B) Dorsal view of head, (C) Ventral view of head, X 1.1. (D) Preanal region, X 1-5.
44
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
Fig. 20. Preanal Region of Ameiva undulata to illustrate two of the pattern types: (A) KU 87417; East of San Andreas de la Cruz, Guerrero, Mexico; Preanal scales in a single row except for paired terminal scales, X 2.7. (B) KU 38232; Chacalapa, Oaxaca, Mexico; Preanal scales in a single row including single terminal preanal, X 1.5.
scales variable in size and orientation; GAB 108-202 (149.5); PV 31-67 (47.4) for males, 33-65 (46.4) for females; PV/GAB 0.21-0.40 (0.32) for males, 0.23-0.43 (0.31) for females; GOR 186- 315 (246.0); GAB/GOR 0.44-0.77 (0.61) for males, 0.49-0.74 (0.60) for females; SAT 18-26 (22.3) for males, 18-27 (22.2) for females; total preanal scales 4-13 (7.6) for males, 5-13 (7.5) for females; terminal preanal scales 1-3 (2.0); one (usually) or two rows of en- larged prebrachials; postbrachials en- larged, in one row; preantebrachials in two rows, one row or two rows prox- imally and one row distally; prefemoral scales enlarged, three to five rows prox- imally reducing to two (usually) or three rows distally; infratibials enlarged, in two rows, femoral pores 29-48 (37.4) for males, 22-45 (35.0) for females; sub- digital lamellae 24-38 (30.2); longitu- dinal rows of ventral scutes 26-34 (30.1); transverse rows of ventral scutes 8-12 (8.2). For details of scutellation see figures 18-20.
Color and Pattern in Alcohol. — Throughout its broad range, Ameiva undulata shows geographic variation in color and pattern unmatched by other Middle American species of Ameiva and ontogenetic variation, especially of males, greater than all, except possibly A. ameiva. The following description of color and pattern in alcohol, as well as the description of color and pattern
in life, is based on specimens of undu- lata from the vicinities of La Ceiba and Trujillo along the north coast of Hon- duras (Localities 17 and 19, Fig. 21).
Juvenile Males: Top of head brown; side of head black above, grading through brown to gray on supralabials; a white patch beneath and slightly be- hind eye; ear surrounded by white markings; middorsal field yellowish- brown or greenish-brown with black flecking, bordered laterally by fine white cream-colored dorsolateral stripes which originate above eye and extend well onto tail; dorsolateral dark stripe ex- tending from behind eye to base of tail, velvet black with no light bars or blotches, bounded below by an inter- rupted white ventrolateral light line originating behind the ear and extend- ing to the groin, thence as an uninter- rupted white line well onto tail, ventro- lateral dark field gray-brown; dorsal surfaces of limbs mottled black and brown, hind limbs with white spots above; white postfemoral stripe joining caudal extension of ventrolateral light stripe; tail brown with black flecks above; ventral surfaces immaculate or blue-gray.
Through ontogeny (Fig. 21) the dorsolateral light stripe disappears or is retained only on the anterior half of the body. The ventrolateral light stripe is usually lost entirely. Light blue vertical bars appear on the flanks, extending from the border of the middorsal field or near it to or nearly to the enlarged ventral scutes. Shorter bars and blotches may appear between the primary bars. The light spots around the ear are lost and the venter may become light blue
Fig. 21. Ontogenetic change in pattern of Avieiva undulata from northern, coastal Hon- duras. Males: (A) KU 101247, SVL 71 mm; (B) KU 101238, 92 mm; (C) KU 101237, 110 mm. Females: (D) KU 101242, 39 mm; (E) JRM 2318, 88 mm; (F) JRM 2554, 109 mm. Small males resemble small females as exemplified by KU 101242 (D above).
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
45
46
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
overall except for black patches in the lateral-most scales and on the enlarged plates on the anterior surface of the femur. Traces of red pigment are visi- ble in the gular region of some large males.
Females: Juvenile females are simi- lar to juvenile males. Through ontog- eny, females generally retain the juve- nile pattern, except that the dorsolateral light stripe may be lost on the pos- terior one-half of the body and regu- larly spaced light brown blotches ap- pear in the dorsolateral dark field (Fig. 21).
Color and Pattern in Life. — The fol- lowing is based on field notes and color photographs of Ameiva tindulata col- lected during July, 1967, near La Ceiba and Trujillo on the north coast of Hon- duras.
Adult Males: Top of head gray- brown to light brown; middorsal field gray-green on neck becoming bright leaf-green on shoulder and extending one-half to two-thirds length of trunk, then becoming golden-brown on pos- terior part of body; entire middorsal field irridescent in sunlight; dorsal sur- face of tail gray-brown with black fleck- ing; middorsal field from shoulder to tail lightly to moderately blotched with black which sometimes appear paired; remnant of dorsolateral light stripe cream, if present, and only visible on anterior two-thirds of body; lateral pat- tern of seven to ten vertical blue-white or greenish-white bars which may touch middorsal field but which usually are separated from it and which taper so that they are wider than or equal to the interspaces dorsally but much narrower than interspaces ventrally; short blue bars extending upward between the prominent lateral bars, sometimes con- necting with them; lateral ground color black above (in area of dorsolateral dark field of juveniles), reddish-brown below. Some individuals retain a rem- nant of the cream-colored ventrolateral light stripe so that the dorsolateral dark
field may be defined. Dorsal surfaces of limbs brown with indistinct black mot- tling; ventral coloration of limbs, body and tail blue-white with black flecking and blotches, especially on the hind limbs and laterally on the body; chin, gular region and neck range in color from white to brick red with no indica- tion of yellow pigmentation.
Adult Females: Dorsal pattern simi- lar to that of males except that cream- colored dorsolateral light stripes usually are retained; dorsal coloration similar to that of males but several shades darker overall; ventrolateral light stripe cream, continuous, broken or broken anteriorly, entire posteriorly; dorsolateral dark field black with brown or cream blotches, each wider than or equal to width of interspaces; ventrolateral field brown or gray-brown with darker mottling; dor- sal surfaces of limbs as for males; ven- tral surfaces immaculate or with slight bluish tint laterally and on hind legs. Females lack the bright vertical bars on the flanks, never show reddish pigment on the ventral surfaces of the head and neck and never develop blue ventral coloration as markedly as do males.
Geographic Variation. — The follow- ing samples were removed during STP analysis of sexually dimorphic charac- ters because of inadequate sample sizes: For characters of males, samples 3, 9 and 48; for characters of females, sam- ples 7, 9, 23, 25, 26, 34, 38, 46 and 48. In addition, sample 3 was removed in the analysis of number of loreals of females and both 3 and 22 were re- moved in the analysis of SO-SC values of both sexes, again for reasons of inade- quate sample size.
The most obvious patterns of geo- graphic variation for meristic characters of Ameiva undulata are clinal trends along the northeast and northwest coasts of Mexico and along the Pacific coast of Middle America from the Isthmus of Tehuantepec to Nicaragua and Costa Rica ( Fig. 22 ) . In some cases one sam- ple along such a cline is "out of phase,"
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
47
interrupting what otherwise is a rather smooth trend. Such samples are indi- cated in parentheses in the following account.
Those characters showing an in- crease in sample means from Tamaulipas to the Isthmus of Tehuantepec include number of supraoculars (2), number of paravertebral granules of males, num- ber of paravertebrals of females (3), total preanal scales of males, total pre- and scales of females (4) and number of femoral pores of males. Shorter clines in the same direction, but involv- ing only samples 1 through 3 or 4 are apparent for PV/GAB of males and SAT of males (the size of sample 3 is too small for inclusion in the analysis for both characters), as well as number of femoral pores of females. The cline for total preanal scales of males and of females extends southward across the isthmus and thence eastward along the Pacific coast as far as locality 33 for males and locality 36 for females. Two characters, COF values and number loreals, show clinal increases from north to south across the isthmus and along the coast to varying degrees. For both characters, lowest mean values are for sample 5. COF values increase through samples 28 and 29 and through samples 6, 30 and 33 to 36. The number of loreals increases through samples 28, 27 and 26 and through samples 28, 29 and 30. Mean numbers of supraoculars show the same trans-isthmian pattern as COF values, but in the opposite di- rection and with the lowest value at sample 35, not 36. CAB increases from locality 5 through 28 and 29 or 6 and 30 to 33 and 36.
Decrease in mean values from Ta- maulipas to the Isthmus of Tehuantepec is notable for the number of supralab- ials of females. Shorter clines in the same direction but involving only sam- ples 1 through 3 or 4 are apparent for CAB and CAB /COR of males (sample 3 is too small for inclusion in the analy- sis of CAB /COR of males) and number
Fig. 22. Major patterns of geographic varia- tion in Ameiva ttndulata. Arrows represent geographical trends involving several charac- ters. Samples on one side of the heavy, dashed line differ from adjacent samples on the other side of the line in a statistically significant manner for five or more characters but do not differ in such a way from adjacent samples on the same side of tlie line. See text for further elal)oration.
of subdigital lamellae. Both COP and SO-SC values increase from south to north across the isthmus and a west to east and northeast increase is noted for SO-SC values.
Northwest to southeast clines are apparent along the south coast of Mex- ico also. Characters showing increased sample means from Colima to the Isth- mus of Tehuantepec or at least to the vicinity of Chacalapa, Oaxaca (Sample 26) include number of paravertebral granules of males and of females, PV/ CAB of males and of females, total pre- anal scales of females and SAT of males (inadequate sample size precludes in- clusion of sample 48 in analysis of these characters). Although the clinal trend terminates at the isthmus for the other characters mentioned, that for total pre- anals of females continues along the Pacific coast to locality 36 in Cuatemala. A short trend in the same direction is noted for number of transverse rows of ventral scutes from localities 22 and 23 through 48 to 24 and 25.
Trends along the same coast, but showing a decrease in sample means from Colima to the Isthmus of Tehuan-
48
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
tepee are noted for number of siipra- labials of males (sample 48 excluded from the analysis) and females (sam- ples 26 and 48 excluded), number of infralabials, GAB/GOR of females, number of tenninal preanals and num- ber of subdigital lamellae. These trends generally involve only coastal samples (samples 22, 24, 26, 27 and 29), but in- clude inland samples (samples 23, 25 and 48) for supralabials of males and total number of infralabials.
A number of characters show clinal tendencies for increase or decrease in sample means between Nicaragua or Costa Rica and localities 33, 37 and especially 36 to the northwest. There is a marked tendency for clines to either begin or end in the area of southeastern Mexico and southern Guatemala with no clines running through to the south- east or northwest along the coast. Num- bers of supraoculars show an increase from sample 40 (.x = 6.13) in Nicara- gua to sample 36 (x = 9.97) in Guate- mala, at which point there is an abrupt decrease in sample mean and a clinal increase from sample 33 (x = 6.28) to sample 5 ( x = 6.70 ) on the Gulf side of the Isthmus of Tehuantepec in Mexico. Samples 34 (Ghiapas) and 35 (Guate- mala) are significantly different from the nearest samples to the east in Guate- mala for this character, but sample 36 (Guatemala) is not significantly differ- ent from sample 33 (Chiapas). A simi- lar trend for increase is noted between sample 39 (x = 2.07) in Nicaragua and sample 37 (x = 2.29) in Guatemala for COF values. Sample 36 (x = 3.77), located just west of sample 37, is signifi- cantly different from it and is the end of a cline beginning at sample 5 (x = 2.13) and which progresses across the Isthmus of Tehuantepec through sam- ples 28 and 29 or samples 6, 30 and 33. Samples 33 and 36 have very high mean values for COF and are statistically iso- lated from samples to the north. Fem- oral pore means for both males and females also show increases from sam-
ple 40 to sample 37, but the trend may be traced from samples 41, 42 and 43 in Costa Rica as well. For this character in females the cline continues to sample 15 and may also be traced from the Costa Rican samples or sample 40 in Nicaragua through samples 39 and 44 to sample 16 in western Honduras. The number of supralabials of males in- creases from sample 37 north and west- ward to sample 29, and GAB/GOR for females increases from sample 37 to sample 33 and from sample 15 to sam- ple 33, both trends running to the west or northwest. Mean numbers of sub- digital lamellae increase from samples 20 and 44 through samples 38 and 37 to sample 36 on the coastal plain of Guate- mala and to sample 33 by way of sam- ples 15 and 35. Total number of pre- anal scales of females follow the same trend. These two characters are the only ones involved in clines running through the area of samples 36 and 37. Mean numbers of supralabials of fe- males increase from west to east, be- ginning at sample 30 and terminating at samples 14 and 37 by way of samples 33, 35 and 15. Sample 36 (x = 13.5), with a mean for this character lower than that of sample 30 (x ^ 13.6) is not part of the trend.
Mean numbers of paravertebral granules of males and of females and PV/GAB of females show trends for in- crease from sample 40 in Nicaragua or sample 20 in Honduras southwestward and along the Pacific coast to sample 36 in Guatemala. With respect to number of paravertebrals of females, sample 36 is significantly different from sample 37. For both number of paravertebrals of females and PV/GAB of females, the cline may be traced to either sample 36 or sample 15 in Guatemala.
Mean SO-SC values and mean num- ber of loreals show increased values from north to south along a line be- tween sample 9 on the Yucatan Penin- sula and samples 37 and 38 in south-
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
49
eastern Guatemala and El Salvador, respectively.
There seems to be a trend toward increase in mean numbers of paraverte- bral granules of females, GAB/GOR of females and total number of preanal scales of males from sample 8 in south- western Campeche through sample 10 in Yucatan to sample 46 on Isla Mujeres. There is a similar trend beginning at sample 7 in Tabasco and terminating at sample 10 for mean numbers of femoral pores of males. For all of these charac- ters, however, sample 9 in central Cam- peche ( Dzilbalchen ) is too small for meaningful comparison (n = 3) where females are concerned and such trends might owe their existences to lack of data from this locality. A clinal de- crease from sample 8 through sample 9 and 10 to sample 46 is apparent for mean number of rows of transverse ven- tral scutes. An increase along the same line is noted for mean number of loreals.
Most samples differ from adjacent samples in a statistically significant manner for at least one character. For purposes of this discussion, a character judged sexually dimorphic would count as two characters because separate con- sideration is given to males and females. If, for example, two samples were sig- nificantly different with respect to num- ber of femoral pores of males and fe- males, they would be said to differ sig- nificantly for two characters, not one. Viewing all characters at once gives one the impression that each sample is more-or-less statistically isolated from its neighbors. Although this may be true, it renders analysis and formation of meaningful generalizations extremely difficult. However, if only pairs of sam- ples differing statistically by (arbi- trarily) five or more characters are examined, some interesting patterns emerge. These are discussed below. It must be emphasized that comparisons are, in most cases, being made only be- tween adjacent samples. Two samples
adjacent to one another in Honduras may be statistically different from one another on the basis of one or more characters, but neither may differ from a sample in, for example, northern Mexico. Of course, the opposite is also possible.
The population of Ameiva iindulata on Isla del Maiz Grande (sample 21) differs from samples 42, 43 and 44 for five characters each, from samples 39 and 40 six characters each and from sample 20 for ten characters. With the exceptions of samples 40 and 41, which differ from each other for five charac- ters, none of these mainland samples differ from adjacent samples among those mentioned by more than four characters. This strongly emphasizes the divergence of the Ameiva on Isla del Maiz Grande from the nearest main- land populations, or at least from the available samples of those populations, and is indicated by the heavy, dashed line in figure 22. Although sample 20 differs from sample 21 for twice or more the number of characters separating sample 21 from other mainland samples, there is still little basis for concluding that A. undulata on the Islas del Maiz were derived from elsewhere because the Honduranian population from which sample 20 was drawn may have been modified itself through introgression of characteristics of populations to the north and west (samples 17, 18 and 19).
The latter three samples represent a population quite distinct from neigh- boring samples on the basis that none of the three differ among themselves by more than one character ( 17 and 18 dif- fer from 19 for SO-SG values; other characters show non-significant differ- ences among sample means), but all differ from other nearby samples by more than five characters (see Fig. 22). Samples 16, 20 and 44, which encircle samples 17, 18 and 19, do not differ among themselves, but all show con- siderable divergence from samples 17, 18 and 19 for meristic and, as will be
50
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
discussed later, color and pattern char- acters. Samples 17, 18 and 19 differ from sample 11 (El Cayo and vicinity, British Honduras) for only three, one and one characters, respectively. Sam- ple 16, on the other hand, differs from sample 11 for four characters, none of which are among those few which dis- tinguish sample 11 from samples 17, 18 and 19.
Samples 9-16, 31, 34, 35, 37, 38 and 46 all show statistically significant dif- ferences involving fewer than five char- acters when compared among them- selves with respect to adjacent samples. However, sample 8 differs from samples 12, 13, 31, 33 and 34 for ten, ten, eight, ten and seven characters, respectively. In addition to sample 8, sample 31 dif- fers from samples 6, 7, 30, 33 and 35 for six, seven, nine, ten and nine characters respectively. Sample 33, in addition to differing from samples 8 and 31 by more than five characters, differs from samples 34 and 35 by six and seven characters, respectively. Sample 36 dif- fers from samples 15, 35 and 37 for seven, eight and seven characters, re- spectively. Samples 8, 7, 6, 30, 33 and 36 constitute a series of localities be- ginning in the Tres Brazos-Balchacaj area of Campeche, progressing west along the Gulf coast and thence south across the Isthmus of Tehuantepec and eastward along the Pacific slopes to the vicinity of Finca La Trinidad, Depto. Jutiapa, Guatemala. These samples, as indicated above, all differ from adjacent more inland samples to the south, east or north by five or more characters. If a line is drawn to emphasize the diver- gence of these "peripheral" samples from the other, more inland, samples (Fig. 22), none of the samples falling within the curvature of the line differ from adjacent samples by as many as five characters with the exception of sample 15, which differs from sample 35 for six characters.
Sample localities from the northern Gulf region of Mexico (samples 1-5) are
isolated from sample localities along the southwest Pacific coast (samples 22-27, 48) by the Mexican Plateau. In gen- eral, however, samples on the one coast differ for five or more characters only from the most distant samples on the other coast. For example, sample 1 (Tamaulipas) differs from sample 27 (south-central Oaxaca) for seven char- acters, but differs from sample 22 (Co- lima) for only three characters. Sample 5 ( San Andres Tuxtla and vicinity, Veracruz) differs from sample 27 for onlv four characters, but differs from sample 22 for nine. Samples 1-5 and 22-27, 48 do not differ from adjacent samples on the same coast by more than two characters, with the exception of samples 25 and 26, which differ for four characters.
Sample 27 (Tres Cruces, Oaxaca), however, seems to be strongly diver- gent from adjacent samples to the east. Sample 27 differs from samples 6, 28 and 29 for ten, six and eight characters, respectively. This, coupled with the isolation afforded by the Mexican Pla- teau, is the basis for separating off the samples from the Pacific slopes of Mex- ico northwest of the Isthmus of Tehuan- tepec as shown in figure 22.
The groupings of samples set apart according to the method just described fits well with the general pattern of clinal trends discussed earlier. There is also a generally good fit by these group- ings with the placement of highland regions with elevations above 1500 meters, an elevation above which Amei- va rarely occur in Middle America.
Using the 14 samples (8, 15-21, 27, 31, 33, 35, 37 and 44) which were found to differ from adjacent samples by five or more characters, it is possible to ar- rive at some idea of which characters are most discriminatory in distinguish- ing adjacent samples. The 14 samples were compared with adjacent samples to give 68 different comparisons. It was found that in 53 per cent of these com- parisons (36 out of the possible 68),
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
51
mean GAB differed in a statistically significant manner. GAB is apparently the primary meristic indicator of sam- ple differences, followed by GOR (43 per cent of the comparisons), SO-SC values (31%), number of subdigital lamellae (31%) and COP values (29%). All other characters were involved in 26 per cent or less of the comparisons each. Two characters, number of su- pralabials of males and number of trans- verse rows of ventral scutes, were found not to distinguish any of the 68 pairs of samples compared. All five of the char- acters which seem most important in this respect do not appear together in distinguishing two adjacent samples, but they do appear in groups of 2, 3 or 4 in various comparisons. GAB and GOR, as might be expected, occur to- gether most often in distinguishing pairs of adjacent samples, appearing in 26 of the 68 comparisons (38%). GAB and number of subdigital lamellae occur to- gether in distinguishing 15 pairs among the 68 (22%), whereas GAB and SO-SC values, and GAB and COP values both occur together in 14 of 68 comparisons (20%). Other combinations of the five characters taken two at a time occur less frequently.
The condition of the preanal scales of Ameiva tindtilata can be analyzed in a number of ways. Two of these, the total number of preanals and the num- ber of terminal preanal scales have been defined and discussed with respect to STP analysis. These characters, along with a third which combines the two, warrant further discussion.
For STP analysis, the number of terminal preanal scales was coded 1, 2 or 3, indicating one, two or three scales in the terminal position. An inherent weakness in this system is that a sample mean or range gives no indication of the actual frequency distribution of single, double and triple terminal pre- anals in the sample. A sample of two individuals, one having one terminal preanal, the other having three, would
have the same mean number of terminal preanals as another sample of two in- dividuals where each animal had two terminal preanal scales. This is an over- simplification complicated by larger sample sizes. Even statistics, such as standard deviation, are of little help in visualizing the frequency distribution. For this reason, the number of terminal preanal scales was coded so that three terminal preanal scales received the same value as one. This system has some validity in that only scales in the midline are being compared. Individ- uals having two terminal preanals con- tinue to receive a code value of two. If this is done, most samples still have a mean number of tenninal preanals near or approaching 2.00, but certain sam- ples have much lower values. The low- est of these are samples 1 (x = 1-67), 8 (1.60), 17 (1.67), 22 (1.42), 23 (1.61), 25 (1.76), 26 (1.66), 27 (1.33) and 48 (1.78). Means of other samples ranged from 1.80 (sample 9) to 2.00 (many samples). It will be noted that most of the low sample means are en- countered in samples 22-27 and 48 be- tween Colima and Tres Cruces, Oaxaca, in Mexico. Sample 24, consisting of 14 individuals each with two terminal pre- anal scales, does not fit the pattern.
Coupled with the number of termi- nal preanal scales is the pattern of these scales. They may be (A) mostly (>50%) single (Fig. 20B), (B) mostly paired, (C) single except the terminal scales paired (Fig. 20A), or (D) mixed. The latter category includes individuals not fitting into any other category if the single or paired condition is based on three or more scales or pairs of scales in a row. Most individuals of Ameiva tin- dtilata fall into class C (71.6%; n r= 1003), whereas fewer are referable to classes A (18.2%), B (4.5%) and D (5.7%). Two series of samples, how- ever, show marked deviation from this distribution. If samples 22-27 and 48, from the Pacific slopes of Mexico north and west of the Isthmus of Tehuan-
52
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
tepee, are eombined, there are about equal numbers of individuals referable to classes A (33.9%), B (27.3%) and D (30.6%), but only 8.3 percent are refer- able to class C (n = 121). Individual lizards in samples 41-43 from Costa Rica are distributed as follows: class A (44.2%); class B (3.9%); class C (35.1%); class D (16.9%) (n = 77). On the basis of this character alone samples 41-43 differ markedly from samples to the north (samples 39-40 in Nicaragua), but the divergence of sam- ples 22-27 and 48 parallels that noted for number of terminal preanal scales and appears to reinforce the evidence for a break in gene flow between popu- lations of iindiilata east and west of a line drawn between sample 27 and samples 28 and 29. If samples 21-27, 41-43 and 48 are removed, the distribu- tion of individuals in the remaining samples becomes: class A (13.4%); class B (1.1%); class C (84.6%); class D (0.9%) (n = 805).
Another character not analyzed by STP, but warranting further mention is the pattern of the median gular scales. Median gular scales may be either ir- regular in arrangement or oriented longitudinally in a more-or-less regular row. If irregular, they are usually at least slightly" enlarged (39.0%; 404 of 1026 individuals), although a few in- dividuals have midgular scales no larger than surrounding scales (0.4%). If longitudinally oriented and enlarged, they may be paired (2.1%), in a row of single scales (34.8%) or mixed paired and single (23.1%; at least three scales or pairs of scales in a row were deemed necessary to classify a lizard as having longitudinally oriented scales). Sam- ples 3-8, 15, 22-30, 33, and 35-36 are made up of individuals more than 70 percent of which have enlarged, longi- tudinally oriented gulars. Samples 5, 7, 8 and 28 contain only such individuals. These sample localities are distributed along the Gulf coast of Mexico east and west of the Isthmus of Tehuantepec
from northwestern Veracruz to the Tres Brazos-Balchacaj region of Campeche, across the Isthmus and along the Pacific coast from Colima to southeastern Guatemala (Fig. 17). Samples 9-14, 17- 19, 31 and 34 from the Yucatan Penin- sula, eastern Mexico, northern Guate- mala and northern Honduras are com- posed of individuals more than 70 per- cent of which have enlarged, irregularly arranged gulars. Samples 13 and 34 consist of only such individuals.
These patterns are generally re- flected in the lines between adjacent samples depicting significant differences involving five or more characters (Fig. 22). Gular pattern tends to place sam- ples 17-19 from northern Honduras with samples from northern Guatemala and Yucatan Peninsula. The line separating samples 6-8, 30 and 33 from samples 12- 13, 31 and 34 is represented, but sam- ples 15 and 35 are not differentiated with respect to gular scale pattern and these samples do not differ from sample 36. Samples 15, 35 and 36 differ from one another for at least five of the char- acters analyzed by STP. With respect to gular pattern, sample 36 differs sharply from sample 37 and most other samples to the south and east. Sample 36 contains only one individual out of 30 (3.3%) with irregularly arranged gulars, whereas sample 37 contains 13 of 31 (41.9%) individuals with irregular gular scales.
Fig. 23. Selected geographic variants of color pattern of adult Ameiva undulata. (A) KU 101300, male, SVL 126 mm, (B) KU 101301, female, 109 mm, both from Isla del Maiz Grande, Depto. Zelaya, Nicaragua; (C) KU 87409, male, 93 mm, from 5 km E Vista Her- mosa, Oaxaca, Mexico; (D) UMMZ 41435, female, 99 mm, from Cuautlapam, Veracruz, Mexico. (E) KU 95645, male, 92 mm, from 3 km E Mapastepec, Chiapas, Mexico; (F) KU 95643, female, 82 mm, from 4.5 km E Mapa- stepec, Chiapas, Mexico. (G) KU 101260, male, 93 mm, (H) KU 101267, female, 90 mm, both from El Zamorano, Depto. Fran- cisco-Morazan, Honduras.
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
53
54
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
Samples 16 ( 19% with irregularly arranged gvilars; n = 16), 40 (20%; n = 15) and 2 (73%; n = 15) stand out as having unusually high or low num- bers of individuals with irregularly arranged gular scales relative to adja- cent samples. Sample 2 would seem to belong with samples from the Yucatan Peninsula, northern Guatemala and northern Honduras, whereas samples 16 and 40 are more like samples involved in the Isthmus of Tehuantepec "X."
In samples other than those spe- cifically mentioned above, the percent- age of irregularly arranged gular scales ranged from 33 to 63.
Discussion of geographic variation in color and pattern over the broad range of Ameiva undulata is compli- cated by the presence of both onto- genetic variation (especially among males) and sexual dimorphism (Stuart, 1942). The magnitude of ontogenetic change encountered is apparent in fig- ure 21, and sexual dimorphism in pat- terns of adults from selected popula- tions is represented in figure 23. That geographic variation is considerable is also apparent from figure 23. There are two basic pattern types (Stuart, 1942). The first of these is characterized by an emphasis on longitudinal stripes rather than blotches or bars on the flanks, by the small amount of sexual dimorphism evident, and by its re- stricted geographic range. This pattern type occurs only in samples 7 and 8 of the present study, but it apparently grades into other pattern types to the west and south. Smith and Laufe ( 1946, PI. IB) have illustrated an individual exhibiting this color pattern from Teno- sique. Tabasco, Mexico. These authors cite specimens having this pattern from Palenque and San Ricardo, Chiapas; Frontera, Teapa and Tenosique, Ta- basco; Balchacaj and Tres Brazos, Cam- peche. I have examined two specimens from San Ricardo, Chiapas (FMNH 106715, male; UIMNH 26157, female). The male has 10 light-colored blotches
in the dorsolateral dark field, a pattern not characteristic of the striped forms of undulata. Nine males (FMNH 106715; UIMNH 8437, 8439, 30229, 39233-4, 39237-40) from Finca San Bar- tola, ca 12 mi SW of Cintalapa and near San Ricardo, also show prominent bars and/or blotches on the flanks. The specimens from Finca San Bartola and San Ricardo comprise sample 30 of this study; on the basis of both meristic and pattern characters, this sample is dis- tinct from those to the north. The striped pattern type seems to be re- stricted to the lowlands of Tabasco, ex- treme western Campeche and northern Chiapas. The pattern is not unlike that of quadrilineata (Fig. 16) of Nicaragua, Costa Rica and Panama or chaitzami (Fig. 27) of eastern Chiapas, Mexico and Guatemala. Some undulata having the striped pattern share with chaitzami and, to a much lesser extent, quadrili- neata, the presence of a secondary dark stripe in the middorsal field medial to the dorsolateral light stripes.
The second basic pattern type is characterized by a highly variable pat- tern of bars and /or blotches on the flanks of adult males often accompanied by loss of the ventrolateral light stripe or its reduction to a series of dashes or spots. However, some adult females, as well as juveniles of both sexes, retain a pattern of longitudinal stripes ( Figs. 21, 23). Specimens of Ameiva undidata included in samples 4, 6 and 28 possess this pattern type (Fig. 23 C and D; Smith and Laufe, 1946, Plate I C and D). The dorsolateral dark field is re- tained in both sexes, but the ventro- lateral light stripe may be completely or partially lost in adult males such that the ventral border of the dark field is indistinct. The ventrolateral light stripe is usually retained in females. The dor- solateral light stripe, which is usually retained in females may be completely lost in males or retained only anteriorly. In many specimens the middorsal field contains dark patches which lie para-
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
55
vertebrally iind may be paired. This is especially notable in specimens from sample 6. A characteristic of individuals from these samples is a light stripe with indistinct borders lying in the dorso- lateral dark field and extending from the ear to the groin (Fig. 23 C and D). This stripe is usually continuous but, if broken, the interspaces are much nar- rower than the remnants of the stripe. The stripe is white or blue-white in males, pale brown in females and is nar- rower and less distinct in females. The ventrolateral field contains vertically oriented bars or blotches on a dark background in adult males. Some males in sample 5 show a tendency for the blotches in the dorsolateral field to be continuous with those in the ventrolat- eral field. The ventrolateral field of females is mottled. Females in samples 6 and 28 occasionally have a secondary dark stripe between the dorsolateral light stripes and the lighter brown area of the middorsal field, the same as de- scribed for individuals from samples 7 and 8.
Adult male A. wulnhta in samples 1 and 2 are similar in pattern to those just described, but the dorsolateral dark field contains a series of large, quad- rangular, white or blue-white blotches which are two or three times as wide as the interspaces. These blotches are never fused into a solid stripe. The dorsolateral dark field of females con- tains similar, but less prominent blotches. The dorsolateral fight stripe, which is retained in its entirety in some adult females, terminates at one-half to two-thirds of the distance from occiput to groin in juvenile males and may be present only on the neck in adults.
Adult males from samples 22, 23 and 48 are similar to those of samples 1 and 2, except that the blotches in the dorso- lateral dark field may be subequal in width to the interspaces. Samples 24 and 25 contain adult males with dorso- lateral blotching as in samples 22 and 23, but also contain some males in
which the blotches are fused into a con- tinuous stripe as described for samples 4, 6 and 28. In addition, there is a greater tendency for the dorsolateral light stripe to be retained in males from samples 24 and 25 than in males from samples 22 and 23.
With few exceptions, other samples of Ameiva undulata from Pacific drain- age localities in Mexico present a variety of lateral patterns ranging from pres- ence of blotches in the dorsolateral dark field and spots, blotches or bars in the ventrolateral field to having the entire flanks barred or a combination of both patterns. No one pattern type prevails among males of any one sample. Adult males in sample 44 (El Zamorano and vicinity, Depto. Francisco-Morazan, Honduras) are unusual in that both males and females possess numerous (12-13 per side) small blotches in the dorsolateral dark field. These blotches are tear-drop shaped, and some extend ventrally (Fig. 23 G and H). This pat- tern is not apparent elsewhere.
Samples 9-14, 16-19, 21, 34, 39-43 and 46 are characterized by a lateral pattern of white or blue-white bars which extend from the dorsolateral light stripe or just beneath it to or almost to the enlarged ventral scutes ( Figs. 21 C and 23 A; Taylor, 1956, Fig. 67). The bars show little or no differentiation into dorsolateral and ventrolateral com- ponents. The dorsolateral dark field is usually present in adult males, but it is difficult to define because it grades grad- uallv into the ventrolateral field. In such males the dorsolateral and ventro- lateral light stripes are retained only anteriorly and the ventrolateral light stripe may be completely lacking in large adults. Sexual dimorphism is marked. With the exception of those from Isla del Maiz Grande (sample 21), adult females retain a juvenile pattern with the addition of blotches in the dorsolateral dark field (Fig. 21 D-F). Adult females from Isla del Maiz Grande resemble adult males in pattern
56
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
TABLE 2. — Total number of lateral bars for some samples of male Ameiva undulata. Arranged in descending order according to sample mean.
|
Sample Number |
X |
S.D. |
N |
Range |
|
46 |
29.7 |
1.11 |
7 |
28-31 |
|
12 |
25.4 |
3.47 |
9 |
20-31 |
|
9 |
24.6 |
3.58 |
5 |
20-28 |
|
10 |
24.3 |
3.20 |
14 |
19-29 |
|
34 |
21.6 |
1.67 |
5 |
20-24 |
|
42 |
21.0 |
0 |
4 |
21 |
|
21 |
20.0 |
1.95 |
20 |
16-23 |
|
41 |
20.0 |
1.81 |
12 |
16-22 |
|
43 |
19.8 |
1.34 |
7 |
18-22 |
|
18 |
19.6 |
2.25 |
16 |
16-23 |
|
14 |
19.5 |
1.77 |
8 |
17-22 |
|
13 |
19.4 |
1.81 |
9 |
16-21 |
|
39 |
18.6 |
1.78 |
10 |
17-22 |
|
40 |
18.3 |
1.97 |
6 |
16-21 |
|
17 |
17.8 |
1.28 |
8 |
16-20 |
|
19 |
16.8 |
1.92 |
5 |
15-20 |
(Fig. 23 A-B), but not in color. The total number (sum of number on both sides) of vertical bars of males varies considerably among these samples (Table 2). The largest sample means are those of samples on the Yucatan Peninsula and Isla Mujeres. The sample from Isla Mujeres (sample 46) has a mean of 29.7, the largest recorded. Lowest sample means occur in the two northern coastal samples from Hon- duras (samples 17 and 19), with means of 17.8 and 16.8, respectively. Males from samples 9-14, 17-19, 34 and 46 may be differentiated from barred males from other samples in that they have little or no secondary spotting or blotching between the bars in the ven- trolateral field. Taken as a whole, barred males from other samples have such markings to a greater degree, al- though individuals from these samples may show few secondary markings.
The general ontogenetic changes ex- hibited by the barred populations of Ameiva undulata are illustrated in Fig- ure 21. Neill (1961, 1965), reported that young undulata from parts of British
Honduras lack lateral light stripes, the ontogenetic progression being from un- striped hatchlings through light-striped subadults to barred adults. Presumably this refers to ontogeny of males only. Neill ( 1965 ) also called attention to two small specimens (MCZ 71623-24) from Augustine, Cayo Dist., British Honduras. MCZ 71623 has four light lines on each side, whereas MCZ 71624 has three. Augustine is in the ecolog- ically distinctive Mountain Pine region of British Honduras.
Among samples of Ameiva undulata there is considerable geographic varia- tion in amount of dark pigment scat- tered as flecks or blotches in the mid- dorsal area. These dark markings usual- ly take the form or more-or-less paired or alternating, paravertebral spots, al- though in heavily pigmented specimens a reticulate pattern may be evident. The latter is especially evident in speci- mens from Isla del Maiz Grande (Fig. 23 A-B). Both sexes may exhibit such markings. Although they are sometimes evenly distributed, the markings are of- ten restricted to or heaviest on the pos- terior one-half or two-thirds of the back. I have not attempted to quantify this aspect of pattern but feel justified in making the generalization that the samples along the Pacific slopes from southeastern Chiapas, Mexico through El Salvador (samples 33, 35-38) and northward into Honduras south of the northern coastal ranges (samples 16, 20 and 44) are the most heavily marked of the mainland populations sampled. Specimens from Isla del Maiz Grande (sample 21) have very heavy middorsal marking, as noted above, but this marking takes a different form, appear- ing to be a continuation of the lateral pattern. Samples 39 and 40 are variable with respect to this character and sam- ples 41-43 show very light markings. A few specimens of Ameiva undulata are available from the Meseta Centrale of Costa Rica. This population was named by Cope (1894) as Cnemidophorus
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
57
Fig. 24. Aiijeiva undulata (Holotype of Cnemidopliorus amivoides Cope; AMNH 16316) from La Caipintera, Cartago Prov.?, Costa Rica, illustrating the heavy blotching in the middorsal area and the divided inter- parietal; a male, SVL 64 mm.
amivoides (see Remarks). These speci- mens all show a high degree of marbling or blotching in the middorsal region (Fig. 24), a pattern sharply divergent from that of other populations of undu- lata at the southern end of its range. Further study of this population is de- sirable. Samples other than those spe- cifically mentioned above show variable markings from light to moderate in extent.
There is little geographic variation with respect to color other than that di- rectly associated with pattern. Studies of such variation are hampered by a paucity of good color-in-life descrip- tions. Such descriptions are available for samples from Michoacan, Mexico (Duellman, 1961), the Yucatan Penin- sula (Maslin, 1963; Duellman, 1965a), British Honduras (Neill, 1965) and Chiapas (Alvarez del Toro, 1960). Other descriptions have been gleaned from my own field notes and those of William E. Duellman.
The most striking aspect of geo- graphic variation in color has to do with color in the gular region of adult males. As noted for Ameiva festiva and quad- rilineata, the gular coloration of males may be associated with reproductive condition. In many specimens through- out the range, gular coloration (and ventral coloration in general) is white or cream. In other specimens, the en-
tire venter is pale blue. A number of individuals, however, show either red or yellow pigment in the gular region and over the anterior part of the chest and ventral surfaces of the forelimbs. The red coloration ranges from orange or brick-red to scarlet. The yellow pig- ment is a bright lemon yellow. Whether one color precedes the other in a dy- namic system of change is not known, but I have observed no specimens in- termediate in coloration. The specimens upon which this discussion is based were collected during June, July or August between 1959 and 1966 by a number of collectors. Yellow pigment has been observed to the exclusion of red at only three localities in Mexico: along the Rio Frio, Tamaulipas; Finca Orizaba, Chiapas; and Tapanatepec, Oaxaca. At some localities, individuals may have either red or yellow pigment in the gular region. Thus, Duellman (1965a) reported that about half of his sample from Piste, Yucatan had orange throats, whereas half had yellow throats, corroborating an observation by Maslin (1963). Other localities from which both colors have been reported are Felipe Carillo Puerto, Quintana Roo (Duellman, 1965a), British Honduras ( Neill, 1965 ) and the Tepalcatepec Val- ley of Michoacan (Duellman, 1961). Specimens from a number of other localities in Veracruz, Tabasco, Cam- peche, Yucatan, Quintana Roo, Chiapas, Jalisco, Honduras and Isla del Maiz Grande, Nicaragua had only red- throated individuals.
There seems to be some geographic variation in coloration of the middorsal area. This area is iridescent moss- green anteriorly grading to yellowish- brown posteriorly in specimens col- lected by me at La Ceiba, Coyoles and Trujillo, Honduras (samples 17-19, re- spectively). It is reported as "olive" for specimens from British Honduras (Neill, 1965) and Chiapas (Alvarez del Toro, 1960). Specimens from the Te- palcatepec Valley, Michoacan (Duell-
58
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
man, 1961); Piste, Yucatan (Maslin, 1963); along the Rio Frio, Tamaulipas; 16 km. SW Frontera, Tabasco; Mapas- tepec, Chiapas; vicinity of Melaque, JaHsco; El Zamorano, Depto. Francisco- Morazan, Honduras (personal observa- tion) have brown middorsal pigmenta- tion. Few descriptions of lateral ground color of living lizards are available, but there may be considerable geographic variation for this character as well.
Ecological Notes. — Despite its broad range and relative abundance in many parts of that range, there have been no thorough ecological investigations of Ameiva undulata. Literature on this subject is limited to miscellaneous ob- servations. Thus, Stuart (1951) found undulata altitudinally distributed to about 2000 m in the "moderately hu- mid" region near Panajachel, Depto. Solola, Guatemala, but commented that in more moist areas the altitudinal limit is reached about 400 m lower. Else- where, the species seems to occur below 1500 m, and its vertical distribution may be limited to below the level of occa- sional frosts (1200-1500 m, depending upon humidity), a factor considered by Stuart (1966) to be of possible bio- logical significance.
Throughout most of its range, Ameiva undulata is found in forest or forest-edge situations (Alvarez del Toro, 1966; Duellman, 1965a; Echter- nacht, 1968; Rand, 1957; Stuart, 1935, 1950, 1954b, 1958) or, in drier regions, dense thickets and scrub (Maslin, 1963; Schmidt and Stuart, 1941; Stuart, 1948).
Stuart (1950), Duellman (1966), and Echternacht (1968) commented on the sciophilic nature of Ameiva iindu- lata, although I pointed out that even the more shade-tolerant species of Am^eiva may utilize open areas in the absence of competition from more he- liophilic macroteiids like Cnemidopho- rus. Stuart (1950) reported undulata to be abundant in the grassland and ri- parian habitats of Depto. Alta Verapaz, Guatemala, and Echternacht ( 1968 )
found the species to be common in open situations near Panajachel, Depto. So- lola, Guatemala. Although the lizards shun the open beach, I found the species occupying clearings, paths and other open areas as well as thickets and piles of debris in the coconut groves of Isla del Maiz Grande, Depto. Zelaya, Nicaragua. In northern Honduras, western Nicaragua, northwestern Costa Rica, and in those parts of Mexico with which I am familiar, undulata occurs in shaded areas; the more open parts of the habitat are occupied by various species of Cnemidophorus.
Virtually nothing has been reported on the reproduction of this species of Ameiva. As I have remarked elsewhere, research on this topic may reveal the sequence and basis for changes in gular pigmentation of adult males.
Stuart (1958) collected 56 juveniles and only four adults of Ameiva undu- lata at Tikal, El Peten, Guatemala, be- tween February and May, 1956, but he cited a collection made by Gosner at the same locality at unspecified dates in 1947 and 1949 that contained a "fine series of adults." This, along with my own observations, suggests that adult and subadult Ameiva may have differ- ent seasonal activity periods. However, there have been no observations of a quantitative nature to substantiate this.
Remarks. — I have examined four specimens of Ameiva undulata (ANSP 15438-15441) from "Wounta" (= Hua- unta) Haulover, Depto. Zelaya, Nica- ragua. This locality is at least 290 kilo- meters from other known localities for the species. Duellman (1958) exam- ined three specimens (ANSP 15445- 15447) of the snake Leptodeira annu- lata rhomhifera Giinther from the same locality, far removed from the re- mainder of the range of the species. Al- though it is possible that a snake might have been carried to Huaunta by man, it is unlikely that a moderately large, active lizard could be so transported. Ameiva undulata shuns deep forest, and
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
59
it is unlikely to have arrived at Huaunta along rivers from the west, but it is pos- sible that the species exists in isolated or semi-isolated populations along the Atlantic coast of Nicaragua northward. An alternative hypothesis would be that the A7newa and Leptodeira populations at Huaunta are relict and that a more careful survey of the region would un- cover other such species as well. This part of Central America certainly de- serves more attention from biologists than it has received.
Such an explanation does not apply for records of Ameiva undidata from Panama (Brennan and Yunker, 1966). All specimens so reported are lepto- phrys.
Taylor (1956) placed Cnemidopho- rus amivoides Cope from the Meseta Central of Costa Rica in the synonymy of Ameiva undidata, and I see no rea- son to dispute this judgment. I have called attention to the unusual color pattern of specimens from the Meseta Central (Fig. 24), but so few are avail- able that further comment is unjustified at this time.
Dunn (1940a) and Villa (1968) in- cluded the population of Ameiva on the Islas del Maiz, Depto. Zelaya, Nica- ragua (sample 21) in the species undu- lata. The population was described as a subspecies of festiva by Barbour and Loveridge (1929a). I have substan- tiated the assignment of this population to undulata ( Echternacht, 1970).
Ameiva chaitzami Stuart
Ameiva chaitzami Stuart, Proc. Biol. Soc. Washington, 55:143-145, 1942 [Holotype: UMMZ 90638. Type Locality: "Along Cahabon-Languin trail about 2 km north of Finca Canihor . . . , Alta Verapaz, Guatemala." Collector: L. C. Stuart]. Stuart, Misc. Publ. Mus. Zool., Univ. Mich- igan No. 122, p. 77, 1963. Echternacht, Breviora, 354: 7, 1970.
Ameiva undidata stuarti — Stuart, Occas. Papers Mus. Zool, Univ. Michigan No. 471, p. 21, 1943.
Ameiva undulata thomasi Smith and Laufe, Univ. Kansas Sci. Bull. 31(2):47-50, Pi. lA, 1946 [Holotype: FMNH 100006. Type
92»
' "\ —
5°
I
',•
s
10 0 50
H 1
KILOMETERS
92°
90°
89°
88°
Fig. 25. Map showing locality records of Ameiva chaitzami. Samples used in statistical analyses were ( 1 ) Comitan, Chiapas, Mexico, (2) Vicinity of San Antonio Huista, Depto. Huehuetenango, Guatemala, (3) combined sample of those specimens originally desig- nated by Stuart (1942) as holotype and para- types from the Languin-Cahabon road near Finca Canihor, Depto. Alta Verapaz, Guate- mala, and specimens from the vicinity of Poptun, Depto. El Peten, Guatemala, later collected by Stuart and identified as A. chaitzami.
Locality: La Libertad, Chiapas, Mexico, near Rio Cuilco where it crosses Guate- malan border. Collector: H. D. Thomas]. Smith and Taylor, U. S. Natl. Mus. Bull. 199, p. 173, 1950.
Distribution. — Known from the val- leys of the upper tributaries of the Rio Grijalva in Chiapas, Mexico and west- central Guatemala, from along the Lan- guin-Cahabon trail near Finca Canihor, Alta Verapaz, Guatemala, and near Poptun, El Peten, Guatemala (Fig. 25).
Diaanosis. — Ameiva chaitzami can be distinguished from other Middle American congeners by the following combination of characters: Small size (maximum observed SVL of males 85 mm, of females 75 mm); central gular scales enlarged, in longitudinal arrange- ment; gradual reduction in size of scales radiating outward from central gular scales; prefrontal scales in contact with
60
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
postnasal scales; three parietal scales (four if median parietal scale divided); a transverse row of abruptly enlarged mesoptychial scales; usually eight trans- verse rows of ventral scutes; relatively narrow middorsal stripe (mean PV =: 37.1); no narrow, well-defined vertebral stripe; usually a dark secondary stripe medial to the dorsolateral light stripes; dorsolateral blotches of males fused to the dorsolateral light stripe such that the latter has a well-defined dorsal bor- der and an irregular ventral border.
Adult male Ameiva chaitzami can be distinguished from other Ameiva readily on the basis of pattern. Juvenile individuals and females are difficult to differentiate from some undidata, espe- cially those from the coastal lowlands of eastern Veracruz, Tabasco, Chiapas and extreme western Campeche, Mex- ico. However, there are no signs of intergradation or hybridization of chait- zami with other species of Ameiva. The only Ameiva with which chaitzami comes in contact is the large form of undulata having a color pattern as shown in figure 21, and there is no dif- ficulty in distinguishing the two species.
Description. — Maximum observed SVL 85 mm for males, 75 mm for fe- males; supralabials 13-16 (14.2); infra- labials 9-13 (10.1); supraoculars 6-8 (6.2); COF value 2; COP value 10-13 (11.6); SO-SC value 2-8 (2.3); loreals 2-4 (3.7); nostril in prenasal-postnasal suture; prefrontals in contact with post- nasals; central gular scales enlarged, longitudinal in arrangement; GAB 83- 134 (108.6); PV 26-51 (37.1); PV/GAB 0.27-0.42 (0.34); GOR 174-243 (203.6) for males, 171-255 (205.1) for females; GAB/GOR 0.44-0.65 (0.54); SAT 20-24 (21.2); total preanal scales 5-9 (6.8); terminal preanal scales 1-3 (2.0); one row of enlarged prebrachials; one or two rows of enlarged postbrachials; en- larged preantebrachials in two rows or two rows proximally and one row dis- tally; prefemoral scales enlarged, four (usually) or three rows proximally re-
FiG. 26. Head and preanal scutellation of Ameiva chaitzami (UMMZ 90643, paratype, from along the Cahabon-Langiiin trail about 2 km N Finca Canihor ) : ( A ) Lateral view of head, ( B ) Dorsal view of head, ( C ) Ventral view of head, X 2. (D) Preanal region, X 2.7.
ducing to three (usually) or two rows distally; infratibials enlarged, in two rows; femoral pores 30-41 (35.0) for males, 26-36 (32.7) for females; sub- digital lamellae 24-33 (28.6); longitudi- nal rows of ventral scutes 27-32 (30.0); transverse rows of ventral scutes 8-10 (8.1). For details of scutellation see figure 26.
Color and Pattern in Alcohol. — De- scriptions of specimens in preservative have been presented by Stuart ( 1942 ) and Smith and Laufe ( 1946 ) . Juvenile males possess a broad middorsal field bounded laterally by narrow, white dor- solateral stripes which originate just be- hind the eye and extend well onto the tail. Between the blue-gray or gray- brown middorsal pigment and the dor- solateral light stripes there are often secondary stripes, velvet black in color, which extend from the occiput to the
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
61
base of the tail. The medial border of the secondary stripe is irregular. The niiddorsal field may be flecked witli black or be nearly uniform in colora- tion. A white ventrolateral light stripe, which may be complete or interrupted, extends from the posterodorsal corner of the ear to the groin, beginning again lieliind the leg and terminating on the tail. A white postfemoral stripe is con- tinuous with the ventrolateral light stripe behind the leg. The black dorso- lateral dark field is prominent and con- tains faint, light-colored blotches or bars. The ventrolateral field contains wliitc mottling or bars on a black back- ground. The same pattern is evident on dorsal surfaces of the limbs, but the mottling is more pronounced. Ventral surfaces are cream or white in color.
Juvenile females resemble juvenile males except that there are no light- colored blotches or bars in the dorso- lateral dark field and the ventrolateral and limb mottling is less well de- veloped.
Through ontogeny (Fig. 27) the light ventrolateral and dorsal limb mot- tling or barring becomes much more pronounced and the ventrolateral light stripe becomes interrupted such that it merges with the ventrolateral mottling. The dorsolateral blotches become white or blue-white in color and fuse with the dorsolateral light stripe in males, giving that stripe an irregular ventral border. Adult females resemble juveniles of both sexes in color and pattern, ap- parent ontogenetic changes being slight. Color and Pattern in Life. — There are no descriptions of the coloration of living Ameiva chaitzami, although Stuart's (1942) description of the holo- type may have been written before ex- tensive fading had occurred in preserva- tive. He stated that the middorsal field and top of the head are olive-brown and that the legs and arms are olive-brown above with black and bluish-white mottling. The anterior surfaces of the thighs are said to be black with blue
spots, the venter l:)hiisli and the head blue mottled with black. There is no mention of yellow or red pigment in tlie gular region of this species, the only Middle American Ameiva with which I have had no field experience.
Geographic Variation. — The follow- ing characters show no significant dif- ferences among sample means: Femoral pores of males, number of supralabials, number of infralabials, number of ter- minal preanals, number of transverse rows of ventral scutes, SAT and COF, COP and SO-SC values.
The three samples of Ameiva chait- zami (Fig. 25 ) were collected in eastern Chiapas, Mexico (Sample 1), western Guatemala (2) and central to eastern Guatemala (3). Samples 1 and 2 are much closer geographically than either is to sample 3 and sample 1 lies slightly northwest of sample 2. This geographic arrangement is reflected in the results of STP analysis of characters which show significant differences among means. Five characters show samples 1 and 2 as constituting one subset, sample 3 a second subset significantly different from the first. These five characters are numbers of loreals and subdigital lamel- lae, GAB /COR, numbers of longitudi- nal rows of ventrals and total preanal scales. For two characters, GAB and PV/GAB, samples 2 and 3 form a sub- set distinct from that containing only sample 1. For both PV and GOR of females, all three sample means fall into separate, significantly distinct subsets. This is true of samples 1 and 2 for num- ber of femoral pores of females and GOR of males; sample 3 is too small for inclusion in STP analysis. Only for one character, number of supraoculars, is there overlap of sample means: there is a gradient from sample 3 through sam- ple 2 to sample 1, samples 3 and 2 con- stituting one subset and samples 2 and 1 another.
Because of the inadequate number of samples and the geographic localities from which the samples were drawn,
62
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
63
detection of clinal trends must await accumulation of additional large sam- ples as the range of the species becomes more fully known.
There seems to be no notable geo- graphic variation in color, pattern or arrangement of gular scales among the three available samples. A comparison of color patterns of equal-sized (68 mm) males from samples 1 and 3 may be seen in Figure 27 (B and G, re- spectively ) .
Ecological Notes.— Stuart ( 1942; personal communication) provided the only insight into the ecology of Ameiva chaitzami. He is of the opinion that the species is restricted to pine savannas on the lowlands of central Guatemala where its population density may be kept at a low level by regular burning- off of ground cover. The species is found with Ameivo tmdulata, but never with Cnemidophorus. Other Middle American species of Ameiva are sym- patric with species of Cnemidophorus over at least part of their range. Stuart (1942) mentioned a chaitzami with a
SVL of 66 mm having "well formed eggs."
Smith and Laufe (1946) stated that these lizards ". . . probably occur in all the dry, hot valleys of the upper tribu- taries of the Rio Grijalva in the interior of Chiapas and of western central Guatemala."
Remarks. — Less is known about the biology of Ameiva chaitzami than of any other Middle American Ameiva. The overall range of the species has yet to be defined.
The divided interparietal (median parietal), which was used by Stuart ( 1942 ) to diagnose Ameiva chaitzami, is a variable character found through- out the genus in Middle America with varying frequency ( Echternacht, 1970). Likewise, I have found the lateral en- largement of the gular scales cited by Smith and Laufe ( 1946 ) to be an unre- hable character. Color pattern and size remain the most useful characters for distinguishing this probable sibling spe- cies from its closest relative, undulata.
DISCUSSION
INTRASPECIFIC RELATIONSHIPS
In my opinion the Ameiva of Mid- dle America belong to three natural groups the relationships of which will remain obscure until their affinities with South American representatives of the genus are known. The three groups contain one, two and three species ( Fig. 28).
Fig. 27. Ontogenetic change in pattern of Ameiva chaitzami. A-D and F, UMMZ 94905, E, UMMZ 94901, all from the vicinity of Comitan, Chiapas, Mexico. Field Tag Num- bers and SVL are: Males, (A) 473, 45 mm, (B) 482,68 mm, (C) 461, 73 mm. Females, (D) 483, 55 mm, (E) 365, 68 mm, (F) 501, 75 mm. (G) For comparison, UMMZ 90640, a male paratype 68 mm SVL from along the Cahabon-Languin trail about 2 km N Finca Canihor, Alta Verapaz, Guatemala.
Ameiva ameiva, the only species in Group I, can be distinguished from the others on the basis of a number of scale and pattern characteristics. Chief among these are the presence of five parietal scales, 10-12 enlarged ventral scutes, a large number of scales around the tail, high GAB and GOR counts and a pattern which emphasizes spots in- stead of stripes in adults. Additionally, ameiva is the only species included in the present study having the bulk of its range in South America. The Pana- manian portion of that range is but a minor and disjunct part of the whole. Group I is farther removed phylogenet- ically from Groups II and III than these are from each other.
Group II includes Ameiva festiva and leptophrijs. Both species are essen- tially lower Central American in dis-
64
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
Fig. 28. Relationships of the Middle American species of Ameiva. Group I: A, A. ameiva. Group II: F, A. f estiva; L, A. leptophrys. Group III: Q, A. quadrilineata; C, A. chait- zami; U, A. undulata. Lengths of the lines and positions of branches do not reflect a time scale.
tribution. Ameiva festiva occurs in Co- lombia, and leptophrys is expected there. The group is characterized by the relatively high number of femoral pores and a tendency for the enlarged, posterior scales of the head to become disrupted (Fig. 7). Disruption of head scales is more noticeable in leptophrys than in festiva but is a more commonly encountered feature of these species than of other Middle American species of Ameiva.
Group III consists of Ameiva quad- rilineata, chaitzami and undulata. The group is entirely Middle American in distribution, and is characterized by similarities in pattern and scale charac- ters. The pattern emphasizes stripes, rather than spots, and features a broad, middorsal field without a vertebral stripe. Prefrontal and postnasal scales are in contact. Except for an occasional longitudinal division of the interparietal scale, there is little disruption of pos- terior scales of the head. Ameiva chait- zami is very closely related to undulata, from which it is distinguishable pri- marily on the basis of the color pattern of males. However, the two species are sympatric in central Guatemala. Clari-
fication of the relationships of chait- zami, its range and biology remain problems of major importance.
Elucidation of the relationships of the Middle American species of Ameiva with those of South America may show that Groups II and III are much more closely related than I have indicated. These may constitute a single group. I should emphasize that these three groups of Middle American Ameiva are not the equivalents of the species groups of Cnemidophorus established by Burt ( 1931 ) . Nevertheless, it seems apparent that my Group I would belong to a difterent species group (sensu Burt) than the other species of Middle American Ameiva, regardless of whether these other species are found to belong to one or several species groups. The relationships indicated without substan- tiating comment by Barbour and Noble (1915) are suspect. The taxonomic im- portance of extensive division of head scales in Ameiva bridgesi Cope, orcesi Peters and septemlineata Dumeril and Dumeril (Peters, 1964) and the rela- tionships of these species to leptophrys and festiva require further study. The relationships of niceforoi also need clari- fication ( Echternacht, 1970).
INTERSPECIFIC PATTERNS IN GEOGRAPHIC VARIATION
Assessing interspecific trends in vari- ation is considerably more difficult than determining intraspecific trends. This is so for two reasons: 1) The narrow lin- earity of lowland Middle America vir- tually dictates that clines will run paral- lel to the coastlines along a northwest to southeast axis and 2) samples of two or more species from the same locality and of sufficient size for statistical anal- ysis are rarely available. Samples from only ten localities where sympatric oc- currence of two or more species was found were large enough for analysis (Table 3), and only three of the six Middle American species of Ameiva were present in more than one sample.
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
65
TABLE 3. — Localities from which samples of two or more species were available for statistical
analysis.
Locality
Species
Costa Rica: Limon: Limon, Los Diamantes and vicinity
Costa Rica: Limon: Tortuguero
Costa Rica: Puntarenas: Golfito Region
Costa Rica: Puntarenas: Osa Peninsula
Guatemala: El Peten: Piedras Negras
Honduras: Northern Coastal Localities
Panama: Bocas del Toro: Almirante and vicinity
Panama: Canal Zone
Panama: Isla Escudo de Veraguas
Panama: Eastern Darien
|
festiva. |
quadrilineata |
|
festiva. |
quadrilineata |
|
festiva. |
quadrilineata, leptophrys |
|
festiva. |
(juadiilineata, leptophrys |
|
festiva. |
tmdidata |
|
festiva. |
itudulata |
|
festiva. |
quadrilineata |
|
ameiva |
festiva, leptophrys |
|
festiva. |
quadrilineata |
|
festiva. |
leptophrys |
Ameiva undtdata was present with fes- tiva in but two samples, and festiva was found with quadrdineata in two. Al- though no rigorous statistical analysis was attempted, some parallel trends are evident. For seven characters (infra- labials, supraoculars, COP, GAB, GOR, subdigital lamellae and scales around the tail) an increase or decrease in sample mean for festiva was paralleled by a like shift in mean for the same character for tindulata when the two localities of sympatry (Piedras Negras, El Peten, Guatemala, and northern Hon- duras) were compared. Similar trends were noted for festiva and quadrdineata for ten characters (GAB, GOR, GAB/ GOR, longitudinal rows of ventral scutes, transverse rows of ventral scutes, femoral pores, subdigital lamellae, total preanal scales, temiinal preanal scales, and scales around the tail) when step- wise pair comparisons of samples were made between localities beginning with Tortuguero in Costa Rica and running through the Canal Zone to Darien, Panama, and thence to the Golfito and Osa Peninsula regions of Costa Rica. Perfect agreement was noted only for GOR, but all other characters men- tioned showed parallel mean-shifts for at least two-thirds of the possible pair comparisons. Similar parallel changes were noted for festiva and leptophrys ( supralabials, infralabials, COF, COP, transverse rows of ventrals and scales
around the tail with COP and COF showing perfect agreement) and lepto- phrys and quadrilineata (infralabials and supraoculars). Such parallel shifts in sample means would tend to indicate that these characters are responding in similar ways to environmental differ- ences between sample localities. Con- siderable additional study will be neces- sary before any light will be shed on the question of which environmental factors are most important in determining the direction of shift of sample mean for a character from one sample locality to another. Because only increase or de- crease in the value of the mean were noted from sample to sample, and not the magnitude of the increase or de- crease, it is probable that some shifts are only the result of small sample size and that additional data would show that there has been, in fact, no shift at all.
An important, although likewise un- explained, trend is that of insular popu- lations of Ameiva to have higher mean GAB and/ or GOR counts than main- land populations (Table 4). Samples are available from Isla Mujeres, Quin- tana Roo, Mexico (undidata), Isla del Maiz Grande, Zelaya, Nicaragua (undu- lata), Isla Escudo de Veraguas, Panama (festiva, quadrilineata) and Isla San Jose, Archipelago de las Perlas, Panama (ameiva). Samples from the Osa Penin- sula, Puntarenas, Costa Rica of festiva,
66
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
TABLE 4. — Relati\'e positions of island samples with respect to mean Granules Around the Body (GAB) and Granules Occiput to Rump (GOR) when sample means are ordered from largest
to smallest.
|
Total Number |
Relative |
Position |
|||
|
Species |
of Samples |
GAE |
; |
GOR |
|
|
ameiva |
7 |
1 |
2*' " 1" |
||
|
festiva: Isla Escudo de Veraguas |
20 |
1" |
1" |
V 1" |
|
|
fcstica: Osa Peninsula |
20 |
4" |
4" |
15" 16" |
|
|
k'pto))hnjs |
4 |
2 |
2 |
||
|
quadrilineata: Isla Escudo de Veraguas |
10 |
7 |
5 |
||
|
quadrilineata: Osa Peninsula |
10 |
1 |
1 |
||
|
imdulata: Isla Mujeres |
45 |
2 |
10 |
||
|
uudulata: Isla del Maiz Grande |
45 |
1 |
2 |
" Males where sexual dimorphism is evident.
" Females where sexual dimorphism is evident.
'' Tied with sample 3 (See Appendix B) for this position.
leptophrijs, and quadrilineata also have relatively high dorsal granule counts; the peninsula may be effectively an is- land as far as Ameiva are concerned inasmuch as it is separated from the mainland by a swampy area. It will be obvious from an examination of table 4 that not all insular samples have high doi-sal granule counts, but the tendency is evident in a remarkable number of instances. There is no clear relationship between high GAB and GOR counts on the one hand and the ecology of the islands upon which the lizards are living on the other. It would seem that Amei- va adapted to life on dry islands would have fewer, but larger, dorsal granules than Ameiva of the same size living on wet islands as a selective strategy for prevention of excess water loss. Fewer and larger granules would mean less ex- posed skin between the granules through which water might be lost than more and fewer scales on animals of equal size. Such an hypothesis is con- traindicated by the observation that the islands from which samples of Ameiva are available range from very dry (Isla Mujeres) to very wet (Isla Escudo de Veraguas and the Osa Peninsula). It would be instructive to have samples of one species of Ameiva from a number of localities for which weather records are available and ranging from rela-
tively dry to relatively wet. Such sam- ples are not presently available and, until they are, this problem is likely to remain unsolved.
Sample mean snout-vent length (SVL) data are not presented in this report because collections are usually strongly biased in favor of large, bright- ly colored males. The relatively drab juvenile and adult female Ameiva are often overlooked by collectors who ob- tain Ameiva incidental to those genera they themselves are working on. A sea- sonal bias also acts to reduce the num- bers of hatchling Ameiva in collections in that, in areas having pronounced wet and dry seasons, reproduction may be temporally restricted with hatchlings present only at certain times of the year. Collections made at other times do not include hatchlings. All of the insular samples of Ameiva available to me were obtained in June, July, or August, and all are the result of but one or two field trips. With the exception of my own collection from Isla del Maiz Grande, all of these collections were made by individuals primarily interested in other genera. For these reasons, I have little faith in the mean SVL of samples used in this study as an estimate of the true population means. Nevertheless, it is apparent that the insular samples have generally higher mean SVL than main-
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
67
TABLE 5. — Correlation between snout-vent length (SVL) and granules around the body (GAB) or granules from occiput to nunp (GOR) for five species of Middle American Ameiva. Cor- relation coefficient followed by sample size in parentheses.
ametva
festiva
leptophrys quadrilineata undulata
SVL-GAB Males SVL-GAB Females SVL-GOR Males SVL-GOR Females
0.1101(112) 0.0897(243) 0.1345 ( 101 ) -0.0289 ( 180) 0.1606(409)
-0.0403(133) 0.1531 (233) -0.1040(98) -0.0950(114) 0.2445(561)
0.0737(114) 0.1857(268) 0.2262(188) 0.1629(507)
-0.0888(138) 0.0682(264) -0.2283(126) 0.1550(448)
land samples, making it necessary to ask the question whether GAB and GOR counts are in any way correlated with SVL. Correlation analysis was carried out on a species-by-species basis (Table 5). There is no indication that either of the tsvo dorsal granule counts are highly correlated with SVL and the high mean GAB and GOR values of island samples is not an artifact resulting from biased estimates of sample mean SVL.
HISTORY OF AMEIVA IN MIDDLE AMERICA
There is no fossil record of macro- teiids in Middle America. Therefore, any discussion of the history of Ameiva in that region must lean heavily on what is known of present distribution patterns and on geological and faunal summaries relating to Middle America and northern South America. Geological theories concerning Middle America have undergone an evolution of their own over the past 50 years; geological data were summarized by Lloyd ( 1963 ) , Vinson and Brineman (1963), and Mal- donado-Koerdell ( 1964 ) . Duellman (1966), Savage (1960, 1966) and Stuart (1935, 1951, 1954a, 1954b, 1957, 1964, 1966) provided thoroughly documented discussions of the ecology, origins, and history of the herpetofauna of Middle America, whereas Haffer ( 1967, 1970 ) presented an important assessment of the geologic-climatic history of the criti- cal Gulf of Uraba-northwestern Colom- bia region. Although the following dis- cussion is based in large part on these works, it is a highly speculative and
personal view and should be read with that in mind.
Dunn ( 1931 ) considered the family Teiidae to be a part of his South Amer- ican element, having its origins in South America with some genera invading Middle America and one, Cnemido- phorus, reaching North America. The South American element itself was de- rived from a generalized tropical Amer- ican herpetofauna which ranged over much of North and South America dur- ing the early Tertiary ( Savage, 1960 ) . Subsequent authors have followed this arrangement (Duellman, 1965b; Savage, 1960, 1966; Schmidt, 1943; Stuart, 1957, 1964), although Savage (1966) sug- gested that the ancestral stock of Cnemi- dophorus was isolated north of the Isthmus of Panama and that of Ameiva south of the Isthmus when the Pana- manian portal was open throughout most of the Tertiary. Northward move- ments of the South American element would have been possible after the closure of the portal in the late Pliocene or early Pleistocene (Lloyd, 1963; Mal- donado-Koerdell, 1964). As outlined by Stuart (1957) and Savage (1966, Fig. 23), Ameiva, with the exception of A. ameiva, probably followed humid, low- land routes from South America into Middle America. Movement through the Pacific lowlands progressed only as far as the Golfito-Palmar-San Isidro del General region of Costa Rica. An At- lantic route was followed northward in Mexico, dispersal to the Pacific lowlands occurring at several points north of what is now Lake Nicaragua. Some
68
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
southward movement along the Pacific lowlands of Nicaragua and Costa Rica at a later time is indicated by the pres- ent distribution of A. tinchilota ( Fig. 17). These routes may have been broader than at present because of low- ered sea levels, especially in areas where the continental shelf is extensive (Bennett, 1968).
Of all of the Ameiva found in Mid- dle America, A. ameiva has made the smallest inroads there. Heatwole (1966) placed its arrival in the Upper Pliocene at the earliest. As an alternative hy- pothesis, Bennett (1968) suggested that the species dispersed into Panama during the period of human occupation of Darien. The A. ameiva of Panama resemble those of western Venezuela; Bennett suggested than an open corri- dor between the two areas for the length of time suggested by Heatwole ( over one million years ) v^^ould have re- sulted in formation of clines. Subse- quent closure of the corridor would have isolated opposite ends of such clines, resulting in morphologically di- vergent populations. Other than a study emphasizing color pattern ( Miiller, 1929), no detailed study of the geo- graphic variation of South American A. ameiva has been reported. Donoso- Barros ( 1968 ) characterized the west- ern population of the species in Vene- zuela as having a light-colored vertebral stripe. Such a stripe is evident in those A. ameiva from eastern Panama, but lacking in those from the western part of that country (Fig. 4). With the examination of additional samples, rem- nants of a cline as postulated by Ben- nett may be found for this and other characters (but see Remarks, p. 19). If remnants of a cline are found, it would indicate that entry of the species into Middle America occurred at a time in- termediate to the extremes suggested by Heatwole and Bennett. Recent evi- dence of alternating dry interglacial and humid glacial periods during the Pleis- tocene so convincingly presented by
Haffer (1967, 1970) would allow for several intervals during which suitable conditions prevailed for the northward movements of A. ameiva. Haffer (1970) suggested that the humid forests of gla- cial times were forced southward dur- ing interglacials and during post-Pleis- tocene dry periods and that central and eastern Panama, as well as the Gulf of Uraba-lower Rio Atrato region of Co- lombia, were not heavily forested. Thus, it would have been possible for A. ameiva to have moved into Panama prior to the period of human occupancy of the Darien and yet not long enough for establishment of clines. Indeed, if Haffer's scheme is followed, the corridor between central Panama and Colombia would have been interrupted and re- established several times, mediating against formation of clinal variation. In any case, it is certain that the Pana- manian A. ameiva originated from a population of South American Ameiva similar to the present population of the species in western Venezuela. Dispersal into Middle America was contained by the availability of non-forested habitats. The species probably reached the is- lands of the Archipelago de las Perlas at a time when a land connection existed between these islands and the mainland during a period of lowered sea level (Bennett, 1968).
With the exception of Ameiva amei- va already discussed, none of the species of Ameiva presently inhabiting Middle America would have required a non- forested corridor for northward disper- sal. Although the species show a graded dependence upon forested habitat, all are found in or at the edge of forested areas. Ameiva f estiva and leptoplinjs may, upon completion of studies now in progress, be found to be closely allied to the A. septemlineata — hridgesi — orcesi assemblage of northwestern South America. Ameiva leptophrys, with its present range limited to eastern Pana- ma and the Pacific slopes of western Panama and southeastern Costa Rica
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
69
( Fig. 5 ) may have evolved ( along with several other species of reptiles and am- phibians) on insular isolates during the Middle Tertiary (Savage, 1966). This seems possible in view of both the pres- ent distribution of the species and its broad sympatry with festiva. The ap- parent absence of populations of Jepto- phrys on islands at present is a minor set-back to a theory of insular origin in that suitable habitat may have disap- peared from the islands in the relatively recent past, or that further collecting will reveal the presence of the species. Ameiva festiva occurs in both Mid- dle and South America ( Fig. 9 ) , as does A. ameiva, but festiva is continuously distributed over its continental range. Populations of festiva on the west and east ( or north ) slopes of the Andes have diverged markedly in color pattern ( Fig. 12 ) . On this basis, the population found in the Choco of Colombia seems to have diverged relatively early from the an- cestral stock. It may possibly have been isolated in the Choco region during a glacial period during the Pleistocene. Haffer ( 1967 ) documented similar iso- lation in birds and fomially proposed the name "Choco-Refuge" for the area. I have seen no specimens intermediate in pattern between those of the Choco (Pattern Type C; see page 31 for de- scriptions of pattern types) and speci- mens from populations to the north and east. Because of similarity in color pat- tern of festiva from eastern Colombia and those of the San Bias and Golfito- Puerto Armuelles regions of lower Mid- dle America, it seems likely that the latter are derived from the former, rather than from the Chocoan popula- tion. The species moved northward, through lowland forests along the At- lantic slopes. The evolution of Pattern Type A probably occurred in situ as a result of environmental selection pres- sures. The present distribution of fes- tiva in Middle America is limited to lowland forests. Expansion of its range in northern parts of the region has been
prevented by lack of suitable liabitat and possibly by competition from Amei- va undtdata and Cnemidophorus. Such expansion in South America has prob- ably been limited by the lack of suit- able habitat in Venezuela and eastern Colombia and possibly by competition from Ameiva hridgesi in western Co- lombia.
The early history of the Ameiva un- didata-q uadrdineata-chaitzami complex in Middle America may have been simi- lar to that of A. festiva. Having origi- nated from an unknown South Ameri- can ancestral stock, the forerunners of the extant species moved into Middle America and along both Atlantic and Pacific routes. Extensive northward movement along the Pacific lowlands was probably prohibited by ecological barriers, possil^ly the climatic filter bar- rier illustrated by Savage ( 1966, Fig. 23). Northward migration along the Atlantic slopes progressed to Nuclear Central America and thence, along two routes, on into what is now Mexico. Presumably, ecological divergence in Costa Rica and adjacent Panama, asso- ciated with exploitation of beach and open-ground habitats, resulted in the southern segment of the basal stock be- coming A. quadrdineata. The range of this species is now disjunct (Fig. 13), but in the past it probably extended along the Atlantic and Pacific slopes of the Western Sierra ( terminology of Ben- nett, 1968) in Panama and was joined in the lowlands of the isthmus. This contention is supported by the simi- larity of samples from opposite slopes in the eastern part of the ]3resent range of the species, a similarity not found be- tween samples from opposite slopes farther west. Populations of quadrdi- neata on Isla del Cano, Costa Rica, and Isla de Cebaco and Isla Escudo de Veraguas, Panama, apparently owe their presence to jiast land connections with the mainland during the Pleistocene (Bennett, 1968).
Ameiva widulata, which continued
70
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
to inhabit more forested areas than quadrilineata, migrated northward to fill its present range (Fig. 17). Smith and Laufe (1946), having to rely on Schnchert ( 1935 ) for geological back- ground, discussed the subsequent his- tory of the species. They postulated a broad distribution of tinchilata in Mid- dle America during the Miocene, with four secondary centers of dispersal, two on either side of the Isthmus of Tehuan- tepec ( Smith and Laufe, 1946, Fig. 4A ) . Further differentiation was promoted by the opening of the Tehuantepec portal during the Upper Miocene and Lower Pliocene and by subsequent migration, isolation or ecological divergence of populations. If we accept the South American origin of the family Teiidae (Dunn, 1931), and the timing of the reconnection of Middle and South America at the Isthmus of Panama as Early Pliocene (Lloyd, 1963; Maldo- nado-Koerdell, 1964), we must reject the hypothesis of Smith and Laufe and can eliminate from consideration the presence or absence of an open portal at the Isthmus of Tehuantepec. It seems probable that one line of undulata stock moved through Nuclear Central Amer- ica along the southern lowlands, cross- ing to the Atlantic lowlands through the Isthmus of Tehuantepec depression. Movement along the Pacific slopes was limited by xeric conditions to the north. Similar conditions in northeastern Mex- ico prevented further northward migra- tion on the Atlantic slopes. Eastward migration along the Atlantic lowlands from the Isthmus of Tehuantepec may have been blocked by unsuitable eco- logical conditions in the form of swamps, or by competition from a sec- ond line of undulata stock that had migrated along the Atlantic lowlands of Nuclear Central America to reach Guatemala and the Yucatan Peninsula. This second line would have given rise to the extant populations of undulata on Isla de Cozumel off the Yucatan Penin- sula and on Islas del Maiz east of Nica-
ragua when these islands were con- nected with the mainland. Migration of the ancestral stock southward from Nu- clear Central America into what is now Costa Rica would account for the pres- ent populations along the southern Pacific slopes and on the Meseta Cen- tral of Costa Rica. I have postulated that the first of the two lines gave rise to those populations having individuals with color patterns emphasizing stripes or combinations of stripes and lateral bars or blotches. The second line gave rise to those subpopulations character- ized by having barred flanks.
Ameiva chaitzami apparently evolved rather recently from a striped fonn of undulata, possibly because of ecological isolation in the arid valley of the Rio Grijalva and its tributaries. If further samples linking populations of chait- zami with undulata are found, it is pos- sible that study will show that chaitzami is a form of undulata, as inadvertently proposed by Smith and Laufe ( 1946 ) and that the sympatry of the two seen today in Guatemala is the result of sec- ondary contact.
Although I think it unlikely, studies of South American Ameiva may show that I have underestimated, rather than overestimated, the amount of diver- gence of the undulata complex from other species. If so, this complex, like Cnemidophorus, may be found to have its center of radiation in northern Mid- dle America and may be considered a component of the Middle American herpetofaunal element ( = "Hanging Relicts," Schmidt, 1943; "Autochthonous Middle American Element," Stuart, 1950 ) proposed by Savage ( 1966 ) . Such an arrangement would mean that the complex was isolated in Middle Amer- ica with the opening of the Panamanian portal and, with some adjustments in timing, would support the hypothesis of Smith and Laufe ( 1946 ) concerning evolution of Ameiva undulata itself in Middle America.
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
71
SUMMARY
The genus Ameiva is broadly dis- tributed in Middle and South America, as well as on islands throughout the Caribbean. The genus is represented in Middle America by six species: Aiiwiva ameiva (Linnaeus), leptophrys (Cope), f estiva ( Lichtenstein and Von Mar- tens), quadrilineata ( Hallowell ) , tin- dulata (Wiegmann) and chaitzami Stuart. These species are variously dis- tributed throughout the lowland tropics of Middle America from Nayarit and central Tamaulipas, Mexico, through Panama. Ameiva ameiva and f estiva, and possibly leptophrys occur in South America as well. Insular populations are known of all species, except lepto- phrys and chaitzami.
Geographic variation is extensive in scutellation and, in some species, color pattern. Sexual dimoiphism and onto- genetic change in color pattern is pro- nounced in some species. Certain char- acters, such as total number of supra- labials, number of dorsal granules from occiput to rump and scales around the tail, are sexually dimorphic in some species, not in others. On the other hand, several characters show no sexual dimorphism at all. Ontogenetic changes are most striking among males. Males of some species undergo marked changes in ventral coloration, especially in the gular region; these may be asso- ciated with reproductive state. Clines are apparent in a great many characters and, as might be expected in the context of the essential linearity of the Middle American land mass, most clines nm parallel to the coastlines. This is best illustrated by undulata, the species having the greatest range in Middle America and for which the most samples were available. In this species, the number of granules around the body and the number from occiput to rump proved the most sensitive indicators of differences between adjacent samples. These two characters are relatively
highly correlated. The numbers and arrangement of posterior head scales and of supraoculars were found to be unstable. The former is especially true of leptophrys and, to a lesser extent, festiva. The condition of a longitudinal- ly divided interparietal, occasionally used to diagnose some species of Amei- va, was found to be of wide occurrence in several species.
In Middle America, the species of Ameiva run an ecological gamut from the savanna-dwelling ameiva to festiva, which inhabits lowland tropical rain forest. In the presence of Cnemido- phorus, Ameiva tend toward less open habitats, although the same species may show more latitude in habitat choice if Cnemidophorus is absent. As many as three species of Ameiva occur sym- patrically in lower Central America. Competition among them is partially averted by partitioning of the available habitat. Typically, one species is a forest-dweller, another inhabits open areas and the third is intermediate. Size differences are also important in this context, and partitioning by subadults may be different than that by adults. Little is known of the demography of the Middle American Ameiva. All are bisexvial.
The species of Ameiva of Middle America have been assigned to three groups. These groups cannot be given formal taxonomic recognition until their relationships with South American spe- cies of Ameiva have been ascertained. The composition of the groups may have to be modified when these rela- tionships are better understood. Group I contains only Ameiva ameiva. This group has diverged considerably from Groups II and III, and can be distin- guished on the basis of a large number of characters of scutellation and color pattern. The group is primarily South American in distribution. Group II con- sists of festiva and leptophrys, both of
72
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
which have ranges centered in lower Central America. These two species have in common such characters as a high number of femoral pores and a tendency for the posterior head scales to be disrupted. Group III is made up of unchdata, quadrilineota and chait- zami and is entirely Middle American in distribution. The group is character- ized by certain details of scutellation and, especially, color pattern.
The genus Ameiva is thought to have originated in South America and to have invaded Middle America fol- lowing closure of the Panamanian por- tal in Early Pliocene. Of the Middle American species, Ameiva ameiva now occupies a disjunct range in central and western Panama and in South America east of the Andes. Ameiva festiva oc- curs on eastern and western sides of the Andes; it is thought that the eastern population gave rise to the Middle American population. The species pres- ently has a continuous range along the Atlantic lowlands to Tabasco, Mexico, and along Pacific slopes to southeastern Costa Rica. Based on present distribu- tion, leptophrys may have evolved as an insular isolate in the Panamanian portal. Ameiva quadriUneata is thought to be a lower Central American derivative of early iimhdata stock, and chaitzami a recently evolved offshoot of a striped form of umlulata. Forerunners of un- dulata itself are thought to have mi- grated northward on either side of Nu- clear Central America. Those moving along the Pacific routes gave rise to the populations in which striped, barred and blotched patterns are emphasized. Those moving through the Atlantic low- lands have given rise to those popula- tions in which lateral bars are empha- sized. Populations of tindulata from the Atlantic lowlands of Mexico between
central Tamaulipas and Campeche are derivatives of the line which moved along the Pacific routes and crossed to the Atlantic slopes through the depres- sion at the Isthmus of Tehuantepec. Without exception, insular populations of Ameiva in Middle America occur on islands of the continental shelf. These populations were isolated from main- land populations when rising sea levels severed connections with the mainland during the Pleistocene.
The genus Ameiva is the most wide- ranging of the exclusively terrestrial Neotropical lizards. Over its entire range, the genus has exploited a broad spectrum of available habitats. For these reasons, a better understanding of the ecological relationships of Ameiva may lead to a better understanding of tropical vertebrates in general. To date, very little ecological research has dealt with Ameiva. Interesting problems in- clude the relationships of characters of scutellation and color pattern to envi- ronmental parameters, the cause(s) and significance of color changes in adult males, and the ecological relationships of sympatric species of Ameiva and Cnemidophonis, regardless of whether or not these two genera are shown to be congeners. Taxonomic problems of im- portance to complete understanding of the Middle American Ameiva include elucidation of their relationships with South American species (and the rela- tionships of the South American species to one another), the relationships of chaitzami to undtdata, and the status of the population of undulata on the Me- seta Central of Costa Rica. These are but a few suggestions for future lines of research. There seem to be sufficient problems to keep a number of investi- gators occupied for years to come.
RESUMEN
El genero Ameiva esta ampliamente distribuido tanto por America Media y Sur America, como por las islas del
Caribe. El genero esta representado en America Media por seis especies: Am- eiva ameiva (Linnaeus), leptophrys
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
73
(Cope), f estiva ( Lichtenstein y Von Martens), qiiodrilineata ( Hallowell ) , y chaitzami Stuart. Estas especies se en- cuentran variablemente distribuidas al traves de las tierras bajas tropicales de America Media, desde Nayarit y Ta- maulipas central, Mexico, hasta Pana- ma. Ameiva amewa y f estiva, y posible- mente leptophrys ocurren tambien en Sur America. Se conocen poblaciones insulares de todas las especies excepto de leptophrys y chaitzami.
La variacion geografica de la escu- telacion y, en algunas especies, el patron de coloracion es amplia. El dimorfismo sexual y el cambio ontogenetico del patron de coloracion es pronunciado en algunas especies. Ciertos caracteres, como el mimero total de supralabiales, numero de granulos dorsales, desde el occipito a la anca, y las escamas alrede- dor de la cola, son sexualmente dimor- ficos en algunas especies pero no en otras. Por otro lado, algunos caracteres no presentan dimorfismo sexual del todo. Los cambios ontogeneticos son mas sorprendentes en los machos. Los machos de algunas especies sufren cam- bios marcados en la coloracion ventral, aspecialmente en la region gular, y estos se pueden asociar con el estado de re- produccion. Gradaciones son aparentes en muchos caracteres y en su mayoria, corren paralelas a las lineas costera, como se puede esperar del contexto del arreglo linear de la masa terrestre de la America Media. Esto puede ser ilus- trado mejor con undulata, la especie que tiene la extension mas larga por America Media y de la cual se tienen el mayor numero de muestras. En esta especie, el numero de granulos alrede- dor del cuerpo, y el numero desde el occipito a la anca probaron ser los mas sensibles indicadores de diferencias en- tre muestras adjacentes. Estos dos car- acteres estan relativemente muy corre- lacionados. El numero y arreglo de las escamas supraoculares y las posteriores a la cabeza es inestable. Lo primero es especialmente cierto en leptophrys y,
en menor grado, festiva. La condicion de poseer la escama interparietal divi- dida longitudinalmente, empleada oca- cionalmente en las diagnosis de algunas especies de Ameiva, fue encontrado siendo de gran ocurrencia en varias especies.
En la America Media, las especies de Ameiva pasan por una gama ecologica desde la ameiva que vive en las sabanas hasta la festiva que habita en las tierras bajas del bosque tropical lluvioso. En presencia de CnemidopJiorus, Ameiva tiende a vivir en habitats menos abier- tos, aunque la misma especie puede presentar mayor latitud en la escogen- cia del habitat si Cnemidophorus esta ausente. Hasta tres especies de Ameiva ocurren simpatricamente en la baja Cen- tro America. La competencia entre ellas se previene por medio de repartimiento del habitat utilizable. Tipicamente, una especie habita el bosque, otra habita las areas/ abiertas y la tercera es inter- media. Las diferencias en tamano tam- bien son importantes en este contexto, y la reparticion del habitat por los sub- adultos puede ser diferente de la de los adultos. Se sabe muy poco sob re la demografia de Ameiva de America Me- dia. Todas son bisexuales.
Las especies de Ameiva de America Media han sido asignadas a tres grupos. A estos grupos no se les puede dar re- conocimiento taxonomico formal hasta que sus relaciones con las especies de Ameiva de Sur America se determinen. La composicion de los grupos podra tener que modificarse cuando estas rela- ciones se comprendan mejor. El grupo I contiene solo a Ameiva ameiva. Este grupo se ha apartado considerablemente de los grupos II y III, ye se puede dis- tinguir en base al gran numero de caracteristicas en la escutelacion y pa- tron de coloracion. Este grupo es prin- cipalmente suramericano en distribu- cion. El grupo II consiste de festiva y leptophrys, ambas con distribuciones concentradas en la baja Centro Amer- ica. Estas dos especies tienen en comun
74
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
caracteres como un numero grande de poros femorales y la tendencia a poseer escamas rotas en la parte posterior de la cabeza. El grupo III se compone de undulata, qiiadrilineata y chaitzami y esta distribuido enteramente por Amer- ica Media. Este grupo esta caracteri- zado por ciertos detalles de escutelacion y, especialmente, el patron de colora-
cion.
Se cree que el genero Ameiva se origino en Sur America y que invadio la Ameria Media despues del cierre del paso panamefio en el Plioceno tempra- no. Ameiva ameiva posee una distribu- cion discontinua en Panama central y oeste, y en Sur America al este de los Andes. Ameiva f estiva ocurre en los lados este y oeste do los Andes; se cree que la poblacion del oeste dio origen a la poblacion de America Media. Esta especie ahora tiene una distribucion continua a lo largo de las tierras bajas del Atlantico hasta Tabasco, Mexico, y a lo largo del Pacifico hasta el sureste de Costa Rica. Basado en la distribu- cion presente, leptophrys pudo haber evolucionado de un aislamiento insular en el paso panamefio. Se cree que quad- rilineata se derivo, en la parte baja de Centro America, de una linea temprana de undulata y chaitzami evoliciono recientemente de una forma de undu- lata listada. Se cree que los precursores de undulata misma emigraron hacia al norte por ambos lados de Centro Amer- ica unclear. Esos que tomaron las rutas del lado del Pacifico dieron origen a las poblaciones en las cuales las listas, barras, y manchas son enfatizadas. Las que se desplazaron por las tierras bajas del Atlantico dieron origen a las pobla- ciones en las cuales se enfatizan las barras laterales. Las poblaciones de undulata de las tierras bajas del Atlan- tico de Mexico, entre Tamaulipas cen- tral y Campeche, son derivadas de la linea que se movio a lo largo de las
rutas del Pacifico y que crusaron la de- presion del istmo de Tehuantepec para dirigirse a la region Atlantica. Sin ex- cepcion, las poblaciones insulares de Ameiva en America Media ocurren en islas de la plataforma continental. Estas poblaciones fueron aisladas de las pob- laciones del continente cuando los ni- veles del mar subieron durante el Pleis- toceno y separaron las conecciones con tierra firme.
El genero Ameiva es el mas amplia- mente distribuido de las lagartijas Neo- tropicales terrestres. Sobre su total dis- tribucion, el genero ha expotado un gran espectro de habitats utilizables. Por estas razones, una mejor compren- sion de las relaciones ecologicas en Ameiva puede llevar a una mejor com- prension de los vertebrados tropicales en general. Hasta esta fecha, muy poca investigacion ecologia a tratado sobre Ameiva. Problemas interresantes inclu- yen las relaciones de los caracteres de escutelacion y el patron de coloracion con parametros ambientales, la causa(s) y significado de los combios de colora- cion en machos adultos, y las relaciones ecologicas entre especies simpatricas de Ameiva y Cnemidophorus, indiferente- mente de presentar o no a los dos ge- neros como si fuesen congenericos. Problemas taxonomicos de importancia para completar la comprension de Am,- eiva de America Media incluyen la aclaracion de sus relaciones con las especies de Sur America (y las rela- ciones de las especies suramericanas entre si), las relaciones entre chaitzami y undulata, y la posicion relativa o status de undulata en la Meseta Central de Costa Rica. Estas son no mas que algunas sugerencias para futuras lineas de investigacion. Parece haber sufici- entes problemas para mantener ocupa- dos a un numero de investigadores durante varios aiios.
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
75
APPENDIX A
An example of summarized data as used in this report, along with results of Gabriel's Simultaneous Test Procedure analysis for number of femoral pores of Pana- manian Ameiva amewa. For a complete set of the statistical results of this study, order NAPS Document #01660 from CCM Information Corporation, 866 3rd Ave- nue, New York, New York 10022; remitting $2.00 for microfiche or $7.00 for photo- copies.
Ameiva amewa: Femoral Pores
|
Sample Number |
X |
Standard Deviation |
Sample Size |
Range |
STP |
||
|
6 |
35.0 |
2.76 |
23 |
29-40 |
|||
|
5 |
34.1 |
1.64 |
11 |
31-36 |
|||
|
7 |
32.4 |
1.85 |
20 |
29-37 |
|||
|
3 |
32.3 |
2.07 |
24 |
28-36 |
|||
|
2 |
31.8 |
1.93 |
21 |
29-35 |
|||
|
1 |
31.7 |
2.61 |
40 |
25-38 |
|||
|
4 |
30.1 |
1.28 |
13 |
27-32 |
|||
|
All |
32.4 |
2.69 |
270 |
25-41 |
76
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
APPENDIX B
Specimens Examined
Ameiva ameiva
Sample 1. PANAMA: Canal Zone: An- con, MCZ 19805-11, UMMZ 98415-16, 98418; Balboa, AMNH 37129, FMNH 13441, UMMZ 98411-14; Bruja Point, FMNH 13445; nr Corozal, MCZ 19813; Diablo Heights, UMMZ 98417; 6.4 km E Lacone, KU 62701; Mira- flores, FMNH 16584-5. Panama: nr City of Panama, MCZ 9924-27, 9929-30, 9933-41; Panama City, El Cangrejo, KU 96742-43; Tapia, AMNH 28005-6, 28008.
Sample 2. PANAMA: Canal Zone: Gatiin, FMNH 16555-57, 16572-74, 16580, 16582-83; Cristobal, FMNH 34163-64. Colon: Achiote, KU 76109-18, 96737.
Sam^ple 3. PANAMA: Panama: Bejuco, KU 107546; Nuevo Gorgona, AMNH 89925- 46; ca 12 km NW San Carlos on road to El Valle, KU 107547.
Sample 4. PANAMA: Panama: Archi- pelago de las Perlas, Isla San Jose, AMNH 66491, USNM 120611-23.
Sample 5. PANAMA: Herrera: Las Minas, ANSP 22452; Parita, USNM 127291-94, 127296; 8 km SE Parita, KU 107538-39; Pese, KU 107540-45.
Sample 6. PANAMA: Los Santos: Gua- nico, KU 107521-35; Playa Guanico, near mouth of Rio Guanico, KU 107909; Las Pal- mitas, USNM 148207-11, KU 110731-34.
Sample 7. PANAMA: Panama: S Slope Cerro Campana, KU 76094-106, 107548-54.
Additional Specimens Not Used In STP Analysis. COSTA RICA: No definite locality, BMNH 80.6.21.1-2. PANAMA: Canal Zone: Alajuela, UMMZ 76104(4); Ancon, KU 110740; Bas Obispo, USNM 54129; Camp Chagres, KU 76107-8; Chagres, USNM 5519; Culebra, FMNH 16559-61; Culebra Cut, FMNH 16554; Fort Kobbe, KU 110739; Fort Kobbe, Venado River, KU 110742-43; El Vigia, UMMZ 76012; Frijoles, UMMZ 63760; Gamboa, AMNH 32818; Juan Mina, Chagres River, KU 108258-61; Lower Chagres, AMNH 37073-4, 37077; Madden Dam, UMMZ 76015(2); Madden Forest Preserve, KU 96740-41; Rio Abajo, FMNH 130685; Rio Frijoles, USNM 8394; Road K9, FMNH 130972; Rosseau, KU 62702; Summit, FMNH 130683, 131309; nr Summit Experimental Gardens, UMMZ 95484; Tabernilla, FMNH 16585-87, USNM 54130. Chiriqui: 2.5 km E La Concepcion, KU 96736; Remedios, AMNH 28390; 3 km W Rio Chorcha, KU 95540-44; 16 km W Rio Chorcha, KU 95545; Rio Salado, KU 95546-47. Cocle: Bank of Rio Chico de Nata, KU 95548-52; El Valle, AMNH 89924; 16 km S, 9 km W Penonome, KU 107536; 19 km S, 9 km W Penonome, Puerto Posoda, KU
107537. Colon: Portobello, USNM 65122; Quebrada Bonita, KU 96738-39; Santa Rosa, AMNH 71705. Herrera: Tres Puntas, ANSP 22470-74. Los Santos: Jobero, KU 110735- 36; Pocri, KU 107520; Punta Mala, AMNH 71704. Panama: Cafiitas, UMMZ 124148-50; Cerro Azul, AMNH 89947-48; Chorrera, AMNH 71703, FMNH 57597-98; Flora de Lagunas (Rio Caiiitas), UMMZ 124146-47; Gatuncillo, UMMZ 76013; Isla San Miguel, MCZ 9942-47; Isla Taboga, FMNH 16576, USNM 54197, 54205, 102754-56; Isla Tabo- guilla, USNM 102809-10; Juan Diaz, KU 110737-38; Nuevo Emperador, FMNH 130937, 130939-45, KU 110471; Pacora, AMNH 32817, FMNH 120956; Rio Bayano, USNM 53980; Rio Mamoni, FMNH 16566-68; Rio Mamoni, 5 km E Chepo, KU 76119-25; Rio Tapia, 24 km E Panama City, AMNH 42953. Veragita^-: 26 km W Santiago, KU 95553-55; Sona, USNM 51972-73.
Ameiva chaitzami
Sample 1. MEXICO: Chiapas: Comitan, UMMZ 94905(24), 94906(5), 94907.
Sample 2. GUATEMALA: Htiehuete- nango: San Antonio Huista, UMMZ 120192; 1 km N San Antonio Huista, UMMZ 120189, 120190(4), 120191(2); 1 km W San Antonio Huista, UMMZ 120195; 1-2 km W San An- tonio Huista, UMMZ 120193(18); 1-2 km E San Antonio Huista, UMMZ 120194(4).
Sainple 3. GUATEMALA: Alta Verapaz: Languin-Cahabon road, MCZ 52170; On Lan- guin-Cahal^on road ca 2 km N Finca Canihor, UMMZ 90638-41, 90643. El Peten: 1 km N Popti'm, UMMZ 124355; 2 km SW Poptun, UMMZ 124356; 2 km NE Poptun nr source Rio San Juan, UMMZ 124357.
Additional Specimens Not Used In STP Analysis. MEXICO: Chiapas: La Libertad, MCZ 56007, USNM 137216; La Libertad, nr Rio Cuilco where it crosses the Guatemalan border, FMNH 100006, 100712-15; unknown, but pro1ia]:)ly near La Libertad, FMNH 100716.
Ameiva festiva
Sample 1. GUATEMALA: El Peten: Piedras Negras, FMNH 115429-43, USNM 133826-36, UIMNH 11349-62.
Sample 2. GUATEMALA: Alta Verapaz: Chinaja, KU 55839, 55849-50, 59566; ca 3 km S Chinaja, KU 55841-48, 55857. El Peten: 6 km NNW Chinaja, KU 55836; 11 km NNW Chinaja, KU 55837-38; ca 15 km NW Chinaja, KU 55851-56, 59567-73; Toocog, 15 km SE La Libertad, KU 55840.
Sample 3. GUATEMALA: Alta Verapaz:
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
77
Finca Chama, UMMZ 91319, 91321(2), 91322(2), 91323(3), 91324(2), 91325(2), 91326(4), 91327, 91328(3), 91329(2), 91330, 91331(8), 91332.
Saiyiple 4. HONDURAS: Atlantida: ca 15 km E La Ceiba, mtns above Corozal, JRM 2299, 2418-21, 2439-40, 2466-69; Along Rio Cangrejal ca 12 km SSE La Ceiba, KU 101195-97; Tela, Lancetilla, UMMZ 70324(3).
Sample 5. NICARAGUA: Zelaya: Bo- nanza, KU 84870-72, 86039-40, 101201-12; 6 km E Bonanza, KU 101198; On banks of Rio Tunkey 4 km E Bonanza, KU 101199-200; Eden Mine, ANSP 2133-35.
Sample 6. NICARAGUA: Zelaya: Cukra, AMNH 16587-96; Kanawa, AMNH 16597, 16605-16; Between Cukra and Kanawa, AMNH 16600, 16602-4.
Sample 7. NICARAGUA: Zelaya: El Recreo, 16 km W Rama on Rio Mico, UCLA 9769-70; El Recreo, S side Rio Mico, KU 101874; 16 km above El Recreo on Rio Mico, UMMZ 79838(12); 11.2 km above Rama, UMMZ 79840(5), 79841(2); Junctions Rio Mico and Rio Siquia, UMMZ 79839(4).
Satnple 8. NICARAGUA: Rio San Juan: Tuli Creek, AMNH 16666-74, 16676-700.
Sample 9. COSTA RICA: Heredia: Puer- to Viejo, KU 67298-309, 91787, 104081-82; UMMZ 125505; 1.5 km N Puerto Viejo, KU 67310-11; 2 km NE Puerto Viejo, KU 95558- 68; 3 km S Puerto Viejo, KU 95556-57; ca 12.8 km NE Puerto Viejo along Rio Sarapiqui, UMMZ 125506(2); Rio Sarapiqui at Puerto Viejo, UMMZ 123582-83; Finca La Selva on Rio Puerto Viejo, KU 100683-85.
Sample 10. COSTA RICA: Cartago: Tur- rialba, FMNH 101906, 103107, MCZ 55091- 92, 55094, USNM 37737, KU 25633-34, 25636-39, 34783-89, 34826-27, 40552, UMMZ 117492; 5 km SE Turrialba, KU 25174-75; 6.4 km SE Turrialba, KU 34778-79; ca 4 km SW Moravia de Turrialba, KU 67312; 4.8 km W Turrialba, Hacienda Florencia, KU 34790- 91.
Sajnple 11. COSTA RICA: Limon: Vi- cinity of mouth of Rio Tortuguero, AMNH 89189-203, 89205-15.
Sample 12. COSTA RICA: Limon: Gua- piles, MCZ 15335-36; La Lola, KU 34798-801, 34807-13, UMMZ 117493(9); Los Diamantes, FMNH 101250, 101252-54, 101256-59, 101261-63, 101904-5, 101907, KU 25581-86, 25619-32, 34772-76, 34802-6; Puerto Limon, KU 34797; Zent, MCZ 15376-81.
Sample 13. PANAMA: Bocas del Tow: Almirante, FMNH 154445-48, 154451, MCZ 19785-87, KU 80207-11; 3.2 km NW Almi- rante, KU 96745-46; 4.8 km NW Almirante, KU 96747; 5.6 km NW Almirante, UU 5162; 9.6 km NW Almirante, KU 96748; 12.8 km NW Almirante, KU 96749-52.
Satnple 14. PANAMA: Bocas del Tow:
Isla Escudo de Veraguas, KU 108262-71, UU 5153-54.
Sample 15. PANAMA: Canal Zone: Bar- ro Colorado Island, AMNH 6.3401, 89949, FMNH 13342-43, MCZ 20612-13, 28066-67, KU 76126-27, UMMZ 63746-55, 63766, 63779, 124156.
Sample 16. PANAMA: Darien: Rio Mono, ca 5 km above junction with Rio Tuira, KU 96781-84; Rio Tuira at Rio Mono, KU 96785-804.
Sample 17. COSTA RICA: Puntarenas: Agua Buena, Caiias Gordas, KU 40545-48, 40553-58; Golfito, KU 67313; 2.5 km ESE Piedras Blancas, KU 93978; Rio Zapote, 8 km E Palmar Norte, KU 93977, 95569.
Sample 18. COSTA RICA: Puntarenas: 3 km NW Rincon de Osa, KU 101487-88; 7.2 km SE Rincon de Osa, KU 101489-90; 9.6 km NW Rincon de Osa, KU 101483-86; 10 km SSE Rincon de Osa, KU 101473-80; 11.7 km SSE Rincon de Osa, KU 101481-82.
Sample 19. COLOMBIA: Cordoba: Cafio Guarmal, FMNH 165260-66; Upper Rio Ure, FMNH 165225, 165227-30, 165748-67, 165769.
Sainple 20. COLOMBIA: Choco: An- dagoya, USNM 124246; Andagoya-Condoto Area, UMMZ 121463; Condoto, UMMZ 121465(13); Tado, ANSP 26265-66.
Additional Specimens Not Used In STP Analysis. BRITISH HONDURAS: Cayo: Valentin, UMMZ 92372; Stann Creek: Mid- dlesex, FMNH 4460; Toledo: 1.6 km E Swa- zey Branch of Monkey River, MCZ 71606-7; District Unknown: Silk Grass Creek State For- est, FMNH 4459; Bobowina, FMNH 49313-14. COLOMBIA: Boyaca: Muzo, MCZ 56233; Choco: Boca de la Raspadura, AMNH 18266- 68; El Valle, Bahia Solano, Mutis, USNM 151512; El Valle, Pizarro, USNM 151513-14; Pizarro, FMNH 43816-28; Rio San Pado, 1.6 hn below Boca de la Raspadura, AMNH 18274; Si- erra Bauda, ANSP 25526; Sierra de Darien, ANSP 25527-28; Upper Rio San Juan, FMNH 15652-53; Santander: El Centro Santander, ANSP 25075, 25427; Departamento Unknown: E base Cordillera de Bogota at extreme limit of llanos, ANSP 24165; "Tambo S. Monica, Atlantic Slope," AMNH 18265; Region of Truando, ANSP 9071. COSTA RICA: Ala- juela: 3.2 km NE Muelle Canal, UCR 23-26; Isla Bonita, FMNH 103103-6; Cartago: nr Peralta, Lake Bonilla, KU 43886-8; nr Peralta, Tunnel Camp, KU 34828, 38825; Reventazon, USNM 37492; Rio Pacuare, nr Pacuare, KU 34777; Suiza, FMNH 101251, 101255, KU 25635, 34792; Turrialba, FMNH 101909; 6.4- 8.0 km S Turrialba, FMNH 101260; Guana- caste: Laguna de Arenal, UCR 232-33, 268- 69; Tenorio, KU 34814-24; Tilaran, ANSP 24561; 5 km ENE Tilaran, KU 40551; Tron- cadero, nr Lake Arenal, KU 40539-40; Here-
78
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
dia: Barro Colorado, KU 100679-80; Behveen Puerto Viejo and Cariblanco, KU 40537-38; Limon: Bataan, KU 34780-82; Pandora, KU 86573-76, 100681-82; Suretka, MCZ 19777, KU 40541-44, 40559-63; Puntarenas: Agua Buena, Caiias Gordas, KU 40549-50; San Jose: San Isidro del General, FMNH 101908, 101910, KU 34793-94; 8 km E San Isidro del General, FMNH 120236; 24-32 km NW San Isidro del General, San Juan Mtns., KU 34795- 96; Province Unknown: Guayabo, FMNH 6167; Rio Frio, USNM 19534-36; Locality Unknown, USNM 30558, 32612-13. GUATE- MALA: Alfa Verapaz: Cubiliquitz, UMMZ 91334, 91335(2), 91336(2); Finca Chama, MCZ 28452-54, UMMZ 91320(2), 91329, 91331; Panzos, UMMZ 9137; Samazana, UMMZ 91333(4); El Quiche: Finca Tesoro, UMMZ 89197(4); Izabdh "Babos Pi., Mt. Playita," FMNH 20081, 20109-11; Quiriqua, FMNH 20487-88; El Peten: Pacomon, USNM 71394; Poptim-San Luis Road, ca 15 km S Poptim, UMMZ 124359; 12 km S PopUm, UMMZ 124360; 10 km NW Poptun, UMMZ 124.361(5); Ramate, USNM 71406, 71817, 71823, 71826-27, 71829-31, 71836; Localiiij Unknown: MNHN 5480. HONDURAS: Co- pan: 19.2 km ENE Copan, JRM 2895, 2897; Cortes: Hacienda Santa Ana, San Pedro Sula, FMNH 5067-71; Mtns W of San Pedro Sula, FMNH 5074; Garcias a Dios: Rio Segovia, USNM 24527-28; Santa Barbara: Jaral, FMNH 5073; Yoro: Mataderos Mountains, FMNH 21781, MCZ 38924; Portillo Grande, MCZ 38925; Subirana Valley, MCZ 38926; De- partamento Unknown: Carmelina, USNM 62972; Lake Ticamaya, FMNH 5072. MEX- ICO: Chiapas: Palenque, USNM 133818-25, UIMNH 11343, KU 94107-8; Tabasco: Teapa, USNM 47454-56. NICARAGUA: Rio San Juan: Greytown, USNM 19533-34, 19640-1; Zelatfa: 3-4 km W Bluefields, KU 101213-15; Cupitna, AMNH 16633-51; Rio Escondido 70 km from Bluefields, USNM 19735, 19871. PANAMA: Boca,s- del Toro: 11 km NW Al- mirante, FMNH 68171; Bastimentos, USNM 150021; Boca del Drago, USNM 142332-33; Cayo Agua, USNM 150014-15; Rio Cahuita, ca 5 km above mouth, KU 108278-79; Rio Changena, FMNH 130971, KU 110744-45; Canal Zone: Cerro Azul, FMNH 130684; Cristobal, FMNH 34165; Gamboa, FMNH 131308; Gatun, FMNH 16654-56, USNM 54134; Chiriqui: Boquete, UMMZ 58296; El Volcan, USNM 129924; Finca Palosanto 7 km NW El Volcan, Rio Chiriqui Viejo, KU 96808- 11; Finca Santa Clara, KU 110753-54; Volcan Chiriqui, UMMZ 101787; Code: El Valle, FMNH 47456; Finca El Valle de Anton, FMNH 60174; CoUn: Achiote, KU 76129-30; Candelaria Hydrographic Station, FMNH 68172-73; Portobello, FMNH 16652-53, USNM 54072-73, 80044; 3.5 km SE Puerto Pilon,
KU 110755; Rio Candelaria nr Hydrographic Station, FMNH 68174-76; Darien: Cafia, USNM 50134, 50136, 50138-39, ,50142-43; El Real, KU 80529; Laguna, KU 76131; Rio Cupe, ca 12 km SSW Boca de Cupe, KU 96805; Rio Esnape, Sambu Valley, MCZ 17224-29; Rio Jaque, 1.5 km above Rio Imamado, KU 110751-52; Rio Jesuito, Sapo Mtns, MCZ 17179-80, 17189; Rio San Antonio nr base of La Jarcia, KU 110750; Tacarcuna, KU 76132-39; Tacarcuna Village, USNM 141819-20; Panama: S Slope Cerro Campana, KU 76128; Cerro Tefe, KU 96806-7; E Slope Cerro Jefe, KU 80530-37; NW Slope Cerro Prominente, KU 80538-39; Chepo, FMNH 16656; El Valle de Anton, trail to La Mesa, FMNH 68177-79; Rio Abajo, FMNH 57599; Rio Siluganti, UMMZ 124151-55; San Bias: Camp Sasardi, KU 110748-49; Camp Summit, KU 110746-47; Veraguas: Mouth of Rio Concepcion, KU 108272-77; Province Uti- known: Cabima, USNM 48504; Chico, UMMZ 76011; Nombre de Dios, MCZ 26944; Punta de Pena, USNM 38693-98; "Veragua," ZMB 881a, 881b.
Ameiva leptophrys
Sample 1. COSTA RICA; Puntarenas: Camp Seattle, 6.4 km inland from Rincon de Osa, USC(CRE) 752; Vicinity of Rincon de Osa, KU 101491-528, 101825.
Sample 2. COSTA RICA: Puntarenas: Goto, Colfo Dulce, UMMZ 71999-72002; Es- quinas Forest Preserve between Golfito and Palmar, KU 34766-70; 3.2-4.8 km W Palmar Norte on road to Puerto Cortez, USC(CRE) 7101c-7101f; 6 km W Palmar Norte, KU 93982-83, 95570-82; 6 km SE Palmar Norte, KU 67314-16; Rio Zapote, 8 km E Palmar Norte, KU 93980-81; Villa Neilly, nr town of Corredor, USC(CRE) 179a; 12.3 km WNW \'il]a Neilly, KU 67317.
Sample 3. PANAMA: Canal Zone: Barro Colorado Lsland, AMNH 47018-19, 63403(2), 89953, ANSP 23021-22, FMNH 13307, 1.3344- 47, 13375-76, MCZ 28063-65, 29386, KU 76140-45, 80540, UMMZ 63756-59, 63780; nr Coro/al, MCZ 19814; Corozal Hill, USNM 102743.
Sample 4. PANAMA: Darien: Rio Chu- cunaque, ca 7 km above Rio Morti, KU 107561-63; Rio Cupe, ca 12 km SSW Boca de Cupe, KU 96815; Rio Tuira at Rio Mono, KU 96816-25; Rio Ucurganti, ca 7 km above mouth, KU 107564-65; Tacarcuna, KU 76158- 61; Tacarcuna Village, USNM 141817-18; Three-Falls Creek, AMNH 37870.
Additional Specimens Not Used In STP Anahisis. COSTA RIGA: Puntarenas: Par- rita. La Tulieta, Unca La Ligia, USC(CRE) 8256, 8260]>8260c; 2.5 km ESE Piedras Blan- cas, KU 93979; Rincon de Osa, RDS 1122-23, 1170, 1177; San Jose: San Isidro del General, FMNH 103111, 103117, KU 25640-42; Prov-
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
79
ince Unknown: Pozo Azul de Parris, ANSP 21470; Santo Domingo de San Mateo, USNM 37754. PANAMA: Canal Zone: Alajuela, UMMZ 76009; Camp Chagres, KU 76146-48; Corozal, AMNH 67073; Gatim, USNM 54142; 0.5 km S Juan Mina, FMNH 68159; Madden Dam, UMMZ 76010; Miraflores, FMNH 16650; Rio Cardenas, Corozal, USNM 53868; Chiri- qui: Progreso, MCZ 19865; 13 km N Puerto Armuelles, KU 96814; Code: El Valle, AMNH 89950-52, USNM 129906, KU 76162-63; Colon: Achiote, KU 76149-51; Portobello, USNM 92593; Quebrado Bonita nr Buenavista, KU 96812; 1 km W Rio Piedras, KU 96813; San Bias Point, AMNH 71706; Darien: Cana, USNM 50135, 50137, 50140-41; Canclones, UMMZ 124975; nr mouth of Rio Canclones, UMMZ 124974; Taque, KU 110767-68; La- guna, KU 76156-57; Mt. Sapo, ANSP 22963, MCZ 17201-3; Rio Jaque, 1.5 km above Rio Imamodo, KU 110769-70; Rio Jesuito, Sapo Mtns, MCZ 17196-97, 17199; Herrera: Las Minas, ANSP 22450; Los Santos: N Slope Cerro Cambutal, KU 107555-56; Guanico Abajo, KU 110771; Panama: Cerro Azul, USNM 54136, 54147; Cliffs W of Cerro Cam- pana, FMNH 60178-79; S Slope Cerro Cam- pana, KU 76152-53, 107558-59; Chepo, FMNH 16645; Las Cumbres, KU 96827; Nuevo Em- perador, FMNH 130938; nr Panama City, MCZ 10925, 10927; 2.8 km ENE Panama City, Rio Pacora, KU 96826; 9 km NNE Pa- cora, Rio Pacora, KU 107557; Rio Mamoni, 5 km E Chepo, KU 76154-55; Rio Siluganti, UMMZ 124144-45; Rio Tocumen, MCZ 19816; San Bias: Camp Sasardi, KU 110756-65; Camp Summit, KU 110766; Veragtias: Mangillo, ANSP 22450.
Ameiva quadrilineata
Sample 1. NICARAGUA: Rio San Juan: Greytown, USNM 19537-45, 20741-42, 32230; Rio San Juan, USNM 24984.
Sample 2. COSTA RICA: Limon: Colo- rado Bar, AMNH 16726-28; Tortuguero, USC(CRE) 2653(2); Vicinity of mouth of Rio Tortuguero, AMNH 89217-46.
Sample 3. COSTA RICA: Liinon: Los Diamantes, UMMZ 117491; La Lola, USC (CRE) 128(3), 8063, 8070(3); Zent, MCZ 15382-85.
Sample 4. PANAMA: Bocas del Tow: Almirante, FMNH 60171-72; Careening Cay, KU 96840-44; Cayo Zapatillo Grande, KU 96845-51; Isla Bastimentos, Bastimentos, KU 96831-39; Isla Bastimentos, Punta Vieja, USNM 150016-20, 150022-23; Isla Colon, road from Bocas to la Gruta, KU 96829-30.
Sample 5. PANAMA: Bocas- del Tow: Isla Escudo de Veraguas, KU 108280-85, UU 5166.
Sample 6. COSTA RICA: San Jose: San Isidro del General, FMNH 103115, KU 25643-
50, 34685-89, USC(CRE) 2742; 4.8 km NE San Isidro del General, USC (CRE) 2843; 5.4 km SW San Isidro del General, UMMZ 117490(2); 3.2 km W San Isidro del Gen- eral, KU 25651; 8 km SE San Isidro del General, KU 95583-90.
Sample 7. COSTA RICA: Ptmtarenas: 1.5 km E Palmar Norte, KU 95592-93; 6 km W Palmar Norte, KU 95633-40; 3.2-4.8 km W Palmar Norte on road to Puerto Cortez, USC (CRE) 7101a-7101b; 5 km SE Palmar Sur, KU 67318-19; 6 hn SE Palmar Sur, KU 67320-25; Rio Zapote, 8 km E Palmar Norte, KU 95607-20, 95622-26, 104083-85.
Sample 8. COSTA RICA: Puntarenas: Rincon de Osa, RDS 1136-38, 1144, 1147, 1152, 1154-63, KU 101463-69, USC (CRE) 7239(14); 3 km NW Rincon de Osa, KU 101470-72; 11.7 km SSE Rincon de Osa, KU 101459-62.
Sample 9. PANAMA: Chiriqui: Comarca del Baru, Puerto Annuelles, FMNH 68166-67, Progreso, UMMZ 58186-202; 8 km S Progreso, KU 96852; Puerto Annuelles, MCZ 45668-71; 13 km N Puerto Annuelles, KU 96853-65.
Sample 10. PANAMA: Chiriqui: Bo- quete, ANSP 21941-43, UMMZ 58183-85, 58209-18.
Additional Specimens Not Used In STP Analysis. COSTA RICA: Li7n6n: Colorado Bar, AMNH 16729-75; La ForUina, AMNH 99679; Old Harbor, USNM 32614-16; Suretka, KU 40578; Puntarenas: 0.8-1.6 km E Bani along N bank of Rio Baru, USC (CRE) 7096; Buenos Aires, FMNH 2519; Isla del Cano, UMMZ 71194; Goto, UMMZ 72005-6; El Gen- eral (Viejo) between Palmar Sur and Golfito, KU 34690; Esquinas, KU 34691; Golfito, USC(CRE) 7111(5), 7233(3); Km 47 on railway to Golfito, USC (CRE) 176(8), 177 (2), 178; Finca El Helechales, 15 km NE Potrero Grande, USC (CRE) 8268, 8271; 27 km E Palmar Norte, Rio Punta Nuevo, KU 95596-606; 30 km E Palmar Norte, Rio La Vieja, KU 95627-32; 2-5 km ESE Piedras Blancas, KU 93984-86; Puerto U\dla, UMMZ 72004; Rincon de Osa, RDS 1139-40, 1142, 1145-46, 1148-51, 1153; San Isidro del Gen- eral, FMNH 103116; 22 km SE San Isidro del General, USC (CRE) 2648; 57 km SE San Isidro del General, KU 95591; Villa Neilly, USC(CRE) 179; 1.6 k-m E Volcan de Buenos Aires, Cone Finca, UMMZ 117573; San Jose: San Isidro del General, 3.2 km W Intabee, FMNH 103102. NICARAGUA: Rio San Juan: Greytown, USNM 6053(2). PANAMA: Chiriqui: SW Slope Cerro Homito, KU 96867; 11 km NE Gualaco Road to Valle Homito, KU 96866; El Hato, edge of lava flow, FMNH 130686; Llano del Volcan above Paso Ancho, FMNH 60175-77, 68168-70; Herrera: Las Minas, ANSP 22454-55; Veraguas: 8 km SW El Maria, KU 107566; Isla Cebaco, KU 96828.
80
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
Ameiva undulata
Sample 1. MEXICO: Tamaulipas: Go- mez Farias, UMMZ 110802, 110839; nr Gomez Farias, UMMZ 111145; 8 km NE Gomez Farias along Rio Sabinas, UMMZ 101476-78, 101479(6), 101480-82, 101515-16, 104059 (7), 105489, 111142-44; Pano Ayiictle, 8 km NE Gomez Farias, UMMZ 98982.
Sample 2. MfiXICO: San Luis Potosi: Axtla, AMNH 67341; Hiiichihuayan, USNM 133859; Tamazimchale, AMNH 66065-7, FMNH 38611, UIMNH 27360, 51207; 3.2 km NE Tamazunchale, TCWC 4090-92, UMMZ
119819, UIMNH 16804; Xilitla Region, KU 24056-7.
Sample 3. MEXICO: Veracruz: Tierra Colorado, FMNH 75793, 126933-41, MCZ 56004.
Sample 4. MfiXICO: Veracruz: Cuaut- lapam, KU 105823, UMMZ 41423-32, 41434- 42, 41444-51, 41453-54.
Sample 5. MfiXICO: Veracruz: Cate- maco, UMMZ 118727; 0.8 km W Laguna de Catemaco, UMMZ 126412; 4 km NE Cate- maco, UMMZ 126409; Goyame, UIMNH 36877, 39217-18, UMMZ 111456, 114794; 1 km S Dos Amates, between Sontecomapan and Catemaco, UMMZ 126413; Laguna En- cantada, 3 km NE San Andres Tuxtla, UMMZ
119820, 126407; Rancho El Tular, ca 8 km N San Andres Tuxtla, UIMNH 39212; San An- dres Tuxtla, USNM 46899-904, UIMNH 24691, 27347, 28054, 28056; nr San Andres Tuxtla, FMNH 126709; 2 km NE San Andres Tuxtla, UMMZ 121156; ca 4 km N San An- dres Tuxtla, UIMNH 39212; San Martin Mountains, UIMNH 35460; Sontecomapan, UMMZ 114795, 126408.
Sample 6. MfiXICO: Veracruz: Hacienda La Oaxaquena, AMNH 62331-33; 20 km S Jesus Carranza, KU 23953, 24186; 25 km SE Jesus Carranza, KU 26956-57, 26959, 26962- 63, 26965-71; 35 km SE Jesus Carranza, KU 23950; 20 km ENE Jesus Carranza, KU 26955, 26961.
Sample 7. MfiXICO: Tabasco: Frontera, USNM 25091, 46659, 47453; 16 km SW Fron- tera, KU 95669-78.
Sample 8. MfiXICO: Campeche: Bal- chacaj, FMNH 106716-18, 106721, UIMNH 26167-70; Tres Brazos, FMNH 106719, UIMNH 26172-75, UMMZ 81924.
Sample 9. MfiXICO: Campeche: Dzil- balchen, KU 75589-98.
Sample 10. MfiXICO: Yucatan: Piste, KU 70598-600, 70604-5, 70609, 70611-15, 70617, 70620-26, 70628-29, 70631-32, 70634- 35, 70637-41.
Sample 11. BRITISH HONDURAS: Cayo: Central Farm, MCZ 71608-9; El Cayo, UMMZ 75012; 1.6 km NW El Cayo, USNM 71372-73; Xunantimich Ruins, MCZ 71619-21.
Sample 12. GUATEMALA: El Peten:
Tikal, UMMZ 117875(8), 117876(7); Uaxac- tim, AMNH 68507-16, 70938-42, UMMZ 70415-19.
Sample 13. GUATEMALA: El Peten: Piedras Negras, MCZ 66965, UIMNH 11363- 93.
Sample 14. GUATEMALA: Alta Verapaz: Canihor, UMMZ 91305(21), 91306(3), 91307 (4), 91308; Finales, UMMZ 91316.
Sample 15. GUATEMALA: Progreso: El Rancho, UMMZ 106994(11), 106995(6); Finca Bucural, UMMZ 107000(9), 107001(4).
Sample 16. HONDURAS: Copdn: Go- pan, UMMZ 83035(16).
Sample 17. HONDURAS: Atlantkla: 1 km W La Ceiba, KU 101227; 2 km SE La Ceiba on Rio Cangrejal, KU 101228-38; 8 km SE La Ceiba on Rio Cangrejal, KU 101239-42; 12 km SSE La Ceiba on Rio Cangrejal, KU 101252; ca 15 km E La Ceiba in mountains above Corozal, JRM 2318, 2441-42, 2460-61, 2470.
Sample 18. HONDURAS: Yoro: 0.5 km N Coyoles, JRM 2004-7; 2 km S Coyoles along Rio Aguan, KU 101216-26, 101254-59, 107911-14, 109973-74; 5 km E Coyoles, JRM 2082-89; 25 km WSW Coyoles, Rancho San Lorenzo, JRM 2116.
Sample 19. HONDURAS: Colon: 0.5 km SW Trujillo, JRM 2554; 1 km SSW Trujillo, KU 101249-51; 0.5-1.5 km W Trujillo, JRM 2511; 1-3 km W Trujillo, KU 101243-47; 2 km E Trujillo, jRM 2533-36; 2 km W Tru- jillo, KU 101248.
Sample 20. HONDURAS: Olancho: 0.5- 1.0 km WNW Catacamas, JRM 1433-37; 1 km NW Catacamas, JRM 1551; 1.5 km NW Catacamas, JRM 1554-55; 2-3 km NW Cata- camas, JRM 1526-29; 4.5 km SE Catacamas, Esquela Agricola Nacional de Agricultura, JRM 1586-87, 1622, 1628-30; 6.5 km SE Cata- camas, JRM 1638-39.
Sample 21. NICARAGUA: Zelat/a: Isla del Maiz Grande, AMNH 97045-67, 97637, MCZ 26970, 26972, 26974, USNM 94053, KU 85988, 85997, 85999, 101268, 101273-74, 101276-77, 101281-86, 101292, 101299-301, 101305-6, 101309-10.
Sample 22. MfiXICO: Colima: Hacienda Paso del Rio, FMNH 75794, 126436, 126438- 41, 126443-47, 126449-51, 167438, UMMZ 80118(2), 80119; Manzanillo, FMNH 100054; 1.6 km W Pascuales, UMMZ 80113(3); 0.4 km E Pascuales, UMMZ 80114; 6 km SW Tecuman, UMMZ 80110; 8 km SW Tecuman, UMMZ 80111(3), 80112(5), 80120.
Sample 23. MfiXICO: Colima: Colima, AMNH 5185, 15448-49; NW of Colima, AMNH 15838-42; 0.8 km SW Colima, AMNH 12634-35; 2.4 km SW Colima, AMNH 12643; 6.4 km SW Colima, KU 29529; near Salva- dor, MCZ 52172.
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
81
Sample 24. MfiXICO: Guerrero: Mtns N of Acapulco, FMNH 126351, 126355, UIMNH 27344; 8 km E Coyuca, USNM 133849-53; Between km 431-432, camp nr El Triente, FMNH 126353; 2 km N El Triente, UMMZ 119564; 1.6 km N Organos, S of El Triente, UIMNH 27345; E San Andreas de la Cruz, KU 87417.
Sample 25. MEXICO: Guerrero: Aca- huizotla, TCWC 7780, 9587-88, 11320, UMMZ 119563; Agua del Obispo, UMMZ 119565, TCWC 7575, KU 87413-16; nr Agua del Obispo between Rincon and Cajones, FMNH 126352, 126354, MCZ 56005; Mtns near Agua del OlMspo, TCWC 7576; 8 km SW Chilpancingo, TCWC 9590; 1.6 km SW Cot- otlipa, TCWC 9591.
Sample 26. MEXICO: Oaxaca: 6.4 km S Candelaria, KU 38234; Chacalapa, KU 38230- 33; 3.2 km W El Soledad, UIMNH 8421-27.
San^ple 27. MEXICO: Oaxaca: Tres Cruces, 32 km SW Tehuantepec, UMMZ 81897(2), 81898(3), 81899(2), 81900(3), 81901(5), 81902(4), 81903(5), 81904(6).
Sample 28. MEXICO: Oaxaca: 3.2 km E Tollocito, KU 39722-32, 39734, 44656-57.
Sample 29. MEXICO: Oaxaca: Tapa- natepec, AMNH 80002-3, MCZ 27283-300; 2.4 km S, 7.2 km E Tapanatepec, TCWC 17047-48; 4 km W Tapanatepec, KU 59577- 80; Rio Novillero, 4.3 km W Tapanatepec, UIMNH 39204-8.
Sample 30. MEXICO: Chiapas: Finca San Bartola, ca 19 km SW Cintalapa, UIMNH 8437-41, 39229-41; nr San Ricardo, FMNH 106715, UIMNH 26157.
Sample 31. MEXICO: Chiapas: Abel- lanal, 2 km SE Florida, TCWC 19764-65; Florida, 50 km E Altimirano, TCWC 19758- 63, 19766-86; Las Tazas, 51 km E Altimirano, TCWC 19757.
Sample 33. MEXICO: Chiapas: Esquint- la, LACM 9527-28; Distrito Soconusco, 6 km NE Esquintla, UMMZ 86862-71, 86873-88.
Sample 34. MEXICO: Chiapas: Sabana de San Quintin, KU 94109-17, 94119.
Sample 35. GUATEMALA: Solold: Fin- ca Santa Buenaventura, 1 km NW Panajachel, UMMZ 98215(2); Panajachel, MCZ 27515-20, KU 59581, 95653, UMMZ 98201(3), 98214 (3); 1 km E Panajachel, KU 95657; 3 km E Panajachel, KU 95654-56; 3 km NE Pana- jachel, UMMZ 98202; 1 km NW Panajachel, UMMZ 98203(2), 98204; 2 km NE Pana- jachel, UMMZ 98212; 4 km NE Panajachel, UMMZ 98213; Solola, UMMZ 98205(2); On road to Solola, 1 km from Panajachel, UMMZ 98206.
Sample 36. GUATEMALA: Jutiapa: Fin- ca La Trinidad, UMMZ 107421, 107422(8), 107423(5), 107424(3), 107425(6), 107427 (3), 107428, 107429(3).
Sample 37. GUATEMALA: Jutiapa: Ha-
cienda Mongoy, UMMZ 107002(6), 107003, 107004(3); jutiapa, UMMZ 106997(4), 106998, 106999(16).
Sample 38. EL SALVADOR: Libertad: 2 km SE Colon, KU 62608-70; Laguna de Chanmico, UMMZ 117486(2); 0.8 km N, 16 km W La Libertad, TCWC 17059; Rancho Belmar, nr La Libertad, LACM 9370. San Salvador: Instituto Tropical, San Salvador, KU 62066-67; 0.4 km NW Instituto Tropical, UMMZ 117487; 0.8 km NW Instituto Tropical, UMMZ 117488; 1.6 km NW San Salvador, KU 42262.
Sample 39. NICARAGUA: Chinamle^a: 4 km N, 2 km W Chichigalpa, KU 86036; Chinandega, MCZ 9540, 9546; Finca San Isidro, 10 km S Chinandega, KU 86030-31; Foothills, N Slope Volcan San Cristobal, KU 86032-35; Hacienda Bellavista, Volcan Casita, KU 101875-77; San Antonio, KU 86008-26.
Sample 40. NICARAGUA: Matagalpa: Guasquali, UMMZ 116420(3); Matagalpa, UMMZ 116414, 116418(5); 0.8 km E Mata- galpa, UMMZ 116416; 1.6 km W Matagalpa, UMMZ 116417, 116419; 2.4 km E Matagalpa, UMMZ 116415(2).
Sample 41. COSTA RICA: Guanacasie: Hacienda Coyolar, 4.8 km N, 4.0 km W Li- beria, USC(CRE) 8206(15), 8207(2); Haci- enda La Norma, 5 km N, 4.5 km W Liberia on Rio Colorado, USC(CRE) 100(2), 103 (3), 104, 105(3), 106(8), 107(4), 111, 251.
Sample 42. COSTA RICA: Guanacaste: Canas, USNM 80894; 1.5 km E Canas, USC (CRE) 208-9; 1 km S Canas, USC(CRE) 201; 7 km N, 3 km E Cai^as, USC (CRE) 270- 77; Finca Taboga, 11 km S Caiias, USC (CRE) 631, 7166(2); Rio Javillo, 2 km S Las Cafias, USC (CRE), 202-3.
Sample 43. COSTA RICA: Guanacaste: On first river N Santa Rosa, KU 40531-36; Road from Silencio to Tilaran, 1.8 km from Tilariin, USC (CRE) 6240; Road from Silen- cio to Tilaran, 2.6 km from Tilaran, USC (CRE) 6239(2); Tilaran, ANSP 24566-67; Rio Santa Rosa drainage 1.2 km by road SW Tilaran, USC (CRE) 7161; 1.6 km SW Tila- ran, USC(CRE) 6270; 2 km W Tilaran, USC (CRE) 732; 4.3 km by road NNE Tilaran, USC(CRE) 8082; 4 km ESE Los Angeles de Tilaran, KU 40527-28; 5.8 km from Tilaran on road to Los Angeles, USC (CRE) 6267; Troncadero, Laguna de Arenal, KU 40529-30.
Sample 44. HONDURAS: Francisco- Morazdn: El Zamorano, Esquela Agricola Panamericana, AMNH 70368, MCZ 49765-67, 10001-3, KU 101260-67; Rio Yeguare, AMNH 70369-79; Valley of Rio Yeguare S of Teguci- galpa, AMNH 70470-73.
Sample 46. MEXICO: Quintana Roo: Isla Mujeres, USNM 47568-70, 47651, KU 70585-88.
82
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
Sample 48. MEXICO: Michoacdn: Apat- zingan, FMNH 38977-85.
Additional Specimens Not Used In STP Analysis. BRITISH HONDURAS: Belize: Aml:!ergris Cay, 3.2 km N San Pedro, MCZ 71611; Belize, USNM 51879-80, 58373-78; Caijo: Augustine, Mountain Pine Ridge, MCZ 71622-24; Corozal: 0.3 km N Corozal, MCZ 71610; Orange Walk: Gallon Jug, MCZ 71612-18; Stann Creek: Mango Creek, MCZ 71627-28; 8 km SW Mango Creek, MCZ 71626; 6.4 km S Waha Leaf Creek, MCZ 71625; Toledo: 1.6 km W Svvazey Bridge, Monkey River, MCZ 71629. COSTA RICA: Alajuela: San Mateo, USNM 37491; "Auto- pista-AJajuela," UCR uncatalogued; Cartage: La Carpintera, AMNH 16314, 16316; Guana- caste: Bebedero, USC(CRE) 7162; 1.6 km N Guayabo de Bagaces, USC(CRE) 7024; Playa Samara, outer coast of Nicoya Peninsula, USC(CRE) 8252; Parrita, La Julieta, Finca La Ligia, USC(CRE) 8255(4), 8260, 8262; 1.6 km S Santa Cruz, USC(CRE) 8218; 1.8 km N Santa Cruz, USC(CRE) 8217(2); Puntarenas: 4.3 km E Esparta, UCR 547; San Jose: Barrio Mexico, San Jose, KU 40526. EL SALVADOR: La Paz: 3 km E San Ra- fael Obrajuela, KU 62065; San Miguel: San Pedro, MCZ 57081. GUATEMALA: Guate- mala: El Rosario, FMNH 68708-10; Izabal: 1.6 km NE York, KU 59574-75; San Marcos: Finca Carolina, MCZ 22144-48, 27301-5; San- to Rosa: Finca Las Vinas, UMMZ 107073-74; Departamento Unknotvn: MCZ 5831, 99840- 41. HONDURAS: Colon: Belfate, AMNH 58607-9; Comaijagua: La Libertad, MCZ 38923; Copdn: Hugito, ANSP 22195-98; Cor- tes: E side Lago de Yojoa, KU 67332; Rancho Agua Azul, Lago de Yojoa, MCZ 49964; 3.2 km W San Pedro Sula, MCZ 29387-88; El Paraiso: Arenal, 25 km E Jalapa ( Nicaragua ) , UCLA 14759; Valle de Jamastran, AMNH 70333-35; Francisco-Morazdn: Cantarranas. ANSP 22199-212; Gracias a Dios: Tancin, ca 15 km NW Puerto Lempira, JRM 1732, 1752. MEXICO: Campeche: 3.2 km E Barro San Pedro Y Pablo, AMNH 88877; Champoton, UMMZ 72956-57; Chiapas: 3.2 km E El Real,
KU 43663-64; Palenque, MCZ 66966; Across Rio Usimiacinta from Piedras Negras (Guate- mala), USNM 108600; San Jnanita, Palenque, USNM 108601; Colima: Queseria, UMMZ 80109; nr Queseria, FMNH 126437, 126442, 126448; Guerrero: nr Rincon, FMNH 100080; Oaxaca: La Concepcion, nr Tehuantepec, AMNH 66932; Cosolapa, USNM 133920; Ju- quila Mixes, AMNH 89761-65; Mitla, AMNH 91023-24; 9.6 km N Putla, UIMNH .52952-53; Rio Cacahuatepec, UIMNH 529,50; Rio Canoa, 26 km W Pinotepec Nacional, UIMNH .52951; Sta. Lucia, Tehuantepec, AMNH 65077-78, 65085; Tehuantepec, USNM 30163; Zacatepec, UIMNH 52948-49; Puehla: Necapa, Rio Ne- capa, AMNH 76441-43; nr Vegas de Xuchil, AMNH 58217; Quintana Roo: 8 km inland from Vigia, Ascension Bay, UMMZ 78586; San Luis Potosi: Ciudad Maiz, FMNH 126942; nr Cuidad Maiz, FMNH 126946; Palictla, AMNH 67358; 8 km S Valles, FMNH 126947; Tabasco: La Venta, USNM 117350; Tamau- lipas-: S of Antiguo Morelos, FMNH 126944- 45; Hacienda La Clementina, FMNH 126943, 126948, MCZ 56006, USNM 106141-42; Rancho Santa Ana, 12.8 km E Padilla, MCZ 43634; Rio Guayalejo, nr Magiscatzin, MCZ 45557, UMMZ 88232; 11 km W Victoria, FMNH 100050; Veracruz: Jalapa, FMNH 126706; Orizaba, USNM 133839; Panuco, MCZ 19255-59; 4 km E Papantla, KU 24189, 24194; San Andres Tuxtla, FMNH 100020; 16 km WNW Temapache, KU 61791-92; Yuca- tan: Chichen Itza, MCZ 52171; Culebra Cay, Ascencion Bay, UMMZ 78587; Progreso, FMNH 100030; Estado Unktwwn: "Mexico," ( Restricted to Tehuantepec, Oaxaca, by Smith, 1940), ZMB 868. NICARAGUA: Esteli: Finca Daraili, 5 km N, 14 km E Condega, KU 85984; Madriz: 1.6 km SE Yalaguina, TCWC 17051; Managua: 12 km E Managua, UCLA 14787; 6.4 km E San Lorenzo, KU 42249; Rio San Juan: Tuli Creek, AMNH 16675; Rivas: within 3 km of Moyogalpa, Isla Ome- tepe, KU 86029; Zelaija: Huaunta Haulover, ANSP 15438-41; Departamento Unknown: ANSP 9133-34.
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
83
LITERATURE CITED
Alvarez Del Toro, M.
1960. Los Reptiles De Chiapas. Institute Zoologico del Estado, Tuxtla Gutier- rez, Chiapas, Mexico, 204 pp.
Barbour, T. and Loveridge, A.
1929a. Vertebrates from the Corn Islands. Reptiles and Amphibians. Bull. Mus. Comp. Zool., 69:138-146.
1929b. Typical reptiles and amphibians.
Bull. Mus. Comp. Zool., 69:205-360. Barbour, T. and Noble, G. K.
1915. A revision of the lizards of the genus
Ameiva. Bull. Mus. Comp. Zool.,
59:417-479.
BaSKIN, J. N. and WlLLL«iMS, E. E.
1966. The Lesser Antillean Ameiva (Sau- ria, Teiidae). Re-evaluation, Zoo- geography and the Effects of Preda- tion. Studies Fauna Curacao and other Carib. Isls., 23:144-176.
Bennett, C. F.
1968. Human influences on the zoogeog- raphy of Panama. Ibero-Americana, Univ. California Press, No. 51, vii + 112 pp. Boulenger, G. a.
1885. Catalogue of the lizards in the British Museum (Natural History), vol. II. London, xiii -f 497 pp. Brennan, J- M- AND Yunker, C. E.
1966. The chiggers of Panama (Acarina: Trombiculidae ) . In: Ectoparasites of Panama ( R. L. Wenzel and V. J. Tipton, Eds.). Field Mus. Nat. Hist., Cliicago. pp. 221-266.
Burt, C. E.
1931. A study of the teiid lizards of the genus Cnemidophorus with special reference to their phylogenetic rela- tionships. Bull. U. S. Natl. Mus., 154:1-286. Cochran, D. M.
1946. Notes on the herpetology of the Pearl Islands, Panama. Smithsonian Misc. Coll., 106:1-8. Cope, E. D.
1875. On the Batrachia and Reptilia of Costa Rica. ]. Acad. Nat. Sci., Phila- delphia, ser. 2, 8:93-157.
1894. Third addition to the knowledge of the Batrachia and Reptilia of Costa Rica. Proc. Acad. Nat. Sci. Phila- delphia, 1894:194-206. Donoso-Barros, R.
1968. The lizards of Venezuela (check list and key). Carribean J. Sci., 8:105- 122.
DUELLMAN, W. E.
1958. A monographic study of the colubrid
snake genus Leptodcira. Bull. Amer- ican Mus. Nat. Hist., 114:1-152. pis.
1-31. 1961. The amphibians and reptiles of
Michoacan, Mexico. Univ. Kansas
Publ. Mus. Nat. Hist., 15:1-148. 1965a. Amphibians and reptiles from the
Yucatan Peninsula, Mexico. Univ.
Kansas Publ. Mus. Nat. Hist., 15:
577-614. 1965b. A biogeographic account of the
herpetofauna of Michoacan, Mexico.
Univ. Kansas Publ. Mus. Nat. Hist.,
15:627-709. 1966. The Central American herpetofauna:
An ecological perspective. Copeia,
1966:700-719. Dunn, E. R.
1931. The herpetological fauna of the
Americas. Copeia, 1931:106-119. 1940a. New and noteworthy herpetological
material from Panama. Proc. Acad.
Nat. Sci. Philadelphia, 92:105-122. 1940b. Some aspects of heipetology in lower
Central America. Trans. New York
Acad. Sci., 2:156-158. 1943. A new race of Arneiva festiva from
Colombia. Notulae Naturae of the
Acad. Nat. Sci. Pliiladelphia, no. 126,
pp. 1-2.
EcHTERNACHT, A. C.
1968. Distributional and ecological notes on some reptiles from northern Hon- duras. Herpetologica, 24:151-158.
1970. Taxonomic and ecological notes on some Middle and South American lizards of the genus Ameiva (Teii- dae). Breviora, 354:1-9. EsTES, R.
1969. Relationships of t\vo Cretaceous liz- ards (Sauria, Teiidae). Breviora, no. 317:1-8.
Gabriel, K. R.
1964. A procedure for testing the homoge- neity of all sets of means in analysis of variance. Biometrica, 20:459-477.
Gabriel, K. R. and Sokal, R. R.
1969. A new statistical approach to geo- graphic variation analysis. Syst. Zool., 18:259-278.
Gorman, G. C.
1970. Chromosomes and the systematics of the family Teiidae (Sauria, Reptilia). Copeia, 1970:230-245.
Haffer, J.
1967. Speciation in Colombian forest birds west of the Andes. Amer. Mus. Novitates, 2294:1-57.
84
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
1970. Geologic climatic history and zoo- geographic significance of the Uraba region of northwestern Colombia. Caldasia, 10:603-636.
Hallowell, E.
1860. Report upon the Reptilia of the north Pacific exploring expedition, under command of Capt. ]ohn Rogers, U.S.N. Proc. Acad. Nat. Sci. Phila- delphia, 1860:480-510.
Heatwole, H.
1966. The eff^ect of man on distribution of some reptiles and amphibians in eastern Panama. Herpetologica, 22: 55-59.
Heatwole, H. and Torres, F.
1967. Distribution and geographic variation of the Ameivas of Puerto Rico and the Virgin Islands. Studies Fauna Curagao and other Carib. Isls., 24: 63-111.
HiLLMAN, P. E.
1969. Habitat specificity in three sympatric species of Ameiva (Reptilia: Teii- dae). Ecology, 50:476-481.
HiRTH, H. F.
1963. The ecology of two lizards on a tropical beach. Ecol. Monogr., 33: 83-112.
1965. Temperature preferences of five species of Neotropical lizards. Her- petologica, 20:273-276.
LiCHTENSTEIN, H. AND VoX MaRTENS, E.
1856. Nomenclator Reptilium et Amphibi- orum Musei Zoologici Berolinensis . . . Berlin, Konigl. Akad. Wissensch., iv -f 48 pp. Lloyd, J. J.
1963. Tectonic history of the south Central American orogen. Mem. American Assoc. Petrol. Geol. 2:88-100.
Maldoxado-Koerdell, M.
1964. Geohistory and paleogeography of Middle America. In: Handbook of Middle American Indians, vol. 1 ( R. Wauchope and R. C. West, Eds.). Univ. Texas Press, Austin, Texas, pp. 3-32.
Maslin, T. p.
1963. Notes on a collection of herpetozoa from the Yucatan Peninsula of Mex- ico. Univ. Colorado Studies, Ser. Biol., no. 9, pp. 1-20.
Mayh, E.
1963. Animal Species and Evolution. Bel- knap Press, Cambridge, Massachu- setts, xiv -j- 797 pp.
MiJLLER, V. L.
1929. Uber einige Rassen der Ameiva ameiva aus Venezuela. Zool. Anz., 83:97-112, 193-211.
Neill, W. T.
1965. New and noteworthy amphibians and reptiles from British Honduras. Bull. Florida State Mus., 9:77-130.
Neill, W. T. axd Allex, R.
1961. Further studies on the herpetology of British Honduras. Herpetologica, 17:37-52.
Peters, J. A.
1964. The lizard genus Ameiva in Ecuador. Bull. So. California Acad. Sci., 63: 113-127.
Peters, J. A., Burt, W. H., Sibley, C. C, Bogert, C. M., Hubbell, T. H., Clexch, W. J., AND Rogers, J. S.
1954. Symposium: Subspecies and CUnes. Syst. Zool., 3:97-125. Power, D. M.
1970. Geographic variation of red- winged blackbirds in central North America. Univ. Kansas Publ. Mus. Nat. Hist., 19:1-83. Raxd, a. S.
1957. Notes on amphibians and reptiles from El Salvador. Fieldiana: Zool- ogy, 34:505-534.
Savage, J. M.
1960. Evolution of a peninsular herpeto- fauna. Syst. Zool., 9:184-212.
1966. The origins and history of the Cen- tral American herpetofauna. Copeia, 1966:719-766.
Savage, J. M. and Heyer, W. R.
1967. Variation and distribution in the tree-frog genus PhijUomedusa. Bei- trage zur Neotropischen Fauna, 5: 111-131.
Schmidt, K. P.
1943. Corollary and commentary for "Cli- mate and Evolution." Amer. Midi. Nat., 30:241-253.
Schmidt, K. P. and Stuart, L. C.
1941. The herpetological fauna of the Sa- lama Basin, Baja Verapaz, Guate- mala. Zool. Ser. Field Mus. Nat. Hist., 24:233-247.
Schneider, C. R.
1965. Besnoitia panaviensis, sp. n. (Proto- zoa: Toxoplasmatidae ) from Pana- manian lizards. J. Parasit., 51:340- 344.
Schuchert, C.
1935. Historical geology of the Antillean- Caribbean region. John Wiley and Sons, New York, xxvi + 811 pp. Schwartz, A.
1965. The Ameiva (ReptiHa: Teiidae) of Hispaniola I. Ameiva lineolata Du- meril and Bibron. Carib. J. Sci., 5: 45-57.
ECHTERNACHT: MIDDLE AMERICAN LIZARDS
85
1967. The Aineiva (Lacertilia, Teiidae) of Hispaniola. III. Ameiva iaenixira Cope. Bull. Mus. Comp. Zool., 135: 345-375.
1968. Two new subspecies of Ameiva ( Lacertilia, Teiidae ) from Hispan- iola. Herpetologica, 24:21-28.
Schwartz, A. and Klinikowski, R. F.
1966. The Ameiva (LacertiUa, Teiidae) of Hispaniola. II. Geographic variation in Ameiva chnjsolaema Cope. Bull. Mus. Comp. Zool., 133:425-487.
Schwartz, A. and McCoy, C. J.
1970. A systematic review of Ameiva aiiheri Cocteau (Reptilia, Teiidae) in Cuba and the Bahamas. An. Carnegie Mus., 41(4) :45-168.
Smith, H. M.
1940. Descriptions of new lizards and
snakes from Mexico and Guatemala.
Proc. Biol. Soc. Washington, 53:55-
64. 1946. Handbook of Lizards. Comstock
Publ. Co., Ithaca, New York, xxi +
557 pp.
1967. The perspective of subspecies in ani- mal evolution. The Biologist, 49:43- 51.
Smith, H. M. and Laufe, L. E.
1946. A summary of Mexican lizards of the genus Ameiva. Univ. Kansas Sci. Bull., 31:7-73. Smith, H. M. and White, F. N.
1956. A case for the trinomen. Syst. Zool., 5:183-190.
Smith, R. E.
1968a. Experimental evidence for a gonadal- fat body relationship in two teiid lizards (Ameiva festiva, Ameiva qiiadrilineata). Biol. Bull., 134:325- 331.
1968b. Studies on reproduction in Costa Rican Ameiva festiva and Ameiva qiiadrilineata (Sauria: Teiidae). Co- peia, 1968:236-239.
Stuart, L. C.
1935. A contribution to a knowledge of the herpetology of a portion of the savanna region of central Peten, Guatemala. Misc. Publ. Mus. Zool. Univ. Michigan, 29:1-56, 4 Pis.
1942. Comments on the imdidata group of Ameiva (Sauria). Proc. Biol. Soc. Washington, 55:143-150.
1943. Comments on the herpetofauna of the Sierra de los Cuchumatanes of Guatemala. Occas. Papers Mus. Zool. Univ. Michigan, 471:1-28, 1 Pi.
1948. The amphibians and reptiles of Alta Verapaz, Guatemala. Misc. Publ. Mus. Zool. Univ. Michigan, 69:1-109.
1950. A geographic study of the herpeto- fauna of Alta Verapaz, Guatemala. Contrib. Lab. Vert. Biol. Univ. Michigan, 45:1-77, 9 Pis.
1951. The herpetofauna of the Guatemalan Plateau, with special reference to its distribution on the southwestern highlands. Contrib. Lab. Vert. Biol. Univ. Michigan, 49:1-71, 7 Pis.
1954a. A description of a subhumid corridor across northern Central America, with comments on its herpetofaunal indicators. Contrib. Lab. Vert. Biol. Univ. Michigan, 65:1-26.
1954b. Herpetofauna of the southeastern highlands of Guatemala. Contrib. Lab. Vert. Biol. Univ. Michigan, 68: 1-65, 4 Pis.
1955. A brief review of the Guatemalan lizards of the genus Anolis. Misc. Publ. Mus. Zool. Univ. Michigan, 91: 1-31.
1957. Herpetofaimal dispersal routes through northern Central America. Copeia, 1957:89-94.
1958. A study of the herpetofauna of the Uaxactun-Tikal area of northern El Peten, Guatemala. Contrib. Lab. Vert. Biol. Univ. Michigan, 75:1-30.
1964. Fauna of Middle America. In: Handbook of Middle American In- dians, vol. 1 ( R. Wauchope and R. C. West, Eds.). Univ. Texas Press, Austin, Texas, pp. 316-362.
1966. The environment of the Central American cold-blooded vertebrate fauna. Copeia, 1966:684-699. Taylor, E. H.
1956. A review of the lizards of Costa Rica. Univ. Kansas Sci. Bull., 38:1-322.
Terry, R. A.
1956. A geological reconnaissance of Pan- ama. Occas. Papers California Acad. Sci., No. 23. 91 pp.
Vanzolini, p. E. and Valencia, J.
1965. The genus Dracaena, with a brief consideration of macroteiid relation- ships (Sauria, Teiidae). Arq. Zool., 13:7-45.
Villa, J. D.
1968. A new colubrid snake from the Corn Islands, Nicaragua. Rev. Biol. Trop., 15:117-121. Vinson, G. L. and Brineman, J. H.
1963. Nuclear Central America, hub of the Antillean Transverse Belt. Mem. American As.soc. Petrol. Geol. 2:101- 112. Wilson, E. O. and Brown, W. L.
1953. The subspecies concept and its taxo- nomic application. Syst. Zool, 2:97- 111.
86
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
YUNKER, C. E. AND RaDOVSKY, F. J.
1966. The Dermanyssid mites of Panama. In: Ectoparasites of Panama (R. L. Wenzel and V. J. Tipton, Eds.). Field Mus. Nat. Hist., Chicago, pp. 83-103.
ZWEIFEL, R. G.
1959. Variation in and distribution of liz-
ards of western Mexico related to CnemidopJiorus sacki. Bull. Amer- ican Mus. Nat. Hist., 117:57-116.
The manuscript for this publication was completed in March, 1970, and submitted as a dissertation in partial fulfillment of the re- quirements for the PhD at the University of Kansas.
Shipment ID _Harvard MCZ 2/13.
-ted to
Was
u'le re-
Item #s
This bound volume contains more than one item.
Divide here for ^ '^ ^ ' '
See loader or manager w/ any questions.
tiiii^ilKti_,
MISCELLANEOUS
PUBLICATION
No. 56
pteran
^ L^-\VV IVLINV^C
wrzr
March 10, 1972
132
MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY
I'interieur de Bresil execute par ordre de S. M. le Roi de Baviere dans les annees 1817, 1818, 1819, 1820. Mo- nachii, 4:viii + 72 pp.
St.\rrett, a., axu R. S. Casebeer
1968. Records of bats from Costa Rica. Contrib. Sci., Los Angeles Co. Mus., 148:1-21.
Steel, R. G. D., and J. H. Torrie
1960. Principles and procedures of statistics. McGraw-Hill, New York, xvi + 481 pp.
Thomas, O.
1892. A preliminary list of the mammals of Trinidad. lour. Field Nat. Club, 1: 158-168.
Trouessart, E. L.
1897. Catalogue mammalium tam viventiun quam fossilium. Nova edito (prima completa). Berlin, R. Friedlander und Sohn, 1:1-218.
Tomes, R. F.
1863. On a new genus and species of leaf- nosed bats in the museum at Fort Pitt. Proc. Zool. Soc, London, pp. 81-85.
Tuttle, M. D.
1970. Distribution and zoogeography of Peruvian bats, with comments on nat- ural history. LIniv. Kansas Sci. Bull., 49:45-86.
Van Gelder, R. G.
1959. A taxonomic revision of the spotted skunks ( genus Spilogale). Bull. Amer. Mus. Nat. Hist., 117:229-392.
Vax Valex, L.
1966. Deltatheridia, a new order of mam- mals. Bull. Amer. Mus. Nat. Hist., 132:1-126. Vaughan, T. a.
1959. Functional moiphology of three bats: Eumops, Myotis, Macrotus. Univ. Kansas Publ., Mus. Nat. Hist., 12:1- 153.
Vaughax, T. a., and G. C. Bateman
1970. Functional morphology of the fore- limb of mormoopid bats. Jour. Mamm., 51:217-235.
ViLLA-R., B., and a. Jimexez
1960. Acerca de la posicion taxonomica de Mormoops megalophijlla senicula Rehn, y la presencia de virus rabico
en estos murcielagos insectivores. An.
Inst. Biol., 31:501-509. Wagxer, J. A.
1840. Die Affen und Flederthiere. Supple-
mentband, erste Abtheil., in J. Ch. D.
von Schreder, Die Siiugthiere in Ab-
bildungen nach der Natur mit
Beschreibungen. xiv -)- 558 pp. 1843. Diagnosen neuer Arten brasilischer
Handflungler. Arch. Naturgesch., 9:
365-368. 1847. Beitriige zur Kenntniss der Saugthiere
Amerika's. Abhandl. Math.-Physik. k.
Bayerischen Akad. Wiss., Mi.inchen,
5:121-208. Walker, E. P., and others
1964. Mammals of the World. The Johns
Hopkins Press, Baltimore, l:xliii-|-
646 pp. Walton, D. W., and G. M. Walton
1968. Comparative osteology of the pelvic
and pectoral girdles of the phyllosto-
matidae (Chiroptera: Mammalia).
Jour. Grad. Research Center, 37:1-35. Wenzel, R. L., AiNd V. J. Tipton
1966. Some relationships between mammal
hosts and their ectoparasites. Pp.
677-723, in Ectopara.sites of Panama.
(R. L. Wenzel and V. J. Tipton, eds.),
Field Mus. Nat. Hist., Chicago, xii +
861 pp. Wenzel, R. L., V. J. Tipton, and A. Kiewlicz 1966. The Strebhd batflies of Panama ( Dip-
tera Calypterae:Streblidae). Pp. 405-
575, in Ectoparasites of Panama. (R.
L. Wenzel and V. J. Tipton, eds.),
Field Mus. Nat. Hist., Chicago, xii -}-
861 pp. Werner, H. J., and W. W. Dalquest
1952. Facial glands of the tree bats, Lasiii-
rtis and Dasypterus. Jour. Mamm.,
33:77-80. Werner, H. J., W. W. Dalquest, AND J. H. Roberts
1950. Histological aspects of the glands of
bats, Tadarida cijnocephala ( Le
Conte). Jour. Mamm., 31:395-399. Winge, H.
1892. Jordfundne og nulevende Flagermus
(Chiw))tera) fra Logoa Santa, Minas
Geraes, Brasihen, Copenhagen, Museo
Lundii, 2:1-65.
3 2044 093 361 616
Date Due
BOUND
1972