• TABLE OF CONTENTS
HIDE
 Front Cover
 Table of Contents
 Title Page
 Front Matter
 Table of Contents
 Main
 Bibliography
 Back Cover






Group Title: Bulletin of the Florida State Museum
Title: Revision of North American salamanders of the genus Plethodon
CITATION THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00001562/00001
 Material Information
Title: Revision of North American salamanders of the genus Plethodon
Series Title: Bulletin of the Florida State Museum
Physical Description: 236-367 p. : maps, diagrs., tables. ;
Language: English
Creator: Highton, Richard Taylor, 1927-
Publisher: University of Florida
Place of Publication: Gainesville
Publication Date: 1962
 Subjects
Subject: Plethodon   ( lcsh )
Amphibians -- North America   ( lcsh )
Genre: bibliography   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
non-fiction   ( marcgt )
 Notes
Bibliography: "Literature cited": p. 359-367.
General Note: Cover title.
 Record Information
Bibliographic ID: UF00001562
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: ltqf - AAA0870
notis - ACK0936
alephbibnum - 000440470
oclc - 05069619
lccn - a 63007205

Table of Contents
    Front Cover
        Page 1
        Page 2
    Table of Contents
        Page 3
    Title Page
        Page 5
        Page 6
    Front Matter
        Page 235
    Table of Contents
        Page 236
    Main
        Page 237
        Page 238
        Page 239
        Page 240
        Page 241
        Page 242
        Page 243
        Page 244
        Page 245
        Page 246
        Page 247
        Page 248
        Page 249
        Page 250
        Page 251
        Page 252
        Page 253
        Page 254
        Page 255
        Page 256
        Page 257
        Page 258
        Page 259
        Page 260
        Page 261
        Page 262
        Page 263
        Page 264
        Page 265
        Page 266
        Page 267
        Page 268
        Page 269
        Page 270
        Page 271
        Page 272
        Page 273
        Page 274
        Page 275
        Page 276
        Page 277
        Page 278
        Page 279
        Page 280
        Page 281
        Page 282
        Page 283
        Page 284
        Page 285
        Page 286
        Page 287
        Page 288
        Page 289
        Page 290
        Page 291
        Page 292
        Page 293
        Page 294
        Page 295
        Page 296
        Page 297
        Page 298
        Page 299
        Page 300
        Page 301
        Page 302
        Page 303
        Page 304
        Page 305
        Page 306
        Page 307
        Page 308
        Page 309
        Page 310
        Page 311
        Page 312
        Page 313
        Page 314
        Page 315
        Page 316
        Page 317
        Page 318
        Page 319
        Page 320
        Page 321
        Page 322
        Page 323
        Page 324
        Page 325
        Page 326
        Page 327
        Page 328
        Page 329
        Page 330
        Page 331
        Page 332
        Page 333
        Page 334
        Page 335
        Page 336
        Page 337
        Page 338
        Page 339
        Page 340
        Page 341
        Page 342
        Page 343
        Page 344
        Page 346
        Page 347
        Page 348
        Page 349
        Page 350
        Page 351
        Page 352
        Page 353
        Page 354
        Page 355
        Page 356
        Page 357
        Page 358
    Bibliography
        Page 359
        Page 360
        Page 361
        Page 362
        Page 363
        Page 364
        Page 365
        Page 366
        Page 367
    Back Cover
        Page 368
Full Text




BULLETIN

OF THE .

FLORIDA STATE MUSEUM


BIOLOGICAL


SCIENCES


VOLUME 6


UNIVERSITY OF FLORIDA
Gainesville


1961-1962













Numbers of the BULLETIN OF THE FLORIDA STATE MUSEUM are pub-
lished at irregular intervals. Volumes contain about 300 pages and are not nec-
essarily completed in any one calendar year.















WILLIAM J. RIEMER, Managing Editor
OLVER L. AUSTIN, JR., Editor













Communications concerning purchase or exchange of the publication and all man-
uscripts should be addressed to the Managing Editor of the Bulletin, Florida State
Museum, Seagle Building, Gainesville, Florida.


Publication
No. 1 28
No. 2 22
No. 3 31


date Price
June 1961 .............------.................- $1.60
December 1961 _----.----.--------.. 1.55
December 1962 .--------------........ 1.55













TABLE OF CONTENTS


Postcranial Osteology of the Waterfowl.
GLEN E. WOOLFENDEN . . . . . . 1.

Osteology of Living and Fossil New World Quails.
J. ALAN HOLMAN .. .. .. ..... .. 131

Revision of North American Salamanders of the Genus Plethodon.
RICHARD HIGHTON . . . . . .... 235







BULLET


N


OF THE


FLORIDA STAT


BIOLOGICAL
Volume 6


MUSEUM


SCIENCES
Number 3


REVISION OF NORTH AMERICAN SALAMANDERS


OF THE GENUS


PLETHODON


Richard Highton


UNIVERSITY
Gaine


OF FL
sville


1962


ORIDA







Numbers of the BULLETIN OF THE FLORIDA STATE MUSEUM are pub-
lished at irregular intervals. Volumes contain about 300 pages and are not nec-
essarily completed in any one calendar year.















WILLIAM J. RIEMER, Managing Editor
OLIVER L. AUSTIN, JR., Editor



Consultant for this issue:
Wilfred T. Neill






















Communications concerning purchase or exchange of the publication and all man-
uscripts should be addressed to the Managing Editor of the Bulletin, Florida State
Museum, Seagle Building, Gainesville, Florida.




Published ) December 1962 Price for this issue $1.55











REVISION OF NORTH AMERICAN SALAMANDERS OF
THE GENUS PLETHODON


RICHARD HIGHTON1






SYNOPSIS: Systematic study of the North American salamander genus Plethodon
reveals three major groups in the genus, designated the western plethodons, the
eastern small plethodons, and the eastern large plethodons. The western
plethodons occur in the northwestern United States and adjacent Canada and
include five species (vandykei, larselli, dunni, vehiculum, and elongatus) in three
species groups. Five species in three species groups also comprise the eastern
small plethodons. Four of these (weller, dorsalis, cinereus, and richmondi) occur
in eastern North America; the fifth (neomexicanus) is known only from the Jemez
Mountains of New Mexico. The eastern large plethodons include six species
(wehrlei, yonahlossee, ouachitae, caddoensis, jordani, and glutinosus) in three
species groups and occur only east of the Rocky Mountains. The eastern large
and small plethodons are more closely related to each other than either is to the
western plethodons, but of the two, the small group appears to be closer to the
western plethodons.
Special emphasis has been placed on a study of geographic variation in
pigmentation and segmentation characteristics of the eastern species. The use of
the trinomen to describe geographic variation has been found unsatisfactory be-
cause of nonconcordance of geographically variable characters in species formerly
considered polytypic. Seven isolated allopatric populations are still recognized as
subspecies because of their apparent close relationship to the nominotypical sub-
species, but the taxonomic status of each is still uncertain. Some of these forms
are weakly differentiated and all are in need of further study.






1The author is Associate Professor of Zoology at the University of Maryland.
He did much of the work reported here while a graduate student at the University
of Florida, and an earlier version was accepted as a doctoral dissertation in 1956.
Although primarily a student of salamanders, the author contributed a paper in
this journal on the snake genus Stilosoma (1956). Manuscript submitted 2
April 1961.-ED.


Highton, Richard. 1962. Revision of North American salamanders of the genus
Plethodon. Bull. Florida State Mus., vol. 6, no. 8, pp. 235-367.




b O. 2z Lz




236 BULLETIN OF THE FLORIDA STATE MUSEUM Vol. 6




TABLE OF CONTENTS

Introduction .................................................... 237
Acknowledgments ........................................ 238
Methods ......................................................... 240
Pigmentation in the genus Plethodon .............................. 240
Variation in number of vomerine teeth ............................. 242
Treatment of infraspecific variation ................................ 243
Systematics .................................................. .. 245
Plethodon ........................................ 245
Key to salamanders of the genus Plethodon .......................... 253
The western plethodons .......................................... 255
Vandykei group .............................................. 257
Plethodon vandykei vandykei ................................ 257
Plethodon vandykei idahoensis ............................... 259
Plethodon larselli ........................................... 260
Vehiculum group ............................................. 263
Plethodon dunni ............................................ 263
Plethodon vehiculum ........................... ............ 265
Elongatus group .............................................. 267
Plethodon elongatus ......................................... 267
The eastern small plethodons ...................................... 269
Neomexicanus group ......................................... 272
Plethodon neomexicanus ...................................... 272
Welleri group ............................................. 274
Plethodon welleri ........................................... 274
Plethodon dorsalis dorsalis .................................... 277
Plethodon dorsalis angusticlavius .............................. 280
Cinereus group ................................................. 285
Plethodon cinereus cinereus ................................. 285
Plethodon cinereus serratus ................................... 288
Plethodon cinereus polycentratus .............................. 289
Plethodon richmondi richmondi ................................ 304
Plethodon richmondi hubrichti ................................ 307
Plethodon richmondi nettingi .................................. 307
The eastern large plethodons ...................................... 314
W ehrlei group ............................................... 318
Plethodon wehrlei ........................................... 318
Yonahlossee group ........................................... 322
Plethodon yonahlossee ....................................... 322
Plethodon ouachitae .......................................... 324
Plethodon caddoensis ........................................ 325
Glutinosus group ..................... ...................... 328
Plethodon jordani ............................................ 328
Plethodon glutinosus glutinosus ............................... 342
Plethodon glutinosus albagula ............................... 343
Literature cited .................................................. 359







HIGHTON: REVISION OF PLETHODON


INTRODUCTION
The North American salamander genus Plethodon is the type genus of
the lungless family Plethodontidae, today the most successful group
of salamanders, for it contains more than half of the known living
species of the Order Caudata. The family's center of distribution is
North America; one genus, Hydromantes, also occurs in southern
Europe, and representatives of several genera enter the Neotropical
Region. Plethodontid salamanders occupy widely varied habitats. Some
are strictly aquatic cavernicoles (Typhlomolge and Haideotriton);
others live only in mountain streams (Leurognathus); and still others,
like Plethodon, are terrestrial even to the extent of depositing their
eggs on land and dispensing with the aquatic larval stage.
Here recognized in the genus Plethodon are 16 species. Most live
on the forest floor in the leaf mold and under stones and rotting logs;
ten occur in wooded areas of eastern North America, five more in the
Pacific Northwest, and one is known only from the Jemez Mountains
of New Mexico. Of the 25 genera in the family Plethodontidae, only
two others show as great or greater disjunctions in their distribution:
Hydromantes, with two species in Europe and three in California, and
Aneides, with four species in western North America and one in the
Appalachian Mountains of the eastern United States.
The last revision of the entire genus was that of Dunn (1926),
when only a fraction of the presently known forms and material was
available. His review of the family led him to consider Plethodon the
most primitive existing genus in the attached-tontge branch of the
Plethodontidae. He considered Hemidactylium, Ensatina, Batracho-
Sseps, and Aneides as more specialized derivatives of a Plethodon-like
ancestor.
Grobman (1944) studied the distribution and systematics of the
eastern species and divided them into two groups, the large plethodons
and the small plethodons. He suggested that these groups might
actually represent distinct genera or subgenera, but he reserved judge-
ment until their relationships with the western species could be
determined. Since the revisions of Dunn and Grobman, many new
species and subspecies have been described. A review of these forms
offers a valuable opportunity to study speciation in this important
group of North American salamanders and to clarify the relationships
of the eastern and western species.
An earlier draft of this paper based mainly on material in the Uni-
versity of Florida Collections was presented to the University of
Florida as a degree requirement in 1956. Since then I have studied


1962


237






BULLETIN OF THE FLORIDA STATE MUSEUM


the Plethodon material in all of the major eastern museums, and have
borrowed additional material from a number of other collections. It
thus has been possible to expand the study into a more nearly com-
plete generic revision. Several changes have been made in the system-
atic conclusions and nomenclature of the earlier draft, particularly in
the treatment of infraspecific variation.
Most recent workers have emphasized the importance of studying
the pigmentation characters of plethodons in life. Many species, partic-
ularly the large eastern forms, are extremely difficult to identify after
years of preservation. Few other vertebrates exhibit after preservation
so few differentiating characters. Living specimens of all 16 species
have been examined, plus representatives from many geographic areas
for most of the widely ranging eastern forms. Living specimens repre-
senting all of the previously recognized geographic races have also
been studied.

ACKNOWLEDGMENTS
More than 100 persons contributed living and preserved materials for
this study. To each of these individuals I express sincere thanks. Most
of the material is deposited in the University of Florida Collections.
The remainder is retained for further study (listed as RH), but
eventually will be placed in permanent museum collections.
A list of abbreviations used in the text follows. For the loan of
personal materials or specimens in their care, and for all courtesies
extended, I wish to thank the following institutions and individuals:
AMNH, American Museum of Natural History, Charles M. Bogert and Richard
Zweifel
ANSP, Academy of Natural Sciences of Philadelphia, James M. Bihlke
BCB, Bryce C. Brown
BDV, Barry D. Valentine
CA, Chicago Academy of Sciences, Howard K. Gloyd and Laura Brodie
CAS, California Academy of Sciences, Alan E. Leviton
ChM, Charleston Museum, Albert Schwartz
CJS, Charles J. Stine
CM, Carnegie Museum, Neil D. Richmond and Graham Netting
CNHM, Chicago Natural History Museum, Robert Inger and Hymen Marx
CSNH, Cincinnati Society of Natural History, Ralph Dury and Jack L. Gottschang
CU, Cornell University, Edward Raney, Bruce Collete, and Frederick Gehlbach
CWM, Charles W. Myers
DC, Donald Cooper
DMB, Douglas M. Burns
DMNH, Dayton Museum of Natural History, James MacMahon
DR, Douglas Rossman
ERA-WTN, Ross Allen Reptile Institute, Wilfred T. Neill


238


Vol. 6








HIGHTON: REVISION OF PLETHODON


GSMNP, Great Smoky Mountains National Park, Arthur Stupka
HAD, Harold A. Dundee
HB, Herbert Barden
HCM, Henry C. Moski
INHS, Illinois Natural History Survey, Philip W. Smith
JAF, James A. Fowler
JSM, John S. Mecham
JWC-FC, John W. Crenshaw, Jr.
KNH, Ken N. Hovde
MCZ, Museum of Comparative Zoology, Arthur Loveridge and Ernest Williams
MMNH, Minnesota Museum of Natural History, William J. Breckenridge
MUOZ, Museum of the University of Oklahoma, Zoology, Arthur N. Bragg
MVZ, Museum of Vertebrate Zoology, University of California, Robert C. Stebbins
NMC, National Museum of Canada, A. W. F. Banfield and Francis R. Cook
OC, Oberlin College, Warren F. Walker, Jr.
OS, Ottys Sanders
PLC, Pacific Lutheran College, Jens W. Knudsen
PSM, Paul S. Martin
RC, Roger Conant
ROMZ, Royal Ontario Museum of Zoology, E. B. S. Logier
SAM, Sherman A. Minton, Jr.
SM, Strecker Museum, Baylor University, Bryce C. Brown
SRT, Sam R. Telford, Jr.
TNHC, University of Texas Natural History Collection, W. Frank Blair
TU, Tulane University, Fred R. Cagle and Harold Dundee
UA, University of Arkansas, Herndon G. Dowling
UF, University of Florida Collections, William J. Riemer
UG, University of Georgia, Bernard Martof
UIMNH, University of Illinois Museum of Natural History, Hobart M. Smith
UKMNH, University of Kansas Museum of Natural History, Edward H. Taylor
and John M. Legler
UKy, University of Kentucky, Roger W. Barbour
UMMZ, University of Michigan Museum of Zoology, Charles F. Walker, Norman
Hartweg, Thomas M. Uzzell, Jr., and James A. Organ
USC, University of Southern California, Jay M. Savage and David Wake
USNM, United States National Museum, Doris M. Cochran
WVBS, West Virginia Biological Survey, N. Bayard Green

Cornelius Barry aided by counting the vomerine teeth of some speci-
mens. Clifford H. Pope lent his original data on vomerine tooth counts
from his three published studies. Vincent Schultz contributed valuable
statistical advice. My wife, Anne, contributed by collecting specimens,
typing and proofreading parts of the manuscript, and especially by
taking color notes on geographic variation in P. glutinosus and P.
jordani. The Chairman of my Graduate Supervisory Committee,
Arnold B. Grobman, has been a constant source of intellectual stimula-.
tion, encouragement, helpful advice, and criticism.
Part of this work was supported by grants from the National Science


1962


239







BULLETIN OF THE FLORIDA STATE MUSEUM .


Foundation, the General Research Board of the University of Mary-
land, the Florida State Museum and the Graduate School of the
University of Florida, and the Highlands Biological Station. Collecting
in the Great Smoky Mountains National Park, in the Shenandoah
National Park, and along the Blue Ridge Parkway was permitted
through the cooperation of the National Park Service.

METHODS
The necessity of studying living material made a considerable amount
of field work imperative, particularly on the eastern large plethodons.
Specimens were kept alive in the laboratory at low temperatures for
long periods for later comparison with other living material.
Detailed descriptions of many living and preserved specimens were
made during the course of this study. Similar data for many forms are
available in Bishop (1943), Stebbins (1951), or in the original descrip-
tions. These are not repeated here as the concern is with variation in
populations, not detailed descriptions of individual specimens.
Many recent workers have given considerable weight to variation
in the number of costal grooves. Analysis of published data often
reveals lack of agreement in the number of grooves in some forms. A
method for determining body segmentation by counting costal grooves
consistently has been suggested (Highton, 1957).
Snout-vent length is measured always from the tip of the snout
to the anterior end of the vent.
The distribution maps are based on literature records (hollow sym-
bols) and specimens examined by the writer (solid symbols, unless
otherwise noted). Many new records and several range extensions are
included.
During the course of this study I examined more than 40,000
specimens, about 25 percent of them of my own collecting. I have seen
all type specimens known to exist with the exception of the holotype
of P. neomexicanus.

PIGMENTATION IN THE GENUS PLETHODON
The pigmentation of living specimens of all previously recognized
forms in the genus was studied under a dissecting microscope. The
great variety of coloration in these salamanders is produced by only
three different types of pigment cells. Following Hertzler (1958), these
are referred to as melanophores, iridophores, and red or yellow chro-
matophores. In his histological and biochemical study of the pigmen-
tation of Plethodon cinereus, Hertzler identified black melanophores,


240


Vol. 6







HIGHTON: REVISION OF PLETHODON


iridophores, and red chromatophores. He presents evidence that the
black pigment cells in the red-backed salamander contain melanin, and
that the white or brassy pigmentation is produced by pigment cells
that contain no guanine, although herpetologists have long referred
to these cells as "guanophores." As the granule-bearing cells responsi-
ble for the white pigment of these salamanders have not yet been
analyzed chemically, Hertzler prefers the nonspecific term, iridophore.
Finally, the red chromatophores, usually referred to in the literature
as "lipophores," are not lipid in nature, but are chemically related to
melanin, although they differ from melanin in having smaller
amounts of the elements magnesium, calcium, and iron.
Although no biochemical studies have yet been conducted on the
pigmentation of other species of the genus, their pigments as com-
pared under a dissecting microscope show no significant difference
from those of Plethodon cinereus, other than in distribution of the cells.
Color and black and white photographs, and drawings of most of the
forms of the genus are available in the literature (see especially
Bishop, 1943; Conant, 1958; and Stebbins, 1951 and 1954).
Melanophores are present in all members of the genus and are
responsible for the dark ground color of these salamanders. The other
pigments generally occur in gaps in the melanophore background.
Partial albinos of five species, glutinosus, jordani, cinereus, dorsalis,
and vehiculum have been reported (literature summarized by Hensley,
1959).
Iridophores are responsible for the white spotting and brassy
flecking present on the dorsum, sides, and belly of most species. Little
difference is apparent in the structure of the differently colored
iridophores, though the amount of light reflected by the spots varies
considerably. Hertzler (1958) finds that small brassy iridophores lie
deeper in the dermis than do the larger white ones and suggests that
the color is altered because light must pass through the pigmented cori-
um twice. Possessing iridophores is a condition that has been described
as "brassy flecking," "metallic golden spotting," "golden blotching,"
"bronzy mottling," and "frosting." Brassy iridophore pigmentation is
present in the iris of most species, though it is often reduced in the
eastern large plethodons. It is responsible for the small brassy flecks of
some glutinosus, jordani, ouachitae, caddoensis, wehrlei, cinereus, rich-
mondi, welleri, dorsalis, neomexicanus, larselli, and vehiculum. The
larger white spots are present on the dorsum of at least some popula-
tions of most of the forms in the genus, but in some, especially welleri,
hubrichti, caddoensis, and some glutinosus, even the largest spots have


241







BULLETIN OF THE FLORIDA STATE MUSEUM


a brassy appearance. The dorsum of a single glutinosus may show
every type of intermediate between the brassy and the white type.
Lateral and ventral iridophore pigmentation, when present, is almost
never of the brassy type.
"Red chromatophores" of Hertzler may vary from yellow to red,
but red is the prevailing hue. They are present in yonahlossee,
ouachitae, cinereus, dorsalis, elongatus, vehiculum, dunni, vandykei,
larselli, and in some populations of richmondi, jordani, and wehrlei.
The variety of coloration in the genus seems due entirely to variation
in the presence, location, and abundance of these three pigments. Some
forms lack the red chromatophores or iridophores or both. Pigmenta-
tion characters may vary widely within a species-individually, onto-
genetically, and geographically-but are constant enough within most
populations so that a person familiar with the variations can identify
most living specimens easily by color pattern alone. The phylogenetic
significance of the distribution of these pigments is discussed in the
accounts of the various groups.
VARIATION IN NUMBER OF VOMERINE TEETH

The number of vomerine teeth is often used as a systematic character
in the genus Plethodon. In a series of three papers on glutinosus,
yonahlossee, and ouachitae published in 1949, 1950, and 1951, Clifford
and Sarah Pope demonstrate that the total number or vomerine teeth
of all three species is correlated with size. The study of variation in the
number of vomerine teeth within and among species of Plethodon
must therefore consider the rate of ontogenetic change in number of
teeth as well as the absolute number of teeth at any given body length.
The Popes fitted rectilinear regression lines to their data on the
three species of eastern large plethodons. The teeth of the specimens
of glutinosus counted by Pope and Pope (1949:253) were recounted
and it is found that our counts are comparable. The data for yonah-
lossee and ouachitae are taken directly from the Popes' original data
sheets. Using the same statistical technique, I have fitted regression
lines to data on the relationship between the total number of vomerine
teeth and snout-vent length for every species in the genus. Geographic
variation in the number of vomerine teeth was studied in seven species
previously considered polytypic, vandykei, dorsalis, cinereus, richmon-
di, wehrlei, jordani, and glutinosus. Significant geographic variation
occurs only in glutinosus.
Two values must be calculated to obtain a regression line, the
Y-intercept and the slope of the line. The slope estimates the rate of


242


Vol. 6







HIGHTON: REVISION OF PLETHODON


increase in the teeth-size relationship; the Y-intercept is an estimate of
the initial number of teeth the animal possesses early in development.
Rectilinear regressions are used rather than curvilinear regressions
because graphical plots of the data show that linear regressions
describe the variation adequately. No species shows significant sexual
dimorphism in vomerine tooth counts. The number of teeth is con-
sidered the dependent variate and the snout-vent length the inde-
pendent variate.
The ontogenetic change in the relationship of vomerine teeth and
snout-vent length in different species is often similar, particularly in
closely related forms. The rate of increase tends to be greater in the
larger species, with the exception of wehrlei, elongatus, and neomexi-
canus. The last species has a negative slope, but it does not differ
significantly from zero. (A real negative slope would be produced by
a loss of teeth in old individuals.) The eastern small plethodons have
the lowest rate of increase in the number of vomerine teeth during
growth.
Regression lines for each species are calculated by the method of
least squares. The data for the several geographic samples of each
of the seven polytypic species listed above are combined to obtain a
single regression line for each species. Several factors may lead to
bias. Geographic samples of each species were not chosen at random,
but were often selected because of availability of material, and in some
instances because of indications of geographic variation in other
characters. Sizes of the various geographic samples were seldom equal.
P. glutinosus shows such significant geographic variation that the pool-
ing of the 12 geographic samples does not reflect the vomerine teeth-
snout-vent length relationship in any one region. In spite of these
shortcomings, the estimates of the slopes and Y-intercepts of the
regression lines, together with their 95 percent confidence intervals,
have value in illustrating patterns of variation in the genus and are
included in table 1.
TREATMENT OF INFRASPECIFIC VARIATION

Several types of geographic variation occur in the genus Plethodon.
The gradual clinal type occurs in some species. In others abrupt
changes are apparent over short distances where the ranges of two
nominal subspecies meet. Some populations are isolated from their
nearest relatives by unoccupied regions. In this event the nature of
intergradation cannot be used to determine whether or not the species
level of differentiation is attained.


1962


243












TABLE 1. Slopes and Y-intercepts of regression lines for number of vomerine teeth on snout-vent length in the species of
Plethodon.
Slope Y-intercept

Sample
Species Size Estimate 95% confidence interval Estimate 95% confidence interval
Lower limit Upper limit Lower limit Upper limit

vandykei 51 .237 .156 .318 5.22 1.87 8.57
larselli 25 .292 .142 .442 0.54 -5.14 6.23
dunni 49 .146 .067 .224 5.43 1.27 9.59
vehiculum 68 .151 .058 .243 3.18 -0.83 7.19
elongatus 49 .028 -.024 .080 9.42 6.73 12.12
neomexicanus 20 -.027 -.235 .180 14.78 4.17 25.39
dorsalis 182 .057 -.005 .118 9.10 6.83 11.37
wellei 31 -.054 -.221 .113 11.18 5.09 17.28
cinereus 398 .054 .025 .084 8.10 7.01 9.18
richmondi 237 .101 .069 .134 7.16 5.82 8.50
wehrlei 183 .065 .029 .101 10.27 8.52 12.02
yonahlossee 125 .410 .372 .447 0.88 -1.49 3.25
ouachitae 125 .267 .213 .321 6.67 3.83 9.52
caddoensis 43 .168 .011 .326 13.00 6.66 19.34
jordani 683 .232 .214 .251 4.97 4.08 5.87
glutinosus 1117 .197 .184 .210 6.25 5.56 6.94








HIGHT,QN: REVISION OF PLETHODON


All three types of geographic variation have been used in the past
as a basis for naming subspecies of Plethodon. The large amount of
material now available permits more detailed analysis of geographic
trends than have heretofore been possible. Results indicate that inde-
pendent characters in Plethodon often do not show concordance in
their geographic variation. This is probably because environmental
factors modifying the phenotype directly or modifying the genotype
through natural selection rarely affect the independent characters of
the organism in the same way.
Systematic literature reflects considerable controversy about the
usefulness of the trinomen (see Wilson and Brown, 1953, and Mayr,
1942, for opposing viewpoints).
Nomenclatural recognition of geographic subspecies based on
characters showing nonconcordant variations has little objective value
because the use of different sets or combinations of characters results
in different infraspecific classifications. Geographic variation in Pletho-
don is sometimes great; spectacular examples occur. Even so it is
necessary to be so arbitrary in the choice of characters used to demar-
cate subspecies that I fail to see the advantage of their continued use.
Therefore I have not used subspecies to describe geographic variation.
The treatment of isolated populations I still recognize as subspecies
remains to be considered. It is not known whether reproductive isola-
tion would be established or maintained if the area between the
disjunct ranges became habitable and isolated forms came into contact.
Seven forms presently recognized as subspecies-P. v. idahoensis, P. r.
hubrichti, P. r. netting, P. d. angusticlavius, P. c. polycentratus, P. c.
serratus, and P. g. albagula-are clearly more closely related to the
species with which they are associated nomenclaturally than to any
other form. In all there is morphological overlap in the systematic
characters used to differentiate the two subspecies. However, this may
also occur between full species of Plethodon. As each isolated form is
apparently closely related to its nominotypical subspecies and was
probably isolated from it relatively recently, I think that the recogni-
tion of these forms as subspecies should be continued. An alternative
some herpetologists might choose would be to recognize some or all as
full species, but this would not reflect their relationships so well.
Further research on each of these cases is greatly needed.

SYSTEMATICS
Plethodon Tschudi
Plethodon Tschudi (1838:58). Generic type by original designation: Salamandra
glutinosa Green.


1962


245








BULLETIN OF THE FLORIDA STATE MUSEUM


Phatnomatorhina Bibron in Bonaparte (1839:256). Substitute name.
Sauropsis Fitzinger (1843:33). Non Sauropsis Agassiz, 1832, Jahrb. f. Min., p.
142. Generic type by original designation: Salamandra erythronotus Rafin-
esque.
DIAGNOSIS. Plethodontidae with tongue attached in front, pre-
maxillae separate, teeth on posterior portion of maxillae, tail not con-
stricted at base, five toes on hind feet, no palmar tubercles, terminal
phalanges normal, and no aquatic larval stage.
REMARKS. According to Dunn (1926), the closest relatives of Pleth-
odon are Batrachoseps, Ensatina, Aneides, and Hemidactylium. All
four of these genera differ from Plethodon in important characteristics.
Batrachoseps and Hemidactylium have only four toes on the hind feet.
Ensatina and Hemidactylium have a basal constriction of the tail, a
provision for urotomy. Ensatina has palmar tubercles. The premaxillae
are fused in Aneides and Batrachoseps (except B. wrighti. Hemi-
dactylium is the only genus in the group with an aquatic larval state,
but this may not represent a primitive condition for its larvae differ


FIGURE 1. Distribution of the western plethodons (stippling), the eastern large
plethodons (vertical lines), and the eastern small plethodons (horizontal lines)
in North America.


Vol. 6


246







1962 HIGHTON: REVISION OF PLETHODON 247

from those of most other plethodontids. Aneides lacks teeth on the
posterior portion of the maxilla and has expanded terminal phalanges.
Most of the characters by which these genera differ from Plethodon
appear to be specializations. For this reason Dunn (1926:22) considers
Plethodon the most primitive of this group of genera, with the possible
exception of Ensatina. Plethodon is the largest genus in number of
species, and shows within itself more divergence than most related
genera.
The genus Plethodon has three major natural subdivisions: the
western plethodons, eastern small plethodons, and eastern large pleth-
odons. The distribution of these groups is shown in figure 1. These
might be considered subgenera in some animal groups, but herpetolo-
gists have rarely used subgenera in their classifications, so only common
names are given them in keeping with current herpetological practices.
Each of these subdivisions is further divided into groups of closely
related species. A few species are distinct enough to warrant mono-
typic grouping.
Grobman (1944:266) suggests that the relationship between the
eastern large and small plethodons is not close and lists size and costal
groove differences between the two groups. Actually variations in these
characters overlap somewhat, but there are other important differ-
ences. The mental gland of the adult male is usually better developed
in the eastern large plethodons (fig. 2). Most eastern small plethodons
have characteristically more elongated bodies and shorter legs; they
have fewer vomerine teeth; and the number of their trunk vertebrae













A B C. D
FIGURE 2. Chins of adult male Plethodon showing variation in the presence
and shape of the mental gland. (A) P. larselli, (B) P. glutinosus, (C) P. rich-
mondi, (D) P. dorsalis. Drawings not to same scale.








BULLETIN OF THE FLORIDA STATE MUSEUM


varies greatly within the group (16-24), whereas the eastern large
plethodons show little variation (16-19, usually 17-18). The small forms
usually have more webbing on the toes than the large ones. Eastern
large plethodons usually have an unpigmented parietal peritoneum;
in the small ones the peritoneum is pigmented with melanophores.
Premaxillary teeth of adult males of some eastern small plethodons
are highly modified (fig. 3).











A B C D
FIGURE 3. Modifications of premaxillary teeth of adult male Plethodon. (A)
P. vehiculum, typical of western plethodons. (B) P. glutinosus, typical of eastern
large plethodons. (C) P. dorsalis; P. welleri similar. (D) P. cinereus; P. rich-
mondi similar. Anterior cusp shown at right, posterior cusp at left.
Plethodon wehrlei, usually included with the eastern large pletho-
dons on the basis of size, is intermediate between the two groups in
several ways. It has more costal grooves and fewer vomerine teeth
than other eastern large plethodons. Its toes are webbed between the
proximal phalanges, as in the eastern small plethodons, and the
peritoneum has some melanophore pigmentation. The intermediacy of
this species suggests the two groups have not become sufficiently dis-
tinct to justify taxonomic separation and shows their close relationship.
Despite the characters linking it with the eastern small plethodons,
P. wehrlei is probably more closely related to the larger species, and
is therefore listed as a separate species group under the eastern large
plethodons.
Series of skeletons are available for, most forms, and the vertebrae
of all the species of the genus have been examined. The two eastern
groups show differences in their body vertebrae. The height of the
trunk vertebrae is proportionately less in the eastern small plethodons
and their vertebrae usually lack the neural spines present on those of
the large eastern species (fig. 4). P. wehrlei closely resembles the other
eastern large plethodons in this regard.


248


Vol. 6









HIGHTON: REVISION OF PLETHODON


1mm


I mm


I mm
FIGURE 4. Variation in the trunk vertebrae of Plethodon. (A) P. glutinosus, (B)
P. welleri, (C) P. dunni, representing large eastern, small eastern and western
groups respectively.


249







BULLETIN OF THE FLORIDA STATE MUSEUM


Comparison of these differential characters also clarifies the rela-
tionship of the western to the eastern species. As in the eastern forms,
size varies within the western group, but 4 (vandykei, dunni, elonga-
tus, and neomexicanus) of the 6 western species are as large as most
eastern large plethodons. The number of their trunk vertebrae varies
considerably (14-20). ,Vertebrae of the 5 species from the Pacific
Northwest are proportionately lower and longer and have longer trans-
verse processes than any of the eastern species (fig. 4). The vertebrae
of neomexicanus are like those of the eastern small plethodons. The
peritoneum of all western plethodons is pigmented with melanophores,
and the mental gland is absent or poorly developed in all adult
males. Body form varies from short and stout in vandykei to elongate
in elongatus and neomexicanus. The vomerine teeth are few in num-
ber, and the premaxillary teeth of males are unmodified except in
neomexicanus. The toes of dunni and vehiculum are slightly webbed,
while the other 4 species have webbing between the proximal
phalanges. Two, larselli and neomexicanus, have the number of
phalanges in the fifth hind toe reduced, presumably a first step in
its evolutionary loss, a characteristic of some related genera.
As a group the western species vary morphologically more than
either of the eastern groups. In vertebral structure the 5 species from
the Pacific Northwest are distinct from the rest of the genus. Although
the eastern and western groups probably share a common ancestor,
their degree of differentiation suggests they have been separated a
long time. In assessing the relationships of these salamanders I place
more weight on vertebral differences than on other characters. As
they are more stable evolutionarily, vertebrae are more reliable indica-
tors of phylogenetic relationship.
P. neomexicanus, known only from the Jemez Mountains of New
Mexico, is the most geographically isolated species. In their original
description Stebbins and Riemer (1950:77) consider the species prob-
ably close to P. cinereus, an eastern small plethodon. Blair (1958:450)
postulates a late Pleistocene connection between the New Mexican
species and the nearest cinereus localities in eastern Oklahoma. Though
this hypothesis appears more plausible than a recent connection with
the Plethodon populations in the Pacific Northwest, the amount of
differentiation in neomexicanus suggests it has been isolated much
longer than since the Wisconsin stage of the Pleistocene. Still, resem-
blances in structure of the trunk vertebrae and in body proportions
support the contention that neomexicanus is most closely related to
the eastern small plethodons. It is not possible to say which eastern


250


Vol. 6








HIGHTON: REVISION OF PLETHODON


species it most closely resembles, as it has differentiated greatly from
them all. It is larger and has a reduced fifth toe. In some adult neo-
mexicanus males the premaxillary teeth have two cusps as in the east-
ern plethodons, but the mental gland is poorly developed as in the
western plethodons.
In several ways the eastern small plethodons resemble the west-
ern group more closely than do the eastern large plethodons.
The red or yellow dorsal stripe is more frequent. Both show
greater variation in the number of trunk vertebrae and degree
of body elongation. Both have low counts of vomerine teeth
and a pigmented parietal peritoneum. The common ancestor of
the genus probably resembled these two groups more closely
than it did the eastern large plethodons. It is unlikely that con-
vergence can account for so many similarities between the two.
The eastern large plethodons therefore have diverged most radically
from the ancestral type, becoming specialized by an increase in size,
increase in number of vomerine teeth, development of the mental
gland, and in the loss of pigmentation in the peritoneum. Yet similar
vertebral structure in the two eastern groups, as well as the inter-
mediate position of wehrlei, suggests that both are more closely related
to each other than either is to the western plethodons. Dunn (1926:23)
suggests that yonahlossee, one of the eastern large plethodons, is the


EASTERN
SMALL PLETHODONS
dorslis rlchmondl EASTERN
WESTERN PLETHODONS Wme Wi cinereus LARGE PLETHODONS
Wneomexcom OcOR.
vehiculum yonahlmesse ddoensis
elongatus dunni glutinosus
laWrelli we \
vandykel jord ni











FIGURE 5. Dendrogram showing the suggested phylogeny of the species of
Plethodon.


251


1962








BULLETIN OF THE FLORIDA STATE MUSEUM


most primitive member of the genus, but I consider it one of the most
highly specialized. Morphological relationships suggest the phylogeny
shown in figure 5.
The relationships of the western plethodons with the other pleth-
odontid genera in western North America need further study. West-
ern plethodons may be more closely related to Batrachoseps, Ensat na,
or western Aneides than to any of the eastern plethodons. The western
genera probably have been isolated from their eastern relatives for
a long time. The two geographic groups now included in the genus
Plethodon, one in eastern and one in western North America, both
retain many primitive characteristics. This does not necessarily mean
that they are congeneric. Each group has more species displaying more
diversity than any other plethodontid genus in its region. A review of
this entire problem is needed, especially an investigation of the rela-
tionships of each group to the genera Aneides, Batrachoseps, Ensatina,










,..e














FIGURE 6. Relationship between trunk vertebrae and coastal grooves. Radio-
graph of a specimen of P. glutinosu from Florida with 17 trunk vertebrae, and
a lateral view of the same individual showing its 16 costal grooves including
two that run together on the groin.


Vol. 6








HIGHTON: REVISION OF PLETHODON


Sand Hemidactylium. Such an inquiry, which is beyond the scope of
this study, should be completed before erecting a new genus for the
western plethodons, a course which seems indicated by the present
Sdata.
The systematic arrangement I propose for the genus Plethodon
is shown in the table of contents.


KEY TO SALAMANDERS OF THE GENUS PLETHODON
The following key is based mainly on the average number of costal
grooves in each form of the genus. The number of costal grooves is
one less than the number of trunk vertebrae if counted to include the
groove over the front limb (even if it is not well developed) and all
dorsolateral grooves anterior to the hind limb (even when two of the
grooves join ventrally on the groin). See figure 6. A small percentage
of some forms may not be correctly identifiable by the number of
costal grooves, but a small series usually suffices to key out correctly.
Ranges are included because locality data may prove more helpful in
identifying preserved specimens than pigmentation characters, which
often disappear after preservation.
1. Costal grooves usually 14; parotoid gland well developed .............. 2
Costal grooves usually 15 or more; no parotoid gland ............... 3
2. Proximal segment of limbs dark in color (northern Idaho) P. v. idahoensis
Proximal segment of limbs red or yellow, similar to pigment in dorsal
stripe (western Washington) ........................ P. v. vandykei
3. Costal grooves usually 15 ................................ ...... 4
Costal grooves usually 16 or more ............................... 5
4. One phalanx in fifth toe of hind foot; belly reddish (Multnomah and
Hood River Cos., Oregon, and Skamania Co., Washington) .... P. larselli
Two phalanges in fifth toe of hind foot; belly without red pigment (west-
ern Oregon and southwestern Washington) ................ P. dunni
5. Costal grooves usually 16 .................................. ... .. 6
Costal grooves usually 17 or more ............................... 15
6. Belly mottled with yellow or red, and white and black; adult size small,
usually less than 50 mm. snout-vent length; often a red, tan, or yellow
dorsal stripe (southwestern British Columbia, including Vancouver
Island, western Washington and western Oregon) ........ P. vehiculum
Belly usually dark, at least posteriorly, or with scattered small white
spots; adult size usually greater than 50 mm. (except weller and cad-
doensis); dorsal stripe usually absent (except yonahlossee and oua-
chitae) (eastern United States) ................................. 7
7. Red pigment on back, legs, or cheeks ............................. 8
No red pigm ent ................................................ 10
8. Red pigment on legs or cheeks; on dorsum only in very young individuals
(southern Appalachian Mountains) ....................P. jordani


253








BULLETIN OF THE FLORIDA STATE MUSEUM


Red pigment largely restricted to dorsum .......................... 9
9. White pigment lacking in dorsal chestnut-colored stripe (Blue Ridge
Mountains of southwestern Virginia, northeastern Tennessee, and
northwestern North Carolina) ........................ P. yonahlossee
Abundant white pigment occurring within the dorsal stripe (Ouachita
Mountains of Arkansas and Oklahoma) .................. P. ouachitae
10. Size small, adults not over 50 mm. snout-vent length; dorsum with large
coalescing brassy spots usually covering at least half the dorsum; dark
parietal peritoneum (Blue Ridge Province of southwestern Virginia,
northeastern Tennessee, and northwestern North Carolina) ... P. welleri
Back without brassy spots, or if present, small in size and occupying less
than one-quarter of the dorsum; parietal peritoneum with a slight
amount of melanophore pigmentation or none .................... 11
11. Body entirely black, completely lacking iridophores; chin usually lighter
than belly (Blue Ridge Mountains of western North Carolina and
adjacent portions of Tennessee and Virginia) .............. P. jordani
Dorsum or sides or both with white or yellow spotting; chin as dark as
belly, sometimes lighter ....................................... 12
12. Dorsum with large conspicuous white or brassy spots ................ 13
Back usually without such spotting; spots present are of very small
size ....................................... ......... P. iordani
13. Melanophore pigmentation on chin much less than on belly ............ 14
Melanophore pigmentation on chin similar to that on belly (eastern
United States from southwestern Connecticut to northern Florida, west
to Louisiana, Missouri, and Illinois, also eastern Oklahoma and eastern
Texas ............................................ P. g. glutinosus
14. Dorsum with small white spots (central Texas) ............ P. g. albagula
Dorsum with large white spots and smaller more numerous brassy
flecks (Caddo Mountains of Arkansas) ................ P. caddoensis
15. Costal grooves usually 17 (Cattaraugus County, New York, south
through western Pennsylvania, adjacent Ohio, West Virginia, and
southwestern Virginia) ................................... P. wehrlei
Costal grooves usually 18 or more .............................. 16
16. Costal grooves usually 18 ........................................ 17
Costal grooves usually 19 or more ................................ 21
17. Size large, to 70 mm. snout-vent length (northwestern California and
southwestern Oregon) ................................ P. elongatus
Size small, to 50 mm. snout-vent length (eastern United States) ....... 18
18. Belly mottled with black and yellow or white; or belly with red, black,
and yellow or white pigment ................................. 19
Belly black with small white spots (Cheat Mountains of West Vir-
ginia) ............................................. P. r. netting
19. Red pigment abundantly present on belly .......................... 20
No red pigment on belly except occasionally between front limbs (Mis-
souri, Appalachian Plateaus, and Catskill Mountains, New York) P. cinereus
20. Dorsal stripe narrow, less than one-third the width of the body (south-
western Missouri, northwestern Arkansas, and adjacent Oklaho-
ma) ......................................... P. d. angusticlavius
Dorsal stripe much wider than one-third the, width of the body, edges


Vol. 6


254






BULLETIN OF THE FLORIDA STATE MUSEUM


the species (table 2) except dunni and larselli. None of the known
forms has a modal number of 18 trunk vertebrae, and they show less
vertebral variation within species than do the eastern forms. Each
species has a mean number of vertebrae close to a whole number, in-
dicating that strong selection restricts variation from the modal num-
ber.

TABLE 2. Number of trunk vertebrae in western forms of the genus Plethodon.
Form Number of trunk vertebrae
14 14/15 15 15/16 16 16/17 17 17/18 18 18/19 19 19/20 20 Mean

P. o.
vandykei 1 54 1 4 15.06
P. v.
idahoensis 29 1 4 15.13
P. larselli 3 22 1 4 16.05
P. dunni 56 1 3 16.06
P. vehiculum 17 2 116 2 17 17.00
P. elongatus 4 41 3 18.98

All the western plethodons except vehiculum and larselli are fairly
large. All have a striped phase, although in larselli, dunni, vehiculum,
and elongatus some individuals of some populations are unstriped.
All but dunni and vehiculum have webbed toes, and their vomerine
tooth counts are low compared to most eastern large plethodons of
similar size; vehiculum and elongatus have especially low vomerine
counts. P. vandykei is unusual in having a parotoid gland and the
fewest vertebrae in the genus. P. larselli is the only one with reduced
fifth hind toes, a character shared by P. neomexicanus.
Of the closely related pair P. dunni and P. vehiculum, the former
attains a larger size, has more vomerine teeth, and is more aquatic.
P. vehiculum usually has one more trunk vertebra than dunni. The
dorsal stripe in dunni is usually greenish yellow, while that of
vehiculum varies from reddish tan to yellow. The toes are slightly
webbed in both; the other western plethodons have more webbing.
Increase in webbing does not appear to be associated with aquatic
habits, as the most aquatic form, dunni, has the least webbing, and
vehiculum and elongatus, both less aquatic, differ greatly in the amount
of webbing. Some eastern forms have webbed toes while others do
not, yet all are terrestrial.
As dunni and vehiculum occur sympatrically over an extensive area
with no evidence of interbreeding, they must be regarded as distinct


256


Vol. 6








BULLETIN OF THE FLORIDA STATE MUSEUM


the species (table 2) except dunni and larselli. None of the known
forms has a modal number of 18 trunk vertebrae, and they show less
vertebral variation within species than do the eastern forms. Each
species has a mean number of vertebrae close to a whole number, in-
dicating that strong selection restricts variation from the modal num-
ber.

TABLE 2. Number of trunk vertebrae in western forms of the genus Plethodon.

Form Number of trunk vertebrae
14 14/15 15 15/16 16 16/17 17 17/18 18 18/19 19 19/20 20 Mean

P. v.
vandykei 1 54 1 4 15.06
P. v.
idahoensis 29 1 4 15.13
P. larselli 3 22 1 4 16.05
P. dunni 56 1 3 16.06
P. vehiculum 17 2 116 2 17 17.00
P. elongatus 4 41 3 18.98

All the western plethodons except vehiculum and larselli are fairly
large. All have a striped phase, although in larselli, dunni, vehiculum,
and elongatus some individuals of some populations are unstriped.
All but dunni and vehiculum have webbed toes, and their vomerine
tooth counts are low compared to most eastern large plethodons of
similar size; vehiculum and elongatus have especially low vomerine
counts. P. vandykei is unusual in having a parotoid gland and the
fewest vertebrae in the genus. P. larselli is the only one with reduced
fifth hind toes, a character shared by P. neomexicanus.
Of the closely related pair P. dunni and P. vehiculum, the former
attains a larger size, has more vomerine teeth, and is more aquatic.
P. vehiculum usually has one more trunk vertebra than dunni. The
dorsal stripe in dunni is usually greenish yellow, while that of
vehiculum varies from reddish tan to yellow. The toes are slightly
webbed in both; the other western plethodons have more webbing.
Increase in webbing does not appear to be associated with aquatic
habits, as the most aquatic form, dunni, has the least webbing, and
vehiculum and elongatus, both less aquatic, differ greatly in the amount
of webbing. Some eastern forms have webbed toes while others do
not, yet all are terrestrial.
As dunni and vehiculum occur sympatrically over an extensive area
with no evidence of interbreeding, they must be regarded as distinct


256


Vol. 6







HIGHTON: REVISION OF PLETHODON


species. A study of possible differences in the niches of the two forms,
a Stebbins (1951:65) suggested, is much needed.
Burns (1954) considers the form larselli a geographic race of
vandykei. I agree with the unpublished conclusions of Douglas Burns
ad David Wake (in letters) that larselli be considered a full species.
The differences between it and vandykei are so great that it is doubt-
ful that the two would interbreed if they were sympatric. In addition
to the pigmentation characters noted in the original description, it
differs from vandykei in having a smaller parotoid gland, 1 instead
of 2 phalanges on the fifth hind toe, and 16 instead of 15 trunk
vertebrae. While I place larselli tentatively in the same species group
as vandykei, its relationships with the other western plethodons need
further study.
P. elongatus differs from the other western plethodons in pig-
mentation characters and in having more trunk vertebrae than the
other forms. It has no close affinities with any of the others and is
placed in a separate species group.

VANDYKEI GROUP
Plethodon vandykei vandykei Van Denburgh
Washington Salamander
Plethodon vandykei Van Denburgh (1906:61). Dunn (1926:151-3); Slevin
(1928:57-9); Bishop (1943:275-8).
Plethodon vandykei vandykei, Lowe (1950:93); Stebbins (1951:80-4), (1954:
56-7).
HOLOTYPE. CAS 6910 (destroyed in the San Francisco fire of 1906),
collected at Paradise Valley, Mount Ranier National Park, Wash-
ington, by E. C. Van Dyke, July 1905.
Slevin and Leviton (1956:535) designate a neotype, CAS 47495,
from Forks, Clallam County, Washington, more than 130 miles from
the type locality and at the extreme opposite end of the known range
of P. v. vandykei. According to the International Rules of Zoological
Nomenclature, a neotype should be designated only when required
for stability or for solving a confused zoological problem, and it should
come from as near as possible to the place at which the holotype
was collected. As the neotype was not properly selected in this case,
later workers do not have to recognize it.
DIAGNOSIS. A western plethodon with a parotoid gland and a modal
number of 15 trunk vertebrae. It differs from idahoensis in having a
wid- Xrl.alstipe, lighter ground color, and yellow pigmenfitei


257






BULLETIN OF THE FLORIDA STATE MUSEUM


proximal segments of the limbs similar to that in the dorsal stripe.
DISTMIBUTON. Western Washington from Clallam County south to
Pacific County, east to Pierce County (fig. 7). The populations in the
Olympic Mountains, the Cascades, and the Willapa Hills may now
be isolated from one another.







0 3


0

P. vandykei
*p vandykei
idohoensis


FIGURE 7. Distribution of P. vandykei in Washington and Idaho

DESCRIPTION. Specimens from Lewis County, Washington, have pale
yellow dorsal stripes in life. The chromatophores vary from reddish
yellow to yellow. Those in the dorsal stripe are unusual in having
small dark granular inclusions. Similar yellow pigment also occurs
on the legs, head, sides, and belly, but is greatly reduced on the chin.
Melanophores are plentiful on the sides, but less so in the stripe region
and on the legs. Melanophore pigmentation on the belly varies; it is
heavy in some black-bellied individuals, while in others it is greatly
reduced. White iridophore spots occur on the head, legs, dorsum, belly,
and sides, and are largest and most abundant on the sides. Brassy
flecking is heavy in the iris, both above and below the pujl. This
is the only species examined that has silvery white pigment in the
iris. It occurs above the pupil. Usually a large area including most
of the chin is almost pigmentless. The dorsal stripe in adults has
straight edges, but often in young specimens its borders are irregular.
Probably the remnants of embryonic dorsolateral spots similar to those


258


Vol. 6








1962 HIGHTON: REVISION OF PLETHODON 259

found in the young of some of the eastern species cause this
irregularity.
Costal grooves usually number 14 and trunk vertebrae 15; vomer-
ine tooth counts range from 8 to 19 (fig. 8). The largest individual
mined is 56 mm. in snout-vent length.

Plethodon vandykei idahoensis Slater and Slipp
Coeur d'Alene Salamander

Plethodon idahoensis Slater and Slipp (1940:38). Bishop (1948:259-61).
Plethodon vandykei idahoensis, Lowe (1950:93); Stebbins (1951:80-4), (1954:
56-7).

HOLOTYPE. USNM 110504, adult male, collected at the northeast
corer of Coeur d'Alene Lake, Kootenai County, Idaho, elevation 2160
feet, by James R. Slater, 13 September 1939.
DIAGNOSIS. A race in which the yellow or orange dorsal stripe is
narrower than in P. v. vandykei and contrasts sharply with the lateral
black ground color. The proximal segments of the limbs are dark.
DISTRIBUTION. Known only from Benewah and Kootenai Counties
in northern Idaho (fig. 7).
DESCRIPTION. The dorsal stripe contains orange or yellow chroma-
tophores. The remainder of the body is pigmented with melanophores,
except for gaps on the chin where there are yellow chromatophores.
The iris has brassy iridophores, and the belly and sides have a few
scattered white ones.



35- 35S
0 R v. vondykei o. P. v. idohoensis


2 g
w tu
a0 2o do do do o ooo



0 o0





of vomerine teeth of P. v. vandykei. of vomerine teeth of P. v. idahoensis.
of vomerine teeth of P. v. vandykei. of vomerine teeth of P* c. "idahoensis.


___






BULLETIN OF THE FLORIDA STATE MUSEUM


Costal grooves usually number 14, the trunk vertebrae 15. Vomer-
ine teeth range from 7 to 22 (fig. 9). The largest specimen examined
is 54 mm. in snout-vent length.

P. v. vandykei is known from the Willapa Hills, Olympic Mountains,
and west central Cascade Mountains of western Washington; a dis-
junct population, P. v. idahoensis occurs in northern Idaho. The three
populations in western Washington now appear to be isolated from
one another. The Idaho population was originally described as a dis-
tinct species, idahoensis, by Slater and Slipp (1940), but Lowe
(1950:98) and more recent workers consider it a subspecies of
vandykei. Although the Idaho form is usually believed to be isolated
from the Washington populations, Savage (1952) suggests that they
may be connected by a corridor of high humid country along the
international boundary between Washington and British Columbia.
The forms idahoensis and vandykei apparently differ mainly in
pigmentation.
Stebbins (1951:80-1) points out the presence of two color phases
of P. v. vandykei. The light phase is rather uniform in color with slight
contrast between the dorsal stripe and the belly and sides; the dark
phase has a greater concentration of melanophores on the belly and
sides closely resembling the pattern in idahoensis. A third color phase
recently discovered in the Willapa Hills is similar to the light phase,
except that the ground color is pinkish rose. Although freshly pre-
served specimens of all three color phases have been examined, only
specimens of the light phase of P. v. vandykei have been studied in
life. Therefore detailed pigmentation comparisons cannot be given.
One specimen, PLC 626, appears to have only one phalanx in each
fifth hind toe, thus resembling the normal condition in neomexicanus
and larselli.
This species usually lives in damp places according to Stebbins
(1951:75, 82), and may replace P. dunni ecologically as well as geo-
graphically. However, Storm (1955) shows that their ranges do overlap
slightly in Pacific County, Washington.

Plethodon larselli Burns
Larch Mountain Salamander
Plethodon vandykei larselli Bums (1954:83-7). 9
HOLOTYPE. USNM 134129, adult male, collected on the north slope
of Larch Mountain, & miles from the summit on the Multnomah Falls


260


Vol. 6







HIGHTON: REVISION OF PLETHODON


Trail, Multnomah County, Oregon, 24 May 1953, by Douglas M.
Burns.
DIAGNOSIS. A western plethodon with only one phalanx in the fifth
hind toe, a modal number of 16 trunk vertebrae, reduced parotoid
gnds, and a cardinal red to reddish-orange venter.
DISTRIBUTION. Known from the type locality and from Wyeth, Hood
River County, Oregon; Wahkeena Falls, Multnomah County, Oregon,
and from Archer Falls, Skamania County, Washington (fig. 10). This-
recently discovered species may have a wider range; it should be
looked for in adjacent areas of both Washington and Oregon.
DESCRIPTION. The dorsal stripe varies from bright red to yellow,
and in some specimens is almost completely obscured by melano-
phores. Many specimens show a concentration of melanophores in the
middorsal region forming a dark black line. Small brassy iridophores
are present on the dorsum of some specimens, particularly those with
a heavy diffusion of melanophores in the dorsal stripe region. Red
or yellow chromatophores similar to those on the dorsum occur on
the sides. On the belly and the ventral side of the tail these chroma-
tophores are bright red, even in specimens with a yellow or dull-red
dorsal stripe. Large white iridophore spots are present on the lower
sides, limbs, and to a variable extent on the chin, but they are absent
from the belly. Most of the iridophore pigment on the upper sides
is of the brassy type, contrasting with the white pigment of the lower
sides. Brassy iridophores are abundant in the iris.
Costal grooves usually number 15 and trunk vertebrae 16; vomer-
ine teeth range from 5 to 17 (fig. 11). The largest individual exam-
ined is 49 mm. in snout-vent length.


1962


261








to





R larselli




0





so 0
*w






I I I I I C I
10 20 30 40 50 60 70 80 c
SNOUT- VENT LENGTH (mm.)


FIGURE 10. Ditribution of P. larselli
in Wasfngton End Oregon.


FIGURE 11. Variation in the number of vomerine teeth of P. larselli.


w I
z

10
0


Or
O
0








HIGHTON: REVISION OF PLETHODON


VEHICULUM GROUP

Plethodon dunni Bishop
Dunn's Salamander

Plethodon dunni Bishop (1984:169-71). Bishop (1943:242-6); Stebbins (1951:
68-72), (1954:54-5).

HOLOTYPE. USNM 95196, adult female, collected just outside the
city limits of Portland, Clackamas County, Oregon, by Stanley G.
Jewett, Jr., 13 January 1934.
DIAGNOSIS. A western plethodon with two phalanges in the fifth
hind toe, a modal number of 16 trunk vertebrae, and a greenish-yellow
dorsal stripe.
DISTRIBTION. Curry County in southeastern Oregon, north to
Pacific County, Washington, east to the western slope of the Cascade
Mountains (fig. 12).
DESCRIPTION. The dorsal stripe is composed of greenish-yellow
chromatophores, in contrast to the red chromatophores usual in
vehiculum. This pigment also occurs abundantly on the sides, but
diminishes on the belly. The lower sides and belly show a few yellow-
ish iridophores, and the brassy type occurs in the iris.
The costal grooves usually number 15, the trunk vertebrae 16.
Vomerine teeth range from 8 to 21 (fig. 13). The largest specimen
examined is 66 mm. in snout-vent length.


P. dunni is closely related to P. vehiculum, but differs in size, colora-
tion, average number of vomerine teeth, and number of body segments.
This species occurs sympatrically with P. vandykei in southwestern
Washington, with P. elongatus in southwestern Oregon, and with
P. vehiculum throughout most of its range. P. dunni is apparently the
most aquatic plethodon; Stebbins (1951:70) states it is almost in-
variably found in places saturated with water and that it often takes
to the water to escape capture.
Most individuals have a greenish-yellow dorsal stripe, but Stebbins
(1951:69) reports melanistic specimens from Benton County, Oregon,
lacking the stripe.















P dunni


* *-


.:7e
*
* f


20 30 40 50 60

SNOUT-VENT LENGTH (mm.)


FIGURE 12. Distribution of P. dunni in
Washington andc Oregon.


FIGURE 13. Variation in the number of vomerine teeth of P. dunni.


I-
w
w
F_
w
z

0
2
0








HIGHTON: REVISION OF PLETHODON


Plethodon vehiculum (Cooper)
Western Red-backed Salamander
Ambystoma vehiculum Cooper (1860:pl. 81, fig. 4). No description accompanies
the figure published by Cooper. The salamander illustrated clearly has the
appearance of the western red-backed salamander, and Bishbp (1934:171)
assigned Cooper's name to this species. However, the salamander Cooper
figures has only 13 costal grooves, a lower number than known in this species
or in larselli or dunni. Occasional specimens of vandykei may have 13 costal
grooves, but Astoria, Oregon, the collection site of Cooper's animal, is out-
side the known range of vandykei. Most likely the specimen figured by
Cooper is a western red-backed salamander with the number of costal grooves
drawn incorrectly. Thus no change in the current nomenclature is necessary.
Plethodon intermedius Cope (1867:209-10). Type locality: Fort Tejon, California;
obviously in error. Dunn (1926:154-6); Slevin (1928:51-5).
Plethodon vehiculus, Bishop (1934:171).
Plethodon vehiculum, Slater (1940:43); Bishop (1943:278-81); Stebbins (1951:
84-7), (1954:57-9).
HOLOTYPE. Apparently no longer extant. The type locality is
Astoria, Oregon.
DIAGNOSIS. A western plethodon with a modal number of 17 trunk
vertebrae.
DISTRIBUION. From Coos County in southwestern Oregon north to
southwestern British Columbia, including Vancouver Island (fig. 14).
DESCRIPTION. Two color phases occur, a striped phase and a uni-
colored phase. The dorsal stripe varies from light yellow through yel-
low, orange, and red to brown. It is never greenish yellow as in dunni.
The chromatophores that color the dorsal stripe are absent from the
sides but present on the belly. The sides are black with a few small
white iridophore spots. These also occur on the belly, and combined
with the melanophores and red chromatophores give it a mottled
appearance. Unstriped individuals have small white iridophore spots
on the dorsum similar to those on the sides and smaller brassy flecks
as well. Brassy iridophores are present in the iris.
Costal grooves usually number 16, trunk vertebrae 17. Vomerine
teeth range from 5 to 14; one unusual specimen has 22 vomerine
teeth (fig. 15). The largest specimen examined is 53 mm. in snout-vent
length.


This species and P. larselli are the smallest of the western plethodons.
P. vehiculum is the most widespread of the western plethodons, and
the most abundant member of the genus over much of its range.
P. vehiculum is superficially similar to the eastern P. cinereus;





































FIGURE 14. Distribution of P. vehicu-
lum in British Columbia, Washington,
and Oregon.


P. vehiculum


w 20.
w

W 15- 0
z 0

2 10- * *** *
O

5- *


0 10 20 30 40 50 60 70 80
SNOUT-VENT LENGTH (mm.)
FIGURE 15. Variation in the number of vomerine teeth of P. vehiculum.







HIGHTON: REVISION OF PLETHODON


both have striped and unstriped color phases and mottled bellies.
These similarities are probably due to convergent or parallel evolu-
tion, because in all other characters P. vehiculum is morphologically
more similar to the other western plethodons. No characters yet found
vary significantly geographically.

ELONGATUS GRoUP
Plethodon elongatus Van Denburgh
Del Norte Salamander
Plethodon elongatus Van Denburgh (1916:216-8). Dunn (1926:156-8); Slevin
(1928:55-7); Bishop (1943:246-9); Stebbins (1951:72-6), (1954:55-6).
Dumeril, Bibron, and Bibron (1854:84) first published the name Salamandra
elongata Valenciennes, which has since appeared in the synonymy of Plethodon
glutinosus (Dunn, 1926:137). As elongata is merely a manuscript name, not pro-
posed validly under the International Rules of Zoological Nomenclature, it is not
available for P. glutinosus. Thus Plethodon elongatus Van Denburgh (1916) is not
a junior homonym and can remain the valid name of the Del Norte Salamander.

HOLOTYPE. CAS 29096, collected at Requa, Del Norte County,
California, by J. R. Slevin, 22-26 May 1911.
DIAGNOSIS. A western plethodon with a modal number of 19 trunk
vertebrae.
DISTRIBUTION. Southwestern Oregon and northwestern California
(fig. 16).
DESCRIPTION. The belly is very dark with a few scattered white
iridophore spots. The orange to reddish-brown dorsal stripe is usually
brighter in juveniles than in adults, where it is usually reduced or
obliterated by melanophore suffusion. Often medial melanophore pig-
ment divides it into a right and left dorsolateral stripe. The head
usually has a few small white iridophore spots, which may occur rarely
on the dorsum. Occasional individuals show a few brassy flecks in the
iris, but most have none.
Costal grooves usually number 18, trunk vertebrae 19. Vomerine
teeth range from 8 to 15 (fig. 17). The largest specimen examined is
70 mm. in snout-vent length.

This species is the most elongated of the western plethodons. Its
greater number of body segments distinguishes it from vandykei,
larselli, dunni, and vehiculum; its webbed toes distinguish it from
the last two also. Known only from Trinity, Humboldt, and Del Norte
Counties, California, and Curry County, Oregon, its distribution is


1962


































FIGURE 16. Distribution of P. elongatus in
Oregon and California.


P elongatus


25F


.- .i-


** b

5 -

C

C
0 10 20 30 40 50 60 70 80
SNOUT-VENT LENGTH (mm.)
FIGURE 17. Variation in the number of vomerine teeth of P. elongatus.

oa







HIGHTON: REVISION OF PLETHODON


rather limited. Stebbins (1951:76) believes that specimens from the
interior of the range differ in several respects from those on the coast.

THE EASTERN SMALL PLETHODONS
As the relationships of P. neomexicanus to the four other species of
eastern small plethodons are not clear, it is considered as a separate
species group. The remaining eastern small plethodons are charac-
terized by their small size. With the exception of welleri, they have
more trunk vertebrae and are relatively more elongated than the
eastern large plethodons. P. welleri has a restricted range in the south-
ern Blue Ridge Mountains, and is known only from a few counties
in northwestern North Carolina and adjacent Tennessee and
Virginia.
P. richmondi is an elongated species whose range centers in the
Appalachian Plateaus Physiographic Province of western Pennsylvania,
Ohio, eastern Kentucky, and West Virginia. It also occurs in adjacent
physiographic provinces in parts of Pennsylvania, Maryland, Virginia,
North Carolina, Tennessee, Kentucky, Indiana, and Ohio. P. r.
hubrichti occurs in the Blue Ridge Province of Virginia, and P. r.
netting in the Cheat Mountains of West Virginia. The range of the
latter subspecies is surrounded by that of P. r. richmondi, and thus ap-
parently represents a Rassenkreis overlap.
P. c. cinereus is widely distributed over most of the eastern United
States and southeastern Canada, except for the region occupied by
dorsalis. Two disjunct populations are recognized as subspecies, P. c.
polycentratus in the Piedmont of western Georgia, and P. c. serratus
in Arkansas and Oklahoma. A third disjunct population in southeastern
Missouri is distinct from the nearest cinereus populations, but is similar
to certain other populations in the central part of the range of P. c.
cinereus.
The range of P. d. dorsalis is centered in the Interior Low Plateaus
Physiographic Province, and an isolated subspecies, P. d. angusticlav-
ius, occurs in the Ozark Plateau.
The four species welleri, dorsalis, cinereus, and richmondi form a
natural group of rather closely allied forms. All have webbed toes, a
pigmented peritoneum, low vomerine tooth counts, brassy flecks on
the dorsum, and white iridophore spots on the belly and sides. Both
cinereus and dorsalis typically show at least as much white as black
pigment on their bellies, and both have two color phases, a unicolored
dark or unstriped phase and a striped phase with a dorsal band of
red or yellow on the body and tail. Both richmondi and welleri







BULLETIN OF THE FLORIDA STATE MUSEUM


resemble the unstriped phase of the other two species dorsally, but
possess much more black pigment ventrally; the white pigment of the
belly is usually limited to small spots. The mental glands of mature
male cinereus and richmondi differ in shape from those of all other
Plethodon including the other eastern small plethodons (fig. 2). The
premaxillary teeth of mature males of these two species are highly
modified (fig. 3), another characteristic they share with no other
species of Plethodon. They have the greatest number of trunk verte-
brae in the genus, and significant geographic variation in this charac-
ter occurs only in these two species. Doubtless they are closely re-
lated. The two are sympatric over more than half of the range of
richmondi, and character displacement occurs in color pattern over
much of the area of sympatry.
P. welleri is the least elongated and has the fewest trunk vertebrae
of the eastern small species. It is unusual in having melanophore pig-
mentation in the ovaries and an extremely dark parietal peritoneum.
The brassy flecks on the dorsum of welleri coalesce to form large
patches, giving it an exceptional appearance found only in one other
form, P. r. hubrichti.
Throughout most of their ranges cinereus and dorsalis are allopatric,
but they overlap in several areas. As they apparently do not hybridize,
they are usually considered distinct species. They differ in several pig-
mentation characters, in the shape of the premaxillary teeth and
mental glands of adult males, and usually in the average number of
trunk vertebrae. Dunn (1926:24) suggests that the striped pattern
of dorsalis is more primitive than that of cinereus, as the paired dorsal
spots of a primitive plethodontid probably pass through a zigzag phase
similar to that of dorsalis before evolving into a straight-edged stripe
as in cinereus.
The elongation and increased trunk vertebrae in richmondi and
cinereus appear as a marked departure from the usual plethodontid
condition. The forms welleri and dorsalis, therefore, have retained
more primitive characteristics.
With mammalian gonadotropic hormone injections I have induced
gravid female cinereus, richmondi, nettingi, hubrichti, and welleri to
deposit eggs in the laboratory. Embryos and newly hatched young of
dorsalis are available in museum collections. Embryonic welleri, rich-
mondi, hubrichti, and netting have abundant red pigment on the
dorsum. In richmondi and netting the red remains after hatching.
The young have a straight-edged stripe, though it is less well defined
than in striped cinereus and dorsalis. Embryos of welleri have a series


Vol. 6















TABLE 3. Number of trunk vertebrae in eastern small plethodons. (See tables 4, 5, and 7 for geographic variation in dorsalis,
cinereus, and richmondi.)


16 16/17 17 17/18 18 18/19 19 19/20 20 20/21 21 21/22 22 22/23 23 23/24


Plethodon
neomexicanus 2 3 16 1
welleri 4 1 154 4 14
dorsalis 25 16 905 34 327 3 4
cinereus 2 114 41 4591 674 18335 672 4571 34 163 2 7 1
richmondi 119 32 592 30 516 25 731 39 839 25 174 7


24 Mean


19.89 z
17.06
19.25
20.00 t
12 20.71 4

0
C;


Species








BULLETIN OF THE FLORIDA STATE MUSEUM


of alternating dorsolateral red spots, similar to those found in
embryonic dorsalis. In welleri these spots disappear. In dorsalis they
enlarge and coalesce to form a stripe in which the original spots appar-
ently produce the lobular edges characteristic of most populations of
dorsalis. Similarities in embryonic dorsal color patterns, mental glands,
and premaxillary teeth of males are the basis for placing welleri and
dorsalis in a single species group.
Grobman (1944) divides the eastern small plethodons into two
species groups, the "cinereus group" including cinereus and dorsalis,
and the "welleri group" including richmondi, nettingi (then considered
a full species), and welleri. This grouping is based mainly on the
presence of the red or yellow dorsal stripe in cinereus and dorsalis
and its absence in the other species, and on the presence of dorsal
brassy flecks in the "welleri group." Rabb (1955) and Muchmore
(1955) correctly point out that the brassy flecks characteristically
present in richmondi and welleri also occur in the unstriped phase
of cinereus. Indeed they occur in many species of Plethodon. More-
over, richmondi, nettingi, welleri, and hubrichti as embryos have con-
siderable red pigment on the dorsum. As pattern dimorphism (stripe
or no stripe) is rather widespread in this genus and occurs in other
genera as well, it is not particularly useful in evaluating relationships.
The presence of two color phases in some western as well as eastern
species suggests that the dorsal stripe has probably been present in the
genus for a long time or has reappeared repeatedly. The two color
phases of cinereus may be due to a single pair of alleles (Highton,
1959).
NEOMEXICANUS GROUP

Plethodon neomexicanus Stebbins and Riemer
Jemez Mountains Salamander
Eurycea multiplicata, Dunn (1926:313-6; part).
Plethodon neomexicanus Stebbins and Riemer (1950:73-80). Stebbins (1951:
76-9), (1954:56).
HOLOTYPE. MVZ 49033, adult male, -collected 12 miles west and
4 miles south of Los Alamos, Sandoval County, New Mexico, altitude
about 8750 feet, by Robert C. Stebbins, 14 August 1949.
DIAGNOSIS. A Plethodon with a modal number of 20 trunk verte-
brae and with a single phalanx in the fifth hind toe.
DIsTmRIBUTON. Known only from the Jemez Mountains of New
Mexico (fig. 18).


Vol. 6'





























FIGURE 18. Distribution of P. neomexi-
canus in New Mexico.


P neomexicanus


- .


6
0 0 60


0 I i I
0 10 20 30 40 50 60 70 8(
SNOUT-VENT LENGTH (mm.)
FIGURE 19. Variation in the number of vomerine teeth of P. neomexicanus







BULLETIN OF THE FLORIDA STATE MUSEUM


DESCRIPTION. Dorsal and lateral melanophore pigmentation is re-
duced so that the ground color appears brown, rather than black as
in most other species. Pigmentation is so reduced on the belly that
the internal organs are visible through the belly wall. Brassy irido-
phore flecking is heavy on the dorsum of the body and tail, particularly
in gaps in the melanophore background. These gaps are more numer-
ous in young animals and often occur most abundantly in two dorso-
lateral longitudinal lines. Brassy flecking is so heavy here that some
young specimens appear to have a brassy dorsal stripe. Lateral irido-
phore spots are yellow and small, not much larger than the dorsal
brassy flecks. A few of these small yellow spots occur on the belly
also. The iris contains abundant brassy pigmentation.
Costal grooves usually number 19, trunk vertebrae 20 (range,
19-21, mean 19.9). Vomerine teeth range from 11 to 18 (fig. 19).
The type is more than 70 mm. in snout-vent length. According to
Stebbins and Riemer (1950:75), sexual maturity is attained at approx-
imately 50 mm. snout-vent length.

WELLERI GROUP
Plethodon welleri Walker
Weller's Salamander
Plethodon welleri Walker (1931:48-51). Bishop (1943:285-7); Grobman (1944:
313).
Plethodon welleri welleri, Thurow (1956a:343-56).
Plethodon welleri ventromaculatum Thurow (1956a:344). Type locality: Mt.
Rogers, Grayson County, Virginia.

HOLOTYPE. USNM 84135, adult male, collected at an altitude above
5000 feet on Grandfather Mountain, near Linville, North Carolina,
by W. H. Weller and Ralph Dury, 27 August 1930.
DIAGNOSIS. A dark-bellied eastern small plethodon with 17 trunk
vertebrae and abundant dorsal brassy spotting. In adult males the
mental gland is round and the premaxillary teeth have small anterior
cusps.
DIsTIBunoN. From Yancey County, North Carolina, northeast-
ward in Tennessee and North Carolina to Mt. Rogers and White Top
Mountain, Virginia (fig. 20).
DESCRIPTION. The dorsal pattern consists of large anastomosing
patches of brassy spots almost entirely made up of brassy iridophores.
Brassy pigment is more abundant in welleri than in any other
Plethodon; in some specimens it occupies more than two-thirds of the


Vol. 6







HIGHTON: REVISION OF PLETHODON


dorsum. White iridophore pigmentation present on the dorsum of most
plethodons is lacking in welleri except on the sides and belly. Small
white spots (0.1-0.2 mm.) are scattered over the belly. Lateral irido-
phore spots (to 1 mm.) are white with an occasional slight brassy
flecking. Four of eight specimens from Mt. Rogers, Virginia, in life
had small red spots on the proximal segment and at the base of the
front legs.
Costal grooves usually number 16, trunk vertebrae 17. Vomerine
teeth range from 7 to 12 (fig. 21). Maximum snout-vent length is
between 45 and 50 mm. Sexual maturity is reached at about 35 mm.
snout-vent length.

P. welleri has the same modal number of trunk vertebrae (17) as
most of the eastern large plethodons and resembles them more in
body form than does any other eastern small plethodon. Yet its small
size, vertebral structure, webbed toes, pigmented parietal peritoneum,
and other pigmentation characters associate it most closely with the
small eastern forms.
This species, long thought to be restricted to high elevations, may
occur in suitable habitats at lower altitudes throughout its range.
Hoffman (1953) collected it at 2500 feet in Johnson County,
Tennessee.
P. welleri has pigmented ovaries. Its closest relative is probably
dorsalis, whose ovaries also are occasionally pigmented. The coloring
of welleri is similar to that of adult P. r. hubrichti, which is not far
separated geographically from welleri and occurs in similar habitats.
Only P. r. nettingi is closer to welleri in number of body segments
than is P. r. hubrichti.
Thurow (1956a) describes geographic variation in the amount of
ventral white spotting in welleri. He shows, as previously noted by
Walker (1934), that individuals from Grandfather Mountain, North
Carolina, the type locality of welleri, have less white spotting on the
belly than those from White Top Mountain and adjacent Mt. Rogers,
Virginia. Thurow examined few specimens from other localities, for
little material is available. It seems wise to await additional informa-
tion on geographic and individual variation before recognizing races
on the basis of such limited material. One specimen from Flat Top
Mountain, Yancey County, North Carolina (CU 4768) has a dark
belly and chin without white mottling, but Snyder (1946) describes
a specimen with conspicuous mottling on the belly from this locality.
Specimens from 5.3 miles north of Carderview, Johnson County,


1962






































FIGmRE 20. Distribution of P. welleri in Virginia, North
Carolina, and Tennessee.


I20
w
w
I-
15
UJ
z
w 10
0


P welleri


0
m 0r
.
Fr


0 10 20 30 40 50 60 70 80
SNOUT-VENT LENGTH (mm.)

FIGURE 21. Variation in the number of vomerine teeth
of P. weUeri.







HIGHTON: REVISION OF PLETHEDON


Tennessee, have a moderate amount of ventral spotting. Specimens
from Mt. Rogers average much more dorsal brassy spotting than those
from Grandfather Mountain. I have not examined living material from
other localities, except for a young salamander from near Carder-
view. Description of a juvenile from this locality by Hoffman (1953)
actually is based on a richmondi (USNM 132325).

Plethodon dorsalis dorsalis Cope
Zigzag Salamander
Plethodon cinereus dorsalis Cope (1889:138-9). Blanchard (1926:269-70);
Bishop (1943:236-9).
Plethodon erythronotus, Garman (1894:38).
Plethodon dorsalis, Stejneger and Barbour (1917:15); Dunn (1926:158-62);
Grobman (1944:308-11).
Plethodon dorsalis dorsalis, Thurow (1956b: 177-82), (1957b:91-9).

LECTOTYPE. USNM 3776A (one of four syntypes listed in the orig-
inal description, here designated as the lectotype), collected at Louis-
ville, Jefferson County, Kentucky.
DIAGNOSIS. An eastern small plethodon with the modal number of
trunk vertebrae 19, usually with a wide dorsal stripe with irregular
borders, a round mental gland in mature males, premaxillary teeth
with small anterior cusps in both sexes, and red chromatophdres on
the belly in addition to black melanophores and white iridophores.
It differs from P. d. angusticlavius in having a wider dorsal stripe.
DISTRIBUTION. From southern Illinois, Indiana, and southeastern
Ohio, south through Kentucky and Tennessee to northeastern Missis-
sippi, northern and central Alabama, the Piedmont of western
Georgia and the French Broad River Valley of North Carolina
(fig. 22).
DESCRIPTON. The dorsum of both color phases of P. d. dorsalis has
small white spots (0.2-0.4 mm.) as well as smaller brassy flecks. The
red-backed phase has abundant red chromatophore pigment in the
stripe region and on the sides and belly. All the chromatophores are
similar to those of P. cinereus. Color differences between the two
species involve mainly the abundance and distribution of the three
pigments. The melanophore background in dorsalis is somewhat re-
duced, giving the animal a lighter appearance over-all than cinereus.
On the belly an increase in the number of red chromatophores pro-
duces a mottling of red, black, and white, instead of black and white
alone as in cinereus. On the head of dorsalis red pigment is concen-
trated in front of the eyes. Lateral iridophore pigment is often yellow.







BULLETIN OF THE FLORIDA STATE MUSEUM


FIGURE 22. Distribution of P. dorsalis in the eastern United States.

The dorsal red stripe varies in the amount of melanophore encroach-
ment and in the degree of irregularity of its edges. Brassy pigment
is present in the iris.
Costal grooves usually number 18, trunk vertebrae 19. Vomerine
teeth range from 5 to 18 (fig. 23). This is a small species; sexual
maturity is reached at about 30 mm. snout-vent length. The largest
specimen measured is 50 mm. in snout-vent length.


Vol. 6










HIGHTON: REVISION OF PLETHODON


R dorsolis
SSW Illinois














10 120 O 4 so 80
SNOUT-VENT LENGTH (mm)


R dorsalis
SE Illinois


I-l






5-
0 ... .-'"




0 10 20 30 7040 50 80 00
SNOUT-VENT LENGTH (nmJ


R dorsolis
Tennessee


R dorsalis
r Arkansas


I .*





0 10 20 30 40 80 60 70 80
SNOUT-VENT LENGTH (mm)


R dorsolis
Georgia


0 10 20 30 40 s0 60 70 80 _0 10 D20 0 40 o0 0 70 80T
SNOUT-VENT LENGTH (mm.) SNOUT-VENT LENGTH (m.)

FIGURE 23. Variation in the number of vomerine teeth in 5 geographic samples
of P. dorsalis.







BULLETIN OF THE FLORIDA STATE MUSEUM


Plethodon dorsalis angusticlavius Grobman
Ozark Red-backed Salamander
Plethodon dorsalis, Black and Dellinger (1938:7).
Plethodon cinereus, Black and Dellinger (1938:7; part).
Plethodon cinereus angusticlavius Grobman (1944:302).
Plethodon dorsalis angusticlavius, Thurow (1956b:177), (1957b:99).

HOLOTYPE. AMNH 40366, adult male, collected at Mud Cave, near
Fairy Cave, Stone County, Missouri, by B. C. Marshall, 1 October
1927.
DIAGNOSIS. A race of Plethodon dorsalis in which the width of the
dorsal stripe is usually less than one-third the width of the body.
DISTRIBUTON. Southwestern Missouri, northwestern Arkansas and
adjacent Oklahoma, north of the Arkansas River (fig. 22).
DESCRIPTION. More than 50 percent of the individuals in some lo-
calities lack the dorsal stripe. The narrow dorsal stripe of the striped
phase may be yellow or red, and it usually widens on the tail. Some-
times the stripe becomes completely suffused with melanophores in
the body region and remains visible only on the tail. The stripe is
sometimes irregular anteriorly as in dorsalis. Pigmentation of the belly
and sides is also similar to dorsalis. Red lateral and ventral pigment is
reduced or absent.
Costal grooves usually number 18, trunk vertebrae 19. Vomerine
teeth range from 8 to 16 (fig. 23). The largest specimen measured is
43 mm. in snout-vent length.


The distribution of P. dorsalis centers in the Interior Low Plateaus
Physiographic Province, but the species ranges various distances into
every adjoining physiographic province. It occurs in the Coastal Plain
of Mississippi and western Tennessee, the Piedmont of western
Georgia, the Cumberland Plateau of Alabama and Tennessee, the Blue
Ridge Province of North Carolina, and the glaciated region to the
north in Indiana and Illinois. A disjunct subspecies, P. d. angusti-
clavius, occurs in the southern section of the Ozark Plateau of south-
western Missouri, northwestern Arkansas, and adjacent Oklahoma.
Grobman (1944) originally described angusticlavius as a race of
P. cinereus, but all available evidence supports its transfer to dorsalis
by Thurow (1956b, 1957b). Number of trunk vertebrae, male second-
ary sexual characteristics, ventral red pigmentation, and habitat all


Vol. 6







HIGHTON: REVISION OF PLETHODON


associate it closely with dorsalis. Thurow (MSb) found that dorsalis in
Indiana is usually associated with wetter habitats than cinereus, an
animal of the forest floor. P. dorsalis is usually associated with
canyons, ravines, and escarpments, in exposures of naked rock around
springs and seepage areas. Both subspecies of dorsalis often are found
in caves.
The species cinereus and dorsalis are so similar in appearance in-
dividual specimens are often difficult to identify, particularly after
preservation. Thurow (1956b) lists a number of differences between
them, but most differential characters show some overlap. The two
species are sympatric in Georgia, eastern Tennessee, Indiana, Illinois,
and Oklahoma. Though hybridization is not reported, studies to de-
termine whether it occurs in these areas are desirable. In Georgia the
stripe of dorsalis often has straight edges, resembling that of the local
cinereus, but a similar condition also occurs in southwestern Illinois,
where cinereus is absent. Differences in number of trunk vertebrae
between the two species are greater in Georgia than in other areas
of sympatry.
In some areas of sympatry ecological differences may isolate
cinereus and dorsalis. The two are generally separated altitudinally
where they occur together in the Great Smoky Mountains of Ten-
nessee, with cinereus at higher elevations (1600 to 5000 feet) and
dorsalis restricted to lower elevations (below 2200 feet) (King,
1939:551, and Grobman, 1944:310-1). The two species occur together
at White Oak Sinks and Little River Sinks. They also occur sym-
patrically in the Piedmont of western Georgia in at least two localities-
3.8 miles north of McDonough, Henry County and 8.5 miles south of
Thomaston, Upson County. Bragg (1955) reports specimens of P. c.
serratus and P. d. angusticlavius from about 4 miles east of Eldon,
Cherokee County, Oklahoma. This is the only record of sympatric
occurrence and of P. c. serratus north of the Arkansas River. Thurow
(MSb) reports that near Bloomington, Indiana, the two species differ
markedly in seasonal activity, in habitat, moisture, and temperature
preferences. Apparently where the two species are geographically
sympatric, they are usually ecologically isolated.
The most useful character for distinguishing between striped in-
dividuals of the two species is the nature of the stripe border. It is
straight in cinereus, but lobular or zigzag in dorsalis. The unstriped
phase of dorsalis differs from that of cinereus in having small amounts
of red pigment on the dorsum, sides, and belly, often enough to out-
line the characteristic dorsal lobular stripe. Unstriped cinereus rarely







BULLETIN OF, THE FLORIDA STATE MUSEUM


have red pigment except in the subspecies polycentratus and
serratus. Adult male cinereus and dorsalis usually show two striking
differences, the shape of the mental gland (fig. 2). and the modifica-
tion of the premaxillary teeth (fig. 3), but these characters are of no
use in distinguishing females and immatures.
The modal number of trunk vertebrae in dorsalis is 19; that in
adjacent and sympatric populations of cinereus is 20, except in Georgia
where it is 21 or 22. Some populations of cinereus also have a modal
number of 19 trunk vertebrae, but none of these occurs in areas of
overlap. As a small percentage of specimens in most cinereus popula-
tions have 19 trunk vertebrae and a similar percentage of dorsalis
have 20 (tables 4 and 5), this character often is not definitive for
individual specimens.
Red chromatophores are present on the belly of living dorsalis,
but absent in living cinereus except occasionally in the pectoral area
between the front limbs. As the red pigment is soluble in preserva-
tives, this valuable character cannot be used to identify museum speci-
mens. Other differences Thurow (1956b) lists are of little value in
identifying individual specimens because of the great variation within
both species.
Dunn (1926:162) states that the unstriped dark phase of dorsalis
is cined to adults and is lighter than the dark phase of cinereus.
Apparently the dark phase is due to the suffusion of melanophores
into the dorsal stripe. In dorsalis this invasion may occur at a very
early age. Five juveniles from Van Buren County, Tennessee (UF
8394) that are only 16 to 18 mm. in snout-vent length have very
dark stripes in contrast to other young and adults of the same series.
I have seen similar young from Monroe County, Indiana (AMNH
52433-6). I previously classified similar specimens of cinereus as
striped (Highton, 1959:34). If the unstriped dark phase of dorsalis
is homologous to the indistinctly striped cinereus, then no truly un-
striped phase occurs in dorsalis corresponding to the lead-backed phase
of cinereus.
Geographic variation in the number of trunk vertebrae is slight
in different populations of dorsalis (table 4). Samples from the south-
ern part of the range (Alabama, Mississippi, and Georgia) have a
slightly lower average vertebral count than those from the northern
parts of the range (Kentucky and Indiana). The difference appears
attributable to a gradual clinal increase from south to north rather
than to hybridization of dorsalis and cinereus in Indiana.
Neill (1957:44) reports dorsalis from Dade County, Georgia, in


Vol. 6







HIGHTON: REVISION OF PLETHODON


TABLE 4. Geographic variation in the number of trunk vertebrae in P. dorsalis.
Subspecies Number of trunk vertebrae
and 18 18/19 19 19/20 20 20/21 21 Mean
State

P. d. dorsalis
Mississippi 8 19.00
Alabama 4 36 1 2 18.97
Georgia 2 1 53 1 3 19.02
North Carolina 6 19.00
Tennessee 7 3 226 5 49 1 19.15
Kentucky 1 57 1 33 19.35
Ohio 1 -
Indiana 5 6 261 22 196 2 4 19.42
Illinois 6 3 164 3 34 19.13
P. d. angusticlavius
Arkansas 1 77 1 10 19.11
Oklahoma 1 10 18.95
Missouri 7 19.00


the Appalachian Plateau region of the extreme northwestern part of
the state. The species also occurs in the Piedmont of western Georgia,
but in this area most specimens have a rather straight-edged dorsal
stripe at least posteriorly, similar to that of cinereus. Only in this char-
acter do they show an approach to cine'reus. Georgia material of the
two forms is easily distinguishable by the costal groove counts, ranging
from 17 to 19 in dorsalis and 20 to 22 in cinereus. Living specimens of
dorsalis have much red pigment on the venter, and the bellies of pre-
served specimens are usually lighter than those of cinereus because
fewer melanophores are present. Georgia dorsalis are unusual in having
few melanophores in the dorsal stripe at the base of the tail, so the
stripe is much brighter there than in the body region. In this char-
acter they resemble many P. d. angusticlavius. A series of 18 specimens
from near Oxford, Cleburne County, Alabama (BDV) approach the
dorsal stripe condition of Georgia dorsalis.
In southeastern Illinois most dorsalis have a typical lobular stripe,
but those in southwestern Illinois usually have a straight-edged stripe,
as in cinereus. Many workers familiar with the typical striped patterns
of dorsalis and cinereus would assign southwestern Illinois material to
cinereus, as did Smith and Minton (1957:347). However, Thurow's
(1957b) analysis of specimens from the area and my own study of
additional living and preserved material indicate that these salaman-
ders resemble dorsalis in all other characters and should be assigned


1962



























FIGURE 24. Geographic variation in P. dorsalis. (Upper) Hatched area rep-
resents region where species has a narrow dorsal stripe. (Lower) Hatched areas
represent regions where species has a straight-edged dorsal stripe.


411







HIGHTON: REVISION OF PLETHODON


to that species. Thurow (1957b:94) regards most of these populations
as P. c. angusticlavius. Others from southern Union County, Illinois,
he considers intergrades between the races angusticlavius and dorsalis
because they have wider stripes than those from farther north. They
also have what he calls "semi-lobed" borders to the dorsal stripe.
Rossman (1960:211) reports the average width of the stripe in a
series from the Pine Hills of Union County, Illinois, to be closer to
P. d. angusticlavius as defined in the original description by Grobman,
but he compares the stripe to the width in cinereus rather than in
dorsalis. The average width of the dorsal stripe when measured be-
tween the lobes is less in P. dorsalis than in P. cinereus.
The southwestern Illinois populations have a straight-edged wide
stripe. Georgia dorsalis material also has a wide straight-edged dorsal
stripe (fig. 24). It seems reasonable to conclude that the straight-edged
stripe has evolved separately in southwestern Illinois and in Georgia,
and that the subspecies angusticlavius developed through an increase
in the frequency of the narrow striped pattern from a population
similar to that now existing in southwestern Illinois. Geographically
the two races are closest in southwestern Illinois and western Ten-
nessee. Southeastern Missouri and northeastern Arkansas may yield
additional intergrade material between the two races. Lack of speci-
mens rather than real isolation is possibly responsible for the apparent
disjunct distribution of the two forms.

CNEREUS GROUP
Plethodon cinereus cinereus (Green)
Red-backed Salamander
Salamandra erythronota Rafinesque (1818:4). Type locality: Hudson Highlands
of New York. Green (1818:356).
Salamandra cinerea Green (1818:356-7). Type locality not listed, but presumably
the vicinity of Princeton, New Jersey.
Salamandra erithronota, Smith in Hitchcock (1833:552).
Plethodon cinereus, Tschudi (1838:58); Dunn (1926:163-80).
Salamandra agilis Sager (1839:322-3). Type locality not listed, but presumably
Detroit, Michigan.
Sauropsis erythronota, Fitzinger (1843:33).
Plethodon erythronota, Baird (1850:285).
Ambystoma erythronotum, Gray (1850:37-8).
Plethodon eryhronotus, Storer (1852:138).
Plethodon erythronotum, Dumeril, Bibron, and Dum6ril (1854:86).
Salamandra puncticulata Valenciennes in Dumbril, Bibron, and Dumeril (1854:
87). Nomen nudum.
Plethodon erythronotus cinereus, Cope (1869:100).


1962







286 BULLETIN OF THE FLORIDA STATE MUSEUM


Plethodon erythronotus erythronotus, Cope (1869:100).
Plethodon cinereus cinereus, Davis and Rice (1883:12); Bishop (1941a: 196-219),
(1943:232-6); Grobman (1944:330-2).
Plethodon cinereus erythronotus, Davis and Rice (1883:12).
Plethodon cinereus erythronota, Britcher (1903:120).
Plethodon huldae Grobman (1949:136). Type locality: Hawksbill Mountain,
Madison County, Virginia.
Many think the names Salamandra cinerea and Salamandra erythronota, based
on the unstriped and striped color phases of this species respectively, were
simultaneously proposed by Green (1818:356). During the 19th century the
two color phases were considered variously as species, subspecies, and varieties,
but Stejneger and Barbour (1917:15) conclude that cinereus should be the
valid name for both, as the first revisor (Tschudi, 1838:58) used this name.
Since 1917 cinereus has been regarded as the valid name of the species, and a
large literature has accumulated using this name. Reed (1960) and Goodwin
(1960) have shown that the name erythronota has clear priority over the name
cinerea. Green (1818) clearly stated that Rafinesque (1818:4) had already
validly described the species as erythronota. Though the name cinereus is thus a
junior synonym, a change in this long established name would be extremely un-
desirable and an appeal has been submitted to the International Commission on
Zoological Nomenclature to validate cinereus and suppress erythronotus (Highton,
1960b). Pending a decision by the Commission, it is suggested that zoologists
continue to use P. cinereus for this species.

HOLOTYPE. Dunn (1926:165) states that the type is not known to
exist. He refers, of course, to the type series of Green's species,
Salamandra cinerea. In the Academy of Natural Sciences of Phila-
delphia collection are 12 specimens from the Green collection (ANSP
1227-38), among them one that is probably Eurycea bislineata
(ANSP 1238) and another Desmognathus fuscus (ANSP 1228). A
third specimen (ANSP 1229) is in such poor condition I cannot iden-
tify it. The other 9 are red-backed salamanders, 5 of the striped
phase and 4 of the unstriped phase. I propose that the unstriped
specimens (ANSP 1232, 1233, 1234 and 1237) be considered the syn-
types of Salamandra cinerea, and here designate ANSP 1232 as the
lectotype. The types of Rafinesque's Salamandra erythronota, orig-
inally supposed to be in the Lyceum Collection, are probably no
longer extant.
DIAGNOsis. An eastern small plethodon with a black and white
mottled belly; a modal number of 19, 20, or 21 trunk vertebrae, de-
pending on geographic locality (fig. 27); and a straight-edged dorsal
stripe in the striped phase.
DISTrIBUTION. In eastern Canada, from Nova Scotia including Cape
Breton Island, Prince Edward Island, New Brunswick, southern Que-
bec and Ontario, southward through the eastern United States to


Vol. 6







HIGHTON: REVISION OF PLETHODON


FIGURE 25. Distribution of P. cinereus in eastern North America.

North Carolina and eastern Tennessee, Ohio, Indiana, and southeast-
ern Illinois; southeastern Missouri (fig. 25).
DESCRIPTION. The dark lead-backed or unstriped phase is charac-
terized by the absence of red chromatophores, the presence of small
white spots on the dorsum (0.07-0.2 mm. diameter), and numerous
smaller brassy flecks on the head, back, and tail. The belly is mottled
with black melanophores and yellow or white iridophores. Lateral irido-
phore pigment is similar to that on the belly and contrasts sharply
with the brassy pigment on the dorsum.
The red-backed or striped phase has sides and belly similar to
the unstriped phase. The white spots and brassy flecks are reduced
in the dorsal stripe region, but not on the head and tail. Chromato-
phores in the dorsal stripe are somewhat variable in color. Some speci-
mens lack red in the stripe. Abundant brassy pigment is present in
the iris of both color phases.
The modal costal groove count varies geographically from 18 to 20,
and the number of trunk vertebrae from 19 to 21. Vomerine teeth











288 BULLETIN OF THE FLORIDA STATE MUSEUM


Vol. 6


range from 5 to 20 (fig. 26). Maximum size is 52 mm. snout-vent

length. Sexual maturity is reached at about 35 mm. snout-vent length.


Plethodon cinereus serratus Grobman

Ouachita Red-backed Salamander

Plethodon cinereus serratus Grobman (1944:306-8). Thurow (1957b:91-9).


R cinertus
4 Jacquea-Cartier County, Quebec


5-


0 0 20 30 40 50 60 70 80
SNOUT-VENT LENGTH (mm.)


P. cinereus
30 Frederick County, Maryland


-




10 20 30 40 50 60 70 eO
SNOUT-VENT LENGTH (mm.)


. cinereus
| Arkansas a Oklahoma


I




I 10 20 30 40 S0 60 70 a
SNOUT-VENT LENGTH (mm.)


P cinereus
4l Somerset County, Maryland


I 10 20 30 40 50 60
SNOUT-VENT LENGTH (mm)


70 80


p. cinereus
Georgia


0 M 20 W 40 B0 60 70 80 0 10 20 30 40 50 60 o70 0
SNOUT-VENT LENGTH (mm.) SNOUT-VENT LENGTH (mm.)

FIGURE 26. Variation in the number of vomerine teeth in 6 geographic samples
of P. cinereus.


P. cinlerlus
Porter County, Indina


I
-


" "
"







HIGHTON: REVISION OF PLETHODON


HOLOTYPE. CNHM 39464, female, collected on Rich Mountain, Polk
County, Arkansas, at 2500 feet, by Karl P. Schmidt and C. M.
,Barber, 23 March 1938.
DIAGNOSIS. A race of Plethodon cinereus in which the edges of the
dorsal stripe have a serration at each costal groove. In the unstriped
phase red pigment is present on the body and sides. The modal
number of trunk vertebrae is 20.
DISTRIBUTION. West-central uplands of Arkansas and adjacent
Oklahoma south of the Arkansas River (fig. 25). Also reported by
Bragg (1955:28) north of the Arkansas River in Cherokee County,
Oklahoma.
DESCRIPTION. In striped individuals the serrations of the dorsal stripe
are present on the body, but not on the tail. Serrations are produced
by the red pigment in the dorsal stripe extending ventrolaterally at
the top of each costal furrow. Absence of melanophores in these ex-
tensions makes the saw-tooth edge of the stripe conspicuous. In all
other ways this form resembles P. c. cinereus except for the presence
of red pigment on the sides and dorsum in the unstriped phase. This
statement is based on preserved specimens; I have examined no living
unstriped P. c. serratus.,
Coastal grooves usually number 19, trunk vertebrae 20. Vomerine
teeth range from 4 to 15 (fig. 26). The largest specimen examined is
46 mm. in snout-vent length.


Plethodon cinereus polycentratus Highton and Grobman
Georgia Red-backed Salamander

Plethodon cinereus polycentratus Highton and Grobman (1956:185-7).
HOLOTYPE. UF 8376, adult male, collected 2 miles northeast of
Palmetto, Fulton County, Georgia, by Albert H. Highton and Richard
Highton, 2 February 1954.
DIAGNOSIS. A race of Plethodon cinereus in which the usual num-
ber of trunk vertebrae is 21 or 22, and with red pigment on the sides of
unstriped specimens.
DISTRIBUTION. The Piedmont of western Georgia (fig. 25).
DESCRIPTION. This subspecies is similar to the nominate subspecies
in color, but usually has more red pigment on the anterior part of
the belly, especially between the front limbs, than do the other two
subspecies. Unstriped specimens have red pigment, unlike unstriped
P. c. cinereus. Trunk vertebrae range from 21 to 23, vomerine teeth







BULLETIN OF THE FLORIDA STATE MUSEUM


from 7 to 15 (fig. 26). The largest specimen is 46 mm. in snout-vent
length. Sexual maturity is reached at about 35 mm. snout-vent length.


Throughout much of its extensive range the red-backed salamander is
the commonest terrestrial urodele. Two distinct color phases occur in
many populations, one with a prominent dorsal stripe (usually red,
but sometimes yellow or colorless), the other unstriped. Geographic
variation in color phase frequencies is discussed below. Recent evi-
dence (Highton, 1959) shows that the difference between the striped
and the unstriped phases may be due to one pair of alleles with the
gene for striped pattern dominant.
An erythristic color phase, similar to the striped phase except for
a reduction of melanophore pigmentation, is reported from several
localities in the northern part of the species' range (Thurow, MSb,
summarizes literature records). In most areas this "scarlet phase" ap-
pears only as a rare variant, but is abundant in some populations in
Litchfield County, Connecticut (Matthews, 1952; Reed, 1955).
Another color variant is a rare spotted phase. This type is similar
to the unstriped phase except that it has small red spots scattered over
the dorsum. In one specimen from Frederick County, Maryland,
these spots vary from 0.3 to 0.8 mm. in diameter. Specimens from some
other localities have larger spots. This phase appears infrequently in
populations sampled so far.
Grobman (1944) recognizes the Arkansas and Oklahoma popula-
tions as a separate subspecies, P. c. serratus. This form differs from
other red-backed salamanders by the presence of regular serrations at
each costal groove along the borders of the dorsal stripe. The unstriped
phase is rather rare in serratus populations, so most specimens can be
identified readily. Occasionally a specimen from outside the range of
serratus exhibits a slightly serrated dorsal stripe, but the serrations
are usually irregular.
Highton and Grobman (1956) point out that Georgia populations
of P. cinereus differ from others in having a greater number of trunk
vertebrae and describe them as a separate subspecies, P. c. polycentra-
tus. It was not known then that many other populations of P. cinereus
have a high vertebral count. Such populations occur in the Piedmont
and Coastal Plain of Virginia and North Carolina, the lower Coastal
Plain of Maryland, the eastern part of Long Island, and probably also
on Staten Island. The average number of trunk vertebrae in most of
these populations is less than in Georgia (table 5). The original


Vol. 8











TABLE 5. Geographic variation in the number of trunk vertebrae in Plethodon cinereus. (Refer to figure 27 for boundaries of the


19-vertebrae areas
Catskill Mts., New York
Appalachian Plateaus
Missouri
Arkansas
20-vertebrae areas
east of Mississippi River
Arkansas and Oklahoma
21 to 22-vertebrae areas
eastern Long Island, New York
Staten Island, New York
eastern Maryland, Virginia,
and North Carolina
Georgia


regions .
Number of trunk vertebrae
17/18 18 18/19 19 19/20 20 20/21 21 21/22 22 22/23 23 23/24 Mean H
0
?5


1 56 6 24
96 28 1389 47 324 4
61 5 41 1
14 5


2 17 12 3056 612 17416 604 3692 20 32
14 2 122 4 5


1 48 11 124 1


1 1 1 353 48 707 10 92 2
33 3 36


19.30
19.13
19.45
19.26 5

1 1 20.03 0
19.95

20.72
20.86

4 20.77
2 21.56







BULLETIN OF THE FLORIDA STATE MUSEUM


description of P. c. polycentratus notes that considerable red pigment
is present on the legs and the sides of the head and body of unstriped
specimens. These red spots are usually lacking on unstriped speci-
mens from all parts of the range except Arkansas and Oklahoma. The
basic diagnostic character of P. c. polycentratus-number of vertebrae
-therefore fails to separate the Georgia populations from those of
several other areas. Another characteristic present only in the un-
striped color phase-red spots-can be used to distinguish Georgia
populations from all others except those in Oklahoma and Arkansas.
The recognition of polycentratus as a valid subspecies is therefore
open to serious question, but it is retained here tentatively as there
is no question that the Georgia populations have differentiated signifi-
cantly from the nearest cinereus populations in North Carolina and
Tennessee.
Red pigment in the unstriped phase of cinereus from two isolated
populations southwest and southeast of the main body of the range is
of interest as this character may likely be a primitive one. Red chroma-
tophores occur regularly in the unstriped phase of P. dorsalis.
The races polycentratus and serratus may be relict populations at the
periphery of the range of a successful species (see Brown, 1957, for
a discussion of centrifugal speciation), though they differ from each
other in number of trunk vertebrae and dorsal stripe shape. Prob-
ably a long period of isolation has permitted their differentiation
from the parent stock.

GEOGRAPHIC VARIATION IN COLOR PHASE FREQUENCIES. Thurow analyzes
gSt^aupbk^u^^k^^^a;talcan toL tk ped and unstriged color
phases and summarizes the results in a distribution map (MSb:217).
His data consist mostly of museum specimens and literature reports.
Many published records are unreliable because P. richmondi is often
wrongly identified as unstriped P. cinereus. Grouping data from large
areas prevented Thurow from evaluating variation among popula-
tions within each region. My field studies on color phase frequencies
in the Chesapeake Bay area show much variation within single
physiographic regions. One extreme example illustrates the type of
microgeographic variation that may occur: of 81 specimens from a
single wood lot near Skipton, Talbot County, Maryland, 80 (99%)
were striped. Only 5 of 49 specimens (10%) collected 14 miles away
(2.3 miles north of Trappe, Talbot County, Maryland) were striped.
The average for the entire Chesapeake Bay region is about 50 percent
of each color phase. About half of approximately 100 series, each of


Vol. 6






HIGHTON: REVISION OF PLETHODON


More than 25 specimens, differ significantly from a 50 : 50 ratio, rang-
ing from 0 to 100 percent striped individuals. Obviously the averaging
of data from museum material and literature reports for several locali-
ties within a physiographi regin n t i et euatey. t6l e + MoV r
phase frequencies in that region.
Series in collections indicate that the unstriped phase is scarce in
parts of Canada, New England, eastern New York, Georgia, Arkansas,
and Oklahoma. Large samples of P. cinereus from many localities
consist entirely of striped specimens. Test (1952) reports this for upper
Michigan, Bogert (1952) for southwestern Virginia. Except for those
few noted, all available individuals from the following areas are
striped: Missouri (115), Minnesota (16), Wisconsin (65), western
North Carolina (438), Tennessee (55), West Virginia except for the
extreme eastern and northern parts (1366), western Maryland (93),
western Pennsylvania except the southwest section from. Lawrence
Butler and Westmoreland Counties southward (505), eastern and
southern Ohio (259), southwestern New York (460). The only
unstriped individuals from these areas are from Yancey County, North
Carolina (AMNH 2212, which may be a young richmondi, but is too
young to be identified with certainty), Trumbull County, Ohio
(USNM 57023, a J. Hurter specimen); and Allegany State Park, Cat-
taraugus County, New York (CU 3515 and CNHM 91807). Also 6 of
19 specimens from the French Broad River, North Carolina, (ANSP
1167-86) are unstriped, but the data are questionable. So few speci-
mens are available from some parts of these regions that large series
from many more localities are needed to confirm the rarity of the
unstriped phase the above data suggest (fig. 27).
The unstriped phase is known to be abundant in only a few scat-
tered areas. From 1% miles southwest of Hutsonville, Crawford
County, Illinois (INHS 5553-73, and others), 32 of 33 specimens are
unstriped. From South Bass Island in Lake Erie, Ottawa County, Ohio
(UMMZ 95998), all of 40 are unstriped, as are 41 of 44 specimens
(UMMZ 95994, 98745) from nearby Catawba Island. The unstriped
phase is also common in eastern Long Island (Weber, 1928:108, and
unpublished data of Ernest E. Williams and Donald Cooper) and
in parts of the southern Coastal Plain of Maryland (Prince Georges,
Anne Arundel, Charles, and St. Marys Counties). High frequencies of
unstriped specimens also occur near Ottawa (Carleton and Russell
Counties, Ontario), and Montreal (Montagres County, Quebec)
(Patch, 1918; Paul S. Martin, personal communication).
In some areas the frequency of the striped phase is less than 5 per-


I., 1962








294 BULLETIN OF THE FLORIDA STATE MUSEUM


D
C
MODAL NUMBER OF
TRUNK VERTEaRAE
\ 19 .. .. .'
/ " ;21 or 22

FIGURE 27. Geographic variation in P. cinereus. Darkened area represents re-
gions where species has: (A) red pigment in unstriped specimens, (B) a ser-
rated dorsal stripe, (C) absence of the unstriped color phase, (D) variation in
the number of trunk vertebrae.

cent. By some current taxonomic procedures these could be recog-
nized as racially different from populations more than 99 percent
striped. Clearly variation of this type cannot be described best
by naming geographic races, for by this interpretation more than half
the range of P. cinereus becomes a zone of intergradation.
Both color phases occur in most of the range of the red-backed
salamander, with the striped phase usually the more plentiful. But
the local variation in some areas is so great that the biological
significance of the species' pattern dimorphism should be investi-
gated. Certain geographic correlations between color phase frequen-
cies and trunk vertebrae variation will be pointed out later.
Test (1952) suggests that the dark phase is of more recent origin


Vol. 6







HIGHTON: REVISION OF PLETHODON


than the striped phase. He bases this on the frequencies of the two
patterns in Michigan populations. It seems unlikely that the unstriped
phase is a recent evolutionary development as it occurs in so many
species of Plethodon. Moreover, some populations in the center of the
range of cinereus lack the dark phase, while others on the periphery
do not. The unstriped phase seems poorly adapted to some areas,
as in the Allegheny Plateau region, although it occurs on both sides
of the area.
The closely related species, P. richmondi, occupies a large part
of the area from southwestern New York southward through Pennsyl-
vania and western Maryland and eastern Ohio into West Virginia
where the unstriped phase of P. cinereus is apparently almost absent.
P. richmondi is known only in the unstriped phase. Probably some in-
teraction between the two species in this area of sympatry results
in increased pattern divergence. Whether color pattern is an isolating
mechanism cannot be determined at present, as nothing is known of
its role in these species' reproductive biology. Character displacement
(Brown and Wilson, 1956) in a zone of geographic overlap between
two closely related species is often due to reinforcement of species
barriers to prevent gametic wastage through hybridization. In many
areas where richmondi and cinereus are sympatric they are ecologically
segregated; where the two occur together in the same habitat, the
rarity or absence of the unstriped phase of cinereus is striking. This
relationship greatly needs quantitative study in the area of species
overlap.

GEOGRAPHIC VARIATION IN NUMBER OF TRUNK VERTEBRAE. In my recent
study of the heritability of the number of trunk vertebrae in P.
cinereus (Highton, 1960a), I induced gravid females from different
regions to deposit eggs in the laboratory and incubated the embryos
under uniform conditions. A significant correlation between the num-
ber of vertebrae in the parental populations and the offspring demon-
strated that much of the geographic variation is due to genetic differ-
ences. Other data obtained from field-collected broods from Giles
County, Virginia, indicate a correlation between the number of.trunk
vertebrae in females and their offspring. An estimate of the percent-
age of the variance due to heredity in these parent-offspring correla-
tions is 61, indicating that more than half the variation in number of
trunk vertebrae within a single population is due to genetic factors.
Variation in the number of trunk vertebrae in cinereus is summarized
in table 5 and figure 27. There is rarely a variation of more than three
vertebrae within any one population, and the class with the medial


1962








TABLE 6. Mean number of trunk vertebrate per county' for all counties with eight or more P. cinereus.b

19-vertebrae areas 20-vertebrae areas 21-vertebrae areas
Mean number of Appalachian Catskill Arkansas & Remainder Eastern Georgia
trunk vertebrae Missouri Plateaus Mts., N.Y. Oklahoma of range Coastal




21.1
21.2 1
21.3
21.4 1
21.5 1
21.6 1

Counties are used rather than samples from single localities in
order to reduce geographic bias. If means of samples from
all localities were used, some geographic areas would be
represented by many series, others by few. 0
b Data for 9 counties were divided into two groups because of a
significant difference in vertebral counts. Those counties
counted twice are: northern and southern Calvert and St.
Marys Counties, Maryland; eastern and western Greene,
Suffolk, and Ulster Counties, New York; eastern and western
Westmoreland County, Pennasylvania; eastern and western Z
Monongalia, Pendleton, and Randolph Counties, West Vir-
ginia. Exceptional county means are listed in footnotes.
SSte. Genevieve County, Missouri
d Logan County, Arkansas.
Clark County, Illinois.
f Emmet and Oakland Counties, Michigan; Franklin County, Ohio;
Crawford and Fulton Counties, Pennsylvania; eastern Pen-
dleton County, West Virginia; Victoria County, Nova Scotia;
and Brome County, Quebec.
SMonmouth and Ocean Counties, New Jersey, Clark and Fairfax
Counties, Virginia.
h Crawford County, Illinois.
Currituck County, North Carolina.







BULLETIN OF THE FLORIDA STATE MUSEUM


number of vertebrae usually includes the majority of the sample.
Therefore the mean number of vertebrae is close to a whole number
in most populations. Grouping all 308 sample means together (table 6)
shows three modes, one just above 19 trunk vertebrae, another around
20, and another a little below 21. The 19 and 21 modes are smaller
than the mode at 20 because the red-backed salamanders with these
counts occupy smaller geographic areas, and a smaller number of
series is available. I know no reason why natural selection should
favor a population with a high frequency of one particular vertebral
number over a population with two equally frequent vertebral num-
bers. As numerous populations have a high frequency of specimens
with 19 trunk vertebrae while many others have specimens mostly
with 20, it is surprising that so few populations do exist that have
equal numbers of specimens with 19 and 20 vertebrae and a mean
around 19.5. The reason for the trimodality might be that three isolates
have diverged sufficiently to produce three geographic races each
with a mean number of trunk vertebrae close to the whole numbers
19, 20, and 21. This thesis is not supported by the fact that the modal
frequency of 19 probably evolved at least three times-in the Catskill
region of New York, in the Appalachian plateaus, and in southeastern
Missouri (tables 5 and 6 and fig. 27). The 21-vertebrae populations
along the eastern coast may have been continuously distributed at
a time when sea level was lower, perhaps during the glacial stages of
the Pleistocene, so there is no need to postulate a separate derivation
for each of the now isolated populations. One exception may be the
isolated Georgia race, which probably is derived independently from
nearby 20-vertebrae populations. It is difficult to account for the
former isolation of 21-vertebrae populations from adjacent ones with
20 vertebrae.
Note that the county means (table 6) are not normally distributed.
Many samples are large and a difference of as little as 0.2 of a verte-
bra is often statistically significant. Distribution of means is not ran-
dom since adjacent populations usually resemble one another in verte-
bral counts. Standard deviations of most of the large samples are
similar and range from 0.4 to 0.6. No evidence indicates that popula-
tions with means midway between 19 and 20 vertebrae show greater
variability than those from the centers of the 19-, 20-, or 21-vertebrae
areas. For example, at a locality near Fort Littleton, Fulton County,
Pennsylvania, 79 of a sample of 176 specimens have 19 trunk vertebrae,
86 have 20, 1 has 21, and 10 are asymmetrical (19 and 20). This
locality is near the boundary between areas occupied by 19- and


Vol. 6







HIGHTON: REVISION OF PLETHODON


20-vertebrae populations. Another intermediate sample of 87 speci-
mens from Reddish Knob, on the boundary between Pendleton
County, West Virginia, and Augusta County, Virginia, is similar in
that 41 have 19 vertebrae, 43 have 20, and 3 are asymmetrical (19
and 20). This series is intermediate between one of 129 specimens
with a mean of 20.41 vertebrae, from Natural Chimneys, Augusta
-County, Virginia (CM 31002-3), and another series of 521 specimens
from the Cheat Mountains of Randolph County, West Virginia, which
has a mean of 19.05 vertebrae. A study of microgeographic variation
in the region where the 20- and 21-vertebrae populations come in
contact in southern Maryland and Virginia is now in progress. The
three areas with a modal number of 19 trunk vertebrae are widely
separated by populations with a modal number of 20 vertebrae.
The mean number of trunk vertebrae in 110 specimens of cinereus
from southeastern Missouri is 19.45. There appears to be some varia-
tion within the area, however. A series of 18 specimens (INHS 8396)
from 6 miles southeast of Weingarten, Ste. Genevieve County, has a
mean of 19.72 trunk vertebrae. Seven specimens from Meramec State
Park, Franklin County (CWM 1306-11; 1443), have a mean of 19.75
trunk vertebrae. The remaining 85 specimens from Bollenger, Butler,
Dent, Iron, Jefferson, Oregon, Perry, Reynolds, St. Francois, St. Louis,
Washington, and Wayne Counties have a mean of 19.36 trunk verte-
brae. The Missouri populations od cinereus are geographicaly situated
between the dorsalis populations of southwestern Illinois and those of
southwestern Missouri (angusticlavius). Since their modal vertebral
number is 19, the same as that of both dorsalis and angusticlavius,
it might be suggested that they actually represent a population of
dorsalis rather than cinereus. The straight-edged dorsal stripe of the
southwestern Illinois dorsalis is very similar to that of cinereus, so this
diagnostic character is of no aid in identification of the Missouri mate-
rial. All of the southeastern Missouri specimens are of the red-backed
phase. There remain only two other good diagnostic characteristics that
may be used to differentiate preserved dorsalis and cinereus and these
are present only in mature males. Only 28 of the available specimens
from Missouri are mature males, and all of these possess the type
of mental gland and modified premaxillary teeth typical of cinereus.
On this basis, all of the southeastern Missouri specimens are tentatively
assigned to cinereus. However, the possibility exists that some popu-
lations from which no males are presently represented in collections
are actually dorsalis. Or, dorsalis may be present in the area, but,
may not have been collected yet. A thorough field study of the eastern






BULLETIN OF THE FLORIDA STATE MUSEUM


small plethodons of southeastern Missouri is greatly needed. Habitat
information and belly pigmentation data on living specimens would
provide valuable additional information relative to the taxonomic allo-
cation of these Missouri populations. The possibility that the eastern
small plethodons of this area represent a primitive relict population
should also be considered, for such a population might possess some
of the characters of both species. The same might also be true of a
population that arose through hybridization of the two species and has
since been stabilized with some of the characters of one species and
other characters of the other species. Probably both species ranged
throughout the entire area during a more favorable climatic period,
judging from their present distributions.
In Garland, Yell, and Logan Counties, Arkansas, the mean of 19
specimens of serratus is 19.26, while the mean of 95 specimens from
Montgomery and Polk Counties is 19.91. The mean of 49 specimens
from LeFlore and McCurtain Counties, Oklahoma, is 19.99. The speci-
men reported by Bragg (1955:28) from Cherokee County, Oklahoma
(MUOZ 30678) has 20 costal grooves (probably 21 trunk vertebrae).
Presently available material does not show a good correlation be-
tween increasing altitude and a reduction in the number of trunk
vertebrae. Almost all samples from the Allegheny Mountains section
of the Appalachian Plateaus Physiographic Province have a modal
number of 19 trunk vertebrae, but populations from still higher alti-
tudes in the Blue Ridge Province of Virginia, North Carolina, and
Tennessee have a modal number of 20. In the Appalachian Plateaus
Province of northeastern Pennsylvania the modal number is 20,. as
it is in the glaciated section of the Allegheny Plateau of New York.
In the Catskill Mountains, however, it is 19.
Barden (MS) reports on a study of the zone of contact between
the high-altitude populations of cinereus in the Catskill Mountains that
have a modal number of 19 vertebrae and consist of almost 100 percent
striped phase, and the lower altitude populations with a modal num-
ber of 20 and both color phases. He finds a narrow zone of intergrada-
tion and fair geographic concordance between change in number of
vertebrae and color phase frequencies in the east, but not in the south
and west. The type of forest vegetation, itself correlated with latitude
and altitude, seems to be the most important factor influencing the
distribution of this zone. Barden discounts the effects of selection or
ecophenotypic variation in influencing the variation in cinereus popu-
lations of southern New York and attributes the variation to a double
invasion and secondary contact of two differentiated populations that


Vol. 6







HIGHTON: REVISION OF PLETHODON


were isolated during the Pleistocene. As there are no nearby popula-
tions with a low vertebral count, it seems likely that selection has pro-
duced the low-count populations at high elevations in the Catskill
Mountains. Whether selection is responsible for variation in color
phase frequencies is not yet clear.
There are few available series from the glaciated section of the
Allegheny Plateau in Ohio, Pennsylvania, and southwestern New
York. A total of 77 specimens from Chautaugua, Erie, Wyoming,
Livingston, Ontario, and Steuben Counties, New York have an average
of 19.2 trunk vertebrae, while 21 specimens from Ashtabula, Trumbull,
and Mahoning Counties, Ohio have a mean of 19.4 trunk vertebrae.
The modal number in the minger YLaes Beg\on of hew %ork is ,
so apparently a rather abrupt change in average number of trunk
vertebrae occurs just west of the Finger Lakes. Variation through the
remainder of northeastern Ohio is not determined because of the
paucity of material.
Populations of red-backed salamanders with a modal number of 21
trunk vertebrae along the Atlantic Coast from Long Island to North
Carolina are now discontinuously distributed. They occur on the
eastern half of Long Island (187 specimens, mean 20.7) and Staten
Island (7 specimens, mean 20.9), but not in the intervening area.
Specimens from western Long Island (316) have a modal number of
20 vertebrae (mean, 20.1). Several series from four coastal counties
of New Jersey (Monmouth, Ocean, Burlington, and Cape May) have
a higher proportion of individuals with 21 trunk vertebrae than other
New Jersey material. In the southern Del-Mar-Va Peninsula, in the
southern Coastal Plain of Maryland, the Coastal Plain and eastern
Piedmont of Virginia, and parts of the Coastal Plain and Piedmont of
North Carolina, the modal number is 21.
The species P. cinereus and P. richmondi exhibit character dis-
placement in color pattern, but a comparison of the number of trunk
vertebrae of the two species does not indicate a consistent increase in
divergence in areas of sympatry (figs. 27 and 30).
The usual taxonomic treatment of geographic variation of the type
found in the number of trunk vertebrae in P. cinereus is to separate
the populations with different modal numbers of trunk vertebrae into
named subspecies. However, the 19-vertebrae populations do not occur
in a single continuous geographic area, but in three (or four if the
Arkansas populations are included) rather widely separated regions.
They do not differ consistently from one another in any other known
character and therefore have to be considered as one polytopic sub-







302 BULLETIN OF THE FLORIDA STATE MUSEUM


species. Of 2108 specimens from the three areas, 1703 (81%) have
fewer than 20 trunk vertebrae; from the regions where the modal
number of vertebrae is 20, 21,897 (86%) of 25,612 specimens have 20
or more trunk vertebrae. Separation would be even greater if speci-
mens from near the boundaries of these areas could be called
"intergrades" and omitted. No evidence indicates that populations of
cinereus with a modal number of 19 trunk vertebrae represent a uni-
form race, but rather that the phenotype has evolved independently in
more than one region. The estimate of heritability of variation in
trunk vertebrae (Highton, 1960a) indicates that only 61 percent of
the variation within one population is due to heredity. If the remainder
is due to geographically variable environmental factors, then part of
the geographic divergence may be ecophenotypic rather than genetic.
Moreover, geographic variation in color phase frequencies is not con-
cordant with that of trunk vertebrae. Though all three areas in which
the modal number of trunk vertebrae is 19 are included in regions
where the frequency of the striped phase is over 99 percent, the
frequency of the striped phase is often equally high in adjoining areas
where the modal number of trunk vertebrae is 20. Hence the factors
that select for the high frequency of the striped phase must not have
the same distribution as those that select for the lower number of
vertebrae.
On Long Island, where abrupt change in the color phase frequencies
occurs, there is concordance in the variation in vertebral number and
color phase frequency. Of 12 specimens from Queens County, 11 (93%)
are striped (mean number of trunk vertebrae, 20.4); of 125 from
Nassau County, 102 (82%) are striped (mean number of vertebrae,
20.1); and of 178 from western Suffolk County, 167 (97%) are striped
(mean number of vertebrae, 20.0). In the eastern part of Suffolk
County only 4 of 199 (2%) are striped (mean number of vertebrae
20.7). A sample of 17 (DC 480-496) from Gardiners Island off the
eastern end of Long Island is 100 percent striped and has a mean
number of trunk vertebrae of 19.6. Thus the concordance between the
unstriped populations with 21 trunk vertebrae and the striped popu-
lations with 20 trunk vertebrae on Long Island is nearly perfect.
In other areas where the frequency of the unstriped phase is high,
as in Crawford County, Illinois, and Catawba Island, Ottawa County,
Ohio, the average number of trunk vertebrae is also high (20.6 and
20.5 respectively), but on South Bass Island, Ottawa County, Ohio, the
mean number of trunk vertebrae is only 19.7. Though many popula-
tions in Maryland and Virginia have a high frequency of the unstriped


Vol. 6








HIGHTON: REVISION OF PLETHODON


phase, these do not occur with any consistently higher frequency in
the area where the modal number of trunk vertebrae is 21 than in
adjacent areas where the modal number is 20. Canadian samples with
a high proportion of dark phase specimens have about the same mean
vertebral counts as adjacent mixed populations. Therefore geographic
variation exists in the correlation between high frequency of the
unstriped phase and presence of a modal number of 21 trunk vertebrae.
One area shows a high correlation (Long Island), while in others
(Maryland, Virginia, Quebec, and Ontario) there is little or no
correlation.

OTHER GEOGRAPHICALLY VARIABLE CHARACTERS. A series of 80 specimens
from Crawford County, Illinois, includes 12 individuals with a type
of belly pigmentation not seen in any other P. cinereus. This series
differs from other Illinois cinereus in being 97 percent unstriped and
in having a high average vertebral count. Instead of the usual ventral
white mottling of cinereus, these 12 specimens have a medial pigment-
free area on the belly. As the belly is unmottled, I thought at one time
they might represent an aberrant isolated population of P. richmondi,
and Conant's (1958) distribution map plots them as a questionable
record of this species. They are probably properly assigned to cinereus,
however, for one red-backed individual has been found, and the other
18 specimens have typical cinereus-type mottled bellies.
In the Blue Ridge Province of Virginia and Maryland and to some
extent in adjacent areas (Bogert, 1952), the usual brassy iridophore
flecking present on the unstriped phase of P. cinereus is highly
developed. This led Grobman (1949) to regard these populations as
related to P. richmondi and he described them as a full species, P.
huldae. Rabb (1955) and Muchmore (1955) present evidence that
P. huldae is based on unstriped specimens of cinereus. Rabb found both
color phases in the offspring of a striped female. These brassy flecked
P. cinereus resemble richmondi dorsally and can be distinguished from
that species only by the lower average number of trunk vertebrae and
by the greater amount of yellow iridophore mottling on the belly and
chin.
Living specimens from Graham, Swain, Haywood, Macon, and
Jackson Counties of western North Carolina have red pigment on the
sides and belly in front of the forelimbs. This pigment fades in pre-
servatives, so the geographic extent of its occurrence is not yet known.
Specimens from New Kent County, Virginia, and a few in
adjacent areas have a median dark band within the dorsal stripe. Some


.1962






BULLETIN OF THE FLORIDA STATE MUSEUM


eastern North Carolina specimens have exceedingly wide stripes.
Sanders and Smith (1949) report a specimen of P. cinereus from
Fern Lake, near Nacogdoches, Nacogdoches County, Texas: This
specimen (OS 556) has 19 trunk vertebrae and lacks the serrated
edges of the dorsal stripe present in the nearest populations of P. c.
serratus. Mr. Sanders writes me that he collected the specimen by dip
net from fruiting sphagnum moss near the lake shore, an unusual
habitat for this terrestrial species. This record may be the result of an
accidental introduction by man and confirmation by additional speci-
mens is necessary before Texas can be included in the natural range
of the species.
All eastern small plethodons I examined from southwestern Illinois
are clearly referable to P. dorsalis except one (UIMNH 16706), an
unstriped female from Alto Pass, Union County, Illinois, collected by
A. Cahn in 1928. Thurow (1957b:96) reports two other specimens
from the same locality (UIMNH 16707-8). Additional material from
southwestern Illinois is needed to confirm these records before
including this area in the range of the red-backed salamander.
Bleakney (1958:15) cites unconfirmed reports of local residents
that P. cinereus occurs in the Lake Melville area of Labrador. Since
environmental conditions there are similar to those where cinereus
occurs farther south, further search is desirable. Four unstriped, speci-
mens from "Hudson Bay Territory" sent to the National Museum in
1861 (-USNM 5963) represent the only record from this northern area.
Grobman (1944:281) points out that cinereus must have extended its
range rather rapidly to penetrate so far north following the retreat of
the last Pleistocene glaciers. No records from Newfoundland or Anti-
costi Island are known.
Most of the literature records of P, cinereus from Kentucky are
probably based on P. richmondi. The only Kentucky specimens of
cinereus examined are from Boone and Kenton Counties in the extreme
northern part of the state. There are no cinereus records from western
West Virginia; the specimen Grobman (1944:299) records from Mingo
County (CM 15998) is a richmondi. P. cinereus should occur in the
mountains of northeastern Georgia and northwestern South Carolina,
but it has not yet been found there.

Plethodon richmondi richmondi Netting and Mittleman
Ravine Salamander
Plethodon richmondi Netting and Mittleman (1938:288). Bishop (1943:272-5);
Grobman (1944:312); Duellman (1954:40-5).


Vol. 6







HIGHTON: REVISION OF PLETHODON


Plethodon richmondi richmondi, Highton and Grobman (1956:187-8).
Plethodon richmondi popei Highton and Grobman (1956:187-8). Type locality:
Comers Rock, Grayson-Wythe County line, Virginia.

HOLOTYPE. CM 14189, adult male, collected in Ritter Park, Hunting-
ton, Cabell County, West Virginia, elevation 600-700 feet, by Neil D.
Richmond and N. Bayard Green, 15 October 1938.
DIAGNOSIS. A dark-bellied eastern small plethodon, usually with 21
or more trunk vertebrae, and lacking the striped color phase.
DISTRIUTION. From Centre County, Pennsylvania, south through
western Maryland, West Virginia, western Virginia, northwestern
North Carolina, and northeastern Tennessee, west to Ohio, southeastern
Indiana, and eastern Kentucky (fig. 28).


FIGURE 28. Distribution of P. richmondi in the eastern United States.

DESCRIPTION. The dorsum is covered with small brassy flecks and
with larger white spots (0.1-0.3 mm. diameter). Lateral white or yellow
iridophore spots are similar in color to those on the belly and chin. The
iris has brassy flecking above and below the pupil. Red chromatophores
are often present anteriorly on the sides and head of adults of eastern
populations, but have not been noted in living material from the
western part of the range. Red pigment is present on the dorsum also
in young individuals.


1962










BULLETIN OF THE FLORIDA STATE MUSEUM


R r. richmondi
Loroin County, Ohio


S1 20 3OUT T L H r0 0 70
SNOUT-VENT LENGTH (mm)


Po Rr.nettingi


5

S10 20 S0 40 LE N 70 ( 0
SNOUT-VENT LENGTH (m.l


Prrid.mond,
C el Cunty, West Virgnia


I.-




0,-- n-- 0- 0 50 0 ?0 60
SNOUT-VENT LENGTH (mm.)

FIGURE 29. Variation in the number of
of P. richmondi.


R r. richmondi
0 Bedford County, Pennsylvania

25-





5-




o o o20 50 40 50 O 70 80
SNOUT-VENT LENGTH (mm.)

35

3o P. r. hubrichti

25










0 10 20 10 40 ,0 60 7`0 o
SNOUT-VENT LENGTH (mm)

35
P r. rlchmondl
3 Graysmon a Wythe Counties, Virginia
25.











rO -L 20 A 0 t S 6L0 7O .80
SNOUT-VENT LENGTH (mm.)

vomerine teeth in 6 geographic samples


The modal costal groove count varies geographically from 19 to
22, and the modal number of trunk vertebrae from 20 to 23. Vomerine
teeth range from 6 to 16 (fig. 29). The largest specimen examined,
from Alleghany County, Virginia, is 60 mm. in snout-vent length, but


Vol. 6







HIGHTON: REVISION OF PLETHODON


Duellman (1954) records a 72 mm. specimen from Ohio. Sexual matur-
ity is reached at about 40 mm. snout-vent length.

Plethodon richmondi hubrichti Thurow
Peaks of Otter Salamander
Plethodon netting, Thurow (1955:102-3).
Plethodon hubrichti Thurow (1957a:59).
HOLOTYPE. USNM 139087, adult male, collected at mile 80.9 Blue
Ridge Parkway, Bedford County, Virginia, at about 3100 feet, by
Gordon R. Thurow, 9 December 1956.
DIAGNOSIS. A race of P. richmondi with a modal number of 20
trunk vertebrae, and with more dorsal brassy flecking in the adult
than in other populations of richmondi.
DISTIBUrION. Known only from the type locality and from Green-
lee, southern Rockbridge County, Virginia (fig. 28).
DESCRIPTION. More than half the adults and subadults have so
much brassy spotting on the dorsum in life that it forms an almost
uniform dorsal stripe. In others the brassy areas are separated into
spots, though many fuse with adjacent spots. Dorsal brassy flecking
is less marked in juveniles, which resemble adults of other richmondi
populations. Newly hatched young have a reddish dorsal stripe. The
amount of white spotting on the belly and chin varies considerably
and is sometimes absent. Within the brassy patches are small (0.1-0.3
mm.) white spots. Larger white spots (0.2-0.9 mm. in diameter) are
present dorsolaterally. Brassy flecks pattern the iris.
The modal costal groove count is 19, and the number of trunk
vertebrae usually 20. Vomerine teeth range from 6 to 19 (fig. 29). The
largest specimen examined is 55 mm. in snout-vent length. Sexual
maturity is usually reached between 40 and 45 mm. snout-vent length.

Plethodon richmondi nettingi Green
Cheat Mountain Salamander
Plethodon nettingi Green (1938:295-9). Bishop (1943:266-9); Grobman (1944:
313); Brooks 1948:239-44); Thurow (1957a:65).
Plethodon richmondi nettingi, Highton and Grobman (1956:187).
HOLOTYPE. CM 10279, adult male, collected on Barton Knob, near
Cheat Bridge, Randolph County, West Virginia, elevation about
4000 feet, by M. Graham Netting, 29 June 1935.
DIAGNOSIS. A race of P. richmondi that has a modal number of
19 trunk vertebrae.


1962







308 BULLETIN OF THE FLORIDA STATE MUSEUM Vol. 6

DISTRIBUTION. Known from altitudes above 3500 feet in the Cheat
Mountains of Randolph and Pocahontas Counties, West Virginia
(fig. 28).
DESCRIPTION. Coloration in life is similar to P. r. richmondi. Costal
grooves usually number 18, trunk vertebrae 19. Vomerine teeth range
from 6 to 17 (fig. 29). The largest specimen examined is 46 mm. in
snout-vent length. Sexual maturity is reached at about 35 mm. snout-
vent length.


Netting and Mittleman (1988:288) and Green (1938:295) respec-
tively described P. richmondi and P. nettingi. Highton and Grobman
(1956:187) named a southern race, P. richmondi popei, and recog-
nized netting as a subspecies of P. richmondi. Thurow (1957a:59) sub-
sequently described another member of this group, hubrichti, as a full
species; it is here considered as a subspecies of P. richmondi. All P.
richmondi have brassy dorsal flecking and are characterized by the
lack of a striped phase except in the young. Most have dark bellies.
These characters show much variation in other eastern small plethodons
and they are not diagnostic for P. richmondi. Variation in the closely
related P. cinereus is so great it is impossible to identify all specimens
of cinereus and richmondi without aid of locality data. Where they are
sympatric the two species usually differ in color pattern and number
of trunk vertebrae. Thus though they cannot always be identified by
morphological criteria, they are certainly distinct species. (See dis-
cussion of character displacement under P. cinereus.)

GEOGRAPHIC VARIATION IN NUMBER OF TRUNK VERTEBRAE. The number
of trunk segments is the most easily analyzed geographically variable
character of P. richmondi. Extremes of variation are found at localities
only 100 miles apart in West Virginia. At Bickles Knob in the Cheat
Mountains of Randolph County, the mean number of body vertebrae
is 18.8 in a sample of 421 P. r. nettingi, while 7 miles north-northeast
of Sandyville, Jackson County, the mean number is 23.6 in a sample of
9 P. r. richmondi, an average difference of almost five vertebrae.
Extremes of variation at the two localities are 18 to 20 and 23 to 24.
Table 7 summarizes geographic variation in samples from different
parts of the range of the species.
Populations from the Cheat Mountains of West Virginia (P. r.
nettingi) have the lowest number of trunk vertebrae. The modal count
of all samples of this form is 19. Leslie Hubricht discovered a high





TABLE 7. Summary of geographic variation in number of trunk vertebrae in Plethodon richmondi.


Geographic region


18 18/19 19


P. r. nettingi
Cheat Mountains, W. Virginia 118
P. r. hubrichti
Bedford County, Virginia 1
P. r. richmondi
Smyth Co. slope of Whitetop
Mountain, Virginia
Southwestern Virginia, north-
western North Carolina, north-
eastern Tennessee, southeastern
Kentucky, and southern West
Virginia
Paddy and Reddish Knobs, West
Virginia-Virginia state line
Eastern Ohio
Western Ohio, southeastern
Indiana, northeastern Kentucky,
and vicinity of Huntington,
West Virginia
Ridge and Valley Province of
Pennsylvania, Maryland, West
Virginia, and Virginia*
Appalachian Plateau of western
Pennsylvania
Appalachian Plateau of West
Virginia


Number of trunk vertebrae
19/20 20 20/21 21 21/22 22 22/23 23 23/24 24


30 537 11 23


2 53 18 244 1 8


218 23 581 18 91


2 1
1 9 1


Modal---


Modal
Mean number


18.85 19


19.82 20


20.07 20





20.87 21
z


28 3 20.96 21 0
46 1 19 1 21.13 21

-S

103 20 428 11 65 3 3 21.95 23


19 12 316 4 42 1 22.05 22


31 3 14


1 22.26 22


31 7 52 3 8 22.75 23


* Includes 3 specimens from Mills Gap, Botetourt-Bedford County line, Virginia, in the Blue Ridge Physiographic Province.








310 BULLETIN OF THE FLORIDA STATE MUSEUM


altitude population with a modal number of 20 between the James
and the Roanoke Rivers in the Blue Ridge Mountains of Virginia.
Thurow (1955) reported this find and referred the specimens to P. r.
nettingi. Thurow (1957a) later described them as P. hubrichti.
All southern populations of P. richmondi-those in northwestern
North Carolina, northeastern Tennessee, southwestern Virginia, and
adjacent areas of Kentucky and West Virginia-have a modal number
of 21 trunk vertebrae with one exception, a population on the Smyth
County slope of Whitetop Mountain, Virginia, with a modal number
of 20. All other samples from Whitetop Mountain have a modal number
of 21. When describing salamanders from this region as P. r. popei
(Highton and Grobman, 1956), we were unaware that the modal
number of trunk vertebrae is also 21 in at least two other areas, in
eastern Ohio, and in the mountains along the Virginia-West Virginia
state line-Paddy Knob and Reddish Knob, west and southwest of
Harrisonburg, Virginia. The salamanders in each of these regions
resemble adjacent 22-vertebrae populations in belly pigmentation,
which may be interpreted as evidence against close relationship of
these disjunct populations with the same modal vertebral count. Hence
I no longer regard popei as a valid subspecies. To recognize the three
isolated 21-vertebrae populations as the same subspecies would be to
ignore differences in belly and chin pigmentation.
The remainder of the range of P. richmondi is occupied by popu-
lations with modal counts of 22 and 23 trunk vertebrae. The highest
numbers occur in the Appalachian Plateau Province of West Virginia
(fig. 30).
Data on vertebral variation in P. richmondi may suffer from
sampling bias. From several regions we have many small samples, but
only one or two large ones. Thus the mean value for a region may be
unduly influenced by only a few samples. The extent of local variation
in many areas is therefore still poorly known.

GEOGRAPHIC VARIATION IN VENTRAL PIGMENTATION. Salamanders from
the Ridge and Valley Physiographic Province of Pennsylvania, Mary-
land, West Virginia, and Virginia show more white mottling on the
belly and chin than do other P. richmondi. This is in sharp contrast to
all adjacent populations of richmondi, which have the darkest bellies
and chins in the species. On belly and chin pigmentation a mixed
group of Pennsylvania specimens sorts easily into those from the Ridge
and Valley Province and those from the Appalachian Plateaus Prov-
ince with more than 95 percent accuracy. A similar separation is possi-
ble when material from southwestern Virginia (formerly P. r. popei)


Vol. 6








HIGHTON: REVISION OF PLETHODON


FIGURE 30. Geographic variation in P. richmondi. (Upper left) Hatched area
represents region where species has a light chin, (upper right) hatched area rep-
resents region where species has a dark belly, (lower left) stippled area represents
region where species occurs at high elevations and is active at the surface in the
summer, (lower right) geographic variation in the modal number of trunk verte-
brae.

is compared with P. r. richmondi from the Ridge and Valley Province
of Virginia. Mottled-bellied P. richmondi occur in other parts of the
range, but nowhere else is the chin so heavily mottled with white as
in the Ridge and Valley Province.
The number and size of the white ventral spots is noticeably
greater in Ohio and northern Kentucky populations than in the rest, ex-
cept for the Ridge and Valley populations which also show much
greater contrast between the amount of mottling on the belly and
chin. Ontogenetic and individual variation in the amount of belly
mottling is so great that it is difficult to measure its geographic varia-
tion quantitatively. Further evaluation is warranted.


1962


- F~








BULLETIN OF THE FLORIDA STATE MUSEUM


GEOGRAPHIC VARIATION IN HABITS AND HABITAT REQUIREMENTS. Brooks'
(1948) study of the ecological distribution of P. r. nettingi shows it
to be restricted to spruce forests at high elevations in the Cheat
Mountains. The lowest altitudinal record is 3550 feet. P. r. hubrichti
is also a high altitude form, the elevation at its type locally is 3160
feet. In the southern part of its range-southwestern Virginia, north-
western North Carolina, and northeastern Tennessee-most specimens
of P. r. richmondi have been collected at high elevations, although the
apparent abundance at high altitudes may be due merely to more
collecting in mountainous areas. Most records from other regions are
from ravine habitats at low elevations.
Few P. r. richmondi are recorded for the summer months, except
from the mountains in the southern part of the range. Apparently over
much of its range this form is not active on the surface in summer.
Near the town of Bedford Valley, Bedford County, Pennsylvania, 9
series of P. gluiinosus and P. richmondi were collected each month of
the year except during the hibernation period. Table 8 illustrates the
relative seasonal abundance of the two species at this locality. Although
glutinosus was abundant during the entire activity period, richmondi
was taken infrequently in summer.

TABLE 8. Relative seasonal abundance of P. glutinosus and P. richmondi at a
locality 2.8 miles south, 0.9 miles west of Bedford Valley,
Bedford County, Pennsylvania.
Date Number of Number of
P. glutinosus P. richmondi

27 April, 1957 44 2
25 May, 1958 81 22
14 June, 1958 59 3
19 June, 1959 4 1
8 July, 1959 18 1
3 August, 1958 68 4
1 September, 1959 65 3
21 September, 1958 151 32
25 October, 1958 32 11

P. r. nettingi is active during the summer. Little is known about
the habits of hubrichti. A clinal change in summer activity is apparent
'om tl' lig'-atit~d' e spruce-forest populations (P. r. nettingi) to the
low-altitude ravine-inhabiting northern P. r. richmondi. Southern P. r.
richmondi are more like nettingi in this respect than they are to
northern P. r. richmondi. This dine corresponds to the variation in the


Vol. 6




HIGHTON: REVISION OF PLETHODON


two geographically variable morphological characters. Number of
trunk vertebrae is lowest in P. r. netting and highest in some northern
P. r. richmondi. P. r. hubrichti and southern P. r. richmondi are inter-
mediate. P. r. nettingi, hubrichti, and southern P. r. richmondi have
dark bellies, while northern P. r. richmondi have mottling on the
belly.
It is clear that nettingi is most closely related to hubrichti, and that
hubrichti is related to the 21-vertebrae populations of P. r. richmondi
in the southern part of the range. All three have almost identical belly
pigmentation. The dorsal patterns of nettingi and southern P. r. rich-
mondi are similar, but hubrichti has abundant brassy flecking. To
recognize nettingi and hubrichti as full species because of their appar-
ent reproductive isolation from distantly related but sympatric popu-
lations of the same stock, would ignore the close morphological
similarity and actual morphological overlap in the only characters that
differentiate nettingi, hubrichti, and southern richmondi. P. r. nettingi
and hubrichti may be regarded as relict mountain populations sur-
rounded by more specialized derivatives of the same parental stock.
P. r. richmondi is considered more specialized because of its more
elongated body.
Of great interest would be a study of the ecological relationships
of P. richmondi in the Cheat Mountains of West Virginia and between
the James and Roanoke Rivers in the Blue Ridge Mountains where
the ends of the richmondi Rassenkreis overlap. Reproductive isolation
may be established, as no hybrids between P. r. richmondi and either
P. r. nettingi or P. r. hubrichti have been found. Records show that
P. r. richmondi occurs in adjacent areas to the east, west, and north
of the Cheat Mountains in West Virginia.
P. r. richmondi is recorded from Mills Gap, 2400 feet, on the crest
of the Blue Ridge, Bedford-Botetourt County line, Virginia, only 8
miles west of the type locality of P. r. hubrichti. Grobman (1949:135)
has recorded P. r. richmondi from Natural Bridge, Rockbridge County,
Virginia, 93 miles north of the type locality of P. r. hubrichti.
I collected eggs and newly hatched young of P. r. nettingi near
the headwaters of Condon Run on Bickle's Knob, Randolph County,
West Virginia, and incubated more than 100 embryos in the laboratory.
The dorsum has reduced melanophore pigment and sparsely dis-
tributed red chromatophores. These young P. r. nettingi differ from
newly hatched unstriped P. cinereus from Virginia, Maryland, and
New Jersey, for the latter lack red pigment in the stripe region. Young
P. r. richmondi from Fulton County, Pennsylvania, and Washington


1962







BULLETIN OF THE FLORIDA STATE MUSEUM


County, Maryland, also have red pigment on the dorsum in life. This
may be a retention during ontogeny of red chromatophores similar
to those found on newly hatched P. r. nettingi. Red pigment also
occurs in newly hatched P. r. hubrichti. Red is not present on adults
of the subspecies nettingi and hubrichti, but is frequently found on
the head and anterior sides of P. r. richmondi from Pennsylvania,
Virginia, and West Virginia. One specimen from Mingo County, West
Virginia, (WVBS 615) may have had some dorsal red spots.
The largest specimens of P. r. richmondi examined are from the
Ridge and Valley Province of Virginia. Populations with the smallest
average adult size are from southwestern Virginia and adjacent North
Carolina. P. r. nettingi is a small animal. Adult P. r. hubrichti are
heavy bodied, but short in snout-vent length. Geographic variation in
size may be similar to the clinal variation in number of body vertebrae,
belly pigmentation, and habits.

THE EASTERN LARGE PLETHODONS
This group, as defined by Grobman (1944:266) on the basis of its
larger size and fewer costal grooves, includes 6 species that inhabit the
eastern United States. Plethodon glutinosus is the most widely dis-
tributed, and all the other species occur within its range. P. yonahlossee
occurs in the southern section of the Blue Ridge Province north of the
French Broad River. P. ouachitac, superficially similar to yonahlossee,
lives in the Ouachita Mountains of Arkansas and Oklahoma. P. cad-
doensis is known from the Caddo Mountains and adjacent areas of
Arkansas. P. jordani occurs in the southern section of the Blue Ridge
Province and enters the adjacent Piedmont Province in North and
South Carolina and the Ridge and Valley Province in Virginia. The
range of P. wehrlei centers in the unglaciated Allegheny Plateau
section of the Appalachian Plateaus Physiographic Province. Two
recently described species, P. dixi and P. jacksoni, are local variants
of P. wehrlei and are here synonymized with it. Grobman (1944) has
analyzed the distributional patterns of the eastern large plethodons.
Much speculation exists on relationships within this section of the
genus. Dunn (1926:23) considers it the most primitive group, with
P. yonahlossee the most primitive form. Dunn as well as Grobman
(1944:276) think the relationship between yonahlossee and wehrlei
close. Hairston and Pope (1948:276-7) suggest that yonahlossee is
closely related to jordani and that the similarity between some jordani
and glutinosus may be due to parallel evolution. Others think the
closest relative of jordani is glutinosus. Bishop (1941b:19) considers


Vol. 6







HIGHTON: REVISION OF PLETHODON


one geographic race (P. j. shermani) a subspecies of glutinosus rather
than of jordani. Grobman (1944) thought that jordani and closely
related forms (then considered full species, but now known to be
conspecific with jordani) are different enough to warrant separation as
the metcalfi group, distinct from all other eastern large plethodons.
Characters studied in determining the morphological similarities
and relationships of these forms include size, number of vomerine teeth
number of costal grooves, degree of sexual dimorphism, and pigmen-
tation. Several characters were found extremely variable and not
diagnostic of any one form. For example, Bishop (1941b:18) and
Grobman (1944:287) think that those forms now considered Plethodon
jordani have fewer vomerine teeth than P. glutinosus. Comparison of
the number of vomerine teeth in large series of each species from
several geographic areas (including localities within the ranges of all
named subspecies) shows that for a given size, the range of variation
is similar, and that greater difference occurs within glutinosus than
between some populations of glutinosus and jordani (figs. 39 and 42).
As average adult size in some sympatric populations of glutinosus and
jordani differs considerably, the two forms often show a difference in
the average number of vomerine teeth. As adults of the same size have
a similar average number of vomerine teeth, this character is not
definitive. This similarity between the two species probably indicates
a close relationship, for the number of vomerine teeth of some of the
other eastern large plethodons differs markedly.
Figures 33, 35-37, 39, and 42 illustrate intra- and inter-species dif-
ferences in the ontogenetic variation in number of vomerine teeth in
eastern large plethodons. P. wehrlei has the lowest rate of increase in
number of vomerine teeth with increasing size, and in this closely
resembles some of the eastern small plethodons. The rate is greater
in jordani and glutinosus, and northern samples of glutinosus are
similar to jordani. The highest rates are found in yonahlossee and
ouachitae. In spite of being a small species, caddoensis has a high
number of vomerine teeth.
Some other supposed differences discussed in the literature are
useless in determining relationships of eastern large plethodons. Except
for wehrlei, all have a similar number of vertebrae (see table 9).
Sexual dimorphism in size is present in several forms that have been
studied critically, but as females in each average slightly larger than
males no species differences exist. The type of lateral pigmentation
varies somewhat within each species both individually and geographi-
cally, but consistent differences between species are rare. Each species








316 BULLETIN OF THE FLORIDA STATE MUSEUM Vol. 6

TABLE 9. Variation in the number of trunk vertebrae in the eastern
large plethodons.
Species Number of trunk vertebrae
16 16/17 17 17/18 18 18/19 19 Mean
Plethodon
wehrlei 16 144 1 5 17.94
yonahlossee 22 2 17.08
ouachitae 1 36 16.99
caddoensis 1 38 2 2 17.05
jordani 41 3 429 1 14 16.94
glutinosus 64 11 855 9 42 16.98

is geographically sympatric and often coexists in the same habitat
with one or more of the other species. Each of the sympatric pairs
appears to be genetically distinct, with the possible exception of glu-
tinosus and jordani, and much evidence indicates that different
species have different ecological requirements. Preserved specimens
that have lost some pigmentation are often extremely difficult to iden-
tify because of the morphological similarity among most of these sala-
manders. Indeed, identification by the field collector familiar with the
animals in life is often more valuable than painstaking study of a
preserved specimen. The importance of studying these animals in life
is readily apparent.



















A B
FIGURE 31. Right hind feet of two eastern large plethodons showing variation
in amount of webbing on toes. (A) P. wehrlei, (B) P. glutinosus.








HIGHTON: REVISION OF PLETHODON


P. wehrlei is distinguishable from the rest of the group by having
the most webbing on the toes (fig. 31), one more trunk vertebra on
the average, a greater development of melanophore pigment in the
parietal peritoneum, and the lowest average number of vomerine teeth.
P. yonahlosse, P. ouachitae, and P. caddoensis may be recognized by
their distinctive color patterns in life. They also differ from the other
species in having a larger average number of vomerine teeth. No
known character will separate all glutinosus and jordani consistently.
The chin of glutinosus is usually darker than that of jordani, but not
always. The morphological similarity among the forms indicates they
are closely related, with only wehrlei separated from the others by
several characteristics.
The young of yonahlossee, some wehrlei, and jordani from the
Great Smoky Mountains and the Nantahala Mountains have dorsal
red spots. These usually disappear in jordani, and in wchrlei too except
in the southern part of its range, and become incorporated into the
dorsal stripe of yonahlossee. No information is available on the very
young of ouachitae, caddoensis, and several geographic populations of
jordani, but young ouachitae may also have red spots (see Pope and
Pope, 1951:145). Red dorsal spots are absent in the hatchlings of P.
glutinosus in Florida (Highton, 1956). Very young specimens from
other parts of the range of glutinosus also lack red spots. Dunn (1926:
1.39) records a glutinosus from Clayton, Georgia, with tiny paired red
dorsal spots, but this may well be a jordani, known to occur within
10 miles of Clayton in the Nantahala Mountains. Cope (1889:141)
records young specimens of glutinosus with red spots from caves in
Montgomery County, Virginia, but these were probably wehrlei.
Melanophore pigmentation on the chin is less than that on the
belly in most species except P. glutinosus. Certain populations of
glutinosus from Texas and the Blue Ridge and Piedmont Provinces
south of the Potomac River also have light chins. Most species have
dark bellies, but northern populations of jordani and southern wehrlei
are light bellied.
P. yonahlossee, and P. glutinosus except for its southeastern Coastal
Plain populations, attain a larger maximum size than do the other
species. P. caddoensis is the smallest eastern large plethodon. Individ-
uals of some northern populations of jordani are also relatively small.
Red pigment occurs in adult yonahlossee, ouachitae, caddoensis,
southern wehrlei, and the populations of jordani from the Nantahala
Mountains and the Great Smoky Mountains. Dorsal iridophores occur
abundantly as white spots in most glutinosus, ouachitae, caddoensis,








BULLETIN OF THE FLORIDA STATE MUSEUM


and wehrlei. P. jordani from the Snowbird and adjacent mountains of
North Carolina have small white spots, while those from Oconee
County, South Carolina, have small brassy flecks abundantly scattered
on the dorsum. Lateral white iridophores usually are present in all
eastern large plethodons except jordani in the northern part of its
range and southern South Carolina glutinosus.
Some P. caddoensis and P. ouachitae have more toe webbing than
is usually present in the glutinosus and yonahlossee groups, thereby
resembling P. wehrlei. All members of the yonahlossee group average
more vomerine teeth than do members of the wehrlei and glutinosus
groups. They resemble the glutinosus group in having 17 trunk verte-
brae, compared to the 18 of wehrlei. It is clear that the glutinosus and
yonahlossee groups are more closely related to each other than either
is to wehrlei, and that glutinosus and wehrlei are the least closely
related.
In summary, P. wehrlei is the most distinct species of the eastern
large plethodons. The others are morphologically similar to each other.
The hypothetical ancestor might have been a moderate-sized animal
with a light chin, red dorsal spots, 17 trunk vertebrae, a short vomerine
series, and webbed toes. P. wehrlei is closest to this hypothetical
ancestor, although its body is slightly elongated by the addition of
an extra trunk vertebra. The chin is still light in all except some glutino-
sus and a few populations of jordani. Red dorsal spots remain only in
adult southern wehrlei and in the young of several other forms. Large
size is attained by yonahlossee and glutinosus, while dwarfing occurs
in caddoensis and northern jordani. Adult yonahlosee have a much
longer vomerine series than the others. The suggested phylogenetic
relationships are shown in figure 5.
WEHRLEI GROUP
Plethodon wehrlei Fowler and Dunn
Wehrle's Salamander
Plethodon wehrlei Fowler and Dunn (1917:23-4). Dunn (1926:133-6); Bishop
(1941a:232-9); (1943:281-4); Grobman (1944:285-7).
Plethodon dixi Pope and Fowler (1949:1-4). Type locality: Dixie Caverns,
Roanoke County, Virginia.
Plethodon jacksoni Newman (1954:9-14). Type locality: 1 mile east of Blacks-
burg, Montgomery County, Virginia.
Plethodon wehrlei wehrlei, Conant (1958:234).
Plethodon wehrlei dixi, Conant (1958:234).
HOLOTYPE. ANSP 19123, collected in the Two Lick Hills, Indiana
County, Pennsylvania, September 1911, by R. W. Wehrle.


Vol. 6







HIGHTON: REVISION OF PLETHODON


DIAGNOSIS. An eastern large plethodon with webbing between the
toes on the hind foot often extending to the joint between the first
and second phalanges (fig. 31), a modal number of 18 trunk vertebrae,
and a moderate amount of melanophore pigmentation in the parietal
peritoneum.
DISTRIBUTION. From Allegany State Park, Cattaraugus County,
New York, south in the Allegheny Plateaus section of the Appalachian
Plateaus Physiographic Province through western Pennsylvania and
West Virginia to southwestern Virginia, where it occurs in the Ridge
and Valley, Blue Ridge, and Piedmont Physiographic Provinces (fig.
32). This species also occurs in Monroe and Washington Counties,
Ohio (Walker, 1933).
DESCRIPTION. The ground color is usually lighter in wehrlei than in
the other eastern large plethodons; the belly has fewer melanophores.
Often large gaps exist between melanophores anteriorly on the belly
and chin; yellow iridophores occupy these gaps. Yellow pigment is


FIGURE 32. Distribution of P. wehrlei in the eastern United States.


1962








320 BULLETIN OF THE FLORIDA STATE MUSEUM Vol. 6

usually abundant laterally and the yellow often persists in preservatives,
unlike that in the other eastern large plethodons. Usually small white
iridophore spots occur on the dorsum. In the southern populations
numerous brassy flecks occur as well, especially among the red dorsal
spots. The red dorsal spots occupy gaps in the melanophore bed.
Brassy iridophores are usually present in the iris.
Costal grooves usually number 17, the trunk vertebrae, 18.
Vomerine teeth range from 6 to 20 (fig. 33). Moderate in size, the
largest specimen examined is 65 mm. in snout-vent length. Sexual
maturity is reached between 45 and 50 mm. snout-vent length.
0 A 0 *

Plethodon wehrlei inhabits the unglaciated Appalachian Plateaus
Physiographic Province in southwestern New York, western Pennsyl-
vania, extreme southeastern Ohio, West Virginia, and adjacent Virginia
(where it occurs a short distance outside the Appalachian Plateaus
Province). The most distinct of the eastern large plethodons, it appears
to occupy an isolated position, differing from the other species in
several respects. Webbing on the toes is usually extensive although
occasional specimens of other species, particularly caddoensis and
ouachitae, show a similar development. P. wehrlei is the only eastern
large plethodon that normally has 18 trunk vertebrae; the others
usually have 17. P. wehrlei has fewer vomerine teeth than any other
eastern large plethodon, and it has the greatest development of peri-
toneal melanophores.
Two populations from southwestern Virginia were described as
distinct species by Pope and Fowler (1949) and Newman (1954).
P. dixi differs from other southern wehrlei only in proportions and
pigmentation, and occurs only in Dixie Caverns and nearby Blanken-
ship Cave, Roanoke County, Virginia. Differences from adjacent
wehrlei populations are slight. Certainly P. dixi is conspecific with
wehrlei, and little is gained by retaining this extremely localized popu-
lation as a named geographic race. One juvenile from Dixie Caverns
(UF 8445) had dorsal red spots in life similar to those in other southern
wehrlei. P. jacksoni, named from a population less than 15 miles from
the dixi locality, differs from other southern wehrlei only in retaining
the juvenile red spots in the adult. Some adults in West Virginia also
retain their red spots (Brooks, 1945). Red dorsal spots are completely
absent in juvenile and adult wehrlei from New York and Pennsylvania
(Bishop, 1941a:238). Geographic differences in the occurrence of red
dorsal pigmentation in wehrlei are significant and clinal. Detailed data











HIGHTON: REVISION OF PLETHODON


on the nature of the dine are not available. I have examined living

specimens from Cattaraugus County, New York; Lewis, Randolph, and

Pendleton Counties, West Virginia; and Alleghany, Roanoke, and

Montgomery Counties, Virginia. Populations from each of these locali-

ties show striking differences in pigmentation and local variation

is marked.


4 P wehrlel
3Centre County, Pennsylvania


10 20 30 40 50 60 70 80
SNOUT-VENT LENGTH (mm.)


R wehrlei
0 Cottorougus County, New York

5 -

:0 -



50

5-


0 10 20 30 40 50 60 70 80
SNOUT-VENT LENGTH (mm)


P wehrlei
SRandolph County, West Virginia



!0




5-
C,




0 10 20 30 40 50 60 70 80
SNOUT-VENT LENGTH (mm.)


35
R wehrlei
Montgomery County, Virginia

25!


5-*


0 10 20 30 40 50 60 70 60
SNOUT-VENT LENGTH (mm.)

FIGURE 33. Variation in the number
of P. wehrlei.


P wehrlei
Roanoke County, Virginia


0 10 20 30 40 50 60 70 90
SNOUT-VENT LENGTH (mm.)
of vomerine teeth in 5 geographic samples


1962











BULLETIN OF THE FLORIDA STATE MUSEUM


Netting (1936:91) compares West Virginia P. wehrlei with
topotypic Pennsylvania specimens. The former have white spotting on
the throat and chest which the latter lack. Lateral white pigment is
also more abundant in West Virginia material. Grobman (1944:287)
suggests that southern wehrlei may grow larger than northern ones.
Two erroneous literature records of wehrlei are based on specimens
of glutinosus. One (CNHM 57029) is from the Blue Ridge Province of
Virginia (Grobman, 1949:136), the other (CU 4132) is from Mercer
County, Pennsylvania (Lachner, 1942).
A specimen (CNHM 60010) from White Top Mountain, Virginia,
typical of southern wehrlei, has 18 costal grooves, webbed toes, dorsal
red spots, pigmented parietal peritoneum, and reduced melanophore
pigmentation anteriorly on the belly. It is the southernmost record of
wehrlei and is the only record of the sympatric occurrence of this
species with yonahlossee. Newman (1954:13) refers a specimen
(USNM 143219) from Smith Mountain Gorge, Pittsylvania County,
Virginia, in the Piedmont Physiographic Province, to P. jacksoni. It is
a typical southern wehrlei. Apparently this species is not plentiful in
the southern part of its range.

YONAHLOSSEE GROUP

Plethodon yonahlossee Dunn
Yonahlossee Salamander

Plethodon glutinosus, Brimley (1912:137-8, part).
Plethodon yonahlossee Dunn (1917:598-603). Dunn (1926:129-33); Bishop
(1943:287-90); Grobman (1944:287); Pope (1950:79-106).

HOLOTYPE. AMNH 4634, collected near the Yonahlossee Road,
about lI miles from Linville, Avery County, North Carolina, at an alti-
tude of 4200 feet, 16 August 1916, by E. R. Dunn and W. S. Nevin.
DIAGNOSIS. A large species with paired dorsal red spots in the
young and a dorsal chestnut-colored stripe in the adult. Iridophores
are lacking in the region of the dorsal stripe, but are concentrated on
the sides to form a white or light gray lateral stripe.
DISTRIBUTION. Northeast of the French Broad River in the Blue
Ridge Province of North Carolina, Tennessee, and southeastern Vir-
ginia (fig. 34). Pope (1950:82) indicates this species has never been
found east of the New River in Virginia, but more recently, Newman
(1954:13) reports it from Kibler Park, Patrick County, Virginia. Possi-
bly it may occur farther north in Floyd and Franklin Counties.


Vol. 6








HIGHTON: REVISION OF PLETHODON


FIGURE 34. Distribution of the species of the yonahlossee group in the eastern
United States.

DESCRIPTION. The belly is black with few to many small white
spots. The throat is usually light. The dorsal chestnut-colored stripe
results from a mixture of red and black pigment. The red pigment
appears similar to that found in other plethodons. Dorsal melanophores
are concentrated around the mucous glands in the skin. The young have
paired dorsal red spots that are often visible after several years of
preservation, unlike the chestnut band of the adults, owing to the
melanophore-free spot that remains. When the red pigment of the
adult disappears in preservative the animal appears uniformly black.
Occasional specimens have reduced red pigment on the dorsum in life,
and Pope (1950:97) reports one specimen that lacks red pigment
completely. Lateral iridophores are so numerous that they form a light
gray or white lateral stripe in most individuals.


1962








324 BULLETIN OF THE FLORIDA STATE MUSEUM


Vol. 6


Costal grooves usually number 16, trunk vertebrae, 17. Vomerine
teeth range from 12 to 43 (Pope, 1950:103), the largest number in
the genus (fig. 35). The largest specimen Pope (1950) records is
a female, 85 mm. in snout-vent length. Sexual maturity is apparently
reached between 55 and 65 mm. snout-vent length.

Plethodon ouachitae Dunn and Heinze
Rich Mountain Salamander
Plethodon ouachitae Dunn and Heinze (1933:121-2). Bishop (1943:269-72);
Grobman (1944:285); Pope and Pope (1951:129-52).

HOLOTYPE. USNM 92484, adult male, collected on the north side of
Rich Mountain, Polk County, Arkansas, 30 May 1933, by Albert A.
Heinze and Dorothy A. Boyer.
45
0

40 yonahlossee
From Pope (1950) **
35

30-
30 *



W I2 -
I- *


Z


10 -
IO



5-


0 10 20 30 40 50 60 70 80 90
SNOUT-VENT LENGTH (mm.)
FIGURE 35. Variation in the number of vomerine teeth of P. yonahlossee. Data
from Pope (1950).







HIGHTON: REVISION OF PLETHODON


DIAGNOSIS. A moderate-sized species with a dorsal chestnut-colored
stripe as in yonahlossee, but with numerous white spots and brassy
flecks on the back. The lateral stripe is often yellow, in yonahlossee it is
usually white or light gray.
DESCRIPTION. Known only from Rich Mountain in Polk County,
Arkansas, and LeFlore County, Oklahoma (fig. 34). Dundee (1947:
118) reports it from McCurtain County, Oklahoma, but Pope and
Pope (1951:134) believe this record is based on an undescribed form,
not ouachitae.
DESCRIPTION. Pope and Pope (1951) describe variation in more
than a hundred individuals. Having examined only three living speci-
mens, I give their data for the frequency of occurrence of pigmenta-
tion characters.
The belly is dark. Melanophore pigmentation on the chin ranges
from heavy in a few specimens to little or none in 27 percent. Of the
Popes' specimens, 93 percent have yellow or white spots on the chin
and more than half have spots on the belly. Lateral white or yellow
pigment is so heavy that in 85 percent it forms a continuous stripe.
On the back, black pigment is concentrated around the glands of the
skin, as in yonahlossee. Red pigment is distributed uniformly between
these glands. Two kinds of iridophores-small white spots ("white
speckling" of Pope and Pope), and smaller brassy flecks ("frosting")-
are also present. White spotting is absent in 2 percent of their speci-
ments and brassy flecks in 19 percent. Variation in the abundance of
both types of iridophores and in the amount of red on the dorsum is
considerable. In 21 percent of their specimens the back was essentially
black, not chestnut. A 26-mm. juvenile showed two rows of faintly pink
spots on the back after preservation, indicating that juveniles probably
have dorsal red spots as do the young of several other eastern large
plethodons.
Costal grooves usually number 16, trunk vertebrae, 17. Vomerine
teeth range from 9 to 30 (fig. 36). The largest specimen, a female, is
67 mm. in snout-vent length. Sexual maturity is reached between 45
and 55 mm. snout-vent length.

Plethodon caddoensis Pope and Pope
Caddo Mountain Salamander
Plethodon ouachitae, Grobman (1944:285, part).
Plethodon caddoensis Pope and Pope (1951:148).
HOLOTYPE. CNHM 61959, a female collected at an altitude of 1200






BULLETIN OF THE FLORIDA STATE MUSEUM


feet on Polk Creek Mountain, Montgomery County, Arkansas, by Sarah
H. Pope, 9 May 1950.
DIAGNOSIS. A small species with melanophore pigmentation almost
entirely absent on the chin and venter anterior to the forelegs. Dorsal
spots are larger and more numerous than in glutinosus. Vomerine teeth
are more numerous than in any other plethodon in its size range except
ouachitae. The toes are slightly webbed at the base.
DISTRIBUTION. Known from the Caddo Mountains of southern
Montgomery County and southeastern Polk County, Arkansas. Re-
ported by Blair (1957) from the south side of Poteau Mountain,
near Oliver, Scott County, Arkansas (fig. 34).
DESCRIPTION. I have examined only 2 living specimens, both from
Bard Spring Recreation Area, Polk County, Arkansas. The belly is
dark, but anteriorly with large melanophore-free spots occupied by
white iridophores. The chin is almost completely devoid of melano-
phores and iridophores. Large white spots and smaller brassy flecks
are scattered over the back, but the latter are most numerous in and
around the white spots. One individual has a little red on the dorsum,
much as in ouachitae and yonahlossee. White pigment is abundant
on the sides.
Costal grooves usually number 16, trunk vertebrae, 17. Vomerine
teeth range from 15 to 25 (fig. 37). The largest of 46 specimens
examined is 52 mm. in snout-vent length. Sexual maturity is reached
at about 40 mm. snout-vent length.


Most previous workers have considered the yonahlossee group closely
related to P. wehrlei, and Grobman (1944:276) suggests that yonah-
lossee and wehrlei may eventually be proved conspecific. I think the
yonahlossee group is more closely related to the glutinosus group than
either is to the wehrlei group. The species in the yonahlossee group-
yonahlossee, ouachitae, and caddoensis-differ from the glutinosus
group mainly in pigmentation and in having more vomerine teeth.
No other differentiating characters are known, although size and some
other characters vary considerably within each group.
Gaps between the ranges of the species of the yonahlossee group
suggest that the prototype was once widely distributed in the eastern
United States, and subsequent environmental changes isolated the
three forms. In external appearance yonahlossee is similar to ouachi-
tae. P. caddoensis, though occupying an area adjacent to ouachitae,
differs from it conspicuously. This suggests that the two forms,


Vol. 6








HIGHTON: REVISION OF PLETHODON


3NI13WOA


* .. *






in
oo


1962


I-


z




S-
>





o0

















- E
SO
0










I-8

IL
00


s>

4







4)



.4-

P
o
S3

4.B





Cd


Cd
o,
>
















4)
4)


S






^j

.4-4









2.


Cd.


nrz
|1
3



.go







BULLETIN OF THE FLORIDA STATE MUSEUM


ouachitae and caddoensis, have been differentiating from one another
for a long time. Pope and Pope (1951) suggest that two additional
undescribed forms of this group occur in Oklahoma and Arkansas.
P. yonahlossee is most abundant at intermediate altitudes in the
southern Appalachians, being rare or absent in the high altitude
spruce-fir forests and below 2500 feet (Pope, 1950:81). In Arkansas
and Oklahoma, ouachitae is known from 1700 to 2800 feet on Rich
Mountain; caddoensis occurs from 950 to 1200 feet. Both ouachitae
and yonahlossee are abundant in some places where glutinosus occurs.
As they are able to coexist under some conditions, competition with
glutinosus is apparently not the sole factor responsible for the absence
of the yonahlossee group at low altitudes in the eastern United States.
Possibly high temperatures at low elevations restrict them, but too
little is known of their life history, physiology, and ecological require-
ments to further speculate on limiting factors. Knowledge of the three
species in this group is summarized by Pope (1950) and Pope and
Pope (1951). Little new information on their variation, ecology, and
life history has been presented in the last 10 years.
Grobman (1944:278) discusses the relationship between ouachitae
and yonahlossee and suggests the need for morphological and experi-
mental studies to determine whether they should be considered
separate species, subspecies, or undifferentiated populations. The
Popes have demonstrated that these forms differ in size at maturity,
and in maximum size, pigmentation, and number of vomerine teeth,
although overlap in the last two characters exists. Consequently, they
certainly are not undifferentiated populations of the same species.
Pope and Pope (1951:149) suggest that the relationship between cad-
doensis and ouachitae may be subspecific. As differences in size and
color pattern are greater than usually found between adjacent geo-
graphic populations of a single species of Plethodon, I consider them
separate species.

GLUTNOSUS GROUP

Plethodon jordani Blatchley
Jordan's Salamander

Ambystoma jeffersonianum, Rhoads (1895:402-8).
Plethodon jordani Blatchley (1901:762). Dunn (1926:145-6); Bishop (1943:
261-4); Grobman (1944:294).
Plethodon shermani Steineger (1906:559-62). Type locality: Wayah Bald, Macon
County, North Carolina. Dunn (1926:146-8); Grobman (1944:294-6).


Vol. 6







HIGHTON: REVISION OF PLETHODON


Plethodon metcalfi Brimley (1912:138-9). Type locality: Sunburst, Haywood
County, North Carolina. Dunn (1926:148-51); Bishop (1943:264-6); Grob-
man (1944:289-92).
Plethodon clemsonae Brimley (1927:73-5). Type locality: Jocassee, Oconee
County, South Carolina. Bishop (1941b:20), (1943:239-42); Grobman
(1944:293-4).
Plethodon glutinosus, Bailey (1937:3-4, part).
Plethodon glutinosus shermani, Bishop (1941b:18-9), (1943:253-6).
Plethodon shermani rabunensis Pope and Hairston (1948:106-7). Type locality:
Rabun Bald, Rabun County, Georgia. Hairston and Pope (1948:274-5).
Plethodon shermani shermani, Pope and Hairston (1948:106-7).
Plethodon shermani melaventris Pope and Hairston (1948:107). Type locality:
Highlands, Macon County, North Carolina. Hairston and Pope (1948:274-5).
Plethodon metcalfi metcalfi, Mittleman (1948:418).
Plethodon metcalfi clemsonae, Mittleman (1948:418).
Plethodon shermani clemsonae, Hairston and Pope (1948:274-5).
Plethodon jordani jordani, Hairston (1950:271).
Plethodon iordani metcalfi, Hairston (1950:271).
Plethodon jordani shermani, Hairston (1950:271).
Plethodon jordani clemsonae, Hairston (1950:272); Schwartz (1957:94-107).
Plethodon jordani rabunensis, Hairston (1950:272).
Plethodon jordani melaventris, Hairston (1950:272).
Plethodon jordani teyahalee Hairston (1950:269-70). Type locality: Teyahalee
Bald, Graham-Cherokee County line, North Carolina.
HOLOTYPE. Dunn (1926:145) states that the type specimen was
originally in the collection of W. S. Blatchley, but that it has been
destroyed. It was collected between 3000 and 5000 feet on the slope
of Mt. Collins or Indian Pass, Sevier County, Tennessee, by L. E.
Daniels.
DIAGNOSIS. An eastern large plethodon with 17 trunk vertebrae, no
dorsal red pigment except in young specimens from the Nantahala
and Great Smoky Mountains, and white spots lacking on the dorsum
except in examples from the Tusquitee, Snowbird, and Cheoah Moun-
tains, where the spots are small.
DISTRIBUTON. The southern Blue Ridge Province from northeastern
Georgia to Mt. Rogers, Virginia. Also recorded from Tazewell, Bland,
and Giles Counties in the Ridge and Valley Province of Virginia, and
in the Piedmont Province from Burke County, North Carolina, and
Greenville, Pickens, and Oconee Counties, South Carolina (fig. 88).
DESCRIPTION. This variable species was previously divided into
seven geographic races, but none is recognized here. The variable
characters are discussed below. Refer to the maps (fig. 40) for data
on pigmentation at any given locality.
Costal grooves usually number 16, trunk vertebrae, 17. Vomerine
teeth range from 6 to 30 (fig. 39). Size is geographically variable; the




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs