THE SYSYEMATICS AND ECOLOGY 6F P0530N~IW
AND THE POESG‘N-OAKS
(TOXICODENDRON, ANACARDIACFAE)

Thesis for the Degree of Ph‘ D.
MICHEGAN STATE UNIVERSITY
WILLEAM T‘ GELUS
1969

 

MW

MlChfgan State

University

 

This is to certify that the

thesis entitled

THE SYSTEMATICS AND ECOLOGY OF

POISON-IVY AND THE PCISCN-OAKS

(TOXICODENDRON , ANACARDIACEAE)
presented by

 

William Thomas Gillis, Jr.

has been accepted towards fulfillment
of the requirements for

__P§& degree in B0 t any

 

flxé/M

/ v Major professor

Date 16 November 1969

 

0-169

 

 

 

 

 

 

q‘a
.—

 

§ 5
:a
6 u

 

_ ,
1,.

ABSTRACT

THE SYSTEMATICS AND ECOLOGY OF POISON-IVY
AND THE POISON-OAKS
(TOXICODENDRON, ANACARDIACEAE)

by William T. Gillis

This study involves taxonomic, nomenclatural, and
ecological investigations of the poison-ivy complex (all
taxa commonly called poison-ivy or poison-oak) in North
America and eastern Asia. It includes sections Toxicoden-
drop and Simplicifolia of the genus Toxicodendron, tribe
Rhoeae of the Anacardiaceae. It attempts to clarify the
taxonomic and nomenclatural status of the genus and to
delimit Species and infraSpecific taxa. Because several
members of the complex exhibit narrow requirements of habi-
tat and also are kept biologically separated by different
environmental preferences, the ecology of these taxa has
also been explored. The poison-ivies and poison-oaks ex-
hibit considerable variation in form, a variation which has
engendered an intricate nomenclatural maze. Since 1753, a
total of 79 epithets have been published for members of
this complex in 130 nomenclatural combinations in 67 liter-

ature sources. Fifty-eight of these are substantiated by

2 'William T. Gillis
known type Specimens, located in 17 herbaria in 8 countries.
These names apply to 12 taxa as treated in this work.

The difficulties in understanding members of the
complex are related to its economic importance, its extreme
genetic variability, and its being a dioecious woody group.
The group includes the most widespread taxa within the
Anacardiaceae. It is indigenous to North America from Can-
ada to Guatemala, including Bermuda and the Bahama Islands,
and to eastern Asia in Japan, the far eastern Soviet Union,
China, Taiwan, Indonesia, and Malaysia. All but two taxa
were studied in the field, generally with cognizance of the
associated vegetation and soil. Approximately 6,000 her-
barium specimens from 62 herbaria were examined. A dispro-
portionate number of Specimens are aberrant ones, probably
collected because they were forms with extra leaflets, fas-
ciations, insect-induced deformities, unusual size, etc.

Intermediate pOpulationS involving ten combinations
of eight taxa have been found. Several have been interpreted
as hybrids whereas others may be either hybrids or examples
of incomplete divergence among taxa, eSpecially in the
Ozarkian region. The suspected evolutionary develOpment of
the taxa iS discussed with reference of geography and the
deveIOpment of landforms. Fossils, pollen, and flower and
fruit morphology are also considered in their relation to
classification.

An ecological life history of the complex iS included.

Among tOpics considered are altitudinal limits, biological

3 William T. Gillis
interactions with insects and fungi, periodicity of flower-
ing and fruiting, and seed germination. Community relations
are discussed in regard to most of the taxa especially the
specificity of I. toxicarium. A transect through a woods
in central South Carolina from an area of abundance of I,
radicans subSp. radicans to one of I, toxicarium abundance
is discussed. Soil Specificities are also examined, eSpeciale

ly for I, toxicarium.

THE SYSTEMATICS AND ECOLOGY OF POISON-IVY
AND THE POISON-OAKS
(TO’XICODENDRON , ANACARD IACEAE)

By

William T. Gillis

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Botany and Plant Pathology

1969

ACKNOWLEDGMENTS

This monograph could not have been completed with-
out the aid of a number of persons. I am sincerely grate-
ful to many persons for their contribution. The patience,
guidance, understanding, and friendship of Dr. John H.
Beaman, my major professor, has contributed much to this
study. He provided numerous suggestions which aided in the
research and writing of the paper, and his help in obtaining
funds and facilities for the investigation has been invalu-
able. He was also co-principal investigator under terms of
the grant funded by the National Institutes of Health which
SUpported a part of this research. DrS. John E. Cantlon,
Jonathan Wright, Eugene Whiteside, Robert Bandurski, and
John I. McClurkin, Jr., and Mr. Jerold Grashoff made help-
ful comments and criticisms concerning the manuscript.

Several persons took time to guide me in the field
to find Specimens in places not familiar to me: Dr. Edward
G. Voss in parts of Michigan, Dr. H. G. Baker in California,
Dr. Tetsuo Koyama in central Japan, Dr. T. Tsujii in north-
ern Japan, Dr. Frank Craighead in southern Florida, Brother
Daniel Lynch in Texas, and Dr. Barton Warnock in southwest-
ern Texas. Others who assisted me when in the field were:
Dr. Daniel B. Ward and the late Erdman West of the univer-

sity of Florida, Mr. Jack Patterson of Nassau, Bahamas,

,ii

 

Mr. John Verdier of Island Beach State Park, New Jersey,
Dr. and Mrs. N. H. Haack of Haren, the Netherlands, M. and
Mme. Henri Paillet of Paris, France, and Dr. and Mrs.
Laurence Jones of Bedford, England.

Dr. Roland L. Fischer collected a larval form of
‘Aggg living on poison-ivy plants and provided determinations
of insects known to visit poison-ivy.

Several persons sent me specimens for study: Dr.
William Harlow, Dr. J. R. Griffin, Dr. Rollin Baker, Mr.
Julian P. Donahue, Mr. Delzie Demaree, Dr. John Thieret,

Dr. Robert Godfrey, Dr. D. C. D. deJong, Dr. Warren P.
Stoutamire, Mrs. L. Kuprianova of Leningrad, Mr. I. W.
Hughes of Bermuda, Dr. Tang-Shui Liu of National Taiwan
University, and Mr. J. A. R. Anderson of Kuching, Sarawak.

Translations were graciously provided by the fol-
lowing persons: Dr. Mladin Kabalin (Czech), Mr. Peter
Krochta and Mr. Robert Parent (Russian), Dr. D. C. D. deJong
(Dutch), and Dr. Shigeo Imamura (Japanese).

Of Special assistance were those young men who ac-
companied me in the field and labored with me in gathering
data: Messrs. Robert Tinker, Bruce McKenzie, Thomas Hicks,
Wyn Wiksell, Edward Davis, Richard Steinhelper, and David
Steensma.

I wish to thank the curators of the herbaria from
which Specimens have been used. These herbaria are listed
at the beginning of the taxonomic treatment. Special thanks

go to Drs. Robert McIntosh and Lloyd Shinners for exchanging

iii

 

 

isotype material with Michigan State University for my
Special study, and to Dr. Rogers McVaugh for dUplicates of
H. N. Bartlett collections.

Special thanks to Dr. Gordon Sabine, Vice President
for Special Projects at Michigan State University for his
inspiration and assistance throughout this project, often
financial. Without the moral SUpport of my late parents
and my Sister, Elizabeth G. Kohler, and of my colleagues
among the graduate students in the Department of Botany and
Plant Pathology, many of the ideas of this paper might not
have matured.

A generous grant from the National Institutes of
Health (No. C-3700) underwrote much of the field and labo-
ratory work accomplished in connection with this research,

and is gratefully acknowledged.

iv

 

 

 

r
l“" V'
I
‘0‘- -7 .
, 0“.“- ~00-‘
. I
\ l
0-. -o-_.-. | .
- r r" - o
‘ . t I 5 -I.‘ o. .
i A
l .‘s‘. -_ -.V ._‘ .
I - o ’ ".~-‘-Q-\ e
e- \ - ' '
wow.-. .__ _ 0-
[1
i ~-‘r-~—-
. u-
:’ ‘ "-'.0-‘
“p‘ 3 O I
It '1
s ._ o
_. , “v
1.} .. "'"~ --.
1 _ yo.
'~'— ~
"V .L‘”
.
.
. ..-
'1‘.-|'.~ A~
.- . ._ ,
i. “ ‘v '
~':-.~.‘_
pH. ‘5-..
~., "' .
.°""'~-..~""" l
.. . -
o
'\\‘.‘-
‘ O|'~. -
.—
v-I‘.- . h “o
'e‘.. ‘
u.
u...
.
:’ -‘ n~~ -
-~~~—\ .1 -s. .
-.
7..
‘A
a‘.
n ‘\ o.‘ .-
‘~\_ \ ‘-
v -
Qflny ..
.‘
.
I
~V...,“.
~‘_ I. . "
'-
'.D s
‘
q.-
. v"
'5‘Il~~ "" --‘
&. .~ . '~.
-.~ H
. o
_-

 

-.
:-w
0
. .
C
u
‘ ~
..
‘
.1
O
.
A
'.~ l'a‘.
‘ ‘v," \. [R
~~\‘~ "‘~,
. I
~ .\ .
§
4-.“
I-
‘r
.‘v‘ .'."
-|. E
‘v ‘— "

 

 

TABL E OF CONTENTS

 

 

ACKNOWLEDGMENTS . . . . . . . . . . . . . . . . . . ii
INTRODUCTION I . . . . . . . . . . . . . . . . . . 1
ECONOMIC SIGNIFICANCE . . . . . . . . . . . . . . . 4
HISTORICAL ACCOUNT OF THE COMPLEX . . . . . . . . . 10
CHEMISTRY O O O O O O O O O O O O O O O O O O O O 0 2O
GENERIC RELATIONS OF THE COMPLEX . . . . . . . . . 21
TAXONOMIC CONCEPTS O O O O O O O O O O O O O 0 O O 33
PHYTOGEOGRAPHY AND EVOLUTION . . . . . . . . . . . 36
CONSTANCY OF CHARACTERS . . . . . . . . . . . . . . 59
POLLEN MORPHOLOGY . . . . . . . . . . . . . . . . . 64
FOSSILS OF TOXICODENDRON . . . . . . . . . . . . . 90
Archeological Remains . . . . . . . . . . . . 105
FLOWER AND FRUIT MORPHOLOGY . . . . . . . . . . . . 108
Inflorescence . . . . . . . . . . . . . . . . 108
Flowers . . . . . . . . . . . . . . . . . . . 108
calyx O O I O O O O O O O O O O O O O O O O O 109
corOll a O O O O O O O O O O O O O O O 0 O O O 109
Androecium . . . . . . . . . . . . . . . . . . llO
Gynoe Cium O O O O O O O O O O O O O O O O O O 111
Pruj-t O O O O O O O O C C O O O O O O O O O O 112
CI-{ROMOSOME NW8 ERS O O O C O O O O O O O O O O O O l 31
TAXONOMIC TREATMENT O O O O O O O 0 O O O O O O O O 134
Toxicodendron Miller . . . . . . . . . . . . . 134
KEY TO SECTIONS, SPECIES, AND SUBSPECIES OF
TOXICODENDRON . . . . . . . . . . . . . . . . . . . 137

 

SECTION SIMPLICIFOLIA . . . . . . . . . . .

l. Toxicodendron borneense . . . . . .

SECTION TOXICODENDRON . . . . . . . . . .

2. Toxicodendron nodosum . . . . . . .

 

3. T.

4. _T_.

lobatum . . . . . . . . . . . .
radicans . . . . . . . . . . . .

a. subSp. radicans
b. subSp. orientale
c. subsp. hispidum

d. subsp. pubens .
e. subsp. barkleyi

f. subsp. negundo .
g. subsp. divaricatum .
h. subsp. verrucosum .
i. subsp. eximium . . .

 

 

 

 

 

5. Toxicodendron rydbergii . . . . . .

 

6. Toxicodendron toxicarium . . . . .

HYBRIDS, CROSSES, AND INTERGRADES . . . . .

Field Studies . . . . . . . . . . . .
Herbarium Studies . . . . . . . . . .

Toxicodendron lobatum X'I. rydbergii

 

I. toxicarium X-T. radicans subsp.
radicans . . . . . . . . . .
. rydbergii X T. radicans subsp.
radicans . . . . . . . . . . .
toxicarium X'I. radicans subsp.
verrucosum . . . . . . . . . .
rydbergii X‘l. radicans subsp.
negundo . . . . . . . . . . .
radicans subsp. divaricatum X
subsp. barkleyi . . . . . .
radicans subsp. radicans X
subsp. pubens and T. radicans

subsp. negundo X subsp. pubens

ha

 

 

ha

 

 

hfi

 

ha ha

Uncertain or Excluded Species . . . .

ECOLOGICAL LIFE HISTORY . . . . . . . . . .

Morphological Limits . .
Altitudinal Limits . . .

Fire

Biological Interactions
Periodicity . . . . . .

vi

0 O O O O

141
141
148
148
157
179
181
216
228
236
245
253
274
284
297

302
333

361

361
367

367
373
375
379
381
389

394
398
401

401
402
407
408
416

, ‘ u
t .
x . . , . o
w . . s
y . I
s t x t
O . ~ . t .. . ,. . 1 I . 1
. c t . . . . n . ~ .
n i , . s , t . r . _
. t , . \ t I u
x . . t . t e I . . . y . o r . . t . a
1 .t . s . . t p .
. . ; a w .. , s u I r .
v ; t . . . I t t . 1 . t
. . . V i . . 1 I x n

 

 
 

Reproductive Biology . .

Community Specificity for Toxicodendron

toxicarium
Soil Studies

LITERATURE CITED

vii

428

434
457

473

Table

l.
2.
3.

10.

11.

12.

LIST OF TABLES

Comparative Treatment of Rhus . . . . . . . .
Specimens Used for Pollen Studies . . . . .

Summary of Pollen Measurements-—Selected
Anacardiaceae . . . . . . . . . . . . . . .

Lindheimer Collections from New Braunfels,
Texas 0 O O O I O O O O O O O O O O O O O

Interspecific Crosses Between Toxicodendron
radicans subsp. radicans and I. toxicarium.

 

 

 

Characters of Parents, Toxicodendron lobatum
and I. rydbergii and Their Putative Hybrid
Offspring O O O O O O O O O O O O O O O O O

 

Characters of Parents Toxicodendron toxi-
carium and T. radicans subSp. radicans and
Their Putative Hybrid Offspring . . . . . .

 

 

 

Characters of Parents Toxicodendron rydbergii
and T. radicans subSp. radicans and Their
Putative Hybrid Offspring . . . . . . . . .

Characters of Parents Toxicodendron toxi-
carium andT T. radicans subSp. verrucosum
and Their Putative Hybrid Offspring . . . .

 

 

Characters of Parents Toxicodendron rydbergii
and T. radicans subsp. divaricatum and
Their Putative Hybrid Offspring . . . . . .

Characters of Parents Toxicodendron R dber ii
and T. radicans subsp. negundo and Their
Putative Hybrid Offspring . . . . . . . . .

Characters of Parents, Toxicodendron radicans

subsp. radicans and subsp. negundo, and
Their Putative Hybrid Offspring . . . . . .

viii

Page

13
65

70

290

365

372

374

378

380

382

385

388

Table

13.

14.

l5.

16.
17.
18.
19.
20.

21.

22.

23.

24.

25.

Characters of Parents, Toxicodendron radicans
subsp. divaricatum and subSp. barkleyi, and
Their Putative Hybrid OffSpring . . . . . .

 

 

Character Clusters Represented in Various
POpulationS Toxicodendron radicans in the
ozarkian Region 0 O O O O O O O O O O O O 0

Elevation Spans for Species of Toxicodendron,
sect. Toxicodendron . . . . . . . . . . . .

 

Fungi Parasitic on Toxicodendron Species . .
Arthropods Feeding on Toxicodendron Species .
Germination of Toxicodendron Seeds . . . . .
Seed Weights of Toxicodendron Species . . . .

Associates of Toxicodendron toxicarium in
Sandhills Vegetation . . . . . . . . . . .

Trees Found on Transect from Hardwood Forest
to Sandhills Community, South Carolina . .

Shrubs Found on Transect from Hardwood Forest
to Sandhills Community, South Carolina . .

Herbs Found on Transect from Hardwood Forest
to Sandhills Community, South Carolina . .

Soil Profile Analysis of Oak-pine Woods,
Columbia, South Carolina . . . . . . . . .

Soil Profile Characteristics of Areas Where

Both Toxicodendron radicans subsp. radicans
and.l. toxicarium Grow. . . . . . . . . . .

ix

Page

392

396

403
410
412
431
433

436

445

446

449

468

469

Plate
I.

II.

III.
IV.
V.
VI.
VII.

VIII.

IX.

XI.
XII.
XIII.

XIV.

XVI.

LIST OF PLATES

Anacardiaceous Pollen. Surface
Sculpturing. I. . . . . . . . . . . . .

Anacardiaceous Pollen. Surface
Sculpturing. II. . . . . . . . . . . .

Anacardiaceous Pollen. Lateral Pore. I. .
Anacardiaceous Pollen. Later Pore. II. .
Anacardiaceous Pollen. Median View. I. .
Anacardiaceous Pollen. Median View. II. .

Anacardiaceous Pollen. Surface Plan
View. I. O O O O O O O O O O O O O O O

Anacardiaceous Pollen. Surface Plan
View. II. I O C O O O O O O O O O O O O

Rhus guercifolia Goepp. from Specimen

 

Shown in Krafisel (1919) . . . . . . . . .

Toxicodendron magnifolium from the Weaver-

 

ville Elora at Redding Creek, California.
Comparison of Fossil and Modern Poison—ivy.

_Quercus applegatei Knowlton (1900) . . . .

Infructescences of Toxicodendron rydbergii,
Fossil and Modern . . . . . . . . . . . .

Longitudinal Section of Flowers of
Toxicodendron radicans subsp. negund . .

Flowers of Toxicodendron radicans . . . . .

Longitudinal Section of Male and Female
Flowers of Toxicodendron radicans subsp.

negundo O O O O O O O O I O O O O O O O 0

X

Page

71

73
75
77
79
81

83

85

93

97
100
103

106

116

118

120

 

 

 

 

 

 

O
Q

I

l

._ ..
-.

-9
l' -

l.
,- .
q_-
‘C

-

-‘ .
.-

v z 5‘

‘Oa.-

.. .-

l

..l

o--.

Plate

XVII.

XVIII.

XIX.

XX.

XXI.

XXII.

XXIII.

XXIV.

XXV.
XXVI.

XXVII.

XXVIII.

XXIX.

XXX.

XXXI.

XXXII.

XXXIII.
XXXIV.

XXXV.

Longitudinal Section of Young Fruits of
Toxicodendron toxicarium . . . . . .

Cross Section of Fruit Wall of
Toxicodendron toxicarium . . . . . .

Longitudinal Section of Nearly Mature
Fruit of Toxicodendron toxicarium . .

 

Meiotic Chromosomes of Pollen Mother
Cell, Toxicodendron radicans subsp.

negundo . . . . . . . . . . . . . . .

Specimens of Toxicodendron borneense .

 

 

Specimen of Toxicodendron nodosum . . .
Specimen of Toxicodendron lobatum . . .
Specimen of Toxicodendron lobatum . . .

Lectotype of Rhus radicans L. . . . . .

Specimen
subsp.

Specimen
subsp.

Specimen
subsp.

Holotype
subsp.

Specimen
subsp.

Specimen
subsp.

Specimen
subsp.

of Toxicodendron
radicans . . . .

of Toxicodendron
orientale . . .

 

of Toxicodendron
hispidum . . .

of Toxicodendron
barkleyi . . .

of Toxicodendron
negundo . . . .

of Toxicodendron
divaricatum . .

of Toxicodendron
verrucosum . . .

radicans

radicans

radicans

radicans

radicans

radicans

radicans

Lectotype of Rhus verrucosa . . . .

Specimen of Toxicodendron radicans

subsp.

eximium . . . .

Specimen of Toxicodendron rydbergii

xi

Page
122
124
126
128
142
151
158

160
182

184

217

229

246

254

275

285
293

298
303

Plate Page
XXXVI. Specimen of Toxicodendron toxicarium . . 334

XXXVII. Lectotype of Rhus toxicodendron L. . . . 349

 

XXXVIII. Gradient Through Forest in South
Carolina . . . . . . . . . . . . . . . 442

xii

Figure

l.

10.

11.

12.

13.

14.
15.

LIST OF FIGURES

Fossil Species Rhus guercifolia Goepp.

and B. aeQOpodiifolia Goepp. . . . . . . .

Drawing of Flowering Toxicodendron radicans

subsp. radicans from South Florida . . .

Drawing of Fruiting Toxicodendron radicans
subsp. radicans from South Florida . . .

Toxicodendron chromosomes . . . . . . . .

Distribution of Genus Toxicodendron,
sect. Toxicodendron . . . . . . . . . .

 

Distribution of Toxicodendron nodosum and
.I. borneense in Malaysia and Indonesia .

Distribution of Toxicodendron lobatum . .

Distribution of Toxicodendron radicans
subsp. radicans . . . . . . . . . . . .

 

 

Distribution of Toxicodendron radicans
subsp. orientale . . . . . . . . . . . .

 

Distribution of Toxicodendron radicans
subsp. hispidum . . . . . . . . . . . .

Distribution of Toxicodendron radicans
subspp. pubens and verrucosum . . . . .

 

Distribution of Toxicodendron radicans
subsp. negund . . . . . . . . . . . . .

Distribution of Toxicodendron radicans
subspp. eximium, divaricatum, and
verrucosum O O O O O O O O O O O O O O 0

Distribution of Toxicodendron rydbergii .

Distribution of Toxicodendron toxicarium .

xiii

Page

91

91

115
130

145

149
165

189

220

232

239

248

257

308
339

Figure Page

16. Distribution of Toxicodendron lobatum and
I. rydbergii in the Columbia River Gorge
of Oregon and Washington and Location of
Intergrades Between Them . . . . . . . . . 368

 

 

17. Distribution of 3 Taxa of Poison—ivy in the
Uhited States and Canada and Location of
Intergrades Among Them . . . . . . . . . . 376

18. Distribution of 2 Taxa of Poison—ivy in
Mexico and Location of Intergrades
Between Them . . . . . . . . . . . . . . . 390

19. Flowering Periodicity of Toxicodendron
rydbergii . . . . . . . . . . . . . . . . 417

 

20. Flowering Periodicity of Toxicodendron
radicans subsp. negundo . . . . . . . . . 419

21. Flowering Periodicity of Toxicodendron
lobatum and I. toxicarium . . . . . . . . 423

22. Flowering and Fruiting Periodicity of
Toxicodendron radicans subsp. radicans . . 425

23. Road Cut in BastrOp County, Texas, Showing
Profile of Soil in Which Toxicodendron
toxicarium is Abundant . . . . . . . . . . 470

 

xiv

INTRODUCTION

An ecological study in which I attempted to germin-
ate seeds of poison-ivy was the forerunner of the present
monograph. A study of the literature at that time revealed
much confusion as to the application of names to the poison-
ivy and poison-oak taxa, not to mention matters relating to
its toxicity to man. Thus it appeared that Toxicodendron
was deserving of a full-scale systematic investigation, even
though it had been revised by Barkley (1937). Two of the
three sections of the genus are treated herein, section
Venenata being considered too large and diverse a grOUp to
include at the present time.

Barkley (ibid.) provided a taxonomic treatment of
poison-ivy and the poison-oaks as part of a larger work on
Bhu§_(§gg§g'latg) in North America. His investigation was
based chiefly on herbarium material, whereas this one has
placed equal emphasis upon field and herbarium studies with
a few laboratory SXperimentS where apprOpriate. Because
of the restriction of this investigation to a small segment
of what Dr. Barkley monographed, it has been possible to
examine the biology, natural history, and ecology of the
taxa to a greater extent than he was able to do.

A number of problems had to be confronted in the

course of this investigation. The plants are notoriously
l

2

poisonous to man. The literature relating to Toxicodendron

 

is voluminous partly because of its economic importance.

The complex includes the most wideSpread taxa within the
Anacardiaceae. It is further complicated by its extreme
plasticity. Whereas the leaf and fruit morphology is highly
variable, the flower morphology is very conservative. The
woody habit and dioecious nature of the taxa complicates
experimental studies Since at least three years are required
for an individual to reach flowering stage from seed. More-
over, in most pOpulationS, it is not possible to determine
the sex of an individual until it flowers.

Variants have previously been treated as Species,
subspecies, varieties, or forms without a unifying philos-
Ophy or concept of these taxonomic units. Often many of
the characters used in differentiating taxa have been incon-
stant, and combined with one another in various ways, irre-
Spective of the total variation of the plant in nature.

Some of the characters which have been used would segregate
plants within a single clone into several taxa. Misinter-
pretation and fractionating of Species in the past have
been due in part to incomplete knowledge of the range of
variation and therefore little concept of which characters
were useful taxonomically.

The position of Toxicodendron in the Rhoeae is a
fortunate one from the standpoint of understanding related

genera. If a guide to the variability of plants in this

3
tribe can be found in Toxicodendron, then the groundwork
has been laid for future study of such vexing genera as
Rhus and its segregates Searsia, Melanococca, and Lobadium,
Comocladia, and indeed for investigation of the third sec-
tion of Toxicodendron, section Venenata.

The present study brings together as much informa-
tion about Toxicodendron (except section Venenata) as is
now available. It is hOped that this monograph will con-
tribute to a better comprehension of the plants for bota-
nists and for those non-botanists who have need to under-
stand them.

All taxa except I. radicans subSp. hiSpidum, I.

 

nodosum, I. borneense and I. magnifolium were studied in

the field, but not all could be investigated with the same
intensity. Although I had a considerably larger number of
Specimens at my diSposal than did Barkley (1937) and, in
addition, benefitted from field experience with the plants,
various problems remain for future eXploration. If this
investigation can serve as a reference point for study of
problems in p0pulation variability, plant migration, modifi-
cation through evolution, and variability in human sensitiv-

ity, then a primary objective will have been met.

ECONOMIC SIGNIFICANCE

The chief importance of members of the poison-ivy
complex to man is the dermatitis they cause. More than
350,000 cases of poison-ivy dermatitis are estimated for
the United States per year (Turner, 1947). This averages
2.5 cases per 1,000 persons, 17.7% of which were bed cases
(averaging 3.5 days in bed), with 1.7 days of disability,'
71.9% attended by professional medical personnel. Of
economic impact are the 600,000 man-days of time lost from
gainful employment due to the dermatitis and its complica-
tions. According to the California State Compensation In-
surance Fund, Western poison-oak is the only weed in the
state which is rSSponSible for occupational injury (Jones,
1955). Statistics for that state, collected by the Depart-
ment of Industrial Relations, indicate for 1953 that there
were 3,658 cases reported as industrially incurred. Jones
(ibig,) estimates the economic loss due to poison-oak derma-
titis throughout California as $160,000 per annum. Extrap-
olation to other states of the Union where Toxicodendron
Species are found is not possible based on these figures
alone, due to the differential in distribution of plant and
of human pOpulation density, but such figures as those above
suggest an enormous loss of gainful work by employees, not
to mention personal misery due to these plants.

4

5
From a medical viewpoint, members of this complex
have had a colorful history. Some of the earliest works on
record include a thesis submitted to the Medical Faculty of
the University of Pennsylvania by Horsfield (1788), and
later, one in Germany by Burse (1811). The reports in the
medical literature during the nineteenth and twentieth cne-
turies are legion, many of the significant ones summarized
by Rostenburg (1955) and Kligman (1958). A homeOpathic
approach was prepared by Delestinne (1956), and other sum-
maries of our modern medical knowledge by Epstein (1958).
Bewildering and awesome are the lists of "treatments" which
Kligman cites as having been used as tOpical therapy:
Qualitatively, they range from the preposterous to
the fantastic . . . A brief listing of some of the
more interesting agents which have been thought
beneficial clearly reveals the profound emotional
effects of therapeutic deSperation: morphine
(tOpically!), bromine, kerosene, gunpowder (the
symbolism here is beautiful), . . . aqua regia (l),
buttermilk, cream and marshmallows (l) . . . strych-
nine (!), etc.
To this, I must add the preposterous idea of drinking a pint
of photographers' hypo!
Kligman continues:
Throughout, there is a remarkable disregard for a
simple biological fact, namely, that poison-ivy
dermatitis is a self-resolving process.
In recent years, there have been attempts to cure
or prevent the dermatitis by use of injections or prepara-

tions taken by mouth of a number of extracts of the poison

under such trade names as Ivyol and Aqua Ivy (Howell, 1947;

 

 

 

 

. vitmuuftu . L’LJ. If? Erin. ..I wahflvufim

  

6
Passenger, Spain, and Strauss, 1956; and Gaillard, 1967).
Dr. Kligman (ibid.) comments on these:
One is confronted with the quixotic fact that many
physicians achieved remarkable results with an ex-
tract biologically equivalent to water!

Currently a study of oral prOphylaxis is under in-
vestigation at the University of Cincinnati under the direc-
tion of Dr. Leon Goldman.

The disease mechanism is basically a delayed hyper—
sensitivity of the contact variety, and is induced by simple
molecular weight haptens rather than proteinaceous antigens
(Crowle, 1964). A sensitizing antigenic complex is presum-
ably created between the poison (S-Q pentadecylcatechol)
and Skin proteins. The antigens thus are believed to be
chemically-altered autologous Skin proteins.

Aside from the medical aspects of the poisoning,
poison-ivy (Toxicodendron radicans) has had other economic
uses. Unbelievable as it may seem, poison-ivy has been
cultivated in gardens in New South Wales, Australia (An-
onymous, 1908), and in Dunedin, New Zealand (Connor, 1951),
and sold as an ornamental by English nurserymen (Anonymous,
1949), probably as far back as 1640 (Miller and Martyn,
1807).1 It seems to possess certain horticultural values:

(1) it has characteristics for the plant connoisseur who

 

1This is not to be confused with another plant
called poison-ivy in Western Australia which is TinOSpora
smilacina Benth. (MeniSpermaceae) (Gardner and Bennetts,
I956).

2.

 

 

—. .u I. o .t i. L. n.. . . 1. 2.. 5.. I. o. C. .\ ~ \.
.aa .. “I A” I. I. .3 Z. .x. h. c . A. ... .8... o.‘ _: a. ’ s»
it p. ”o. .v .. _ . v. ;. .... .. .7 I. .. r.. C. .3 L. a. is .3. .2 .2
I. .w. ”I .. ... . - a. .~. 1.. fi.. .¢ «. :g a». .~.. .~.. \g A: a» .3 a... \.
A. I ... I. .. .2 .7. . . .2 E It. 2. z. E E t AC .2 I. L» z. 2»
. I
. a.

 

7
likes its autumnal foliage; (2) its autumn leaves have been
used to highlight displays of fruit at Shows; and (3) it is
reported not to escape cultivation (1). Since 1919, poison-
ivy has been planted along the dikes of Southwest Friesland
Province in the Netherlands in an effort at dike consolida-
tion. One colony has become established along more than

500 meters of one dike (Van der Ploeg, 1966).

Toxicodendron lobatum is occasionally browsed by
horses and cattle without apparent ill effects (Rostenburg,
1955a). Many birds, eSpecially quail, feed on its fruits.
Bees are reported to make honey from its flowers, an appar-
ent non-toxic variety (Rostenburg, 1955b). Martin, gt 31.,
(1951) list a number of birds which eat poison-ivy and
poison-oak fruits (not Specifying the scientific names for
either of the plants they had in mind), particularly in
winter when other foods are scarce. Among 55 listed bird
Species which have been observed eating the seeds, yellow-
shafted flickers and wren-tits are claimed to be partial to
the fruits of Toxicodendron Species up to 25% of their diet.
Crows have been reported to be fond enough of the fruits to
be important disseminators of the species (Burrows, 1895).
Poison-ivy fruits in the diet of birds during the winter
have been observed by W. R. Overlease in Michigan and Indi-
ana, and by John Verdier for shore birds at Island Beach
State Park, New Jersey (both personal communications).
Additional relations to wildlife are discussed further under

the tOpic "biological interactions."

 

 

 

8

Although the lacquer-bearing prOperties of resins
of these plants have been established by Dawson (1956a and
b) to be chemically almost identical to those of the lacquer
trees of the Orient [Toxicodendron vernicifluum (Stokes)
Barkley and possibly several other Toxicodendron Species to
a lesser extent], the poisons have never been collected
from any member of section Toxicodendron nor Simplicifolia
for commercial use. Only the Pomo Indians Of California
ever used the resin for a product: they dyed basket fibers
with it (Balls, 1965).

Various American Indian tribes had intriguing uses
for poison-ivy and its allies. The Meskwaki, Ojibwe, and
Potawatomi used it (presumably'l. rydbergii) as a poultice
on a swelling to make the skin Open (Smith, 1928, 1932).

The Ramah Navahoes made an arrow poison from poison-
ivy sap mixed with deer's blood and charcoal from a lightning-
struck tree, the latter no doubt a connection with black
magic (Vestal, 1952). Another version of making this arrow
poison added the juices of Phacelia crenulata var. ambigua
Macbr. (wyman and Harris, 1941). The Navahoes also used
poison-ivy (I, rydbergii) for good luck in gambling: they
chewed a small piece of leaf and gave it to an Opponent
(Vestal, ibig.). The Karok Indians used sticks of I, lobatum
to Spit salmon steaks while smoking them, and its leaves to
cover soaproot (Chlorogalum pomeridianum Kunth) when baking
it in an earthen oven. The Concow Indians of northern Cali-

fornia even mixed the leaves of I, lobatum into their acorn

 

 

 

 

 

9
meal when they baked bread. Other northern California
tribes Simply wrapped their meal in its leaves while baking
(Balls, ibig.). Some, too, used its supple stems as the
warp in weaving baskets (Balls, ibig.).

The Yuki tribe of California used sap from I, lobatum
to get rid of warts: they would cut Off the wart and apply
poison-oak sap to the wound. The same treatment was applied
to ringworm and rattlesnake bites (Balls, ibig,). They also
used the sap, mixed with mountain hemlock and suet, for
tribal markings (John N. Taylor, personal communication).
The Yukis also (especially the Tatu or Huchnom branch) used
a Sprig of poison-oak dipped in water to "keep the women in
due subjection." The men, while attempting to conjure up
the devil in their meeting-hall, would paint one of their
peers, strip him, place a Chaplet of leaves over his face
to render him incognito, and send him out through the vil-
lage amid whoops and diabolical yells. As he cavorted
through the village, he would Sprinkle wet poison-oak
branches in the squaws' faces. Screaming with uncontroll-
able terror, the women would fall prostrate on the ground.
Sworn to Silence lest they die while discussing a Spook,
they would never realize who had, in fact, been their at-

tacker (Powers, 1877).

HISTORICAL ACCOUNT OF THE COMPLEX

The first known reference to poison-ivy in written
record dates from the seventh century in China and next
from the tenth century in Japan, according to Toyama (1918).
Inasmuch as the various Species do not grow in EurOpe, the
plants were unknown to Western civilization until explorers
visited the New World, seven centuries after the first
written record of them in the Orient. Captain John Smith
(1609, 1624) is credited with the first reference to the
poison—ivy plant in writings on his findings in North Amer-
ica. It was he who originated the common name to the plant
since he noted a Similarity in its climbing habit of the
North American poison—ivy to English ivy (Hedera helix L.)
or to Boston ivy (Parthenocissus tricuSpidata Planch.) in
the trifoliolate nature of the leaves. He also noted that
the plant produces a dermatitis which "causeth itchynge,
and lastly, blisters." Probably for reasons of Smith's
description, the first published illustration Of the plant,
and presumably the first reference to it in a work that is
even remotely taxonomic, was its appearance in Cornut's
"Canadensium Plantarum" (1635) as Edera trifolia canadensis,
a name which implies a true ivy. (The taxon pictured is
most likely not a collection made by John Smith, but a later

One from more northern latitudes, and is I. rydbergii.)
lO

11

The details of its pre-Linnaean history, including possible
polynomial synonymy, are discussed by Barkley and Barkley
(1938).

Toxicodendron is a pre-Linnaean name, hot accepted
at the generic level by Linnaeus (1753). Tournefort (1700)
made the following distinction when he first used the name
Toxicodendron: that Rhus had unequally pinnate leaves and
a villose berry (Sic) with a globular nucleus, and that
Toxicodendron had ternate leaves and a striate berry (Sic)
with a compressed nucleus. By limiting the genus to ternate-
1eaved plants, Tournefort would have omitted such close
relatives of poison-ivy as poison-sumac and the oriental
lacquer trees which presumably he did not know. He also
did not know the African Rhu§_5pecies nor Metopium. More-
over, the compressed nature Of the "nucleus" (endocarp and
its contents?) is highly variable and certainly not a reli-
able character. In comparing "villose" to "striated" in
reference to the fruit, Tournefort contrasted an exocarp
character with a mesocarp character. Boerhaave (1727) fol-
lowed Tournefort in distributing Speciesin the Rhus complex
in these genera: Rhus, Toxicodendron, and Cotinus, and he
lumped the African Species in Rhus. Dillenius (1732) like-
wise preserved Toxicodendron. He illustrated poison-ivy
and referred to it as Toxicodendron rectum foliis minoribus
glabris; from the illustration and from a Specimen in the
Dillenian Herbarium, one can deduce that the taxon which

Dillenius had in mind was I. radicans subsp. radicans.

12

Parkinson (1640), Boerhaave (1727), and Munting (1713) had
at one time considered poison-ivy to be a grape (Vitis),
and Boerhaave (ibig.) referred to its being called an
Apocynum.

Linnaeus in the Species Plantarum (1753) and Systema
Naturae (ed. 10, 1759) recorded 13 Species of Rhus, although
probably none of them was known to him in the field. As
Barkley (1963) pointed out, 14 more Species had been assigned
to Rhus by the end of the eighteenth century, including
Species in Rhus, Toxicodendron, Cotinus, Metopium, and the
Thezera groups.2 If priority played a role in keeping Spe-

cies in the genera to which they were originally assigned,

 

2Engler (1881) divided Rhus into four sections:
lggchocarpeae, Gerontogeae, Venenatae, and Melanocarpae.
DeCandolle earlier (1825T_had made the following sections:
Sumac to be roughly equivalent in Engler's concept to Iri-
chocarpeae and Venenatae; he also recognized Sect. Lobadium
to contain some Species which Engler later placed in Tricho-
carpae. The type section Rhus, as presently recognized, is
equivalent to section Sumac DC. = sect. Trichocarpeae Engl.
Blume (1850) erected the genus Melanococca for the black-
fruited Species of the southwest Pacific region; it corre-
Sponds to Engler's section Melanocar ae; Barkley (1942)
raised this to generic rank as Duckera, but did not until
1963 reco nize that this name was antedated by Melanococca.
Brizicky 1963) has changed this to a subgenus of’Rhus.
Section Thezera DC. is the earlier name for sect. Gerontogeae
Engl., which Koch (1853) raised to a subgenus, and Barkley
(1942, 1963) renamed as a genus, Searsia. For the sake Of
Simplicity, I shall refer to the Rhus Species of Africa as
the Thezera group, and those black-fruited Pacific taxa as
the Melanococca group. A summary, with synonymy and desig-
nation Oftypes or lectotypes where needed is given in
Brizicky (ibid.). It is further simplified for reference
purposes in Table 1. It is beyond the SCOpe of this study
to make taxonomic judgments concerning these groups.

 

3
l

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mHmNOCH .poom

 

 

 

 

 

 

 

mHONmCH puma CH
mCComesm mfimnmom mmmmowconow .poom mHoNoCH .poom .omezm .voom
moooooCmHmE
mscomesm mnmxoso omOHmOOCmaoz .poom
memoamz
mscmmesm mEmoamE
puma CH
.ommeOCo> .voom
COHCCOUOOHxOH puma CH pHmQ CH pnma CH
mscmmbsm COHCCOCOOHXOH .mmCHmOOCowHH .poom .omECm .pomm .omEOm .pomm
Eswpmnoq puma Ca
mCComQSm Eswpmeoq .omCHmOOCOAHH .DOOm Eswpmnoq .pomm Ezwpmnoq .pomm
uHmQ CH
wDCm mscomnsm .meHmOOCOAHH .pomm Omenm .poom omssm .poom
mscm msnm mscm mscm mscm
Amooav “mooa popcoem flammav Ambmav AnNmHV
>xoa~anm hmoav >meHmm Hmamcm .CCmConE maaopcmooo

 

 

mDIm mo HszH<mmH m>HH<m<QEOOII.H mqm<H

 

 

 

 

14
then Rhus was fast becoming one of the more inclusive of
angiOSperm taxa.

Linnaeus described poison-ivy first in the Hortus
Cliffortianus (1737); he later edited his description of
the plant to include the climbing habit and added a descrip-
tion of Eastern poison-oak when he wrote the Species Plag-
tgrum (1753). The Clifford Specimens include one flowering
branch of I, rydbergii and a sterile branch of I. radicans
subSp. radicans. Linnaeus apparently did not recognize
poison-ivy to be dioecious, since his symbol for dioecism
is missing from the species descriptions.

Eastern poison-oak presents problems that cannot be
unraveled any farther back than Linnaeus. Several authors
use the same polynomial phrase by which Linnaeus describes
the plant known to him as Rhus Toxicodendron, but the Speci-
men at the Linnaean Herbarium is apparently the oldest
Specimen extant, and can be used to verify the Linnaean
concept of that species.

It is with Michaux that much of the nomenclatural
confusion begins. He made vulgare, microcarpgg, and guer-
cifolium varietal names under Rhus toxicodendron, whereas
Rhus radicans, the Linnaean name for poison-ivy, was not
used at all. In the several later editions of Miller's
Gardeners' Dictionary, Toxicodendron and Rhus were segre-
gated. In Miller's view, Bhflé had hermaphrodite flowers
and Toxicodendron was dioecious. Since both taxa have dioe-

cious members, this character is inapprOpriate. Miller

 

 

 

 

15
(1768) discussed ten species of Toxicodendron, of which
five or six can be referred to the poison-ivy complex.
None can be typified by a Specimen from the Chelsea Physic
Garden, although several can be typified in other ways (see
later discussion under the nomenclature of each taxon).
A. P. DeCandolle (1825) altered Miller's work by changes
in nomenclature, reducing some of his species to varieties,
and reverting to Rhus, all the while citing Miller as his
source.

DeCandolle (1825), Michaux (1803), and later Engler
(in DeCandolle, 1883, and in Engler and Prantl, 1897) re-
turned tO the use of Rhus to the exclusion of Toxicodendron
as a generic name. It may be significant that in Engler's
key to the tribe Rhoideae (1892), Rhus is left close to the
end where it appears that it might have been a catch-all.
Moench (1794) and Rafinesque (1819) recognized Toxicoden-
gggg.

Engler (1883) aligned the Species with essentially
the same subdivisions but, like Michaux, treated all poison-
ivies and poison-oaks as varieties of Rhus toxicodendron,
whereas DeCandolle (1825) made them varieties of Bhg§,;gdi-
gags. Engler, though attributing his interpretation of
this species to an emendation by Torrey and Gray (1838),
was nevertheless in error because the Linnaean specimens
were extant and would have demonstrated that his several
varieties were more closely allied to Rhus radicans than

to R. toxicodendron.

16

Work in this complex in the nineteenth century
began to round out knowledge of distribution. Engler,
relying presumably on Specimens then at Berlin, had a geo-
graphically very diverse group of specimens at his disposal,
for he cites Specimens from localities in Japan, Sakhalin,
Canada, Mexico, Texas, California, Central and Eastern
United States, and Bermuda, making a correct assessment of
several controversial Specimens, not correctly interpreted
since (e.g., Berlandier 2035). It is tragic that this
array of Specimens was destroyed during World War II. The
first nomenclatural separation of Asiatic pOpulationS came
about by Lavalée (1877), and the first for those from Mexico
by Sessé and Mocino (1887).

Otto Kuntze (1891) obliterated the genus Rhus as
he made wholesale transfers to Toxicodendron. His work
has little value except that it is the first place that a
number Of transfers are made from Rhus to Toxicodendron.

Edward Lee Greene (1905, 1910), as he was wont to
do, named 24 new species in this complex, separating rela-
ted pOpulationS on SUperficial grounds. He even annotated
isotypes of his own naming years afterward without recog-
nizing them as the Species he had described! He is known
to have assigned different names to flowering and fruiting
collections from the same pOpulation along the Klickitat
River in Washington. .

Greene, the redoubtable Splitter, very sagely point-

ed Out (1905) that for hundreds of years before Tournefort,

17
there was only one Rhus, i.e., R. coriaria, the coriar's
(tanner's) sumac, whose chief use was the source of tannins
for the processing of leather. When 5, hi;$a_(= R. typhina)
and'fl. glabra were first brought from North America to
EurOpe, Greene continued, there was no question as to the
genus of these sumacs. But then poison-ivy appeared on the
scene. Presumably the first member of this taxon to reach
EurOpe was Toxicodendron rydbergii, the northern represent-
ative of the complex. It is unquestionably the first mem-

ber of the grOUp to be illustrated (as Edera trifolia cana-

 

densis in Cornut, 1635). This Species is likely to have a
subterminal inflorescence like traditional Rhus, and from
the illustrative material that was more commonly available
than herbarium specimens in the eighteenth century, likely
to have been the source of the Rhus myth about poison-ivy
(even though Greene incorrectly equates Cornut's illustra-
tion to be that of I. vulgare = I, radicans subSp. radi-

cans).

 

One of the Clifford Herbarium Specimens, which were
the ones which were instrumental in formulating Linnaeus's
early ideas about members of this complex, is Toxicodendron
rydbergii, a non-climbing sub-shrub which bears superficially

much in common with the habit Of Rhus Species.

 

3Greene correctly points out that Miller, in using
'1. vul are, was using the priority of his day: not that
related to Species Plantarum (1753), but that of Dillenius
(1732), who happened not to use the binomial system.

.. A O . e a
—. .. F.
3. v. .

n.- ’4‘ Av
:- .o. P.»

u
.44
s 3
.3 I.
an Av

.na

2..

 

 

.3
..
:
.»~

 

 

... .3 .. : z‘ . . . : L. .1 .t
: .. .C .. i 4 E ..v Q a.
C. .... ... an .3 7. e. :‘ .2 .2 .C
. 2.. C :. 3.. :s .3 as .7. C .2

 

 

18

In Spite of their penchant for Splitting, it is
significant that E. L. Greene and John K. Small were the
first North American botanists who studied the poison-ivy
complex carefully, and among the first who knew it first-
hand in the field. Gray, Torrey, and Nuttall apparently
did not make an intensive study of this group: Britton
changed his mind, using Rhus in his earlier publications
(1894, 1901) and Toxicodendron in his later ones (Britton
and Shafer, 1908 and Britton and Brown, 1913).

The work of James B. McNair in the early part of
this century on the poison-ivy complex is noteworthy for
its volume. There has been no more prolific writer on the
broad scope of taxonomy, poison chemistry, medical treat-
ment, etc., of the poison-ivy complex. In 1923, his col-
lected works were republished in book form under the title
Rhus Dermatitis. In 1925, he published a Short treatise
on the taxonomy of poison-ivy, an inadequate and fairly
incorrect assessment Of its taxonomy and nomenclature.

In 1937, Fred A. Barkley wrote his doctoral disser-
tation on §hu§,(§§g§g,latg) in North and Central America
including the West Indies. His treatment of the Toxicoden-
dron group was a conservative approach to Species and varie-
ties. However, the nomenclatural decisions which he made
have been only partially followed subsequently, largely be-
cause he chose the unpOpular course of action, i.e., to
accept Toxicodendron as a genus. He later further Split

Rhus into Schmaltzia (1940) (= Lobadium according to

...
u—— ». AV

«4
.aa

v-

A?

0,.

.3

o..

v.
p...

:.
...

 

 

 

 

19

Brizicky, 1963, and Barkley, 1963), Duckera (1957) (= Melano-
cocca), and Searsia (1957, 1963).

Fernald in his intensive examination of the flora
of southeastern Virginia (1950), made further nomenclatural
additions to the poison-ivy complex. Although Fernald
worked from Specimens which he himself had collected in
the field, he made some errors of judgment (e.g., naming
a new form [Rhus toxicodendron f. elobata] of Eastern poison-
oak based on a character which is confined to female plants),
therefore not warranting nomenclatural concern.

A.comprehensive summary of the Rhus vs. Toxicodendron
problem, with resolution of numerous nomenclatural and typi-
fication dilemmas, was published by Brizicky (1963). Barkley
(1963) wrote a rebuttal to this discussion, and so the con-

troversy continues.

 

 

 
 

CHEMISTRY

From a chemical viewpoint, a study of the active
poisons began in 1858 with the work of Khittel who thought
he was dealing with a volatile alkaloid. Maisch (1866) and
Pfaff (1897) isolated the poisons for study, but it was
McNair (1916, 1921) who determined the catechol nature of
the poison, and its possession of an unsaturated Side chain.
Majima and Cho (1908) and Majima (1922) recognized the
similarity of poisons in poison-ivy and in the Japanese
lacquer tree, calling these substances urushiols after
urushi, the Japanese name for plants of the Toxicodendron
gIOUp. Majima first synthesized a compound like that of
urushiol; Hill, gt_al., (1934) isolated some of the pure
poison; and Haack (1940) added to our knowledge of the
poison chemistry of related toxic Anacardiaceae from the
Dutch East Indies. The exact structure of the side chains
and of the number of poisons involved in poison-ivy was
determined by Dawson and his students at Columbia University
(Symes and Dawson, 1954; Dawson, 1954, 1956). There are
four toxic substances in poison-ivy, all 31g pentadecyl-
catechol. The four differ only in degree of saturation of

the side chain.

20

GENERIC RELATIONSHIPS OF THE COMPLEX

Toxicodendron is most nearly allied to Pseudosmo-

 

dingium, Comocladia, MetOpium, and the Rhus complex, SSpecial-

 

1y Actinocheita and Malosma. It Shares with Pseudosmodingium,
Comocladia, and MetOpium possession of toxic resins (likely
all catechols) and Odd-pinnate leaves; in the latter two
genera, however, the endocarp is thin and layers of the
fruits do not separate upon ripening. Furthermore, the
flowers Of Comocladia are trimerous, whereas those of Pseudo-

modingium, MetOpium, and Toxicodendron are pentamerous.

 

The drUpes Of Pseudosmodingium Share with those of Toxico-
dendron the prOperty Of separation of pericarp layers as
the fruits mature; however, the drUpes of Pseudosmodingium
possess two, more or less equal, broad lateral wings.
Engler (1881) and Barkley (1957) separated these two genera
on the Shape of the fruit, with those of Pseudosmodingium
being reniform and strongly compressed, while those of
Toxicodendron are subglobose or ovoid and not strongly com-
pressed. There are several Toxicodendron Species, however,
particularly in Asia (I, succedaneum, I, nodosum, and I.
vernicifluum), that have drupes which may be more or less
reniform and are strongly compressed, although New World
Species (plus I, trichocarpum of Asia) have subglobose

fruits. In fruit and flower morphology, Toxicodendron is
21

 

 

 

 

 
 

22
most closely related to Rhus and its allies, Actinocheita
and Malosma, but vegetatively it is more similar to MetOpium
and Pseudosmodingium.

The question of whether Toxicodendron Should be
segregated at the generic level from Rhus has long been
problematical, there being SO much froth of conviction in-
volved that no series of reasonable arguments in support
of either view has much hOpe of victory. It is necessary
at this point to choose a nomenclatural path and defend its
use in this paper. It would appear that first the taxon
Toxicodendron needs to be defined. Secondly, Bhgg needs
to be examined as a composite genus made Up of several seg-
regates. Thirdly, the reasons pro and con for segregating
these components as genera, i.e., what happens to Rhus
(§§Q§g_latg) in each case, must be evaluated.

McVaugh (1945) made eight recommendations to be
followed in segregating genera. Because each Of these
recommendations will be considered in evaluating the status
of Toxicodendron and Eggs, they are rephrased in summary
here:

1. Strong morphological characters or presence-
absence of distinctive attributes should be
emphasized.

2. There Should be no segregation of genera on
the basis of characters of minor import unless
maintained because of long-held botanical tra-
dition.

3. Not the width of the gap between genera but
the taxon's own biological unity Should be
emphasized. Homogeneity, regardless of degree

of overlap Of characters with those of other
genera, is paramount.

23

4. One should be guided by diagnostic features Of
more inclusive genera throughout the complete
range of the group, and not by local variations.

5. Decisions in regard to the prOposed segregate
Should not be based Upon the equality of rela-
tionships between it and some other segregate,
but upon the relationships between the segre-
gate and the more inclusive genus.

6. Segregate genera should be Sharply delimited.

Any Species intermediate between the segregate
and the more inclusive genus should be relegated
to the latter.

7. The security of position of any genus increases
rapidly in prOportion to the number of differ-
entiating characters.

8. The generic position of a segregate group is
strengthened if it has a distinctive geographical
range together with more or less distinctive
morphological features.

With the exception of item 5, with which I am not
in complete agreement, these guides will be discussed and
observed.

Toxicodendron as a taxon (without rank Specified)
may be defined as being made up of plants having the prOper-
ties of the anacardiaceous tribe Rhoeae which have impari-
pinnate leaves, poisonous effluvium, dark brown root hairs,
exocarp wall cream, yellow, tan, or dun-colored, and white
mesocarp that is fatty and waxy with black striations which
are in reality peripheral resin canals. The plants have
axillary, somewhat loose inflorescences which tend to be
pendent under the weight of the ripening fruit when the
clusters are large. Pubescence of the fruit, if any, is

eglandular. These characteristics have been discussed by

Barkley (1937, 1965), Brizicky (1963), and others, but these

 

24
authors have reached diverse conclusions as to the level of
separation of this taxon from any near relatives.
Almost all workers agree that there should be a

segregation of Rhus and Toxicodendron, if only at the sec-

 

tion or subgenus level. The question then is: are the
characters which segregate these two taxa at least as sig-
nificant as those which segregate other taxa within tradi-
tional Rhus, so that justification of the segregation can

be made at this time? If not, what arguments may be advanced
for retaining the genus Rhus in its traditional sense with

a group of five or six sections or subgenera?

First, let us look at the status of Rhus in the
traditional sense. The traits which serve to characterize
BEBE (£22§2,lgtg) are: (l) the tendency of the fruit walls
to separate in various ways, (2) nonenlarging calyx in the
fruiting stage, (3) presence of one basal ovule, (4) apical
position of the stigma and style, (5) staminal number the
same as the petals, (6) presence of a functionally one-
celled ovary with the hypocotyl against the cotyledons,

(7) ovary and fruit not deeply buried within the intrastam-
inal disc, (8) the ring-porous woods with inter-vascular
pitting circular or hexagonal, bordered and alternate, and
(9) vessel parenchyma and vessel-ray pitting composed of
small, round or slightly oblong pit-pairs. Most students
of the Anacardiaceae agree that Cotinus, MetOpium, and
Actinocheita Should be segregated: (Cotinus on the basis of

Simple leaves, long fertile and plumose sterile pedicels,

 

. . lo p. 0..
.r. . .s. m 0. .. c - u o . V. A—.
o. o v. . p .. r” .. .. .3
.3 .1. . . .v p. . ,.. .. w. .u A” . .2
.. u..- .w p .. 3 r. L. ~ A u. i.

 

 

 

 

 

 

 

. . N4: .4 C a :v
... C.— ..J C» . a . a. «. u
.«i .«v ..-.. o. ‘3 .4‘ a»
o.‘ .k. s «x» H- Ca
V. an “C. an a» a: 2» .2 n/Nl :«
q. 3. 2. e. A.» .‘g v. .s. .C .nu

25
lateral styles, and permanently adhering fruit walls;
MetOpium on the basis of permanently adhering fruit walls,
more or less united calyx lobes, marked bands of wood paren-

chyma, and fairly pulpy mesocarp; Actinocheita on the basis

 

of the long Silky hairs of the fruit and the presence of a
gynOphore.

What is left of traditional Rhus when the foregoing
taxa are removed is a genus which occurs in North America,
northern South America, Mediterranean eastern Africa, south-
ern Africa, southeastern Europe through southern Asia to
China and Japan, Malaysia, and Polynesia. It possesses red,
white, black, yellow, or tan fruits which have either gland-
ular or non-glandular hairs or are glabrous. Its mesocarp
is either resinous or waxy. Its resins are either benign
or dermatitis-producing catechols. Its flowers are either
in dense, terminal thyrses, terminal and lateral compound
Spikes, or loose lateral paniculate thyrses. Its fruit
walls separate upon maturity so that there is adhesion of
the exocarp and mesocarp, or of the mesocarp and endocarp.

Such divergent characters do not unify Rhus.

Once segregated (following the Splitting off of
Cotinus, Metopium, and Actinocheita), then Toxicodendron
is a grOUp of plants which Shares the presence of poisonous
resins, axillary inflorescences, pendent clusters of fruit,
tan, cream, or dun-colored exocarps, pollen of less than
32p in length, plus some additional anatomical characters

discussed by Heimsch (1940, 1942) that need further study

26
in some of Asiatic Species. When the fruits of Toxicoden-
grog are pubescent, they are not also glandular. What
remains of Rhus at this point (including the Thezera group,
the Lobadium group,the Malosma group, and the Melanococca
group) is still heterogeneous for a single genus.

The author admits only a cursory acquaintance with
the African Thezera group and the South Pacific Melanococca
group. A new look needs to be made of the Lobadium and
Malosma groups; the most recent studies (Barkley, 1937, and
Brizicky, 1963) involved only herbarium work. To decide
the question Of Splitting or lumping in the Rhg§_complex,
one must examine the complex worldwide, complete with cyto-
logical, ecological, and comparative anatomical studies.
These investigations are beyond the SCOpe of this paper,
yet a decision must be made as to the most correct names
for the poison-ivy complex to be used herein.

Let us eXplore the kinds of characters which are
used to segregate other genera in the Anacardiaceae (Engler,
1883; Barkley, 1957). Comocladia is separated from MetOpium
on the characters of sinuate to Spinose-pinnatifid leaves
and bony seeds vs. entire leaflets and chartaceous seeds.
Actinocheita is separated from Rhus on the basis of presence
of a gynOphore and long, silky hairs on the exocarp vs.
sessile ovary and hairs on the exocarp being short or absent.
Semecarpus is separated from Holigarna on the basis Of a
half-inferior ovary and intact petiole vs. an inferior ovary

and petiole with one to two lateral Splits at the base.

 

E

27
Harpephyllum is distinguished from Koorderisodendron on
the number of locules in the fruit at maturity (two vs.
one). Gluta is separated from Swintonia because the petals
of Swintonia enlarge in the fruit, whereas Gluta petals
are deciduous.4 Pseudosmodingium has reniform, strongly
compressed drupes, and pinnate leaves, whereas Rhus has
subglobose, ovoid, or only slightly flattened drUpes and
trifoliolate, pinnate, or rarely simple leaves. All these
separations of genera within the Anacardiaceae seem to sug-
gest that the number of consistent differences between Rhus
and Toxicodendron is certainly greater than, and of at
least the same magnitude as, some of these others. In some
other families "good" genera have traditionally been made
on the basis of a few characters, e.g., Canavalia is segre-
gated from closely related Phaseoloideae because of its
bilabiate calyx (Sauer, 1965), and Astragalus from OxytrOpiS
on the shape of the tip of the keel petals.
In his revision of the Anacardiaceae, Marchand

(1869) poignantly points out:

Personne ne songe, nous 1e croyons du moins, a

rétablir l'indépendance de ceS sections pour en

faire des genres distincts. Cependant ils sont

séparés par 1es caracteres d'une valeur beaucoup

plus grande, que ceux qui distinguent encore

certains genres de cette famille. Pourquoi cette

double maniére d'interpréter 1a classification?

Pourquoi ne pas établir des divisions comparables

entre elles? Au moment 00 1e tendance a la
division s'accentue a un tel point sur toutes

 

4Current thinking among workers on Flora Melasiana
is to unite these two genera into a single genus.

28
les autres parties de la famille, ici la tendance
contraire continue a se manifester. Nous ne
comprenons pas pourquoi cette difference existe.

One must be cautioned as by Adanson (1757) when he
suggested that one cannot choose diagnostic characters of
genera when they are incompletely known, because new Species
might prove to be exceptional. Admittedly, the Far Eastern
Species of Bhg§_and Toxicodendron are incompletely known,
but not likely soon to become better known.

The most useful characters for distinguishing
Toxicodendron from related taxa, in my Opinion, are pollen
size, presence of poisonous resin in the phloem, exocarp
color and pubescence, pendent position of the fruiting
inflorescence, and absence of red glandular hairs on pedi-
cels and fruits. There are in addition anatomical differ-
ences of lesser importance, but as Barkley says (1965),

". . . an evaluation of the anatomical differences can
await more exhaustive studies, and the separation of genera
is not dependent on such studies."

The question arises whether the characters which
are used to distinguish these genera are consistent in all
Species. Let us look first at the poisonous nature of the
oleoresins.

Literature references to the toxic nature of the
various species of Rhus are not altogether to be trusted.
Bhg§_michauxii (Rhus, sgggg stricto) of the eastern Uhited
States Coastal Plain and Piedmont has been cited by Sargent
(1895) as being highly poisonous, whereas Warren (1909,

 

 

 

 

29
1910) correctly disputes this statement. Rhus trichocarpa
(= Toxicodendron trichocarpum) is declared by Rehder and
Wilson (1914) to be distinct from R. vernicifera (= I.
vernicifluum) in the "absence of poisonous, varnish-yielding
sap." Both of these Species have poisonous catechols in

abundance. The poisonous nature of Rhus punjabensis from

 

India (ChOpra, Badhwar, and Ghosh, 1949 and Brizicky, 1963)
is still suspect. When I requested of Dr. ChOpra a voucher
Specimen of the R. punjabensis which he had in mind when

he labeled it toxic, a sample Ole. wallichii (= I, walli-

chii) was sent. The problem is not uncommon for Toxicoden-

 

dron Species of the Far East Since they have never been
monographed. The best treatment is Rehder and Wilson (1914).
Quite recently a seed exchange sent to me from the Soviet
Union included a packet labeled "Rhus silvestris S. and Z.
(= I, silvestre) which was in reality Rhus punjabensis.
It is not enough for an investigator to indicate that a
certain Species of Rhus is toxic; he must verify his deter-
mination by a voucher Specimen. Otherwise a good ease may
be made for toxicity among Rhus species which are in reality
Toxicodendron species. Mere reference to the toxicity of
Rhus in the literature, therefore, does not constitute
authenticity that they are indeed possessors of the typical-
ly poisonous catechols.

The presence of the poisonous sap is easy for sus-
ceptible persons to detect in the field, but not so detect-

able from herbarium Specimens. When Specimens lack data

30

regarding the toxicity of the plant, one must look else-
where for suggestion of whether or not the plant was toxic
when alive. Inasmuch as the toxic catechols in Toxicoden-
dron turn black and generally Shiny when allowed to oxidize
and polymerize, one may look for evidence of these Spots
where the leaves and stems have been cut, bruised, or eaten
by insects. The tell-tale lacquer-like Spots near such
wounds may be used to indicate the presence of the toxic
substances.

The terminal nature Of the Rhus inflorescence is
Open to some question if the African Species are included.
In'fihug (egggu stricto = subgenus Rhus) the inflorescence
is strictly terminal. Among the lower pedicels, small
leaves or leaflike bracts may intrude, but the inflorescence
is essentially terminal. The occasional pedicels that have
leaflike structures emerging below them cannot reasonably
be considered axillary. The many Rhus species in Africa
(= subgenus Thezera) have distinctly axillary flower clus-
ters, thus distinguishing them from the Rhus species of
North America, southern EurOpe, and temperate Asia. If
the African Species are segregated into a separate genus
(Searsia), then Rhus indeed may be characterized by terminal
inflorescences. The Rhus Species of the western Pacific
islands (= subgenus Melanococca) has a terminal inflores-
cence as a rule, but as Brizicky (1963) pointed out, there
are exceptions. All inflorescences appear to be thyrses

with the terminal flower in each tertiary cluster Opening

first.

 

1, the“, PJ‘IIJcs--. 15“ «III

  

31

As to fruit color, one must be certain that he is
Observing a ripe or (in the case Of herbarium Specimens) a
prOperly dried Specimen. Brizicky (1963) has examined many
fruiting Specimens of Bhu§.§gg§g.latg, and suggests that
there is overlap in color between the "brown":&uits of
.B- coriaria, 5, COpallina, R. javanica, R, punjabensis,
and R. punjabensis, and 5. (Toxicodendron) griffithii.

The decidedly brown fruits of R. coriaria and R. javanica

 

have the character of glandular pubescence linked with the
fruit color. The "brown" of the other true Rhu§_SpecieS

are red or red-brown, except in slowly dried material where-
in the color deepens to a dark brown. In the case of I.
griffithii, the brown is a tan or Olive, consistent with

the fruit color of Toxicodendron.

AS will be demonstrated later, the pollen grains
of Toxicodendron are significantly smaller and Often more
round (smaller length/width ratio) than those of Rhus.
Mrs. Kuprianova (1965) has supported the erection of a
whole family (Pistaciaceae) as a segregate of the Anacard-
iaceae, based on pollen grain morphology. It is perhaps
not too extreme to use pollen grain size and shape as yet
another set of characters for separating genera.

The segregation of Toxicodendron from Rhus appears
to be appropriate in view Of MOVaugh's "laws." There are
several strong characters which separate it from all other
Rhus Species, characters which unify the genus throughout

its range in both Old and New World. These characters are

32
greater in number and of greater variety than those which
separate generally accepted segregates of Rhus. One species
--admittedly too rare for complete knowledge of it--has the
paniculate inflorescence of Toxicodendron, yet otherwise
possesses Rhus characters. It is relegated to the latter
as Rhus paniculata Wall. There are hybrids and evidences
of crossing among species of Toxicodendron, but none between
Toxicodendron and other Rhus species. Their ranges do,
however, overlap considerably in Asia and North America.
By removing Toxicodendron from Rhus, both taxa become more

homogeneous.

TAXONOMIC CONCEPTS

The tendency among taxa in Toxicodendron under in-
vestigation here is to have allOpatric ranges with only one
species pair being sympatric. Among the allOpatric taxa
intergrades, representing hybridization or incomplete diver-
gence in the Speciation process, occur at the fringes of
ranges where two or more taxa come into contact. Strict
genetic isolation does not appear to be Operating to the
same extent that geographical and ecological barriers do
in limiting gene exchange among taxa.

The sections of the genus are recognized chiefly
on the basis of vegetative morphology: growth form, phyl-
lotaxy, and nature of the leaves, i.e., Simple or compound.
The phyllotaxy of section Toxicodendron is 3/8, that of
section Venenata is 5/13, and the leaves of section Simpli-
cifolia are whorled. In terms of reproductive morphology,
the stamens of section Venenata have long filaments and
Short, globose anthers, but those of sections Toxicodendron
and Simplicifolia have short, subulate filaments and elon-
gate anthers. In the poison-ivy complex, I have recognized
as species those pOpulations which are morphologically
distinct by more than three characters and which seldom
interbreed in the field because of major geographical and
ecological barriers to gene exchange. Only two sets of

33

34

Species of the complex are sympatric (one barely so), and
few inteISpecific hybrids are known. Fruits have been pro-
duced by artificial cross-pollination, but Toxicodendron
toxicarium possesses a much larger set of unique morpholog-
ical characters and ecological requirements than does any
other species, so that it rarely hybridizes with anything
else in nature.

The taxa recognized herein as subspecies of Toxico-
dendron radicans are regarded as geographical segregates
with fewer characters distinguishing them from related taxa
than were utilized to distinguish species. These subspecies
occupy fairly large ranges and seem to intergrade in the
area where they come into contact with other subSpecies.

The term "subSpecies" is used rather than "variety"
for the infraSpecific units because each of the taxa has a
distinctive range, and because there is some variation even
within the subSpecieS. Should a subsequent investigator
feel that these variants are worth naming as distinctive
taxa, he may assign varietal names as subdivisions of the
subspecies. The subspecies group themselves into two dis-
tinct categories according to clusters of characters which
they share in common. It was originally my intent to recog-
nize these two groups as subSpecies and the individual taxo-
nomic variants as varieties. The group Of taxa including
T. radicans subspp. radicans, orientale, and hiSpidum would
then have been a subspecies divided in range between two

hemiSpheres, whereas other infraSpecific taxa with ranges

35
adjacent to subSp. radicans in North America have less in
common with it and would have placed in another subspecific
cluster.
Because of the relative abundance of these plants
and of references to them in both scientific and non-
scientific literature, I feel that no good end would be

served by erecting quadrinomial categories at this time.

PHYTOGEOGRAPHY AND EVOLUTION

Toxicodendron is primarily a genus of North America
and eastern Asia. The range of section Toxicodendron extends
from central China and Japan south to Malaysia and from
southern Canada to western Guatemala, Bermuda, and the west-
ern Bahamas. The range Of section Simplicifolia is much more
restricted, having been found only in north central Borneo.
The range of section Venenata (not treated herein) is throughe
out eastern Asia including Japan and Indonesia with a single
species in the united States from southern Quebec to Florida
in the eastern third of the country, and a single species
in Latin America from Veracruz, Mexico to the highlands of
Colombia.

It is possible that Toxicodendron diverged from Rhus
in the early Tertiary in China with the rare Rhusgpaniculata
Wallich of central China a morphological link between the
two genera. This Species possesses aninflorescence that
is loosely paniculate like Toxicodendron, yet terminal like
Rhus. The color of the fruits on herbarium Specimens, which
may be misleading, is light brown like Toxicodendron. They
are also eglandular. The mesocarp is fibrous and barely,
if at all, striate, resembling Rhus, Due to the paucity
of herbarium material in Western HemiSphere herbaria, it

has not been possible to test for the presence of catecholic

36

37
poisons, but lack of lacquer-like Spots at breaks in the
leaves and lack Of similar resin duct striation in the
fruits are strongly suggestive of its being benign.
Section Toxicodendron with its viney habit appears

to be more advanced than section Venenata. Toxicodendron

 

may well be derived from the common Rhus Species of Asia,
with the multiple leaflet taxa being the Older, reduction

in leaflets being advanced. On occasions I, radicans

 

subSpp. produce leaves with more than three leaflets; I,
lobatum does so even_more commonly. This tendency to pro-
duce multifoliolate leaves is interpreted as reversion to
an ancestral condition. Section Simplicifolia demonstrates
reduction to simple whorled leaves, still exemplifying the
liana habit, and is interpreted as being highly advanced.
Toxicodendron nodosum may be a link between sec-
tions Toxicodendron and Venenata. This Species is scandent,
but apparently produces no aerial roots. It will produce
leaves composed of as many as seven leaflets, but normally
gives rise to only five. In one plant grown in the labora-
tory leaf primordia initiated during periods Of water stress
had fewer leaflets than those develOping when the plant was
abundantly watered. The species appears to combine the
multi-foliolate leaf character and lack Of aerial roots Of
section Venenata with the scandent habit and tendency to
reduction in leaflet number of section Toxicodendron. Be-
cause of the importance accorded to habit within the complex,

and also the stamen structure, this plant has been placed

in section Toxicodendron.

38

Toxicodendron nodosum is found today in the Malay
Peninsula, Borneo, Java, Sumatra, and Celebes. Backer and
Van Den Brink (1965) indicate that Malaya and all the
islands of the Sundra Shelf (Java, Sumatra, gt_al.) were a
continuous land mass during the Pleistocene and would have
permitted a free exchange of disseminules. Presumably I,
nodosum had a distribution throughout this region during
the Pleistocene, its distribution having ebbed and become
dissected as the sea rose.

Toxicodendron radicans is a wide-ranging species,
having several genetic expressions through North American
and eastern Asia. Quoting Asa Gray, J. C. White (1873)
notes:

Our Rhus Toxicodendron, or poison ivy, is very
exactly repeated in Japan, but found in no other
part of the world, although a species much like

it abounds in California. Our other poisonous rhus
(Bhg§_venenata) is in no way represented in Western
America, but has so close an analogue in Japan

that the two were taken for the same by Thunberg
and Linnaeus, who called them both Rhus vernix.

Gray mentions poison-ivy (1846, 1858, 1889) as a
typical example of the floristic connection between eastern
Asia and North America, especially of the eastern United
States. Forbes and Hemsley (1886) felt that the trifolio-
late Species which Augustine Henry had collected from
Ichang on the Yangtze River was "evidently allied to Rhus
toxicodendron L." Miquel (1867) further recognized the
identity of the plant called Rhus toxicodendron (ggggu

auctt. plur. non L.) as being in common with Asia and North

America. He further hypothesizes a common ancestor. More

39
complete studies of the poison-ivy complex have served only
to confirm this affinity. The Atlantic Coast taxon (I,
radicans subSp. radicans) is more Similar in morphology to
the taxa of eastern Asia (I. radicans subSp. orientale and
subSp. hiSpidum) than to any other; the three are thus
grouped as morphologically closely related subspecies. Con-
sidering the present abundance of I. radicans in Japan, it
is puzzling that the Species is not mentioned in any of the
older floras of that country (Thunberg, 1784; Siebold and
Zuccarini, 1826-70; Blume, 1849-56).

The North American taxon (Toxicodendron radicans
subsp. radicans) which most closely resembles the oriental
and fossil ones perhaps represents pOpulations which have
been exposed to relatively little environmental change,
except glaciation, since their invasion of this region,
presumably during Eocene time or shortly thereafter. POpu-
lations in other parts of North America have been exposed
to mountain-building, glaciation, and changes in climate
which have been accompanied by changes in leaf and'fruit
morphology, possibly of a non-selective nature. A.major
change has been in the develOpment of dentate, crenate, and
lobed leaflet margins. This latter type of margin is most
common in pOpulations Of I, lobatum, a near relative Of
poison-ivy. It is this form which represents the greatest
departure of the leaflet from the primitive, entire-margined

form.

40
Tatewaki (1958) noted that poison-ivy, like certain
other Japanese plants such as Hydrangea petiolaris, H,

paniculate, and Actinidia kolomikta, has a distribution

 

throughout Japan, the southern Kuriles, southern Sakhalin,
and central China. Nakai (1909) had pointed out earlier
that it does not occur in nearby Korea, Manchuria, and Far
Eastern Soviet Union.5 Phytogeographically, poison-ivy's
oriental distribution may be described as being in the
Temperate Eastern Asiatic and Subarctic regions as defined
by Engler. It follows Miyabe's line in the Kuriles, being
south, but not north, of the Etorof Strait (Miyabe, 1890),
and Schmidt's line in Sakhalin (Schmidt, 1868, and Tatewaki,

1958). Toxicodendron radicans subsp. orientale is found

 

as far south in the Japanese Archipelago aS Yakusima Island,
one of the northermost islands of the Ryukyus (Masamune,
1934). It skips over the more southerly Ryukyus and is
replaced in Taiwan by I. radicans subsp. hiSpidum.

In Japan, poison-ivy is common, not in fields,
hedges, and fencerows (of which there are few in that coun-
try), but in the forests as a denizen of the Old forest
vegetation. It is not the common intruder into ruderal
Sites as it is in North America. It has been collected
more Often in Hokkaido and Honshu than in Kyushu and Shikoku,
for the former two islands have been studied more intensively

than the latter two.

 

5Its occurrence in the Soviet Union is restricted
to the islands east of the Mainland, over which Russia
secured mandate following World War II.

 

Fl

 

l'
I,
J
‘3!
sand/a: -' d;

 

41

From regions of more rigorous growing conditions
of Hokkaido and the Rishiri Islands, a supposedly non-
scandent form has been described as Rhus rishiriensis. For
lack Of any evidence of a genetic difference between this
form and the more typical Japanese poison-ivy, they are
treated here in synonymy. It is possible that this sub-
shrub is an incipient subSpecies.

In the central mountains of Taiwan and in the moun-
tains of central and western China, Toxicodendron radicans

subSp. hiSpidum occurs as a climber or small bush in rocky

 

places. Its distribution in China follows the Yangtze
River drainage system of the Hunnan and Kweichow plateaus
and foothills of the Tibetan Plateau in Yunnan, Szechuan,
Kweichow, Sikang, and HUpeh Provinces.

The Chinese component of Toxicodendron radicans is
found in two widely disjunct localities, central China and
Taiwan. Taiwan is a continental island probably connected
to the mainland until Tertiary times, possibly until the
Pliocene (Hosokawa, 1954). The Formosa Straits, separating
Taiwan from mainland China, are no deeper than 100 fathoms.
The natural initial conclusion to the student Of the Chinese
flora is that the flora of Taiwan would be more closely
related to that of the East Coast provinces south of the
Yangtze River, but according to Li (1963) the affinity of
Taiwan's mountain flora is with that of central and south-
western China. Even the so-called "endemic" genus of Taiwan,

Taiwania, has been found in northwestern Yunnan. The

.
V.‘
Uni--.

¢~.

     

. __. .L

e . .2 I.

. a 2 . C.

0 v 0 » Fb.
..L-1..v?$s)rult Plates-ill

all? I

H

 

‘

42
phytogeographic link was likely the mountain chain from
China and the Tibetan Plateau through the Malay Peninsula
to Taiwan, a thoroughfare that has been closed since the
Tertiary (Wilson, 1920). According to Hosakawa (1954),
Taiwan was connected geologically with continental China
during the Pliocene. Hypothetically, the distribution of
poison-ivy may have ince been contiguous from western China
to the mountains of Taiwan, and has Shrunk in more recent
times with interruptions by the sea. Long-distance dis-
persal cannot be ruled out as the explanation for the dis-

junct distribution of I, radicans subSp. hiSpidum, although

 

no means of tranSport is suSpected at this time.

The mechanism of inter-continental dispersal for
Toxicodendron radicans is now of some importance. One sug-
gestion to account for the present-day distribution is the
shrinking of a once more extensive range with I. radicans,
or an ancestor, a member of the Arcto-Tertiary flora. One
may envision a continuous distribution throughout eastern
Asia and North America across a hypothetical connecting
land-bridge in the Tertiary when the climate of the boreal
lands was probably temperate. This hypothesis would account
for the striking similarity between taxa on these two con-
tinents--even greater similarity between the intercontin-
ental pOpulations of I, radicans subSp. orientale and radi-
gags than within the pOpulationS of the various taxa in
North America. The latter have evolved with the evolution

of land forms in North America since the closing of the

 

 

 

43
land-bridge, eastern North America and eastern Asia having
remained more stable in the interval than the intervening
land mass. No fossils of poison-ivy, however, are known
from this intervening mass except that I. magnifolium has
been found in several Oligocene formations in northern
California where I, radicans is not found today (MacGinitie,
1937, and Jack Wolfe, personal communication).

Another possibility for the two-continent distribu-
tion of this Species is long-distance diSpersal, but analysis
as part of the Arcto-Tertiary flora is a more logical possi-
bility. The prevailing winds and ocean currents are west-
to-east; thus, it is possible that the Species may have
had its origin in Asia with North America the recipient.
However, its closest relatives are Metopium, Pseudosmoding-
ium, and Rhus Species which are New World in distribution.
It seems, therefore, that Toxicodendron radicans, like the
horse, may have had a North American origin and later mi-
grated to Asia.

In North America, the possible Speciation pattern
of the past may be analyzed by present-day distributions.
There appear to be, in general, only subtle barriers to gene
flow. Toxicodendron radicans subSp. radicans is a lowland
subspecies, separated from subsp. negundo to the west by
the Allegheny Ridge in Pennsylvania and New York, and the
Blue Ridge and adjoining mountains to the south. Zones of
overlap and introgression can be demonstrated in the above-

mentioned regions (see discussion under Hybrids, Crosses,

 

 

 

.us

9...

44
and Intergrades.) In the north subsp. radicans is separated
from I. rydbergii essentially along the 44th parallel of
latitude. This phenomenon is eSpecially evident in Nova
Scotia where I. radicans subSp. radicans is found in the
southwest portion of the province, and I, rydbergii is
found farther northeast above the 44th parallel.

Taken together, Toxicodendron radicans subSp.
radicans and subsp. negundo occupy much the same range as
such genera as Nigga and CornuS, which seem to drOp out
on their western limit where the rainfall is significantly
below the 30-inch annual mark (Dressler, 1954). The north-
ern limit of these taxa (and the southern limit of l.
rydbergii) also correspond. Moreover, subSpp. radicans
and negundo, taken togehter, have an almost identical
northern range as that of Panicum miliaceum, ASparagus
Officinalis, Bromus secalinus, and Arrhenatherum elatius,
as recorded by Hultén (1958,1964) as well as a quite sim-
ilar northern range as a number of others. Presumably
when Hultén's work is completed for the dicots, sufficient
additional Species will be Shown to demonstrate a Similar
distribution to suggest a phytogeographic line at this

location in Canada and upper United States.

 

6Bromus japonicus, Setaria viridis, 223 compressa,
Holcus lanatus, Athyrium filix-femina, Phragmitis gommunis
(sensu lato), Osmunda cinnamomea var. cinnamomea, Q.cIay-
toniana, var. claytoniana, Alisma plantagg-aquatica,

Potamo eton nodosus, Cyperus flavescens Rugpia maritima,
Typha Iatifolia, and Carex atherodes. ’

 

 

 

 

 

 

.-

 

. .! rash: u. .l

 

 

45

In further SUpport of the existence of a phyto-
geographic tension zone here, one might note the southern
limit of Toxicodendron rydbergii. It is identical with
that of Myrica gale, Oxalis acetosella, and Goodyera re ens,
all of which, like I. rydbergii, have an extension of their
ranges along summits of the central Appalachians south to
Virginia. Equisetum palustre has the same southern limit
as I, rydbergii, and in addition demonstrates a southward
continuation along the Shore of Lake Michigan as does I.
rydbergii. In addition to these taxa, a number of others
Show a similar, although not identical, southern limit7
(Hultén, 1958, 1964).

This tension zone as eXpressed in Michigan has
been the subject Of several studies involving pollen, plants,
animals, and Indian tribes. Potzger (1948) pointed out
that climatic factors are less varying over most large
geographical distances than they are within the small lati-
tudinal distance Of about sixty miles under consideration
here. Quercus declines in forests north of the zone, where-
as Pings does so south of it. Livingston, too (1903, 1905),
noted Kent County as the northern limit of many typically
southern plants and the southern boundary of the pine forest
region. Livingston tends to believe soil factors are con-

trolling in the distribution whereas Potzger (ibid.) feels

 

7Carex flava, Subularia a uatica, Asplenium viride,
Equisteum pratense, Carex vaginata, g. disperma, Calypso
bulbosa, EriOphorum alpinum, DryOpteris fragrans.

46
that the "master control" is climate. "There is every evi-
dence that the forest tension zone is climatically transi-
tional, where a number of climatic factors, especially
extremes, are closely Spaced latitudinally, and the forest
is merely expressing the biological values of these environ-
mental factors."

He cites these dividing lines at the tension zone:
average depth of forty inches of frost penetration, the
divide of average date of daily normal temperatures above
500 F between 1 April (to the south) and 1 May (to the
north), a similar divide Of the beginning of daily normal
temperatures above 680 F ("beginning of summer") between
15 June (to the south) and 1 July (to the north), boundary
of the number of days with temperature constantly below
freezing in the shade between 30 and 60, and a difference
of a month and a half in length of continuous snow cover.

Cleland (1966) has noted this tension zone as the
boundary between the hunter-gatherer livelihood of Indian
tribes and the agricultural one. Furthermore, north Of the
line, no mastodon fossils nor arrowheads have been found.
Van Faasen (unpublished) has mapped chromosome patterns in
species of Agigr, showing that within one section of the
genus (section Aéigr) with 2_= 5, almost all members occur
in the tension zone or south of it: and all Species where
fl_= 9 (except A, ciliolatus) occur in the southern two-
thirds of the Lower Peninsula of Michigan, essentially in

the tension zone and south of it.

47

This ecotone extends across the southern lobe of
Ontario, upper New York State, and New England at approxi-
mately the 44th parallel of latitude. It is possibly the
same factors discussed above which have contributed to the
distribution patterns Of poison-ivy taxa, collectively
becoming the barriers to gene flow and factors enhancing
Speciation. The range of Toxicodendron rydbergii is in
western United States and Canada under similarly rigorous
environmental conditions, eSpecially that of reduced rain-
fall. It occurs west of a tension zone that runs north-
eastward through western Texas and Oklahoma, central Kansas,
and the southeastern portions of Nebraska, Iowa, and Wis-
consin, influenced chiefly by the rainfall which is less
than 30 inches west of the zone. Curtis (1959) identifies
this tension zone as the boundary between the prairie-
forest margin and the northern hardwoods province, noting
plant, bird, and fish Species which have the edge of their
ranges there. Krause and Kent (1944) discuss the zone in
New England.

Two of the taxa which occur in Mexico have ranges
which follow the major mountain ranges Of that country.
SubSp. barkleyi is found in the Sierra Madre Oriental, ex-
tending south tO the Sierra de los Cuchumatanes in western
Guatemala and the Isthmus of Tehuantepec. SubSp. divaricatum,
on the other hand, follows the Sierra Madre Occidental,
although its northern limit continues into the mountains

of southeastern Arizona. The two subspecies intergrade

48
in the trans-Mexican volcanic belt, especially in Jalisco
and Michoacan.

The taxon which is closest morphologically toulgéi-
codendron radicans subSp. barkleyi is subsp. pubens of the
southern Mississippi drainage in the United States. Both
forms have pubescent leaves, differing only in that subsp.
pubens has pronounced notches on the leaflets and broad
leaflet bases, whereas subSp. barkleyi lacks these features.

Non-productive branches of subSp. barkleyi revert to a

 

lobed form, although maintaining a narrower leaflet base
than does subsp. pubens. There have been studies which
have noted a phytogeographical connection between plants

of the southeastern United States and Mexico (McVaugh,
1943, Crum, 1951, and Sharp, 1966), eSpecially from the
Ozarkian area where subSp. pubens is indigenous. I postu-
late that the ancestors of this pair Of taxa (ggbgp. barkleyi
and pubens) once existed from the south-central United
States continuously down the mountain chain of the eastern
branch of the Mexican Cordillera. As Dressler (1954) has
suggested for disjuncts between Guatemala and the Ozarks,
these plants became separated into two distinct populations
because of decreased rainfall in Texas and northeastern
Mexico in early Pleistocene. The northern portion has ex-
panded its range into the Lower Mississippi Valley into the
former Eocene embayment region, and portions have,in modern
times, formed hybrids with taxa in adjacent regions over a

large area. The southern pOpulation has also expanded its

49
range down from the mountains into various provinces of
eastern Mexico.

Highly variable pOpulationS of subSp. barkleyi in
northern Mexico posed problems of interpretation until
late in this investigation. They Show some variation in
degree of leaf pubescence, but more significantly, tend
to have leaflets which are broad at the base and more deeply
notched on the leaflet margins, both departures from less
variable pOpulations to the south. Of these, one pOpula-
tion is represented by a large number of herbarium Specimens
collected by Palmer near Ciudad Victoria (Tamaulipas) in
1907; another I have examined myself on two visits to a
site near Galeana (southern Nuevo Léon). These pOpulations
are now interpreted as evidences of the common ancestry
with subsp. pubens to the north, a taxon which has broader
leaflets and notched or nearly lobed leaflet margins.
Farther to the south in the range of subSp. barkleyi, there
is greater uniformity in the entire nature of the leaflet
margin, dense leaflet pubescence, and elliptical nature of
the leaflet margin.

There appears to be a Similar tie between the popu-
lation Of poison-ivy from the Edwards Plateau region of
Texas (Toxicodendron radicans subSp. verrucosum) and the
taxon of the western part of Mexico (subSp. divaricatum).
Both have completely glabrous leaves and shoots except for
an occasional pOpulation with small tufts of hairs in major

vein axils on the lower leaflet surface, a character which

50

possibly hearkens back to a more distant ancestral type,
the Asiatic, Atlantic and Gulf Coast cluster of subSpecies.
I therefore postulate that there was an original connection
in the range of the ancestors of subSp. verrucosum and
divaricatum through the late Tertiary in the province of
parallel ridges in northern Mexico, that the population was
separated into two phases during Pleistocene submergence
of northern Mexico, New Mexico, and southern Arizona (see
Eardley, 1962), and that the two pOpulations have evolved
in situ ever Since. The northern Edwards Plateau popula-
tion has become distinctive by the prominent Sharp lobes
on the leaflets; the southern (Mexican) population demon-
strates narrower, entire leaflets. The southern expression,
subSp. divaricatum, has in recent times SXpanded its range
along the Mexican Highland into southeastern Arizona; the
northern expression, subsp. verrucosum, has begun to extend
from the limestone of the Edwards Plateau and Arbuckle
Mountains into adjoining country to the south and east.

Western poison-oak, Toxicodendron lobatum, is a
likely descendent of the ancestral I, magnifolium, or other
early form which is known from Western United States. The
fossil record Shows divergence from other North American
taxa at least as far back as the Pleistocene in the western
part of the country. Fossil seeds of I, lobatum of Pleis-
tocene age have been found in various parts of California

(Potbury, 1932, and Mason, 1934).

. r.
.3 ..
.u. I.
to. v. 3.
.9 2. a v
f.

3. . .. .. . . C. .2
p“ L. .. ... I. .. C. o.
: Lu .3 :. wn .: .... .3
. .. .. . . 4.1.- . . .9. T.

 

 
 

p 7
L.
.:

 

at m. .t 3. .3 .2 t‘ _. —.
.7. a. r T. T. .q .... a. .2 .1
.. ... .3. :L pt. 3. 1.. .C S 1.1

51

Today, all of the North American taxa are denizens
of altitudes below 1220 m (to 1830 m in the Southwest) and
are not able to cross high mountain barriers, notably the
Cascades. Such pOpulations, which grow west of the Cascade
Mountains, the Great Basin, the Sierras, and the Mojave
Desert, have been isolated from their eastern counterparts
long enough--starting in Miocene time--to have undergone
the modification necessary to be recognized as a distinct
taxon, based largely on leaf and fruit morphology. The up-
lift of the Cascades was late Pliocene and early Pleistocene
(Williams, 1957); the block faulting of the Great Basin was
late Tertiary and Pleistocene (Kay and Colbert, 1965); and
the greatest tectonic activity in the Sierras was in the
late Miocene (Kay and Cobert, ibig,). These landforms
separate 1, lobatum and I. rydbergii today. Only where
the Columbia River cuts through the Pliocene basalts of
the Cascade Mountains have Western poison-oak and poison-
ivy mingled and interbred. The hybrid progeny show intro-
gressive variation in leaf shape and fruit pubescene rang—
ing from the western to the eastern taxa. (See discussion
under "Hybrids, Crosses, and Intergrades.")

The origin of Toxicodendron toxicarium is Open to
question. Although the remote ancestry and derivation of
this plant remains obscure, its recent history probably can
be synthesized from what is known or postulated of other
sandhills Species. This plant inhabits chiefly the fall-

1ine sandhills of the Carolinas and adjoining states, but

52
also has moved into pine barrens, has reached up the Missis-
sippi Valley to southern Missouri, west to Texas, and north
to New Jersey. It has passed freely into the xeric soils
of the Piedmont and lower mountains. Duke (1961) describes
plants such as these as being indigenous to the sandhills,
but not endemic there; rather, they are able to become
established in ecologically similar habitats.

The sandhills and pine barrens floras seem to be
determined by the inability of roots to grow rapidly enough
to penetrate the sterile upper soil horizons to reach the
richer soils below (Wells and Shunk, 1931) and to utilize
nutrients in the surface decomposition zone. Toxicodendron
toxicarium is such an example; its roots rarely reach the
horizon of illuviation. Plants which comprise these floras
were probably the first to colonize the available sandy
habitats uplifted after the Cretaceous inundation (Duke,
1961). Where were they before this?

Fernald (1937) postulated a Virginia-centered class-
ification of floral elements that were later distributed
throughout the sandhills area of the fall-line. Duke (ibid.)
feels that a northern Florida center is more meaningful.

For those who believe in the "Orange Island" refugium during
the Cretaceous (Woodson, 1947; Vaughan, 1910; Cooke, 1945;
Coin, 1958), such a northern Florida center seems most reason-
able.

Following late Tertiary uplift, there might have been

a dichotomous radiation of migration: northward along the

53
xeric sands of the Cretaceous Tuscaloosa formation to the
middle Atlantic states and west through the xeric sands
across the Gulf Coastal Plain. The flora which develOped
in the sandhills may have evolved along with the develop-
ment of the sandhills as a geomorphological formation since
the late Tertiary.

Toxicodendron toxicarium might well have been one
such species. There are a number of eXamples of plants
which share this dichotomous distribution (Duke, 1961, and
Gould, 1962): Pigg§_tagga, Prunus angustifolia, Centrosema
virqinianum, Desmodium strictum, Lespedeza striata, Polygala
polygama, Croton glandulosus, LeiOphyllum buxifolium,
Breweria pickeringii, Ruellia carolinensis, Elaphantopus
carolinianum, Eupatorium rotundifolium, Heterotheca sub-
axillaris, Danthonia sericea, Aristida lanosa, Gymnopogon
ambiquus, Panicum aciculare, P, trifolium, Sacciolepis
striata, Andropogon ternarius, A. elliottii, Cyperus £2339-
fractus, ggggx complanata, Scleria nitida, Smilax glauca,
and Toxicodendron toxicarium.

As Harper (1906) has noted, these plants whose
ranges cross the fall-line (as does Toxicodendron toxicarium)
nearly all occur in the pine barrens as well as in the sand-
hills.

If Toxicodendron radicans were an ancestor of 125;-
codendron toxicarium, then I. toxicarium was gradually
modified for characters of selective advantage in the xeric

habitats in which it grew, i.e., pubescent leaves, reduced

 

 

 

 

54
habit, revolute leaflet margins--all considered at one
time or another as water-saving devices. Its ancestry to
poison-ivy may be seen in its trifoliolate condition and
pubescent fruits, which is a primitive character in the
l. radicans group.

Part of the clue to the relations of Toxicodendron
toxicarium may be found in I, radicans subsp. eximium.
Both have narrow ecological requirements. Toxicodendron
toxicarium inhabits scrub oak forests (often with pine)
that are quasi-savannas. The regions are Open, seldom
with trees more than a foot in diameter, often subject to
burning. The soil is generally sand of low nutrient con-
tent. The range of I, toxicarium is typically the Gulf
and Atlantic Coastal Plains, but has extended into adjacent
regions where a sandhills vegetation prevails. Toxicoden-
gggg radicans subsp. eximium occurs in mesic canyons in
southwestern Texas and northern Mexico, but has the dis-
tinction of being the rarest of the taxa here considered.

The affinites of these taxa are obscure. Both
have pubescent leaves; both have predominantly lobed leaf-
lets; both have fruits which are pubescent to a greater or
lesser degree. Several floras have mistakenly included
subsp. eximium in Toxicodendron toxicarium. These taxa
are more like each other than they are like anything else,
and have no close affinities with other taxa in the complex.
Subsp. eximium is completely separated from any taxa in

adjoining areas and hence demonstrates no intergrading or

 

 

.. . a.. .b. .d o.. f. .hv ... ¥. .~_ ‘5 «v V—m .h» -‘ a. a~» «a 0~.\..
. I C. .. .. I E E S I .3 2. I ... .. C. r -- I S 7.

g

I
\- .

 

 

 

55
hybridization that is known. Toxicodendron toxicarium,
on the other hand, can form hybrids with I. radicans subSp.
radicans, although such hybrids are rare in nature. It is
possible that I. toxicarium shared a common ancestor with
the poison-ivies because they behave as not distant rela-
tives in their ability to cross, yet their actual relation

remains obscure. Perhaps l, toxicarium and I. radicans

 

subSp. eximium are more closely related to each other than
either is to other taxa--they do share many morphological
similarities--but their present ranges seem difficult to
reconcile on the basis of anything but chance long-distance
diSpersal.

Since the modern distributions of the poison-ivies
are related to Pleistocene glaciation, we must attempt to
reconstruct their postglacial history and attempt to account
for where the various taxa must have been during the Pleis-
tocene. There is some evidence from present distribution
patterns that several of the taxa in question may have
resulted from separation by the glaciers of fewer once-
wideSpread ancestral forms into separate pOpulations. At
the height of glaciation, subsp. radicans was undoubtedly
restricted to a region of southern and eastern United States
south of the ice sheet. At the same time, subSp. negundo
might have existed in a refugium in the sand plains of the
central states or in the Ozark region; it may have evolved
from subSp. radicans during this separation by ice sheet

and mountains (Appalachians) at this time. SubSpecies

 

 

 

 

56
pubens probably occupied essentially the same pattern as
today in the lower Mississippi Valley.
With release from the ice, the terra'ggva was free
to be invaded by disseminules of plants such as Toxicoden-

dron radicans behind the influx of Spruce, fir, willow, and

 

aSpen. The northern taxon, I. rydbergii, probably invaded
from the west through a northern route, coming to occupy
land north of the 44th parallel of latitude across eastern
United States and Canada. Subspecies radicans migrated
northward along the Atlantic Coast and lower lepes of the
Appalachians while subSp. negundo invaded along a north-
easterly path through the center of the country into western
New York and Pennsylvania. In relatively recent times,

I. rydbergii may have expanded its range southward along
summits of the Alleghenies and the Blue Ridge by long-
distance dispersal or inadvertent introduction; several

of the southerly outlying populations in mountain areas

are associated with road gravels and recent construction.
These patterns of post-Pleistocene invasion follow sugges-
tions of migration of many other plants described by Braun
(1950), Fassett (1941), and Iltis (1965, 1966, and personal
communication), and others.

In the present Ozarkian region, the systematics of
poison-ivy becomes very confusing. The region of southern
Missouri, eastern Oklahoma, northwestern Louisiana, and
most of Arkansas (with the exception of the "Delta" region

of cotton-growing country along the Mississippi River)

 

57

present problems of poison-ivy identity. In this area,
which is complex lithologically and pedologically, there
appears to be extensive variation over a large area, pre-
sumably the result of large-scale introgression among sev-
eral subspecific taxa. POpulations may show a complex of
the leaf pubescence of subsp. pubens, the entire leaflet
margin character of subsp. radicans or the lobing on the
leaflets of subsp. negundo or pubens, tufts of hairs in

the vein axils of the lower leaflet surface as in subSp.
radicans or absence of such hairs as in subSp. negundo or
ubens, the glabrous fruit surfaces of subSp. negundo or
ubens, or the pubescent fruits of subsp. radicans, etc.
This region has had varieties of poison-ivy indigenous for
a longer time than any of the glaciated country, so that
there has been a longer time for migration, introgression,
and exchange of genes. There appear to be no barriers to
gene exchange among the subspecies, neither of the physical
environment nor of biological mechanics. The activities

of man in altering the landscape will probably enhance this
process of gene exchange in the future. No conclusions
other than the probability of hybridization between the two
subspecies seem to be warranted until experimental studies
and additional field work are carried out in this area.

In summary, one might suggest the following as the

pattern of develOpment and evolution of the poison-ivy com-
plex inlknth America. The ancestral form is that which occu-

pies the Atlantic and Gulf Coastal Plain (Toxicodendron

58
radicans subSp. radicans). The combined forces of the early
Cenozoic Uplift of the Alleghenies and Pleistocene glacial
ice separated this pOpulation into two facies which have
produced subSp. negundo to the west and continued subSp.
radicans in the east. The Ozarkian and Mexican pOpulations
were continuous until the deserts of the early Pleistocene
were formed. Likewise, a western counterpart in subspp.
verrucosum and divaricatum exists with separation in the
Pliocene or late Miocene with faulting and uplift in the
American Southwest. The far western pOpulations have been
separated into distinct pOpulationS since the uplift of
the Sierras and Cascades and the block faulting of the Great
Basin. In regions of more rigorous growing conditions, the
evolution of I, rydbergii may have progressed at a more
rapid rate so that this taxon possesses more distinctive
characters than the others in the complex. The Spread of
I. toxicarium is, no doubt, late Tertiary in age.

One more case of poison-ivy distribution bears men-
tioning. In the late 1950's, a case of poison-ivy derma-
titis showed up in the union of South Africa near Cape Town.
This was anomalous Since there was no known poison-ivy on
the African continent. It was traced to a Specimen of‘l.
radicans which was climbing the side of a house in that
area, introduced at some time Since 1950 (Ross, 1959). The

plant has since been removed from South Africa.

 

CONSTANCY OF CHARACTERS

The constancy and dependability of characters ex-
pressed in the poison-ivy complex has relevance to the
determination of which characters Should be used for tax-
onomic purposes. One reason for the confusion among tax-
onomists in the past regarding the number of taxa in the
complex has been the profusion of characters which have
been used for clasificatory purposes. Determination of
character constancy has been based herein on uniformity
within a region, eSpecially when a character is found (or
is absent) in various Specimens from a mass collection of
a Single clone.

Almost all workers have considered the fruits to
be of Special significance, chiefly in regard to presence
or absence of fruit pubescence. Anatomical studies reveal

that the trichomes on the fruits are hollow, septate hairs

that have been produced by evaginations of the exocarp wall.

Small papillae are considered to be incipient or reduced
hairs; therefore either trichomes or papillae on the fruit
walls are considered an expression of the same character,

varying only in degree. Fruits of Toxicodendron radicans

subSp. orientale, for instance, are always either pubescent

or papillose, never glabrous.

59

q,-

-—

 

. .Kn

. HI;

p.

p:

n

u

n?
C.

.C

 

60
Fruit pubescence is considered to be primitive in
this complex, its greatest expression being found in the
pOpulations of China. It may appear secondarily in aafew
pOpulations in advanced forms which typically have lost
their pubescence through evolution, i.e., pOpulationS from

parts of Indiana and Oklahoma in Toxicodendron radicans

 

subsp. negundo. It may also become reduced or lost entirely
in a taxon wherein it has been an important character.
There appears, for example, to be a cline within this char-
acter's expression in I. radicans subSp. radicans, a taxon
normally distinguished by its pubescent fruits. From cen-
tral Georgia to southern Florida, the fruit hairs appear
less and less pronounced and more sparse, until pOpulations
from Key West, Florida, are decidedly glabrous. The char-
acter of fruit pubescence is still considered of importance

within the I, radicans group because it helps unify most

 

pOpulationS of this taxon. When it does vanish from the
pOpulation, it does so over a fairly gradual cline. Except
for extreme southern Florida pOpulations, then, fruits are
pubescent in this subSpecieS, even though occasionally only
sparsely so. Pubescent or papillose fruits are found in
all fruiting populations ofll. lobatum and I, toxicarium.
The fruits of the genus Toxicodendron are typically
laterally flattened. Only in eastern North American taxa
are they globose or subglobose. Toxicodendron striatum of
Central and South America and l, lobatum Show a tendency to

flattening, but the other taxa of North America have

 

61
globose or subglobose fruits. This character then is depend-
able within Species grOUps.

Habit of the plant and presence or absence of aerial
roots are dependable characters. Toxicodendron toxicarium
and I. rydbergii and supposedly I, nodosum never produce
aerial roots in the wild, the former two being sub-shrubs.

A single leaf does not a taxon make! Although most
keys to poison-ivy taxa have been based on fruit characters,
more new Species have been described on nuances of leaflet
morphology than on any other single trait. Many such taxa
have been described on the basis of only one or two speci-
mens. Most of these are undependable characters because
they vary within as well as among pOpulations.

Number of leaflets is fairly standard at three in
Toxicodendron radicans, I. lobatum, and I. toxicarium.
However, a number of pOpulationS are known in which addi-
tional leaflets are produced year after year (Gillis, 1960).
In I, lobatum it is common for a clone to produce a few
leaves consisting of five leaflets. As many as eleven leaf-
lets have been found. In subSpecieS ofll. radicans the
production of multiple leaflets is more Sporadic, in I.
toxicarium, rare. Although multiple-leaflet forms exist,
they do not deserve taxonomic recognition. Toxicodendron
nodosum characteristically has more than three leaflets.

Juvenile leaves are atypical as are shade leaves,
and particularly leaves from procumbent stems. Leaves used

for diagnostic purposes should always, in erect and liana

62
forms, be chosen from the climbing stems or terminal branches.
There is too much variation in Shade leaves, those from
creeping branches, and those buried within the center of a
large shrub to be of diagnostic value.

If mature leaves on mature flowering or fruiting
branches are selected for diagnostic purposes, then the
characters of importance and dependability are: length-to-
width ratio, presence or absence of tufts of hairs in axils
of vein branches from the midrib on the lower surface of
the leaflets, nature of the leaflet apex and base (eSpecial-
ly on the terminal leaflet), nature of the leaflet margin
(especially on the terminal leaflet), pubescence of the
lower surface of the leaflet, and presence or absence of
pubescence on the petiole. The length of the petiole and
prominence of veins on the leaves are critical in a very
few taxa. If a choice of leaves is possible, one Should
select them from both male and female material, for they
often differ in lobation and serration between the sexes
in the same pOpulation. Young, emergent leaves do not
often have all diagnostic characters SXpressed well enough
for accurate determination.

Of little reliability in determining taxa is leaf
thickness, since this character is highly variable, often
related to the thickness of the cuticle and not to layers
of mes0phyll tissue. Leaf Size is more often than not

related to light environment, larger forms being sun (sic)

 

63
leaves. Notching and serration are good only within certain
broad limits. Upper leaflet pubescence is not constant

within taxa.

 

 

.. .C
—,. .. u
u. __. . . .‘
. .s
. . u‘ «M ~\y\

 

POLLEN MORPHOLOGY

Pollen was examined from Species of Rhus, Toxico-
dendron, MetOpium, Actinocheita, and Schinus to determine
the significance of pollen morphology on the classification
of Toxicodendron and related taxa.

The study was an attempt to determine the degree
of variation among genera, Species, and subspecies of the
tribe Rhoeae, as well as to evaluate pollen characters as
a means for distinguishing the taxa.

For most Species studied, anthers were removed from
materials in the Michigan State University Herbarium (MSC),
chiefly from collections made by me. Two samples were taken
from Specimens at other herbaria, as noted in Table 2..
Pollen was obtained from the type species of the genus Rhus

(the south EurOpean and western Asiatic B. coriaria L.),

 

from the two poison-oak Species, from three American Species
of Rhus sensu stricto, and several taxa of poison-ivy from

North America and Japan. Two Species of Toxicodendron sect.

 

Venenata and one of sect. Simplicifolia also were examined.
One sample each was taken from MetOpium, Actinocheita, and
Schinus.

Dried anthers were collected from herbarium Speci-
mens and stored in vials of glacial acetic acid. Acetolysis
was achieved in the following manner. Pollen and anther

64

 

65

 

oomv mflaafiw

 

.dam .UmwanmHo

 

mmmv wHHHHD

mam? mflaaww

 

 

 

 

.o.m ..oo namasoflm
.mHm ..oo mama
.o.m ..oo vamacoflm
.mam ..oU manoma<

 

 

Pd

mamoflvmu .QwQSm mamoflvmn .

 

 

F4

mamoflpmn .QmQSm mamowvmn

 

F4

EDfiHmofixov .

 

F4

ESHHNOHXOP .

 

 

 

 

 

 

 

 

 

 

.mxm .QMMME .maamo ..oo ompmz cmm sapwooa coupcmbooAXOH
.mdmum cmHmHooE mcflno .cmcc:> ..p§ mcoz Gamma mowcwm .Hm> mflmcmomncso am
ommm mwflflwm .mam ..oo manoma< mc#Cmo:mH .Hm> mafiaamomw .m
mmmm mHHHHG .mw .UCmHmH oamqmm mafiaamqoo .Hm> mcflaamooo .m
«flow. mflflm .1132 too amamfi 3% .m
mfll.nnmmwa meocmaxo mmmmmm am
an .m.m.m.D mfinmflaoo mznm

966832 6cm
Houomaaoo >pwamooq mmwomom

 

 

Ammmwnpcmumo cw

coapmfi>munom >6 pmwmoficcw ma Esanmohwz HmchCm mmmac: Ezwamnemm Umzv
.mMHDDHm quqom mom QmmD mszHUmmmuu.N mqm<H

 

66

 

Vahm mHHHfiO

 

omhm wwaaww
mom? wfiaaflw

 

Axv moaa UCme>mI

 

mObbN finmhmfiCmm

 

 

 

 

 

 

oowxms .maomsm .mapxmwdmso

ooflxmz .manmsm .cmomscmH

.mHm ..oo mama

AmQSSV mHmSmHmz .amamm .samamcflx .pz

 

 

mHHoE mscwnom

 

mcfiofiafim magmauocupo<

 

Ennwmwxov Eswoopms

 

mmcmwcnon am

 

 

 

 

 

 

 

 

 

ammw< .Hsaeflcxma .z Hanpummflum.m
Aq<ov bmom mmmm. m>mam2 .xmawm .pznmq Eamopoc .H
vmbv mfiaaww amomh .Smpmbfinonoz Esohmoocofinp .H
mmmm mflafifio .aouz ..oo EmcmcH xficmm> .H
mwmm mflflaflo .5032 ..06 mpama uflmumau>a .H
_flmmw mfiaafiw .LQHE ..oo EmcmcH mmdmmmm .anzm mamowva aH
mflMm mfiaaflw amdmh .vamnHHOQoz mamvcmwuo .Qmosm mamofivma xH
966552 cam
Houomaaoo >pflamooq mmfiomom

 

 

 

nmscflpcoo--.m mam<e

 

 

67

material was centrifuged at 1000 rpm; the acetic acid was
decanted. Acetic anhydride was added to the mixture and
the whole mixture centrifuged. The supernatant liquid was
decanted and 5 cc of freshly prepared acetolysis mixture
was added (one part conc. H2 804 and nine parts acetic
anhydride). The acetolysis mixture was heated quickly over
a Bunsen burner to boiling. It was then centrifuged immedi-
ately and decanted as close to the residue as possible in
order to remove most of the acetolysis mixture. Seven to
10 cc of water were added to reduce the Specific gravity of
the remaining liquid and to wash the grains.

The washed mixture was centrifuged and decanted.
Ten cc of water were added, the mixture shaken, and the
foam broken up with a few drOps of methanol. The mixture
was again centrifuged, the supernatant liquid decanted,
and the residue was suSpended indefinitely in glycerin-water
mixture (1:1). The pollen was again shaken and centrifuged
and then a saturated solution of safranin in ethanol was
added for 30 seconds, and the mixture centrifuged. One
washing with ethanol, followed by centrifugation, was car-
ried out. A.small amount of glycerin jelly (enough to
cover the residue) was added and heated to melt (but not
boil). By means of a pipette, a portion of the mixture was
transferred to each of several clean coverslips. The ma-
terial was smeared evenly with a toothpick and allowed to
dry. The coverslip, when dry, was affixed firmly to a slide

by Canada balsam. Ikuse_(l956) prefers not to use balsam

 

...

p:

.3
ck.

.3

.-~

A»

0‘

ax»
I! .\

t n
a:

L. . ‘
“KM ll:

 

68
because he claims the Shape of the pollen grain changes as
the balsam dries. Inasmuch as all my slides were prepared
in the same manner, any change would probably affect all
of them more or less uniformly.

Photographs were made under the following conditions:
Zeiss research microscope and Exacta camera, full-length
tube, Optovar at 1.60, iris diaphragm at 4, oil immersion
objective (lOOX), 8X ocular, green and neutral density
filters. Because of vibrations due to movement of the
shutter, time exposures were employed. Photographs were
made with Kodak Plus-X film.

Measurements of acetylated pollen grains were made
at a magnification of 1280 X with a calibrated ocular
micrometer in the microsc0pe. A total of 15 grains was
measured for each "pOpulation," three grains on each of
five slides. Sets of microsc0pe slides of acetylated grains
have been filed in the Herbarium and in the Palynological
Laboratory of Michigan State university.

All pollen grains of Anacardiaceae studied are tri-
colporate (tritreme),8 finely reticulate (striate-rugulate)
with long, tapering, sharply defined transverse furrows
(sulci). The furrows become approximate, but do not meet,
at the poles which are Slightly apiculate. In each colpus,

there is a large, oblong germ pore, wider than the furrow,

 

8Terminology follows Faegri and Iversen (1950),
Selling (1952), Erdtman (1952), Erdtman and Vishnu-Mittre
(1958), Erdtman and Straka (1961), and Kremp (1965).

 

 

69

whose equatorial diameter equals at least two meridional
diameters, each possessing a porous vestibulum with centri-
petally develOped lips. The colpi are zonicolporate, i.e.,
arranged parallel to the polar axis. The ektexine is
Slightly raised above the general surface at the pores;
the endexine is also thickened internally at the edge of
the colpus, forming a ridge or rib (distinct costae trans-
versales). The nexine is incrassate toward the apertures.
The exine is heavy, finely reticulate, pitted, or occasion-
ally granular with pits tending to occur in rows. The
reticulum is small-meshed with irregular and varied anas-
tomoses bearing columellae conjunctae in the ornamentation.
The tegillum is thick; its edge in cross section is Slightly
wavy or smooth. The exine at the edge of the pores thickens
from 1.5p of the mesocolpium to about 3.0 - 3.5u. Heimsch
(1940) describes the exine as having "elevated portions
between the pits . . ." which "may be reticulate in certain
grains, or they may be in definite ridges or striations
with the pits in all stages of alignment or fusion."

The pollen grains of Toxicodendron, Actinocheita,
and Schinus are prolate Spheroidal (length-width equals 1.01
- 1.14); grains of MetOpium are borderline between prolate '
Spheroidal and subprolate (l/w ratio = 1.14 - 1.33). Grains
of fihgg‘sgggg stricto are subprolate. All have long polar
axes. See Table 3.

Plates I-VIII Show photomicrographs of anacardiaceous

grains. Four views of each sample grain are shown: (1)

7O

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

oo.H mm.vm A.SN o.Hm ma.vm m.mm m.mm maaoe mscflnpm
mo.a ma.om o.>m «.mm ma.mm o.vm H.mm mcaoaaam mpflmcoocapo<
mH.H mv.mm 4.0m a.ma ms.mm «.mm 0.4m sanmmflxop annaopwfi
mo.a mm.em 0.0m 8.0m ww.»m 4.0m 0.0m wmcmmcaoa Aw
DH.H mm.om 4.Hm m.mH mm.mm a.qm H.4m uagpmmmaam 4w
Hm.H oo.oH H.mm o.ea mm.om m.mm m.om samouoc .H
mH.H SH.SN o.>m m.mm no.5m m.mm 8.0m Esaamooaoaao AH
qo.a 8H.om n.>m 4.0m aH.Sm 0.0m >.mm xacgm> Am
ma.a Ha.qm 0.0m m.mm om.>m o.mm o.mm Hflmwmausa 4w
om.H me.am o.mm o.oH Ha.om m.mm o.mm oncsmmc .amQSm mcmoflumn .H
om.a no.4m m.mm 0.0m SS.SN 0.0m o.Hm mHmpcmHHo .amQSm mamouwmu .H
om.a no.0m 0.6m o.ma mo.mm m.mm o.mm M.o.mw mcmumwma .amQSm mamoawmu sh
NN.H mm.am H.mm 0.0m mo.mm o.mm o.mm .mam mamoflcmy .amazm mamoaumu Aw
HN.H SH.HN m.mm m.oa ma.nm 0.0m m.mm 1.6.mv ESHHmonOS .H
mH.H mm.Hm 0.6m m.ma om.mm m.>u m.mm 1.mamv ssfiumoaxop.uw
0H.H av.am v.>m o.Hm vH.mm 0.0m 0.0m sapmaoa conucmuooflxow
om.a am.mm >.mm a.am ma.am a.¢m m.om museum .Hm> mflmcmnmmcsq 4m
om.a mm.am 0.4m m.mm Hm.ov m.ma 0.5m mgugwosma .Hm> mafiaamaoo .m
0N.H mo.om N.om 0.5N 00.0m o.m¢ H.mm mcwHHmQOU .Hm> mad ono mm
aa.a no.4m 0.0a o.mm o>.om «.ma o.mm mmmmmxm Am
mN.H mv.Nm 0.0m 0.0N no.0v 0.0% 0.0m muawam .m
om.H mo.om H.mm m.>m om.¢m m.ov m.am mflumfinoo warm
mez mmCmm Cam: magmm
oapmm coca; nova; cpmcma

unpmcmq Cmmz

 

 

9mm mma m2H<mw ma .mzomon zH mm< mezmzmmam
L<m2 2H mm< mezmzmmam<m2 m<mo<Ham<o<z¢ omeomamm--mH2m2mmsm<m2 2maaoa mo >m<22:m--.m mam<e

 

..I. III. I|.|.l1 lll‘ll

f‘. a. a»
5.1.1

llil'l'll.‘

.111 .. . 1...‘
?

 

 

 

.hm ——-—-‘ a- ..,_...-~.___—r__

PLATE I.

71

Anacardiaceous pollen. Surface sculpturing
of grains, X 1280.

a. Rhus coriaria
b. B. c0pallina var. c0pallina
c. Toxicodendron radicans subsp. negundo

d. I, radicans subSp. orientale

 

e. I, rydbergii

72

 

'1
w 1
*m
" . o.
: 'mau‘ Elam-Wm"

PLATE II.

73

Anacardiaceous pollen. Surface sculpturing
of grains, X 1280.

a. Toxicodendron lobatum, aberrant grain
b. I. lobatum, normal grain

c. I. toxicarium from Florida

d. I, trichocarpum

e. Actinocheita filicina

 

 

i 1
-
~ -M-IL
ATFKM' m-h

PLATE III.

75

Anacardiaceous pollen. Section through
lateral pores Showing costae transver-
sales, X 1280.

a. Rhus c0pallina var. cogallina

b.'fl. coriaria

c. Toxicodendron radicans subsp. radicians

d. I. radicans subsp. orientale

 

76

 

PLATE IV.

77

Anacardiaceous pollen. Section through
lateral pore showing costae transversales,
X 1280.

a. Toxicodendron rydbergii
b. I. lobatum

c. I. toxicarium

 

d. l. trichocarpum

 

e. Actinocheita filicina

f. MetOpium toxiferum

 

78

 

PLATE V.

79

Anacardiaceous pollen grains. Median section
showing wall structure, X 1280.

a. Toxicodendron rydbergii
b. I. toxicarium from Florida
c. I, trichocarpum

d. Actinocheita filicina

e. MetOpium toxiferum

_ _-—-—__—_——.-——~————

 

 

A

80

 

PLATE VI.

81

Anacardiaceous pollen grains. Median section
showing wall structure, X 1280.

a. Rhus c0pallina var. c0pallina

 

 

b. R, coriaria
c. Toxicodendron radicans subsp. negundo
d. I. radicans subsp. radicans

e. T. radicans subsp. orientale

82

 

83

PLATE VII. Anacardiaceous pollen grains. Surface plan
view showing pore and colpus, X 1280.

a. Rhus coriaria

 

b. Toxicodendron radicans subsp. negundo
c. I. radicans subsp. radicans

d. I. rydbergii

 

 

 

PLATE VIII.

85

Anacardiaceous pollen grains. Surface
plan view Showing pore and colpus,
X 1280.

a. Toxicodendron radicans subsp. orientale
b. I. lobatum

c. l. toxicarium from Florida

d. I. trichocarpum

e. Actinocheita filicina

f. Metopium toxiferum

 

87
plane surface view Showing sculpturing, (2) longitudinal,
equatorial view through costae transversales Showing longi-
tudinal section of a colpus, (3) median Optical view show-
ing section of the exine, and (4) view through the pore
from the inside (lower surface). Surface sculpturing in
all grains resembles the surface of a peach pit, seen
eSpecially clearly in Rhu§_copallina.

There are few differences among the pollen grains
studied which might be useful in classification, except
for Size. All have similar "peach-pit" sculpturing on the
surface. In Rhus OOpallina and B. glabra the internal
ribs (costae) tend to be more tapering toward the poles
of the grain than they do in‘fi. coriaria and R. typhina,
but these differences are only a matter of degree. .lgxi-
codendron rydbergii has ribs which arch more broadly toward
the poles than do those of other members of the genus
which were examined. The ribs of MetOpium appear to arch
considerably less (i.e., more angularly) than those Of other
genera studied.

A major difference in size is noted between grains
of Rhus and Toxicodendron, yet there remains a consistency
within each grOUp. The mean length of theugggs grains stud-
ied here is 39.11p, whereas that for Toxicodendron is 26.6“;
among 285 grains which were measured, only three had lengths
which overlapped the range of grain Size in the other genus.
By inSpection, one may see the clear-cut separation of the

samples. Likewise, the mean width of Rhus grains studied

88
is 31.7lp whereas that for Toxicodendron is 22.84p. Here
18 grains overlapped the range of widths, SO that the
separation, though marked, is not SO clear-cut as that
for the lengths. Kuprianova (1965) found a Specimen of
Eggs semialata (= 5. javanica) for which the pollen polar
axis was 32.4 - 43-2H and the equatorial diameter was
25.2 - 41.8p as contrasted with Toxicodendron orientale
(= I, radicans subSp. orientale) whose length through the
polar axis was 28.8 - 29.3p and equatorial diameter was
21.6 - 28.8u. Wodehouse (1932) measured Rhus typhina pollen
length as 37.5 - 39p and R. glabra as 35.3 - 40p. Erdtman
(1934) measured 5. typhina pollen as having length 53p x
width 35p. Kuprianova (ibid.) further describes Rhus pollen
as having a Slit-type pore vs. the oblong pore Of Toxico-
dendron; she also calls the sculpturing Of Toxicodendron
grains as more patterned (streaming) than those of Rhus
("barely noticeable streaming sculpture"), i.e., there are
more likely to be rows of markings.

The fact that these differences are significant is
of importance to the question Of whether or not Toxicodendron
Should be segregated from Rhus as a separate genus. The
pollen size is certainly not the major factor in settling
this controversy, but it lends some weight to the argument
that the two Should be recognized as separate genera.

Where pollen grains of two different pOpulations
of the same taxon were measured (e.g., Toxicodendron £25;-

carium, and I, radicans subSp. radicans), there was close

89
agreement in Size between the two pOpulationS. The grains
of.l. lobatum, I. vernix, I. trichocarpum, I, borneense,

I. radicans subSp. orientale, and I. rydbergii are larger

 

(longer and broader) than grains of other Toxicodendron
taxa. The smallest grains to be measured were found in
Schinus mgllg; these grains are also most nearly spherical
(l/w ratio = 1.0). Ikuse (1954) found Significantly larger
grains of Rhu§,ambigua ( ='l. radicans subSp. orientale)
than are reported here: 28.5 - 32-5H x 27.5 - 31p. But
Ikuse used glycerin jelly as a mounting medium rather than
Canada balsam, claiming that the glycerin jelly caused the
grains to swell so that the surface markings might be stud-
ied more favorably.

There have been few Rhus or Toxicodendron pollen
grains found in the fossil record. Wodehouse (1932) dis-
covered some 5hu§_pollen in the middle Eocene Green River
formation in Wyoming. These grains are very Similar to
those Oflfl. typhina and R. labra,the latter of which grows
in that region today. Traverse (1955) described a new
Species, Rhus carbogena, based solely on pollen grains
found in the Brandon lignites Of Forestdale, Rutland County,
Vermont. These were Tertiary (Oligocene?) in age and again
resemble those of B, typhina and fl. glabra, both of which
currently inhabit that region Of Vermont. They are smaller
than grains of either extant species, but fall within the

range of Rhus pollen sizes reported here.

FOSSILS OF TOXICODENDRON

As with most taxa, the fossil record for the poison-
ivy complex is incomplete. A study Of some fossils from
this group of plants and the taxa they represent gives some
indication of the past history of the group. The discussion
will center around plants listed in Fossilium Catalogus
(Edwards and Wonnacott, 1935) and more recently described
fossil Species. Each will be discussed, evaluated as a
member of this complex, and its affinities determined, where
possible.

Rhu§_and its relatives are found frequently in the
fossil state in various periods of the Tertiary (Saporta,
1888). They are never represented by more than detached
leaflets and are therefore difficult to determine accurately.
The fossils suSpected Of belonging to this group, or for

which discussion is relevant, will be discussed.

Rhus quercifolia Goeppert (1855). (Fig. 1.)

This name is probably synonymous with R. aeq0podii-
‘fglia Goeppert (Fig. 2). The fossil Specimens come from
the Tertiary of Bohemia and southwestern Russia (Kryshtofovich,
1914). The material illustrated does not resemble anything
in the modern poison-ivy complex, especially that which has

previously been called Rhus guercifolia auct. Q1, and which

90

Fig. 1.

91

Fossil Species Rhus guercifolia Goepp.
and R. aegopodiifolia Goepp. (Goeppert, 1855).

 

Upper - B. queroifolia.

Lower - R. aegopodiifolia.

 

92

is herein called Toxicodendron toxicarium. The palmate
venation shown in Goeppert's illustration reveals five
main veins with the lowest two leading to a pronounced
lobe on each lateral leaflet, unlike present-day poison-
ivy or poison-oak. Specimens which were presumably depos-
ited in Breslau have not been located and are presumed to
have been destroyed during World War II.9

Krausel (1919), however, throws some light on the
subject. He discusses Goeppert's plants (while citing his
own name as author of the binomial) from Silesian collec-
tions by Perusch. His illustration (Plate IX) is more sug-
gestive of Eastern poison-oak (Toxicodendron toxicarium)
than is any illustration by Goeppert. The rounded lObeS
of the terminal leaflet very closely resemble those on
leaves of modern Eastern poison-oak, as do the cuneate
leaflet bases. The venation also agrees with patterns on
modern leaves.

Krausel felt that his fossil plant was not con-
Specific with the plant of North America. Until his speci-
mens can be studied more thoroughly, one must consider this
fossil as a doubtful member of the poison-ivy complex.

There are a number Of genera which presumably occu-
pied the region Of the Tethys Sea, several Of which also

extended to eastern Asia (e.g., Liquidambar, Juglans, Salix,

 

9Personal communication from Dr. Bohnslav Fott,
Head of Katedra Botaniky Prirodovedecke Fakulty university
Karlovy, 1963.

93

PLATE IX. Rhus guercifolia Goeppert from Specimens
shown in Krausel (1919).

 

 

 

m.

 

95
and DiOSperS). Some genera of these flors apparently van-
ished with the tectonic activity that was associated with
the uplift of the Alps. The Pliocene vegetation Of western
EurOpe is distinguished by the occurrence Of a large number
of elements no longer a part of European vegetation, but
identical with species now living in China, Japan, and North
America (Seward, 1931). The close morphological link of
.Rhug coriaria of EurOpe and R. michauxii Of the southeastern
United States may be an example of plants in the Rhus com-
plex that may have once had a range which included both
North America and southern EurOpe. Section Toxicodendron
may also have shared a distribution in Europe and North
America at this time, represented by a fossil which Krausel
illustrated (Plate IX), a plant now extinct in EurOpe pos-
sibly due to mountain building. Current thinking among
European paleobotanists, however, places Rhus quercifolia
Goeppert in the fossil genus MonOQleura (Aceraceae). Whether
this plant is really the same one which Krausel had under
consideration is not certain.

The binomial Rhus quercifolia was listed in synonymy
by Steudel (1821) for Eastern poison-oak, earlier than
Goeppert's publication; but publication in synonymy does
not constitute valid publication according to the Code
(Lanjouw,.g£_gl., 1966). Therefore, Goeppert's name is
valid and is ascribable to him. Unless one regards the
Goeppert fossil as conSpecific with modern Eastern poison-

oak, this binomial combination cannot be used for the modern

96

plants, even by those who believe them to be Rhus Species

and not Toxicodendron Species.

Rhus toxicodendroides Pilar (1883)

This specimen comes from the Miocene of Croatia.
Pilar's plate Shows but a Single lanceolate acuminate leaf
(leaflet?) that is irregularly dentate and slightly asym-
metrical. He described it as recalling the form and vena-

tion Of "Rhus Toxicodendron L. of North America." It is

 

probably not related to any extant Toxicodendron, and is
difficult to place even in the Anacardiaceae. The type can-
not be found at the Geological Paleontological Museum at
Zagreb, where Pilar's Specimens are located, although other

Rhus Species described at the same time are there.10

Toxicodendron magnifolium (MacGinitie) Gillis

This Species, the most convincing member of the
fossil poison-ivy group, is from the Oligocene Weaverville
flora at Redding Creek, California (Plate X). In morphology,
the leaves of this plant with their entire-margined, large
elliptic leaflets resemble more closely the leaves of Japa-
nese poison-ivy (I. radicans subsp. orientale) than North
American forms, an Observation made by MacGinitie in the
original description. MacGinitie (1937) has cited the de-

gree of asymmetry of the lateral leaflets on the fossil

 

10Personal communication from Dr. Ivan Crnolatac,
Curator Of the Geological Paleontological Museum, Zagreb,
Jugoslavia, 1962.

PLATE X.

97

Toxicodendron magnifolium (MacGinitie) Gillis
from the Weaverville flora at Redding Creek,
California. Specimens at the University of
California, Dept. of Paleontology.

a. Specimen Showing three leaflets.

b. Holotype of l. magnifolium, based on Rhus
magnifolia. Presumably the right lateral
leaflet.

0. Left lateral leaflet of I. magnifolium.

98

 

99
specimen as a major difference from extant poison-ivy. It
is probably less different from that Of East Coast forms
(I, radicans subsp. radicans) than it is from northern and
western ones (I. rydbergii). Leaflet venation of this
fossil (Plate X1) is shown for comparison with that of mod-
ern venation. A more recent fossil, which may represent
a later form of the plant mentioned above,has been found
by Dr. Jack Wolfe in California, but is, as yet, undescribed.

The western poison-oak (I, lobatum) has been found
also in fossil forms.

Potbury (1932) first reported Rhus diversiloba in
Pleistocene deposits at San Bruno, California. Her speci-
mens are seeds with obtuse ends and irregular nodules 5.5 x
3 mm. (Univ. Cal. Coll. Pal..Bot. Plesiotype 437). The
modern seed measures 5 - 6 x 3.4 - 4 mm. Fruit characters
of taxonomic significance are all found in the evanescent
exocarp which is not available in the fossils. Wood frag-
ments were reported by Mason and Chaney (1933) from Carpin-
teria, California, in asphalt sand. The accompanying fossil
vegetation was a Monterey pine forest. Such forests are
now found 200 miles farther north, suggesting that the plant
grew during a glacial (vs. interglacial) epoch when the lower
latitudes were colder than they are today. The wood frag-
ments are rare. Both those records were too fragmentary
to determine the materials with certainty.

More seeds were reported by Mason (1934) in the

Tomales formation of the Pleistocene near Inverness Bay

lOO

PLATE XI. Comparison of fossil and modern poison-ivy.

a. Close-up view of venetion pattern on
Toxicodendron magnifolium from holotype.

b. Close-up view of venation pattern on
I. radicans subSp. orientale from collec-
tion made in Chiba Pref., Japan, June
1961 (Gillis 4691).

—-—_—. .—

 

 

 

102
(Mason's locality 609). The soft mesocarp and exocarp
were lost, but the hard endocarp persisted (Univ. Cal. Coll.
Pal. Bot. Plesiotype 514). Other plants found with it (Rubgg

Spectabilis, Symphoricarpos, Pseudotsuga, and Acer macro-

 

phyllum) are frequent associates of I, lobatum today in
northern California, Oregon, and Washington. Based on fos-
sil invertebrates also present, this plant must have grown
during an interstadial period warmer than today, in a vege-
tation related to the closed-cone pine forests now discon-
tinuously relict along the coast (Mason, 1934).

Attention was called to Quercus applegatei Knowlton
(1900) as a possible Anacardiaceous plant related to lgxi-
codendron lobatum. A careful examination was made of Speci-
mens from the type collection (at US*) from the Tertiary
Ashland beds in Oregon. Most specimens are not Anacard-
iaceae and probably are Quercus (except for one which ap-
peared to be neither Anacardiaceae nor Quercus). The leaf-
let margins Of the fossil (Plate XII) appear to be bristle-
tipped; the lobes are somewhat pointed; there is no evidence
that the leaves were compound. This Species therefore can-
not be recognized to be a Rhus, i.e., 5hg§_applegatei (Knowl-

ton) Chaney ex Potbury (1932).

 

*Abbreviations are for herbaria and are those listed
in Index Herbariorum (Lanjouw and Stafleu, 1964).

103

PLATE XII. Quercus applegatei Knowlton (1900) from
the type collection. Now deposited at
the United States National Museum.

 

 

105

Archeological Remains

Special note is made of a unique group of Specimens
which are barely fossils in the strictest sense. They are,
however, a group of fruits and stem and root fragments esti-
mated by the Wetherell Project using radiocarbon methods
to be 700 years Old, found in adobe dwellings at Mesa Verde
National Park, Colorado. The dry desert air and freedom
from disturbance have permitted complete preservation of
the plant fragments, so that they are still recognizable
today. Poison-ivy plant parts are known to be used by mod-
ern Indian tribes in various ceremonial ways, customs orig-
inating perhaps as far back as 700 years ago.

Several of the stem and root fragments were examined
by Dr. Charles Heimsch of Miami University, and determined
possibly to be poison-ivy. Most Of them were decidedly not
poison-ivy. The fruits are unquestionably l. rydbergii, a
taxon which still grows in that vicinity of Colorado.

(Plate XIII)

PLATE XIII.

106

Infructescences of Toxicodendron rydbergii,
fossil and modern.

Upper cluster on the left is Toxi-
codendron rydbergii from Wexford County,
Michigan, 1961.

Lower roots and stems from cliff
dwellings at Mesa Verde National Park,
Colorado (Specimens courtesy of the Wetherill
Archaeological Project, Douglas Osborne,
Director).

 

 

107

Michigan poison-ivy Mesa Verde poison-ivy
Ad). 1961 ‘ ‘kout Ad). 12m
and cluster and olustc' 35821/709

 

I _ 5,. ‘ U _ _ . ‘ .,
,. ,, .11
‘W 11 ' 12%

mn- VON. hum-all Mr], Cole.
man/709

FLOWER AND FRUIT MORPHOLOGY
Inflorescence

The flowers of the genus Toxicodendron are borne in
paniculate or racemose thyrses in the axils of leaves. The
ultimate clusters are determinate, of three or four flowers
each. The inflorescences are variable in number of flowers;
when large, the clusters are pendulous. In fruiting condi-
tion, the inflorescence of a few taxa may be erect in small,
tight clusters, but will become pendulous in those taxa
wherein the inflorescence is large and loose, bending under
the weight of the develOping fruits. Lignification of the
fruiting rachis enables even a heavy cluster to stand erect
in some taxa (I, rydbergii and I. toxicarium). The bracts
of the inflorescence are lanceolate and deciduous. Inflo-
rescences vary from a few cm in length in I, rydbergii to

over 40 cm in I, nodosum.
Flowers

The entire genus has relatively small, pentamerous
flowers which differ more markedly between the sexes of a
single Species than they do within the same sex throughout
the genus. That is to say that one cannot use the flowers
for key taxonomic characters except for separating section

108

109

Toxicodendron from Venenata on stamen characters.ll Miller

 

and Martyn (1807) maintained that the male flowers of
poison-ivy smelled like lily-Of-the-valley, and that the
female flowers bore no scent, but I find no evidence to

substantiate this.
Calyx

The five sepals are united only immediately above
their insertion below the disk. They are broadly lanceolate
or ovate, imbricate in the bud, and do not enlarge Signifi-
cantly once the flowers Open. The bases are green, but the
upper half is cream-colored like the petals, Often with
prominent purplish veins. They alternate with the petals.
There is a thicker epidermis on the abaxial surface of the
sepals than on the adaxial surface, and the outer surface
may be slightly pubescent at the base. Resin ducts pervade
the phloem of the sepals as they do also in other floral
organs. The sepals Shrink to dark brown vestiges in fruit,
but remain at the base Of the staminal disk, and are dis-

seminated with the fruits.
Corolla

The five petals are cream-colored, ascending, gla-
brous, free, and imbricate in the bud, Often with prominent

purplish veins. They are obtuse, ovate to lanceolate,

 

llSee discussion of the androecium.

110
always smaller in the female than in the male flowers.
In male flowers of Asiatic subSpecies of Toxicodendron
radicans the tips are reflexed, longer and more prominently
reflexed--to over half the total length--than elsewhere in
the genus. The Shorter petals in female flowers may be
Slightly recurved, but never reflexed. Poison-carrying
resin ducts are found in the phloem of the petals. There
is a thick epidermis on both surfaces of the petals. The

corolla is attached to the under side Of the staminal disk.
Androecium

The five stamens are borne on a glabrous, sub-
staminal disk, typical Of this and related families and
orders. Anatomy texts do not treat the origin of this
disk, but it appears tO have its origin as an outgrowth
of the torus. The disk surrounds the ovary, being present
to a greater or lesser degree even in female flowers, and
is attached to the ovary base. In ripened fruits, the disk
and attached calyx remain intact, although shriveled, at
the proximal end. }

In flowers of sect. Toxicodendron the stamens pos-
sess subulate filaments with large, elongate anthers, 4-
celled at maturity. They are borne from the edge of the
cushion-like disk, alternating with the petals and lobes
of the disk. The anthers are oblong, dorsifixed, nearly
as long as the filaments, with only the anthers projecting

beyond the corolla. Dehiscence is longitudinal and introrse.

111

The male flowers of sect. Venenata may be used
diagnostically to separate this section from sect. Toxico-
dendron. The former has stamens exerted at maturity with
globose anthers only about one-fourth as long as the enlarged-
based filaments. In other characters, the anthers of this
section are like those of sect. Toxicodendron.

In female flowers reduced, sterile stamens are pres-
ent. MicrOSCOpic sections Show that there is active nuclear
division (meiosis?) in progress in these anthers, but they

produce no pollen.
Gynoecium

The ovary is tricarpellate, with but one carpel
fertile. It is sessile or partially buried in the hypog-
ynous disk. The style is terminal on the ovary, branching
near the tip into three parts. The stigma is capitate,
Simply an expanded portion Of the style just above a Slight
constriction. It may be rough and minutely pitted, but is
never pubescent. It turns black when it ceases to function
in receiving pollen. This color change from white to black
is useful in indicating whether a flower can still be pol-
linated. The single ovule is anatr0pous and partietal,
seemingly basal, raised on a Short, ascending basal funicle.

In male flowers, there is a rudimentary ovary com-
plete with ovule, style, and stigma, although all are re-
duced in Size. The ovary in male flowers begins to develOp

apparently fertile tissue for a time and even produces

112
rudimentary fruit characteristics such as a thickened ovary
wall complete with resin ducts. These parts are deciduous
with the flowers when pollination is complete.

In functional female flowers, the ovary wall will
enlarge along with the adjacent hypogynous disk after
fertilization. In addition to enlargement, the resin ducts
become very prominent in the develOping fruit. It is these
ducts which make Up the striations on the mesocarp of the

ripened fruit.
Fruit

The fruit is a dry drupe with a pericarp that ma-
tures into three distinct layers: a brittle, chartaceous
exocarp, a fibrous and waxy mesocarp in which are embedded
resin canals, and a bony endocarp to which is firmly affixed
the seed wall. The drupe is globose in Toxicodendron
radicans and I, toxicarium, but laterally flattened, bi-
laterally asymmetrical, and with a somewhat acuminate tip
culminating in the stigmatic surface in most other Species.
The bony endocarp develops early in the maturation of the
fruit, but the separation of the exocarp occurs only in the
later stages of ripening.

The exocarp is green until late stages of develOp-
ment when it turns yellow, straw-, dun-, or cinereous-
colored (deep brown in Toxicodendron wallichii). It is
glabrous in a number of taxa, but may produce non-glandular

hairs in others. Any hairs that are produced begin as

113

evaginations Of the outer cell layer. In some populations
the ripened fruit is papillose rather than pubescent. 1
These papillae are interpreted as incipient or reduced
hairs, and are therefore considered along with hairs for
taxonomic purposes. The hairs begin as multi-celled, bul-
bous evaginations, attached by stalk cells, appearing
glandular, but developing ultimately into unbranched,
multicellular, elongate, bristle-like hairs Of varying
length. The hairs are scattered to dense in I, radicans
subSp. orientale, subSp. radicans, and in I. toxicarium.
There are much finer, almost setose hairs in I, lobatum,
but very dense, elongate, and bristly trichomes in I.
radicans subsp. hispidum, reaching a length of nearly a
millimeter in the latter taxon. The trichomes appear to
be hollow at maturity, unlike the leaf trichomes which often
appear to be filled with a dark-colored substance at matur-
ity.12 Wrinkled exocarp, Often seen in herbarium material,
is merely an indication that the fruits are not ripe; the
wrinkling occurs due to differential Shrinking in the drying
process.

The endocarp is a triple-layered tissue, consisting
of a Single basal layer of small, thick-walled cells, and
a double layer of smaller, isodiametric, very thick-walled

cells. Perpendicular to this endocarp layer is a Single

 

12The fruit trichomes of T. trichocarpum are unique
in the genus. Their bases anastomose all over the surface
Of the fruit and finally project outward as hollow, curved,
bristly hairs.

114

layer of thinner-walled elongate seed coat cells. Entire
fruits with shrunken calyces and disks attached constitute
the functional disseminules, although overwintering fruits
clinging to the pedicels Often lose their exocarps before
diSpersal. References in the literature to "seed" char-
acters are interpreted as references to the mesocarp and
its contents. The endocarp may separate from the mesocarp

to some extent in some taxa, but never from the testa.

Fig.

115

3. Drawing of Flowering Toxicodendron radicans
subSp. radicans from South Florida (Drawn—by

Priscilla Fawcett).

Habit Sketch, X 0.6.
9 Flower, X 6.
d Flower, X. 6.

d Flower, X 6.

A.

B.

C.

D.

E. 9 Flower, X 6.

F. Ultimate Branch of Inflorescence, X 6.
G. L.S. of 9 Flower, X 6.
H

- I. Floral Diagram of Alternate Forms of
Aestivaticn, 9 Flower.

J — K. Floral Diagram of Alternate Forms of
Aestivation, 6 Flower.

 

 

 

 

PLATE XIV.

116

Longitudinal section of the flower of
Toxicodendron radicans subsp. negund .

a. Flower in bud, showing stamens, stam—
inal disk, and sterile ovary, X 40.

b. Open flower showing attachment of

petal and sepal to staminal disk,
X 100.

 

117

 

118

PLATE XV. Flowers Of Toxicodendron radicans.

a.

b.

 

Cross section Of flower of I. radicans

subsp. negundo, taken just below mid-
point of flower, X 40.

Longitudinal section Of flower of

I. radicans subsp. orientale, showing
perianth parts, X 100. Notable is
the long, reflexed petal.

 

 

 

119

 

 

PLATE XVI.

120

Longitudinal section of flowers of

I. radicans subsp. negund .

Upper. Male flower Showing fertile
stamens and develOping rudi—
mentary ovary.

Lower. Female flower, sectioned after
petals had fallen, showing
enlarging disk and ovary. Dark-.
staining cells in the ovary are
incipient resin canals.

 

 

 

 

121

 

PLATE XXVII.

122

Longitudinal sections of young fruits
Of Toxicodendron.

Upper. Fruit Of I. toxicarium, showing
enlarged disk, calyx remnant,
resin canals in the young fruit,
and hairs on the outer surface,
X 30.

Lower. Fruit of I. radicans subsp.
negundo, showing ovule develOp—
ment, X 120.

 

125

 

PLATE XVIII.

124

Cross section of fruit wall,
Toxicodendron toxicarium.

 

a. DevelOping wall showing emerging
trichomes, X 120.

b. Later stage than A, showing mature
exocarpal hairs, X 120.

 

 

 

126

PLATE XXIXI' Longitudinal sections Of nearlyrnature
fruit of Toxicodendron toxicarium.

a. Resin ducts are functional; endocarp
has begun to sclerify; embryo is
elongating. X 100.

b. Tissues of seed coat and endocarp.
X 250.

 

 

127

'1
_I
..

nun!-

-_"."."
‘

 

128

PLATE XX. Meiotic chromosomes in anther of
Toxicodendron radicans subsp. negundo,
X 200.

 

 

 

 

 

150

0,“,

‘1‘! ”’1 5

Hm

I B

Fig. 4. Toxicodendron chromosomes.

A.

 

T.

Gillis 431i.

B. T.
G

toxicarium, Alachua County, Florida,

r dber ii, Alger County, Michigan,
ilIis EEEE.

CHROMOSOME NUMBERS

Chromosome numbers were determined from observa-

tions of pollen mother cells. Toxicodendron Species flower

 

for a relatively brief period each year; hence, it was not
possible to collect cytological material for chromosome
counts in most of the taxa in this group. Counts were made
for I. radicans subSp. negundo, I, rydbergii, and I, tgxi—
carium. In all taxa for which counts were available, mei-
osis appeared regular and the chromosome number was the
same. This figure of 22’= 30 agrees with published counts
of g.= 15 for I. diversilobum (= I. lobatum) by COpeland
and Doyel (1940) and for Rhgg toxicodendron by Grimm (1912).
In the latter case, we have no way of knowing which taxon
was under study; from the context of the article, it might
have been I, radicans subSp. radicans. Because Of the small
Size, relative uniformity in Size Of the chromosomes, and
difficulty in differentiating the kinetochore, karyotype
analyses were not undertaken. In one pair of chromosomes
in I. toxicarium, however, there appear to be satellites.

Material Of Toxicodendron used in cytological in-
vestigations:

Toxicodendron radicans subSp. negundo, Gabriele

Mfihling S.n., Michigan, Oakland County, 5 miles north Of

131

132
Farmington on l2-Mile Road between Halstead and Haggerty.
11 May 1959.

Toxicodendron rydbergii, Gillis 5542, Michigan,
Alger Co., T. 49N., R. 14W., SW’% Sec. 7. Devil's Slides
on high sand dune overlooking Lake Superior. 17 June 1964.

Toxicodendron toxicarium, Gillis 45l5, Florida,
Alachua County, T. 98., R. 19E., Sec. 13, 12 miles NW of
Gainesville, San Felasco area. Turkeyoak-pine Woods.

6 April 1961.

In the citation of Specimens, herbarium abbrevia-
tions in the fifth edition of "Index Herbariorum" (Lanjouw
and Stafleu, 1964) have been used. Those herbaria which
are not included in the Index but have been assigned pro-
visional abbreviations are marked with a pIUS‘(+). The
herbaria and their abbreviations are: Arnold Arboretum Of
Harvard University (A); University of Alberta (ALTA); Mohr
Herbarium, Alabama Museum Of Natural History (ALU); Uni-
versity of Arizona (ARIZ); Arizona State College (ASC);
Cranbrook Institute of Science (BLH); British Museum (Natural
History) (BM); Herbarium Bogoriense, Lembaga Biologi Nasional
(B); Brigham Young university (BRY); Central National Herbar-
ium, Calcutta (CAL); Chrysler Herbarium, Rutgers university
(CHRB); Plant Research Institute, Canada Dept. Of Agriculture,
Ottawa (DAO); Dayton Museum of Natural History (DMNH+);
Dudley Herbarium, Stanford university (DS); Duke university
(DUKE); Royal Botanic Garden, Edinburgh (E); Escuela Nacional

de CienciaS Biologicas, Mexico City (ENCB+); Chicago Natural

133
History Museum (F); Florida Agricultural Experiment Sta-
tion (FLAS); Florida State University (FSJ); Conservatoire
et Jardin Botaniques, Geneva (G); University Of Georgia
(GA); Gray Herbarium of Harvard University (GH); University
of Illinois (ILL); Herbarium, Kew (K); Kansas State Uni-
versity (KSC); Botanical Institute, Kiew, Ukranian S.S.R.
(KW); Rijksherbarium, Leiden (L); Catholic University of
America (LCU); Komarov Botanical Institute, Leningrad (LE);
Linnaean Society (LINN); University of Michigan (MICH);
Public Museum, City Of Milwaukee (MIL); University of Min-
nesota (MIN); Missouri Botanical Garden (MO); Michigan State
University (MSC); Universite de Montreal (MT); University
Of Notre Dame (ND); Greene Herbarium, University of Notre
Dame (ND-G); North Dakota State University (NDA); New York
Botanical Garden (NY); Botanisk Museum, Oslo (0); Oregon
State university (OSC); Fielding Herbarium, Oxford Uhiver-
sity (OXF); Museum National d'Histoire Naturelle, Paris (P);
University of Pennsylvania (PENN); Philadelphia Academy of
Natural Sciences (PH); Rocky Mountain Herbarium, University
of wyoming (RM); Hokkaido University (SAP); South Dakota
State College (SCD); University of South Dakota (SDU);
Botanic Gardens, Singapore (SING); Southern Methodist Uhi-
versity (SMU); Sul ROSS State College (SRSC); National Taiwan
university (TAI); University Of Texas (TEX); Botanical Insti-
tute, Tokyo University (TI); Botanical Museum, Utrecht (U);
University of California, Berkeley (UC); united States Na-
tional Museum (US); University of Manitoba (WIN); university

of Wisconsin (WIS); and Wayne State University (WUD).

TAXONOMIC TREATMENT
Toxicodendron Miller

Toxicodendron Tournefort, Inst. Rei Herb., p. 610.
1700; Miller, Gard. Dict., Abr. ed. 4. 1754, non
Toxicodendron = Toxicodendrum Thunb. in Kongl.
Vetensk. Acad. Nya Handl. 7: 188. 1796 (= Hynaen-
55555 Lam., Euphorbiaceae), nec Toxicodendrum
J. Gaertner in Fruct. l: 207. 1788 (= AllOphylus
L. or Schmidelia L., Sapindaceae).

5555 L., Sp. P1. 1: 265. 1753, in part as to Species

4, 6, and 7. Vernix Adanson, Fam. Pl. 2: 342. 1763.

Rhus-Toxicodendron Marshall, Arbust. Am., p. 130. 1785.

POCOphorum Necker, Elem. Bot. 2: 226. 1790.

Philostemon Raf., Fl. Ludoviciana, p. 107. 1817; Steu-

del, Nom. Bot. ed. 2: vol. 2., p. 320. 1841 as 255;:
ostemum.

Deciduous dioecious trees, shrubs, or woody vines
with poisonous effluvium. Leaves alternate, palmately
ternate or imparipinnate, estipulate, thin, glabrous to
tomentose on the lower surface, glabrous to sericeous on
the upper surface; leaflets Opposite except for the terminal
one. Flowers in axillary, lateral, thyrsoid panicles or

racemes, ultimate clusters determinate of 3 or 4 flowers;

134

135
bracts of the inflorescence lanceolate, deciduous; fruit-
ing inflorescence usually pendent, but occasionally erect
by lignification of the rachis. Sepals 5, persistent,
imbricate in the bud, united at base. Petals 5, ascending,
veined, glabrous, imbricate in bud, ovate, usually with
reflexed tips, smaller in female flowers than in male.
Ovary l-celled by abortion, sessile or partially buried in
the hypogynous disk; style terminal, stigma 3-parted, turn-
ing black with age; ovule raised on an elongate ascendent
basal funicle. Drupes subglobular or laterally flattened,
2.5 - 7 mm broad, 1.5 - 5 mm long, whitish, cream, yellow,
to dun-colored, glabrous or pubescent with simple, non-
glandular hairs, less than 1.5 mm long, the bony endocarp
and mesocarp adhering; mesocarp waxy, striate with black
resin canals; exocarp brittle, papery, separating from the
mesocarp at maturity; remnant of the 5-lobed disk becoming
chartaceous. Stones bony, Often with a few ribbon-like
longitudinal ridges, endocarp fused to testa; endOSperm
absent; embryo large, generally transverse with flat folia-
ceous cotyledons; hypocotyl : elongate, incurved upwards
toward the hilum. Stamens 5, haplostemonous, inserted be-
low a lobed, annular, patellior, cupuliform, intrastaminal,
nectiferous disk, reduced in female flowers; anthers Oblong,
dorsifixed, usually Shorter than theli subulate filaments;
dehiscence longitudinal, introrse. Pollen tricolpate,
finely reticulate, ellipsoidal, with tapering transverse

furrows, endexine thickened internally at the edge Of the

136

colpus forming a rib; mean length 26.6n, mean width 230.
Vessels not clustered in the late wood; resin ducts absent
in the xylem rays; septate xylem fibers present; pith solid,
white. Root ribrous; root hairs dark brown. North America
to northern South America, and eastern Asia.

Type Species; Toxicodendron toxicarium (Salisb.)
Gillis (As 5555 toxicodendron L., Sp. P1. 1: 265. 1753,
lectotype of Barkley, 1937).13

 

l3Barkley (1937) named Rhus toxicodendron L. as
lectotype Species of sect. gggoxicodendron (= sect. Toxi-
codendron) and equated it to Toxicodendron radicans. How-
ever, it is now clear that Rhus toxicodendron of Linnaeus =
I. toxicarium, and therefore this Species must be the cor-
rect lectotype Species of the genus.

KEY TO SECTIONS, SPECIES, AND SUBSPECIES
OF TOXICODENDRON

A. Large Shrub or tree, clonal, but without long stolons; an-
thers conspicuously exerted, filaments filiform.....Section
Venenata (not treated herein).

A. Vine, sub-Shrub, or Shrub with long stolons; anthers in-
cluded, filaments subulate at the base.....................B.
B. Leaves simple......Section Simplicifolia; (one Species),

l. I, borneense.
B. Leaves compound......Section Toxicodendron..............C.
C. Leaflets 5 - 9......2..l. nodosum
C. Leaflets 3...........................................D.
D. Fruits glabrous (or with only occasional hairs);
leaflets coarsely toothed, l-lobed, or entire; if
entire, then tufts of hairs concentrated in vein
axils on lower surface absent.....................E.
D. Fruits pubescent or papillose; leaflets entire,
undulate, or round-lobed; if toothed, then tufts of
hairs in vein axils on lower leaflet surface
present...........................................J.
E. Leaflets with entire, undulate, notched, or
serrate margins................................F.
E. Leaflets with lobed or deeply cut margins...4h.

I, radicans subsp.verrucosum.

137

138
F. Leaflets glabrous to scattered-strigose on
the lower surface, or with pubescence con-
fined to the veins, not velvety to the touch;
glabrous or with Sparse pubescence on the
upper surface; pubescence, if any,
appressed...................................G.
F. Leaflets densely pilose to velutinous on
lower surface, velvety to the touch; pubes-
cence on the upper surface; pubescenceeflect.l.
G. Most leaves made up of entire or subentire
1eaflets....4g. I. radicans subsp. 5535;;-
m.
G. Most leaves made up of notched or serrate
leaflets.................................H.
H. Leaflets suborbicular or broadly ovate;
petiole completely glabrous; plant sub-
shrub or shrub........5. I. rydbergii.
H. Leaflets ovate or lanceolate; petiole
puberulent to densely pubescent; plant

Shrub or vine.....4f. I. radicans

subSp. negundo.

I. Leaflets serrate; terminal leaflet
broadly ovate.....4d. I, radicans
subsp. pubens.

I. Leaflets entire or subentire; term-
inal leaflet elliptic or lanceo-

late. . ASL radicans subsp. barkleyi.

139
Leaflets mostly obtuse or rounded
at apex; fruits setose or minutely
puberulent; tufts Of hairs absent
from vein axils on lower leaflet
surface...........................K.
Leaflets acute or acuminate; fruits
puberulent, hispid, or scabrous;
tufts Of hairs present in vein
axils on lower leaflet surface....M.
K. Leaves pubescent...............L.
K. Leaves glabrous to glabrate..3.

I. lobatum.

L. Leaflets undulate or lobed;

leaflet base cuneate......6.

'1. toxicarium.

L. Leaflets deeply incised;
leaflet base rounded......4i.

.1. radicans subsp. eximium.

M. Widest part of terminal
leaflet blade below center;
tufts of hairs in vein ax-
ils usually not red-brown;
terminal leaflet blade
2.5 - 14 cm long; fruits
globose, 3 - 5 mm broad;
terminal leaflet obtuse or

truncateOOOOOO O. .4.

140
I. radicans subsp. radicans.
M. Widest part of terminal
leaflet blade at or above
center; tufts of hairs in
vein axils usually red-
brown; terminal leaflet

6 - 20 cm long; fruits

flattened, often obliquely

deltoid, 5 - 6 mm broad;
terminal leaflet attenuate
or cuneate................N.

N. Fruits short hairy or
papillose only; terminal
leaflet 1.2 - 1.7 times
as long as broad, cune-
ate.....4b. I. radicans
subSp. orientale.

N. Fruits hiSpid with hairs
nearly 1 mm long; termi-
nal leaflet nearly twice
as long as broad, atten-
uate...4c. I, radicans

subSp. hispidum.

SECTION SIMPLICIFOLIA
GILLIS, SECT. NOV.

Vites lignosae deciduae; folia simplicia, sub-
coriacea; frutus elevatus super vegitigium recep-
taculi et disci.
* * * * * *
Deciduous woody vines; leaves Simple, subcora-
ceous; fruit raised from remnant of receptacle and
disk. Northern Borneo (Sabah).
Type Species:b Toxicodendron borneense (Stapf)
Gillis.
Toxicodendron borneense (Stapf) Gillis, comb. nov.
(Plate XXI).
5555 borneensis, Stapf in Trans. Linn. Soc.
London Bot. 4: 142. 1894.
Type: 5, 5, Haviland llgg, 6000 ft. on Mt.
Kinabalu, Sabah, 19 March 1892 (K) male
flowering Specimen.

Scandent shrub; branches slender, brown, Often bear-

ing prominent lenticels. Leaves Simple, subcoriaceous,
elliptic to Oblanceolate, short petiolate, entire with revo-
lute margins, with 14 - 22 pairs of i_prominent veins; gla-
brous, with acuminate to mucronate tips, base attenuate;

blade 6 - 16 cm long, 2.5 - 6.5 cm broad. Inflorescence a

141

142

PLATE XXI. Specimens of Toxicodendron borneense.

 

Left --Holotype of Rhus borneensis Stapf.
Male flowers at top; female flowers
at bottom.

 

Right-—Fruiting Specimen.

 

11+}

 

4/1. 1 I.

144

lateral paniculate thyrse, 6 - 14 cm long, longer in female
than in male, generally in verticillate clusters with leaves
above and below; flowers clustered Often in threes on short,
minutely bracteolulate pedicels; bracts lanceolate, 1 mm
long, 0.2 mm wide, glabrate, gradually tapering to a blunt
tip, narrower in female than in male; corolla lobes oblong
to Oblanceolate, 2.0 - 2.8 mm long, 1.0 - 1.5 mm broad,
narrower in female than in male, blunt; fertile anthers 1.2
mm long, broadly ovate, filaments 0.5 mm long; in female
sterile anthers 0.5 mm long, subulate. Fruiting inflores-
cence glabrous, erect; fruit obliquely globose-reniform,
laterally flattened, 5.5 - 7 mm broad, 5 - 6 mm deep at its
widest point (through the stigmatic region), glabrous, with
mesocarp striate in a white, waxy matrix; fruit raised from
remnant of receptacle and disk. Distribution: Province of
Sabah (Borneo), Malaysia (Fig. 5).

Specimens examined: MALAYSIA: Sabah; Mt. Kinabalu,
6000 ft., Haviland 1125, (K); 6600 ft., Haviland 1254, (K);
Kamberanga, 7040 ft., Griswold 115, (A); Gunong Alab, Tam-
bunan, 4000 - 5000 ft. (K, SAN). ' ‘

Toxicodendron borneense is the rarest of the Species
in the poison-ivy complex. It has been collected only four
times. On one occasion, the collector noted that his col-
lection came from a fresh branch found on the ground, no
parent plant having been discovered from which it may have
come. The range of variation in the Species is therefore

unknown. Although its type locality is Mt. Kinabalu in the

145

Fig. 5. Distribution of the genus Toxicodendron,
sect. Toxicodendron.

 

 

146

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

147

state of Sabah in northern Borneo, a mountain which has
been visited by several plant-collecting expeditions, I.
borneense has been found there only twice.

This Species shares with all other Toxicodendron
Species the prOperty of possessing resins which darken with
age (presumably poisonous) plus their diagnostic fruit and
flowers characteristics. It differs markedly in having
Simple leaves and verticillate inflorescences. The fruits
appear to be raised on a remnant of the receptacle and disk,
unlike other members of the genus. Because of these unique
characters which set I, borneense apart from the rest of
the genus, a new section of the genus has been erected,

Sect. Simplicifolia.

SECTION TOXICODENDRON KOCH,
DENDROLOGIE 1: 580. 1869

Eutoxicodendron Schneider, Illustr. Handb. Laubholzk.
2: 149. 1907.
Toxicodendron Robinson and Fernald, Gray's New Man-
ual, p. 552. 1908, as subgenus.
55m55 DC., Prodr. 2: 67. 1825, as sect. in part.
Deciduous woody vines or small shrubs with compara-
tively slender branches and tendency to proliferate exten-
Sive rhizomes. Flowers in small, axillary thyrses; stamens
with Short subulate filaments and elongate, included an-
thers; fruits usually pendent. Leaves palmately trifolio-
late or pinnately penta- or septi-foliolate, alternate, with
phyllotaxy 3/8. North America from Canada to western Guate-
mala; Asia: Japan, China, Taiwan, Malaysia, Indonesia, and
U.S.S.R.: Sakhalin and Kurile Islands. V(Fig. 6.)
2. Toxicodendron nodosum (Blume) Gillis, comb. nov.
(Plate XXII).
5555 nodosa Blume,Bijdragen tot de Flora van
Nederlandisch Indie, 17: 1164. 1826. Type:
Java, collector unknown (L, holotype).
555§.perakensis Scortechini ex King in J. Asiat.
Soc. Bengal Pt. 2 Nat. Histo. 65: 500.
1896. Type: Scortechini l555, Perak, Malaya

(CAL- 98286, holotype).
148

Fig. 6.

149

Distribution of Toxicodendron nodosum and
I. borneense in Malaysia and Indonesia.

 

 

 

 

150

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

_,r ———<

1
1

 

  

. tandocum

A T. bornoonoo - H -

 

 

ol
“1
o
o
9
c
,1... 1
o
o
_—J

 

 

 

 

 

151

PLATE XXII. Specimen of Toxicodendron nodosum.
Type of Rhus nodosa.

 

152

   
 

\

14
1 \ TYPE!

IIrrh. LII-II. III-II“.

 

 

“I“ “1721' '1"\

153

Scandent shrub up to 15 meters long with occasional
roots; branches slender, red-brown, often bearing prominent
lenticels. Leaves imparipinnate; leaflets 5-7 (-9),
elliptic-lanceolate to oblong with 14 - 20 pairs of veins
which are :_prominent beneath, entire, completely glabrous
or with occasional hairs, with long acuminate tips, rounded
or truncate to obtuse or attenuate at the base, Short peti-
olulate (petiolules less than 1 cm) or subsessile except
for the terminal leaflet; blade 5 - 15 cm long, 2.5 - 5.5 cm
broad, generally longer in Bornean than Mayalan or Javan
pOpulationS; petiole l5 - 30 cm long, glabrous; petiolule
2 - 4 mm long, glabrous; terminal petiolule 7 - 15 mm long,
glabrous or exceedingly finely puberulent. Inflorescence a
lateral or subterminal thyrse, 15 - 40 cm long; flowers
abundant, tightly clustered on Short, minutely bracteolulate,
glabrate or puberulent pedicels; bracts Oblanceolate, 0.2 -
0.3 mm long and broad, glabrate, deciduous; calyx 0.5 - 1.7
mm long, 0.5 - 1.0 mm broad, narrower in female than in male,
glabrous; fertile anthers 1.2 mm long, broadly ovate; fila-
ments linear-subulate, 0.8 - 1.5 mm long, in female sterile
anthers 0.4 mm long, subulate, filament 1.5 mm, pubescent;
pollen 20.8 - 28.5u long, 17.5 - 22.0p broad. Fruiting
inflorescence puberulent or glabrous, pendent; fruit oblique-
ly globose-reniform, 5.0 - 8.0 mm across, 3.0 - 6.0 mm
broad, glabrous with mesocarp striate in white waxy matrix;
exocarp dun-colored. Buds stalked, naked, mucronate, brown,

covered with brown pubescence. Distribution: Indonesia

154

and Malaysia. At elevations of sea level to 1500 meters
growing in shallow soil on limestone, apparently rare. F1.
Aug. - Dec.; fr. Oct. - Mar. Local name: "jahor" (Burkhill,
1935). Specimens examined: INDONESIA: Celebes; Makassar,
collector unknown (BO). Java: Batavia Tijanten, Backer
25958 (BO); Nirmala, Backer 19522 (BI); G. Batoe, Backer
25552 (BO); Josoredja en Dara, Pekalongan, Backer l5215
(BO); Tjidadap, Bangas, 1000 meters, Bakhuizen 255_555_5;;gk,
515 (L); Preanger Regentschappen, Geger Bentang, 55555, 553.
(B0); Preanger G. Karang, Bakhiuzen ygg,§gg,§;ig§l§§§ (BO);
locality and collector unknown (Type) (L); Sumatra; Batang

BaroeS, Tegsmain, 5.5. (B0); Sekintjau Belirang, 1400 m.,
Rappard 22_(BO).
* * * * * *

MALAYSIA: Perak: Larut, King's collector 5551_
(BM, CAL, L); locality unknown, Scortechini l55§ (CAL);
Tupai, Wig! £95 (CAL, SING); Thaipeng, Wm 5&4 (SING).
Sarawak: Bau, Haviland 1255.(CAL); Bau, Purseglove 4449,
(A, L, SING); Kuching, Haviland §_ m §_3_6_6_1== (BO, L); Mt.
Buan, Haviland 1252 (SING); Tambusan, HM 12;;9 (SING); Mt.
Dulit near Long Kapa, 400 m., Richards l45§ (A, L); Stetapok
F.R., 4 miles south of Kuching, limestone hills, Anderson
3.25131 (MSC).

Toxicodendron nodosum is described (as,§55§.pg£-
akensis) by Corner (1940) as being like one of the "poison
ivies of North America." Like poison-ivy, it is a vine or

climber; unlike poison-ivy, it has five to seven leaflets.

155

Its inflorescences are longer than those of poison-ivy. As
in the case of other members of this complex, I, nodosum
has been referred to the genus Eggs by various authors. It
shares with all other Toxicodendrons, however, smaller pol-
len than that of Eggs species, toxic effluvium, absence of
glandular hairs, buff to tan exocarps, striate mesocarps,
and chiefly axillary inflorescences. Some flower clusters
OCCUpY a sub-terminal position, but in all specimens that
have been examined, there is a vegetative shoot continuing
beyond the inflorescences, continuing terminal growth. Be-
cause of its scandent habit, anther size, shape, and posi-
tion, I, nodosum has been placed in section Toxicodendron
of the genus.

Collections made in former Dutch possessions have
been called Eggs nodosa, while those collected in regions
of the British Commonwealth more likely have been named
B. perakensis. On the basis of herbarium material examined,
it is not possible to-distinguish between these two taxa.

This is a rarely collected plant, its rarity prob-
ably due more to inaccessibility than to actual lack of
abundance. It is found in the mountainous areas of northern
Perak in the Malay Peninsula, on limestone ridges bear Bau
and Kuching in the state of Sarawak in Borneo, and scattered
throughout Sumatra, Java, and Celebes. Presumably it grows
immIndonesian Borneo, but no collections from this region
are known. It has been collected only six times since 1920.

Label data on herbarium Specimens are generally scanty, so

156
that it would not be possible to return to the exact site
of a previous collection.

Recent collections were desired to study fresh seeds
inasmuch as some important diagnostic characters are found
in the fruits of this genus. The Purseglove 4449 Specimen
indicates that the seeds were "red, drying buff," yet the
fresh seeds suggest that the exocarps could hardly ever
have been red. They certainly are buff, and show the plant
to be a Toxicodendron. (Acknowledgment is made of Mr.

J. A. R. Anderson, Conservator of Forests in Kuching, who
kindly sent fresh material of this Species for study.)

Like many other Toxicodendron species, I. nodosum
seems to occur chiefly in limey areas. It is generally
found in shallow, rocky soils in Open jungle, climbing
trees up to 15 meters. Its ecology is incompletely known
due to the paucity of collections and collection data,
particularly in recent years. Backer and Van Den Brink
(1965) suggest that the plant in Java occupies the wetter
sites (up to only three dry months per year, where rainfall
greater than 100 mm per month = "wet") at altitudes of
900 - 1200 m. They, in fact, label it as an "everwet cli-
mate" indicator. Flowering season for the plant seems to
vary from July to December, possibly just prior to onset
of monsoon seasons. It appears not to be a heavy seeder.

Backer and Van Den Brink (ibid.) further describe
the petiolules as being winged. From all Specimens I have

examined, I would describe this condition as the attenuate

157
nature of the terminal leaflet base, rather than a true
"winged" nature of the rachis as in Rhus javanica and R.
c0pallina. In the latter the winged condition extends
from the terminal leaflet to the base of the lowest leaf-
let; in Toxicodendron nodosum, it is merely restricted to
the base of the terminal leaflet.

From a plant grown from seed in the laboratory,
one can see that the young seedling has simple leaves for
an appreciable period of time before any compound leaves
appear, well after the stem has begun to lignify. The
first compound leaves appeared on one such laboratory?
grown plant only after 11 simple leaves were produced.
Throughout this time, the cotyledons also remained on the
plant and remained green.

3. Toxicodendron lobatum (Hook.) Gillis, comb. nov.
(Plates man and XXIV).
Rhg§_lobata Hook., Fl. Bor. Am. p. 127. 1830,
non Poiret, Ency. Meth.I§hg§ varielobata
Steudel, Nom. Bot. 2: 452. 1841 (nomen

nudum) Suppl. 5: 264. 1804. Type: Douglas

§,n. British Columbia, Port Vancouver, June,

common on the outskirts of woods in dry soils

(K, holotype).
Rhus diversiloba T. and G., Fl. N. Am. 1: 218.

1838. Rhus toxicodendron subsp. diversiloba
Engler in DC. Monog. Phaner. 4: 395. 1883.

Rhus toxicodendron var. diversiloba Brandegee,

 

158

PLATE XXIII. Specimen of Toxicodendron lobatum.
Type of Rhus lobata Hook.

 

 

 

 

   
  
  

 

 

o-.‘_.-r:.4:.v .41.:
_P..__._
J -A
l).< In 0.“ I " 341w

mama“. nuu Lhwilmvw nun-lu-

”'31

 

.-

I 1.1. A;

PLATE XXIV.

160

Specimen of Toxicodendron lobatum.
Holotype of I. oxycarpum.

 

 

 

161

-.-— umx
I HUMIJLADNIV OH\ (H nun Arm ml
III!" A". A-JJD \ Axum 'A

5.51;“: ugg. g £=1gt§klcb.1’l 3: 5..
erm.
TYPEx c d r - ;
‘ yet. fins. 5 Crit. : 131. 1905.)
‘ I
’:14(4 ’4‘ ”7 ”5]”

A”.-

w... .0 ..

 

NICO q in a‘t-Ke 1 Unnvuuly
Em hnu 0,10th

Noe-an qunblifllm

7X3 .

44”. 1 5, l, 2 q

«at: 572;“

'x 29222::

V, .9

‘ gim:\?§/

WTID ITAYE. NATDONAL "WARM

l-n « xuxlnn
/

   

lea

'o

Ilwulr'lt Iv nu mrunn

TWW 0“" MM» ah“, 7v)!

ITER HOREAIJ-AMEIHCANI'M. In“.

In; nmum- collinn (‘nlifurnim «ua-i-Iz-nulu

Rn! Wood Faint. Saul" (
.l. HALL.

lit-hut Jih-

162
Zoe 2: 345. 1892. Toxicodendron diversilobum
(T. & G.). Greene, Leafl. Bot. Observ. Crit.
l: 119. 1905. Type: Nuttall, §,g, "Cali-
fornia," (NY, holotype).
Toxicodendron comarOphyllum Greene, Leafl. Bot.
Observ. Crit. l: 120. 1905. 'Type: Palmer
35, California, San Diego Co., Tighe's near
San Diego, 1875. (US-19802, holotype; F-
302931, isotype; NO-1773737, possible isotype).
Toxicodendron oxycarpum Green, Leafl. Bot. Observ.

Crit. 1: 121. 1905. Type: .l- Ball £13.,

 

California, Santa Cruz Co., Redwood Forest,
Santa Cruz, 10 July 1894. (US-292229, holo-
type).

Toxicodendron iSOphyllum Greene, Leafl. Bot.
Observ. Crit. 1: 121. 1905. Type: ‘gghg B.
Leiberg 3111, California, Riverside Co.,
river banks near San Jacinto, 600 meters eleva-
tion, 9 March 1893. (US-342019, holotype).

Toxicodendron dryOphilum Greene, Leafl. Bot.
Observ. Crit. l: 121. 1905. Rhus diversiloba
f. radicans McNair, Publ. Field Mus. Nat. Hist.
225. Vol. 4: 61. 1925. Type: Mg. 5. 111.
Austin §,g,, Butte Co., climbing oak trees
in Little Chico Canyon, May 1896. (ND-G-
22098, holotype; MO-l773736, US-28527, iso-

types).

163

Toxicodendron vaccarum Greene, Leafl. Bot. Observ.
Crit. l: 122. 1905. Type: Mn.§-.§é£§£.éfl§
.Fragk Nutting §,g,, 23 May 1894. (ND-G, holo-
type; MO-lll4683, photograph of type).

Rhus diversiloba f. quinquifolia McNair, Bull.
Torrey Bot. Club 63: 476. 1936. Type:
A. Wynn 1866 (NY, lectotype).

Vine, shrub, or rarely tree, with slender, glabrous
to puberulent branches, usually with brown aerial roots;
rhizome creeping; leaves crowded on branches. Bark gray-
brown to red-brown on new wood, becoming gray on older wood;
leaf scars V- or U-shaped with 6 - 10 bundle scars; lenticels
usually conspicuous. Beds stalked, naked, mucronate, brown,
covered with a brown wooly pubescence. Autumn coloration
scarlet to red-violet. Leaflets 3, occasionally 5 (-ll),
ovate, oblong, or suborbicular, entire, undulate, crenate-
dentate or bluntly lobed; apex obtuse or rounded; base
rounded, truncate, or subcuneate; lateral leaflets mostly
inequilateral, 1.0 - 7.0 cm long, 1.0 - 6.0 cm broad; short
petiolulate or subsessile; terminal leaflet 0.8 - 13.0 cm
long, 0.3 - 8.0 cm wide, petiole 1.0 - 10.0 cm long; terminal
petiolule 0.1 - 2.6 cm long; leaflets attached in palmate
fashion when trifoliolate, pinnate when more than 3 leaflets;
glabrous above, occasionally with a line of minute, curly
hairs along the midrib, sub-pilose below or with trichomes
restricted to appressed hairs on major veins. Leaves often

mottled when dry. Inflorescences lateral paniculate thyrses

 

 

. lllfi,
"u|,..‘ . )Z. ., _
..

 

164

up to 1 dm long; pedicels 2.0 - 8.0 mm long; bracts oblanceo-
late, 1 mm long, 0.3 mm wide, glabrate, ciliate, deciduous;
sepals 5, yellow-green, deltoid-ovate, 1 mm long, 0.5 mm
broad; filaments linear-subulate, 1.0 - 1.8 mm long. Drupe
cream-colored, globose-reniform, sometimes obliquely deltoid,
sometimes laterally flattened, 4.5 - 7.5 mm across, 1.5 -
5.5 mm broad with densely setulose exocarp becoming char-
taceous Upon ripening and readily separating from the neso-
carp; mesocarp black striate in white, waxy matrix; endocarp
dun-colored, bony.

Distribution: Northern Baja California, Western
California, Oregon, and Washington, Vancouver Island, British
Columbia. Common name: *Western poison-oak. (See Fig. 7.)

Representative specimens. united States. California.

 

Alameda Co.: Berkeley, Greene, 3,3. (ND-G); Oakland, Eggw,
s13. (F). Amador Co.: locality not cited, Hansen, §,g.

(MO, ND-G, US); Ione, Eastwood 19152 (A). Butte Co.:

Little Chico Creek, Austin, §,g, ND-G, MO, US); Big Chico
Creek canyon, Heller 11211 (A, F, GH, MO, MIN, NY, OSC, ND-G,
US); east of Chico, Stilson Canyon, Heller 12221 (A, F, MO,
NY, OSC, US, WIS); Oroville,1311g£M1QZ§Z (A, F, GH, MO, NY,
US). Chico, Palmer 2QQQ_(NY). Calaveras Co.: Angels Creek,
Eastwood 11622 (A); Calaveras River, Gibbes, Egg, (NY).
Colusa Co.: mountain slopes in chapparral, Gillis 4842

(MSC); Colusa, Eastwood 19226 (A). Contra Costa Co.: Mt.
Diablo, Brewer 1968 (GH); Clayton, Brewer 1QQ§_(US); DelNorte

Co.: Gasquet, Parksug Tracy 11188 (GH, US). El Dorado Co.:

165

Fig. 7. Distribution of Toxicodendron lobatum.

 

 

167
east slope of densely vegetated stream bank, 4 miles NE of
Aulrun, Holloway 41 (ARIZ). Glenn Co.: Stonyford along
Stonyford Creek, Weston, 212. (MSC); Elk Creek, near Stony-
ford, Gillis 4222 (MSC). Humboldt Co.: Humboldt State
Park, Weston, 2,2, (MSC); Willow Creek, 11222_124Q4 (MO,
US); Blacksburg, 11221.12221 (GH, NY). Jackson Co.: "Walker
Creek, Applegate 2222 (US). Kern Co.: Fort Tejon, DeVasey
12_(GH, NY, US); Havilah, Grinnell 222 (US); Caliente Canyon,
Zigmond 122 (F). Lake Co.: 1.1 miles east of Middletown
on road to Lower Lake, serpentine soil, Morlex 222 (CAL,
MSC); Mt. Konocti, Blankinship, 2,2, (MO); Middle Creek,
6 miles south of Elk Mountain, between Lake Pilsbury and
Lower Lake, Everettig'22112.22222 (NY, RM). Los Angeles
Co.: Santa Catalina Island, Millspaugh 4124 (F), Avalon
Canyon, 22112.2222_(F, US); Hamilton Canyon, Fosberg 24221
(A, F, MO, NY, US); Swain's Canyon,Nuttall 222 (F, US);
Claremont, Whited 222 (ND); Topango Canyon, Santa Monica
Mountains, Egling'é Ellison, 212. (OSC); Pasadena, 22222
2222 (A, MSC, MO, NY, US); Mandeville Canyon, Santa Monica
Mts., Egling, 212. (A, MO, OSC, US); Temescal Canyon,
Pacific Palisades, Hastings, 2,2, (NY); Bouquet Canyon,
San Gabriel Mts., Templeton 1122 (NY); Eagle Rock Valley,
22112.4224 (F, US). Madero Co.: North Fork, Griffiths _221
(NO, US). Marin Co.: Taylor State Park, Gillis 4222 (MSC);
Reyes Point, Gillis 4122 (MSC); 6 miles SE of Lagunitas,
west of Fairfax, Gillis 4121 (MSC); Rose Valley, Eastwood,
‘222. (A); Tamalpais, Kountze 22142.(NY); Mill Valley,

168
Eastwood 2222 (A, NY). Mariposa Co.: Mariposa, Congdon,
_,2, (MIN); Sherlocks, Congdon, 2,2, (MIN). Mendocino Co.:
Mendocino, 21222.122_(F, MIN, MO, NY, US); Glen Blair,
Eastwood 2 Howell 2222 (A); Elk Creek, Leiberg 4122_(US).
Monterey Co.: Salinas Valley, 12221 22 (F, US); Tassajara
Hot Springs, 2122£_2112 (MIN, MO, US); Carmel Highlands,
.222. 2222 (MIN); Big Sur, Eastwood 14222 (A); Pacific Grove,
Demaree 12212 (NY). Napa Co.: Calistoga, Greene, 2,2, (ND-G);
Napa Valley, Torrey 14 (NY); Mt. St. Helena, Eastwood 2221
(A). Nevada Co. 120° 39' w. Long., 39° 20' N. Lat., £91-

lector unknown (ND-G). Orange Co.: Santa Ana Mts., Rice,

 

2,2, (MSC). Riverside Co.: San Jacinto, Leiberg 2111 (US).
Sacramento Co.: Sacramento Pocket Road, COpeland, 2,2, (MSC);
Folsom, Copeland 1222 (GH). San Benito Co.: Tres Pinos,
Eastwood 2222 (A, US); The Pinnacles, Eastwood 2122 (A).
San Bernadino Co.: San Bernadino, Wright 221_(GH). San
Diego Co.: San Diego, Palmer 42 (G, GH, MO, US); Cuyamaca
Peak, Hitchcock, 2,2, (US); Alpine, Mearns 4212 (NY, US);
Descamso, Eastwood 2212 (GH). San Francisco Co.: San
Francisco, Lincoln Park, Eastwood, 2,2. (US). San Luis
ObiSpo Co.: Morro Bay, Eastwood 14224.(A); Avila, Eastwood
12122 (A); Cambria, Eastwood 12212 (A). San Mateo Co.:
Higgins Creek Road, east of Half Moon Bay, M2222,'222.
(MSC); Corte Madera Creek, Wiggins, 2,2, (MSC); San Carlos
River, Eastwood 12414 (A). Santa Barbara Co.: 21222.2242
(F, GH, GH, MIN, MO, US, NY); Carpinteria, Dickover, 2,2,
(US); Santa Cruz Island, Eastwood 24Q2_(A); vicinity of

1‘

169
Smuggler's Cove, Santa Cruz Island, Abrams 222 Wiggins 224_
(A, GH); Montecito Valley, Bingham 142 (KSC). Santa Clara
Co.: Stanford University, McMurphy 22 (MIN, MO, ND-G, NY,
OSC, GH, US); Gilroy Hot Springs, Eastwood 222,Howe11 4224_
(A); Clarita Vineyard, Black Mts., 21222,'222. (MIN, MO,
NY, OSC, US); Los Gatos, Heller 1221 (A, F, MO, NY, US, GH);
Beels Station, Goldman 1222_(US); New Almaden, Torre! 14.
(GH, NY); Saratoga Springs, Leeds, 2,2, (F); San Jose,.2222-

 

2222 11222 (A); Mt. Hamilton, Eastwood 12422 (A); Mayfield,
Wiggins 2221 (A); west slopes of Sierra Nevada, east of Palo
Alto, Wiggins, 2,2, (MSC); bank of San Francisquito Creek,
Palo Alto, Wiggins, 2,2, (MSC). Santa Cruz Co.: Santa
Cruz, Ball, 2,2, (US). Shasta Co.: SE slope of Long Valley

 

Mt., Griffin 1142_(MSC); gulch north of Pit River, Griffin
1122_(MSC); Cow Creek Mts., 22222_222 Nutting (ND-G). Sierra
Co.: locality unknown, Lemmon 1222 (GH). Siskiyou Co.:
Hornbrook, Klamath River, Abrams 2222 (M0, RM); Shasta Pine,
Butler 1224 (E); Shasta Springs, Eastwood 2224_(A). Sonoma
Co.: Petaluma, Congdon, 2,2, (MIN); Geyserville, Torre ,

2,2, (NY). Sutter Co.: Yuba City, Marysville Buttes, Ferris
221 (MIN, NY). Trinity Co.: Tangle Blue and Scott Mt. creeks,
Howell 12826 (A). Tulare Co.: Eshom Creek, Clemens 2,2, (RM);
Kaweak River, Woolsey, 2,2, (GH). Ventura Co.: Foster Park,
Eastwood 2212 (A); Sulphur Mt., Abrams 2gMcGregor 42 (A,

ARIZ, GH, NY, OSC, US). Yolo Co.: chamise chapparral along
Buckeye Creek, near Dunnigan, Gillis 4222 (MSC); Putah Canyon,
Dickson 24.(OSC);Wintere, Eastwood 14222'(A)). Oregon.

v'.

 

 

170
Benton Co.: Corvallis, Steward 6863 (NY, 080); McFadden
Marsh, north of Bellfountain, Steward 7011 (NY, OSC);
MacDonald Forest, Ishimoto 3077 (OSC). Clackamas Co.:
Hills, Mulford, 2,2, (GH, MO, NY); Coos Co.: Coos Bay,

House 4746 (US). Douglas Co.: Umqua River, Engelmann'2

 

 

Sargent, 2,2, (A, MO); 5 miles south of Roseburg, Gillis
4122_(MSC); between Tiller and Drew in foothills of Cascades,
1200 feet, Gillis 4122 (MSC); Peel, Agplegate 2122 (US);
Oakland, 22221,44I(A); near Tyee Mt., collector unknown
(OSC). Gilliam Co.: Blalock along Columbia gorge, 22222
12112_(WTU). Grant Co.: Blue Mts., Hooker, 2,2, (GH). Hood
River Co.: between Hood River and Mosier, Suksdorf 2122 (A);
Mt.Hood, Sava e, Cameron, 2_Lenocker, 2,2, (MO); Hood River,
floodplain of Hood River, Gillis 4246 (MSC); 2 miles south
of Hood River, Gillis 2221.(MSC). Josephine Co.: Grants
Pass, Jack, 2,2, (A); Kerby, Thompson 4222 (MO, US). Lane

 

Co.: Cheshire along highway 36, Gillis 4122_ (MSC); Mapleton,
Gillis 4122 (MSC). Marion Co.: Jefferson, south lepe of
Knox Butte, Gillis 4121_(MSC); Salem, near Morningside Grade
School, Matthews, 222. (OSC). Multnomah Co.: Rocky Butte,
collector unknown (F); Bridal Veil, 22112.2111, (F, NY);
Portland, Walgole 44 (US). Wasco Co.: the Dalles, Gillis
4222_(MSC). Washington Co.: Scroggins Creek, Thompsont2fi2.
(MO). Uncertain locality, 2211.22 (A, GH, F, MIN, MO).

Washington. Clark Co.: % mile east and Z mile north of

 

Crawford, Gillis 5274 (MSC). Cowlitz Co.: Kalama, Gillis

 

171
4222 (MSC); Kelso, Palmer 21221 (A, MO). King Co.: Seattle,
Alki Point, Gillis 4221 (MSC); Mercer Island, collector 22-
,22222 (F). Kitsap Co.: Orchard Point, 2122;, 2,2, (A).
Klickitat Co.: terrace of Klickitat River near confluence
with Columbia, Gillis 2222 (MSC). Pierce Co.: between
Steilacoom and West Tacoma, Gillis 4222 (MSC).

******

CANADA: British Columbia. Fort Vancouver, Dou las,

 

.2.2. (K); Vancouver Island, Anderson, 2,2, (GH).

******

MEXICO: Baja Californig Norte: Ensenada, Wiggins 2

 

Gillespie 2221 (A, GH, MO, NY); El Rosario, Brandegee, 2,2,
(GH).

Toxicodendron lobatum is known by most botanists as
2222 diversiloba or Toxicodendron diversilobum. Its first
binomial was 2222 lobata Hooker, but this name was a later
homonym of Poiret'sl2222 lobata, a non-toxic sumac from the
Canary Islands. Therefore, in searching for the next epithet
at the Species level, one finds 2222 diversiloba to be
satisfactory. When the name is transferred to Toxicodendron,
the epithet lobatum has priority, although it has not here-
tofore been used. It is labeled "plentiful" on the type
Specimen from Vancouver, but the Species is rare to the point
of being almost unknown in British Columbia today. Uhder
its name'2222I21versiloba, the plant is occasionally found
in print asi2. diversifolia, a miSprint probably first pub-

lished by Engler in DeCandolle (1883, p. 305).

172

Greene did not Specify where his type for Toxico-
dendrum comarophyllum was located, but most of the type
specimens for poison-ivy and poison-oak taxa he described
are at the Smithsonian Institution; therefore, the Specimen
cited as the type is the one located there. A.probable
isotype at the Missouri Botanical Garden has a different
collection number, but presumably come from the same gather-
ing; its label data are the same. Jepson (1936) clarified
the collection Site by indicating that it was "at the foot
of Ballena grade, San Diego."

In regard to Toxicodendron oxycarpum, Greene places
emphasis on the wrinkled nature of the fruits, but this is
a common phenomenon of herbarium Specimens bearing unripe
fruits; when they have not yet reached full growth, the
exocarps wrinkle in the drying process. Greene's notation
that the branching of the inflorescence is Sparse certainly
calls attention to a chance character of this single speci-
men and refers to a character of no genetic Significance.
Likewise, the importance Greene attaches to the pendulous
nature of the fruits is probably not significant.

Greene's choice of specimens for his Toxicodendron
iSOphyllum leaves much to be desired. His description and
the type Specimen do not agree in most characteristics.

The leaflets are not "all three sessile”; there is no epicarp
present from which to determine that it is "muricate-punctate"
and "pubescent." Perhaps it was available when Greene de-

scribed the Species.

173

As with other types, Greene did not indicate the
location of the type of Toxicodendron dryOphilum. Repre-
sentatives of the type collection are at Notre Dame (Her-
barium Greeneanum) and at the Smithsonian. He wrote "type"
on the Specimen at ND-G even though part of the description
("reported to sustain itself to the height of 20 feet on
oak trees") comes from the label data present only on the
Specimen at US. I have chosen the Notre Dame specimen as
the lectotype. Greene's description "never panicled fruits
seldom more than 2 or 3 to each raceme" is not borne out
by the type collection. The "wrinkling of the epicarp,"
which seems to have been firmly implanted in Greene's mind
as significant in this and several other Species of the
complex, is typical for dried immature fruits.

The type collection of Toxicodendron vaccarum is
indicated as having come from Shasta County, California,
from the Cow Creek Mountains. In order to locate the site
of collection more precisely, I have contacted several per-
sons in California, especially in Shasta County, to ascer-
tain the precise location of these mountains. No one has
been able to give me any information about any mountains of
that name in Shasta County. Such mountains are known in
Mendocino and Lake Counties, but not in Shasta County.

McNair's name 2222 diversiloba f. radicans poses
some perplexing problems. He indicated that it was syn-
onymous with Toxicodendron dryOphilum Greene; therefore it

Should have the same type as 1, drxoghilum, but he proceeded

174
to write on the Specimen (MillSpaugh 4124) at the Field
Museum "forma radicans McNair type." Although he indicated
no type in his publication, he apparently intended the
MillSpaugh Specimen to be a representative one. He was
attempting to recognize nomenclaturally a fonm of western
poison-oak which had the climbing habit, stating that "taxo-
nomic recognition as a form may be desirable" (McNair, 1925).
The ability to climb appears to be general in this species,
so that the climbing forms are not genetically different
from suffrutescent ones and therefore deserve no nomencla-
tural recognition.

McNair later (1936) gave a form name to a five-
1eaflet specimen of Western poison-oak. A.type specimen
again was not cited. Of the seven specimens which he lists
as having the five-leaflet trait, I have chosen the first-
mentioned as lectotype: .2222, 2,2, 1866. This form also
deserves no segregate name since there is variability from
branch to branch even on the same shrub or vine. 'Western
poison-oak, even more than its eastern counterparts, may
frequently have additional leaflets, the author having seen
up to eleven (Gillis 4214, Glenn Co., California, MSC-
165725). Barkley (1936) haS also described a five-leaflet
poison-ivy, but did not give it a name. Vines (1943) did
assign a name to a five-leaflet form (2222 toxicodendron
var. multifolia) which he later abandoned (see discussion

under 1, radicans subSp. 2ubens).

175

‘Western poison-oak was not among the plants mentioned
by the early eXplorers of the American West. Although the
Spanish padres and Conquistadores described the wild life
they encountered in California in the fifteen and sixteen
hundreds, poison-oak is not among the Species discussed.
It is difficult to imagine that this plant could have been
as abundant as it is now, and yet escape detection, particu-
larly in view of its toxic prOperties. A possible conclu-
sion is that the plant was less common several centuries ago.
Being a plant which increases in abundance with disturbance,
it is likely that.1. lobatum has become more abundant with
the activities of man in clearing the countryside, eSpecially
changing fire frequency. Its distribution in Baja California,
where men's activities have been relatively small, is still
Sparse, although in that region low rainfall also plays a
significant role in its scarcity.

The distribution of the plant is throughout much of
California west of the Sierras and the desert. In Oregon
and Washington, it is channeled between the Coast Ranges
and the Cascades. It is known from only one station on the
east flank of the Cascades where James R. Griffin collected
it for me in Shasta County, California, on the southeast
lepe of Long Valley Mountain and on SlOpes above the Pit
River. Local lore in southern Oregon has it that neither
poison-oak nor rattlesnakes occur east of the Rogue River.
The highest elevation in the Cascades from which Tgxicoden-

dron lobatum has been collected is in Douglas Co., Oregon,

176
between Tiller and Drew, at 1200 feet.

Like its counterparts elsewhere in the United States,
Toxicodendron lobatum is highly variable, especially in
leaflet shape. Its leaflets may be as deeply lobed as are
the leaflets of male clones of 1, toxicarium, thus resembling
the leaves of Quercus 2122, but more often they resemble
the leaves of one of the western oaks, such as 2, agrifolia.
At times, they are even without any evident lobes at all
and may be only undulate in leaflet margin. I have been
unable to correlate geographical or regional factors with
variation in leaflet shape except that sun leaves are nor-
mally much larger than Shade leaves (Sic). The fruits are
among the largest of Western Hemisphere Toxicodendron Spe-
cies, with deeply furrowed striations in the mesocarp and
short, abundant trichomes on the exocarp.

Toxicodendron lobatum is far less common in Washing-
ton than farther south. In some counties; it is unquestion-
ably scarce. On disturbed Sites in the Seattle area (Alki
Point, for example), it is abundant enough that municipal
authorities have seen fit to erect a Sign warning the pOpu-
lace that the hillside is covered with poison-oak, rather
than clear away the menace.

In the Columbia River gorge, the ranges of Toxico-
dendron rydbergii and 1, lobatum overlap. Intergrades are
known from Hood River to The Dalles from the washes and
rills of the basaltic hills through which the river has cut;

on the Washington Side of the river, such intergrades are

177

known from the confluence of the Klickitat and the Columbia
east to the vicinity of the John Day Dam. The zones of over-
lap in range and apparent hybridization between these two
Species seems to cover about 23 miles along the Oregon Shore
and about 43 miles along the Washington Shore of the Columbia.
It was from this region that the type Specimens of 1, 22212-
2222 and 1,lobadioides came, presumably from the same popu-
lation (one flowering and one fruiting). In naming these
Species, Greene recognized that they were different from
other named taxa, but did not realize that they might be
hybrids, which they probably are (see discussion under sec-
tion on Hybrids, Crosses, and Intergrades).

Habitats for Toxicodendron lobatum are nearly as
catholic as those of any other taxon in this complex. Jepson
(1936) described its range of habitats in detail:

In San Diego County it iS associated with
characteristic woody species of cismontane Southern
California, all of which are wholly different from
the associated Species in various northern regions,
as the Sierra Nevada foothills or Vaca Mountains,
these in each case being'in the main different from
each other yet again from the associated Species in
the case of the Redwood belt. It has a wider geo-
graphic range than any California shrub and grows
under a greater variety of soil conditions than any
other; though usually preferring good soil or rich
loans, it grows in blue adobe, in saline soils, in
gray clays, in sandy flats, in heavy gravel deposits
and in the crevices of outcrOpping rock piles. It
is also adapted to a greater range of rainfall and
temperature than any other California shrub and is
eSpecially remarkable for its extreme Shade toler-
ance and its tolerance of intense insolation. In
common with all other woody vegetation it is absent
from wide areas on the inner foothills and the plains
of the Great Valley. Its preferred habitats are
moist SlOpes in the hills; in such situations it
is seldom lacking. It is also curious to add that

178

of all native woody vegetation about the great
Keswick smelter on the upper Sacramento River, the
species least injured by smelter fumes was Rhus
diversiloba; at a given station near the smelter

it lived in spite of the conditions, while practical-
ly all other Species were seriously destroyed or in
the main destroyed.

In number of individuals Poison Oak exceeds, in
our judgment, any other shrub Species in California.
While the vast colonies of Adenostoma fasciculatum
or the mesa growths of Larrea tridentata are impres-
sive to the eye, Rhus diversiloba occurs over a much
larger area than either. It 15 often an unobtrusive
Shrub--scarcely noticed in many formations and yet
the individuals are really very numerous. 'When,
for example, a Mendocino hillside wooded with Arbu-
tus menziesii, Quercus dou lasii and various Shrubs
'TS—cleared, the reSponse 0% Poison Odcby crown-
Sprouting tells the story of its abundance. It
grows in the Openly oak-wooded foothills, in Douglas
Fir forest, in the Redwood belt, along river banks
and "bottoms" and even in salt domes, often in chap-
arral and rarely in chamisal. In altitudinal distri-
bution the Species exists from sea-level to about
5000 feet. In the southern Sierra Nevada it grows
aS high as 4760 feet, and it is not found on the
Yosemite Valley floor, but occurs in Hetch-Hetchy and
on the slopes of Francheria Mt. above Hetch-Hetchy
at about 4500 feet; in Butte Co. it is found as high
as 3800 feet.

 

 

No voucher collections have been seen that substanti-
ate some of the extremes to which Jepson alludes, but suf-
fice it to say, Toxicodendron lobatum is very abundant. It
is found through central and western California and in the
Willamette Valley of Oregon. It is less common, even scarce,
north of the Columbia River except for very local pOpulations.
The disturbance due to the activities of EurOpean settlers
in develOping California may account for the tremendous in-
crease in abundance of this plant.

Contrary to the report of Jepson (122,'212.), I have

found that it is now quite abundant, at least locally, in

179
the chamise Chaparral. It grows well with Quercus douglasii,
Ceanothus 222., and 21222 sabiniana. It also occupies
ruderal environments especially those susceptible to fire,
provided some moisture is available during the long, dry
summer. In Oregon it is abundant in almost any lowland or
mesic woods in the Willamette Valley and at elevations up
to 1250 meters in the Cascade foothills. Its common associ-
ates are Pseudotsuga menziesii, Quercus ggrryana, and
Berberis aquifolium.
4. Toxicodendron radicans (L.) Kuntze, Rev. Gen. P1.

Pt. 1, p. 153. P1. 26, fig. 2. 1891. 2222 radicans

L., Sp. Pl. 1: 266. 1753.

Vine, shrub, rarely an epiphyte or tree, with slender,
glabrous to puberulent branches, often with brown aerial
roots; rhizome creeping. Leaflets 3 or rarely 5 (-7), ovate
to elliptic, entire or irregularly serrate, dentate, lobed,
or deeply incised; acute to acuminate or cuSpidate at apex,
glabrous to puberulent above, often with minute curled hairs
on upper mid-vein near the base, glabrous to velutinous below,
often with tufts of hairs in main vein axils below; petiole
2 - 20 cm long; lateral leaflets mostly inequilateral, 2.5 -
18 cm long, 1.0 - 10.0 cm broad, truncate to obtuse, sub-
cuneate or attenuate; petiolules less than 0.1 - 1.2 cm
long; leaflets often subsessile; terminal leaflet blade 3.0 -
20.0 cm long, 1.5 - 13.0 cm broad, truncate, obtuse, subcune-
ate, or attenuate at the base, petiolules 0.3 - 6.0 cm long;

leaflets attached in palmate fashion when leaf is trifoliolate,

180

pinnate when leaf is composed of more than 3 leaflets.
Inflorescence a lateral paniculate thyrse, Up to 1 dm long;
pedicels pilose, 2 - 5 mm long; bracts deltoid to lanceo-
late, 0.7 - 10 mm long, 0.5 - 3 mm broad at the widest point,
glabrate, ciliate, deciduous; sepals 5, deltoidovate, 1 mm
long, 2 - 4 mm broad, glabrate; petals 5, cream to yellow-
green, pinnately dark-veined, Oblanceolate, glabrous, 2 - 5
mm long, 2 - 5 mm broad, reflexed in male, recurved in
female, larger in male than in female, anthers lanceolate,
l - 1.5 mm long, 0.5 - 1.1 mm broad, filaments linear-
subulate, 1.3 - 1.8 mm long. Drupe hard, cream, yellow,
or tan, globose-reniform, sometimes laterally flattened,
0.25 - 0.7 cm across, 0.15 - 0.65 cm broad, with exocarp
glabrous to bristly, becoming papery upon ripening, readily
separating from the mesocarp, later deciduous; mesocarp
black-striate in white, waxy matrix; endocarp dun-colored,
bony, Bark gray-brown to red-brown on new wood, becoming
gray on older wood; leaf scars V- or U-shaped with several
bundle scars; lenticels usually conspicuous. Buds stalked,
naked, mucronate, brown, covered with a brown wooly pubes-
cence. Autumn coloration yellow to orange, occasionally
red or bronze. Common names: poison-ivy, markweed, poison-
mercury, herb 5 la puce, cow-itch, Gift-sumach.

Toxicodendron radicans is one of the two Species of
poison-ivy of the New World and the tsuta-urushi of Eastern
Asia. It is the most common and most widespread Species in

the Anacardiaceae. It is distributed from southern Canada

181
to western Guatemala, the eastern third of the United States
and throughout Mexico, on Bermuda, and in the western Ba-
hamas, in Japan, western and central China and Taiwan, the
Kurile Islands and Sakhalin of the U.S.S.R.

The plant is primarily scandent, but may also grow
as a Shrub. All taxa within this Species produce aerial
rootlets but will develOp into large shrubs if there is
nothing to climb, or if the supporting object upon which
the climbing began is removed. Its closest relatives are
Toxicodendron lobatum and 1, rydbergii. The plant is highly
variable, much of the literature being replete with names
of taxa which have few if any significant genetic differ-
ences. Because the nomenclature is very complex, it is dis-
cussed with each subspecies recognized in this study. The
Species consists of nine subspecies.

4a. Toxicodendron radicans (L.) Kuntze subSp. radicans.

(Plates XXV and XXVI). .

Toxicodendron radicans (L.) Kuntze, Rev. Gen. Pl.
pt. 1, p. 153. P1. 26, fig. 2. 1891. 2222
radicans L., Sp. P1. 1: 266. 1753. ‘2222 radicans
ogacum Aiton, Hort. Kew. 1: 367. 1789. 2222
toxicodendron var. radicans Eaton, Man. Bot.,
p. 400. 1789. 2222_scandens Salisbury, Prodr.,
p. 170. 1796. 2222 radicans var. toxicodendron
Persoon, Syn. Pl. p. 325. 1805. Philostemon
radicans Raf., F1. Ludov., p. 107. 1817. 2222

toxicodendron radicans Torrey, Fl. North. Mid.

PLATE XXV.

182

Type collection of Rhus radicans L.
From the Linnaean Herbarium, London.

 

 

4_———r—-____

 

 

185

 

184

PLATE XXVI. Specimen of Toxicodendron radicans subsp.
radicans. Type of Rhus radicans var.
vulgaris forma intercursa Fernald.

 

 

 

 

1|I'H\ I" ‘IH H” \VIFI(\ \H‘l I\I\ /;,A
H... .. ’"7.

. 54772 .7,
W1 «macaw/hull, .) «t
#5,,“ ”J‘I/u‘uuu J7. all

nu..."
U- vmllly ‘fi‘ ((4.1% I» ‘ \ ‘ M

, v
1L4! swq Mzcknqm

. .1” ”mm. 01197141 M I. III-nu», um... l..-\: ....u u ) \uuu I", ., l'r4

 

186
0.8., p. 324. 1824. Toxicodendron radicans
normale Kuntze, Rev. Gen. Pl. pt. 1, p. 153.
1891. 2222 toxicodendron a radicans Dippel,
Handb. der Laubholzk., p. 376. 1892. 2222
toxicodendron f. radicans McNair, Publ. Field
Mus. Nat. Hist. 225, Vol. 4: 68. 1925. 2222.
toxicodendron subSp. radicans Clausen, Mem.
Cornell Univ. Agric. Exp. St. 291: 8. 1949.
Type: "Habitat in Virginia, Canada." (LINN
378.14, lectotype of Brizicky, 1963; GH, MSC,
photographs of type).

Toxicodendron g1abrum Miller, Gard. Dict. Art.
Toxicodendron, No. 3. 1768, non 1. glabrum (LJ
Kuntze, Rev. Gen. Pl. pt. 1, p. 153. 1891
(= 2222_g1abra L.). ‘2222_humilis (as humile)
Salisbury, Prodr., p. 170. 1796. 2222 radicans
lucidum Aiton, Hort. Kew. l: 367. 1789. Type:
"Dill. Elth., 122, 212." (OXF, lectotype).

.2222 toxicodendron var. microcarpon Michx., Fl.
Bor.-Am., p. 395. 1803. ‘2222 toxicodendron
microcarpum Pursh, F1. Am. Sept. 1: 205. 1814.
2222_radicans var. microcarpon (Michx.) DC.,
Prodr. 2: 69. 1825. 2222 toxicodendron subsp.
microcarpon (Michx.) Engler in DC., Monog.
Phaner. 4: 395. 1883. [2222’microcarpa Steudel,
Nom. Bot. ed. 2: 694. 1841. Toxicodendron

radicans var. microcarpon (Michx.) Farwell,

187
Amer. Midl. Naturalist 12: 125. 1930. Type:
"Hab. in Carolina inferiore et Georgia." (P,
lectotype).

2222 blodgettii Kearney, Bull. Torrey Bot. Club 21:
486. 1894. Toxicodendron blodqettii (Kearney)
Greene, Leafl. Bot. Observ. Crit. l: 126. 1905.
Type: Blodgett, 212., Pine Key, near Key West,
Monroe Co., Florida, (NY, lectotype of Barkley,
1937).

2222 floridana Mearns, Proc. Biol. Soc. Wash. 15:
149. 1902. Type: Hitchcock 22, Hammocks,
Alva, Lee County, Florida, A.S. (US-387083,
holotype; F-101118, GH, MIN-149688, NO-l773781,
NY, isotypes).

Toxicodendron goniocarpum Greene, Leafl. Bot. Observ.
Crit. 1: 125. 1905. Type: 12212 McCulloch 42,
Lake City, Columbia Co., Florida. 25 June 1901.
(US-440443, holotype; FLAS-16l53, isotype;
US-440444, probable isotype).

Rhus toxicodendron f. malacotrichocar222.Moore,
Rhodora 11: 163. 1909. 2222 radicans f. 2212-
cotrichocarpa (Moore) Fernald, Rhodora 43: 592.
1941. Type: 2, 2, Chamberlain 222M2.'2. Dinsmore
222, in and growing over bushes, Pemaquid Beach,
Bristol, Lincoln Co., Me. 9 Sept. 1898. (GH,
holotype).

Rhus littoralis Mearns, Proc. Biol. Soc. Wash. 15:

188
148. 1902. 2222 radicans var. littoralis
(Mearns) Deam, Fl. Indiana, p. 651. 1924.
Toxicodendron radicans var. littoralis (Mearns)
Barkley, Ann. Missouri Bot. Gard. 24: 434.
1937. Type: 22g22 Mearns 233, Newport, Rhode
Island, Sept. 10, 1901. (US-403300, holotype;
US-1317009, US-403299, NY, isotypes).

2222 radicans var. vulgaris f. intercursa Fernald,
Rhodora 43: 592. 1941. Type: Fernald, Lon ,
222 22212 2222, Virginia, deciduous woods,
Curles Neck Farm, Henrico Co., June 21, 1936.
(GH, holotype; NY, isotype).

Terminal leaflet obtuse, subcordate, or truncate at
the base, with widest part generally below center; tufts of
hairs in vein axils of leaflets hyaline or straw-colored,
occasionally red-brown; blade 2.5 - 17.0 cm long, 2 - 13 cm
wide, petiole 2.0 - 20.0 cm long; terminal petiolule 0.5 -

6 cm long; fruits globose, 3 - 6.5 mm broad, puberulent,
scabrous, or papillose. Distribution: United States, states
bordering on the Atlantic Ocean, southern Maine southwest,
west to Appalachians, Ozarks, and East Texas; Bahama Islands;
Bermude. (Fig. 8)

Representative Specimens. UNITED STATES. Alabama.
Autauga Co.: Prattville, Gillis 4222 (MSC). Baldwin Co.:
Loxley Research plots, Stapleton, Grelen, 2,2, (FLAS).

Clarke Co.: Jackson, Gillis 2212 (MSC). Cleburne Co.:
6 miles NNW of Edwardsville, Donahue 242 (MSC). Colbert Co.:

189

Fig. 8. Distribution of Toxicodendron radicans
subsp. radicans.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

191
Sheffield, river bank, Bartlett 2212 (MSC, MICH). Elmore
Co.: Elmore, Gillis 4222 (MSC). Fayette Co.: southern
edge of Fayette Experimental Forest, Gillis 2221 (MSC).
Greene Co.: 2 miles north of Forkland, Gillis 2224 (MSC).
Hale Co.: Talladega National Forest, Payne Lake Recrea-
tional Area, Donahue 242 (MSC). Jefferson Co.: Flat woods
between Mobile Junction and McCalla, Bartlett 2222 (MSC,
MICH). Lee Co.: Auburn, 22212,.222. (NY). Macon Co.:
Tuskegee, Gillis 4222 (MSC). Madison Co.: Huntsville,
Gillis 4222 (MSC). Marion Co.: 9.1 miles south of Hamil-
ton, Gillis 2222 (MSC). Mobile Co.: Chunchula, Shinners
.ZZQQL (SMU); Mobile, Mohr, 2,n. (US). Shelby Co.: a mile

 

SW of Columbiana, Donahue 247 (MSC). Tuscaloosa Co.:
Tuscaloosa on river bluff, Bartlett 3314 (MICH, MSC). Union
Co.: 6.3 miles SE of New Albany, Gillis 6046 (MSC). 'Wash-

 

ington Co.: Chatom, Gillis 6015 (MSC). Arkansas. Columbia

 

Co.: Magnolia, Demaree 22222 (SMU). Craighead Co.: Jones-
boro, Demaree 2221 (SMU). Drew Co.: Monticello, Demaree
12222 (MO). Garland Co.: near Hot Springs, Runyon 1211
(NY). Hempstead Co.: Fulton, Palmer 2412 (A, MO, US).

Hot Springs Co.: Magnet Cove, Demaree 14222 (GH, MO, NY).

Connecticut. Fairfield Co.: Bridgeport, Eames, 2,2, (US);

 

 

Easton, Johnson 441 (F). Litchfield Co.: without definite
locality, Weatherby 2224 (US). Middlesex Co.: 'Westbrook,
Gillis 4112 (MSC); Middletown, Buckley, 2,2, (M)). New
Haven, Safford‘22 (US); Oxford, Harcer, 2,2, (GH); New

London Co.: Groton, Jansson, 212. (OSC); Windham Co.:

192

Thompson, Churchill 2. Lane, 2.2. (MO). Delaware. Kent Co.:

 

 

 

Leipsic, Laisen 714 (GH). New Castle Co.: New Castle,
Tidestrom 11547 (GH); Wilmington, Canby, 2,2, (NY). Sussex
Co.: Fogg 11186 (GH); Milford, Gillis 3848 (MSC). District

 

21 Columbia. Brookland, Bartlett 1788, 1789, 1790, 1791,

 

 

1122, 1122 (MICH, MSC, SMU). Columbian Formation, Boettcher
222 (A, F, GH, M0, RM). Without definite locality, Steele,
_,2, (E. MO, RM, US). Florida. Alachua Co.: Gainesville,
O'Neill, 2,2, (MO); Sugarfoot Hammock, Gainesville, 2222'2
2y222, 2,2, (FLAS); Gainesville, periphery of golf course on
Newberry Road, Gillis'2222 (FLAS, MSC); Devil's Millhopper
sink-hole, 12 miles NW of Gainesville, Gillis 4242 (MSC).
Baker Co.: titi pond, Manning,;2222 & Arnold, S.n. (FLAS).

Bay Co.: St. Andrew's Bay, Benker 2222 (NY). Bradford Co.:
Brevard Co.: Cocoa, Gillis 4222 (MSC). Broward Co.: Hugh
Taylor Birch State Park, Fort Lauderdale, Roestad, 2,2.
(FLAS). Citrus Co.: Red Level,;2222'2 Arnold, 2,2, (FLAS);
Pineola, GilleSpie, 2,2. (FSU). Clay Co.: Goldhead Branch
State Park,‘2221, 2,2, (FLAS). Collier Co.: Everglades,
Miles City near fire tower, Gillis 4422 (MSC); 7.3 miles NW
of COpeland, Fahkahatchee Strand, 2222 2 Botany 21222 2-_6_
(FLAS); Marco, Hitchcock,.2,2. (F). Columbia Co.: Lake
City, McCulloch 42, 42 (FLAS, US). Dade Co.: Homestead,

2,2, (NY); Roberts Island, Small 7387 (NY); Long Pine Key,
Traverse 590 (GH); Cox Hammock, Small 2.Mosier 5508 (M);

193
Paradise Key, Craighead, 2,2, (FLAS); Sykes Hammock, Ever-
glades Keys, 22211_2 Mosier 2422 (FLAS, GH, NY, US). DeSoto
Co.: Hammock, west edge of Arcadia, Botany £12§§.§:§ (FLAS).
Dixie Co.: Hammock, 1 mile NE of Jena, Pasture Survey, 2,2,
(FLAS). Duval Co.: Jacksonville, Curtiss 4242 (MO, US).
Escambia Co.: Santa Rosa Island, 21212,.222. (FLAS). Andu-
lusia, Godfrey 2 Reinert'21142 (MSC). Franklin Co.: St.
Vincent Island, McAtee 11222 (US). Glades Co.: west of
Lake Okeechobee, 22211 2 DeWinkeler, 2,2, (NY). Gulf Co.:
’Wewahitchka, Dead Lake, Demaree 22422 (MSC). Hamilton Co.:
Bayhead 31miles south of Jennings, 2222 2 Arnold, 2.2.
(FLAS). Highlands Co.: Lake Placid, Craighead, 2,2, (MSC);
cypress swamp at west end of Highlands Hammock State Park,
2222 2 Botany 21222 2:1 (FLAS). Hillsborough, Tampa Bay
Shore, 22112,241 (F, MO). Jackson Co.: Blue Springs north
of Marianna, Exploration 2221yI1221, 2,2. (FLAS). Lafayette
Co.: Opposite Branford on Suwannee River,;fl222I2 Arnold,”
2,2, (FLAS). Lee Co.: Alva, Hitchcock 22 (F, GH, MIN, MO,
NY, US); Fort Myers, Hitchcock, 2,2. (F); Sanibel Island,
Cooley 2422 (FLAS, GH, NY, US). Leon Co.: hammock north
of Tallahassee, Crevasse, 2,2. (FLAS); 'Woodville, Godfrey 2’
Morrill 53026 (FSU, GH, NY). Levy Co.: 2 miles south of
Gulf Hammock, Pasture Survey, 2,2. (FLAS). Madison Co.:
Sirmans, Palmer 21222 (A, MO). Manatee Co.: Bradenton,
Wheeler, 2,2,-(F); Long Boat Key, Gillis 2222_(MSC); mouth

of Manatee River, Rugel 12 (BM). Marion Co.: bayhead, 3

194
Eureka on Oklawaha River, Godfrey 2 2222 22222 (MSC); Ocala,
Palmer 22422 (A, MO); Silver Springs, bank of Silver River,
Gillis 2224_(MSC). Martin Co.: Lake Okeechobee, Moldenke
2442.(NY). Monroe Co.: Key West, Blodgett, 2,2. (NY);
Flamingo, Craighead, 2,2, (MSO); Hammock near Pinecrest,
Craighead, 2,2. (FLAS); Big Pine Key, 2212_4222 (FLAS);
northern Key Largo, 21222,2 Chambers 222_(GH, US). Nassau
Co.: Julee, Gillis 2222 (MSC). Okaloosa Co.: Bolton
Creek swamp, 1y222 214 (FLAS). Orange Co.: near Christmas,
Gillis 2224 (MSC); Rock Spring,;2222,.2,2. (FLAS). Osceola
Co.: Kissimmee, Hunnewell 2222 (GH); outskirts of Interces-
sion City, Gillis 2422 (MSC). Palm Beach Co.: Lake Okee-
chobee, 22211 _8._ 212211 4221 (NY); Palm Beach, Garvens, 2.2.
(F). Pinellas Co.: St. Petersburg Beach, Gillis 2241.(MSC).
Putnam Co.: west of Mud Springs, Welaka, Laessle,_2,2.
(FLAS). St. Johns Co.: 10 miles west of St. Augustine,
‘2222.2 Arnold, 2,2. (FLAS); St. Lucie Co.: along Indian
River, 1.3 mile south of junction with F-712, 4 miles south
of Ft. Pierce, 2222 4221 (FLAS); along S-607, 1 mile west
of Indio Road, Gillis 2224, (MSC). Sumter Co.: Rutland,
Taylor Co.: Half Mile Rise on Aucilla River, Exploration
2222y 1222, 2,2.'(FLAS). Uhion Co.: Cypress Pond Road, 2
miles South of Raiford, 2222|2.Arnold, 2,2, (FLAS). ValuSia
Co.: 3 miles west of Stetson University, Prichard 222 (FLAS);
Cypress hammock about 5 miles NE of Deland, Rossbach.2

Prichard 2039 (FLAS). *Wakulla Co.: 6 miles E of SOpchoppy,

195
2222 2 F_o_r_d_ 26_21 (FLAS); St. Marks, Harbison 1212 (A).
Georgia. Bibb Co.: Macon, Gillespie 4222 (A, ND, SMU).
Bryan Co.: Richmond Hill, Gillis 4222_(MSC). Cherokee Co.:
SWof Canton, Donahue 222 (MSC). Clarke Co.: Athens, 22222,
2,2, (MSC). Cook Co.: south of Lenox near Barneyville
Road, Gillis 2222 (MSC); north of Cecil, Gillis 4222 (MSC).
Crisp Co.: 0.5 mile north of Turner Co. line, Gillis 2212
(MSC). Decatur Co.: Bainbridge, 22211,_2,2, (A, F, NY).
Dooly Co.: south of Unadilla, Gillis, 2212 (MSC). Dougherty
Co.: without definite locality, GilleSpi24241 (A, ND, SMU).
Douglas Co.: NE of Villa Rica, Donahue 222 (MSC). Early
Co.: Blakely, Gillis 4222 (MSC). Emanuel Co.: Swainsboro,
Gillis 4222 (MSC). Floyd Co.: Rome, Horseleg Mt., Demaree
2 11222.22122_(MSC). Habersham Co.: between Clarkesville
and Demorest, Gillis 2242 (MSC). Haralson Co.: Tallapoosa,
Donahue 222 (MSC). Houston Co.: 6 miles north of Henderson,
Gillis 2221 (MSC). Jackson Co.: Commerce, Gillis 2211
(MSC). Lumpkin Co.: Chattahoochee National Forest, DeSoto
Falls Campground, Donahue 222 (MSC). McDuffie Co.: Fall-
line sand hills, vicinity of Thomson, Bartlett 1222, 14229
12.8.. m. 1.6.622. 1919.. 19141. 17.01. 129;. 17.4.1 (MICH. MSG.
SMU). McIntosh Co.: Sapelo Island, Gillis 2221, Morgan Co.:
6.6 miles north of Putnam Co. line, Gillis 22422 (MSC).
Pickens Co.: 5 miles west of Emma, Donahue 224,(MSC); 2
miles north of Eatonton, Gillis 2241 (MSC). Turner Co.:
1 mile north of Inaha, Gillis 2222 (MSC). Union Co.: 3
miles NW of Blairsville, Donahue 222 (MSC). Locality

196

uncertain, Lulu Falls, Churchill, 2,2, (M)). Kentucky.

 

Bell Co.: 0.5 miles NW of Fourmile, weathered rock cliff,
Gillis 6540 (MSC). Trigg Co.: 6 miles west of Cadez,

Donahue 211 (MSC). Louisiana. Calcasieu Parish, Lake

 

Charles, Allison 222 (US). Franklin Parish, Glibert, Gillis
5951 (MSC). Morehead Parish, Jones, Gillis 2242 (MSC).
Natchitoches, Palmer 1222 (A, MO). Rapides Parish, Valen-
tine Lake, Thieret 11242 (MSC). Saint Tammany Parish, Cov-
ington, Demaree 22124 (MSC). 22122. Lincoln Co.: Bristol,
Chamberlain 222 (GH); Boothbay, Fassett 22212 (F, MO, SMU,
'WIS); Monhegan Island, Churchill, 2,2, (MO). Sagadahoc

Co.: Five Islands, Rehder 1242 (A, US). York Co.: Isle

of Shoals, Howard 2 Wagenknecht 2222 (A, MSC). Maryland.

 

 

Anne Arundel Co.: Annapolis, Tidestrom 2222 (A, MO, US).
Baltimore Co.: without specific locality, Taylor, 2,2, (F);
Sorrento, Tidestroml2 Bartlett 1222 (SMU). Calvert Co.:

Cony Point, Sey2our 11242 (SMU). Caroline Co.: Horse Point,
Tuckahoe River, Tidestrom 11211 (US). Frederick Co.:
Lewistown, Gillis 2222_(MSC); Hunting Creek, Sullivan 415
(ND). Hartford Co.: Spesukie Island, 22112, 2,2, (US).
Montgomery Co.: Bethesda, Bartlett 1222 (MICH, MSC, SMU).
Prince Georges Co.: Beltsville, Hermann 10866 (F); Laurel,
Togging, 2,2, (MIN); College Park, Tidestrom 1122 (US).

St. Marys Co.: St. Mary, Walker 2224 (US). Talbot Co.:
Claiborne, Tidestrom, 2,2, (US). Worcester Co.: Ocean City,
132gg11424 (GH); Snow Hill, Boettcher, 2.2. (US). @-
chusetts. Barnstable Co.: 'W. Falmouth, Churchill, 2,2,

 

 

197
(MO); Centreville, Williams, 2,2. (GH); Hyannis, Fernald,
Butters, 2.21, 2222.12222 (GH); between Woods Hole and
Falmouth, 22222,'222. (A, GH); Brewster, Fernald 2 2222
11222 (GH);Harwich, Fernald 2 122g 11222 (A, GH); Dorchester,
Churchill, 2,2, (MIN). Berkshire Co.: Williamstown, Gillis
4222_(MSC). Bristol Co.: Dighton, Seymour 18559 (SMU);

Dartmouth, Knowlton, 2,2, (MIN); New Bedford, Greene, 2,2,

 

(WIS); Dukes Co.: Oak Bluffs, Sey2our 1222 (GH, NY, US);
Gay Head, MacKeever 221_(MSC); The Meadows, Nelson 2 Nelson
2627 (RM); Cappaquiddick Island, Bicknell 2222 (NY); Noma-
masset, 22gg_2242,(M)). Essex Co.: Rockport, Rouleau 1222
(ARIZ, E, F, FLAS, GH, MIN, Mo, MSC, NY, 050, RM, US, WIS);
Ipswich, Morong, 2,2, (NY). Hampden Co.: Holyoke, Sey2our
422 (GH, MO, NY); Springfield, Pillsbury, 2,2, (US); Wilbra-
ham, collector unknown (MIN). Hampshire, 3. Hadley, Alelenk,
_,2, (US). Middlesex Co.: Cambridge, Kennedy, 2,2. (GH);
Lexington, Williams, 2,2, (GH), Wayland, 22222 212 (A);
Winchester, Williams, 2,2, (GH); Lowell, Beattie 2,2, (E,
MO, OSC, SMU). Nantucket Co.: Wauwinet, MacKeever, 2,2,
(MSC); Bicknell 2221 (NY). Norfolk Co.: Stoughton, 21222
1224_(US); Brookline, Rohson,‘2,2. (A). Plymouth Co.:
Middleboro, Murdoch 221 (F); Norwell, Knowlton, 2,2, (MO).
Suffolk Co.: Boston, Williams, 2,2, (GH). Worchester Co.:

Bolton, Sey2our 6092 (WIS). Mississip21. Benton Co.: near

 

Union Co. line, Gillis 6047 (MSC). Covington Co.: pine
ridge west of Collins, Gillis 6006 (MSC); along route 35
below Bowie River, Gillis 6003 (MSC). Franklin Co.: road

198
to Clear Springs, Gillis 2212 (MSC); 3 miles from Roxie,
Gillis 2212 (MSC). Itawamba Co.: between Tremont and Fulton,
Gillis 2242_(MSC). Jackson Co.: Magnolia State Park, Demaree
31216 (SMU, TEX); Krede, Demaree 24212_(SMU). Jefferson
Davis Co.: SE of Prentiss, Gillis 2222 (MSC). Lawrence Co.:
Monticello, Gillis 2222 (MSC). Lee Co.: Tupelo, Gillis
2242_(MSC). Lincoln Co.: New Prospect Baptist Church,
Gillis 2212 (MSC). Madison Co.: without definite locality,
McDougall 1222 (US). Marshall Co.: 2.4 miles SE of Holly
Springs, Gillis 2242 (MSC). Oktibbeha Co.: Starkville,
Hill, 2,2. (ILL). Panola Co.: Sardis, Broadhead 22 (SMU).

 

Smith Co.: Marathon Lake Recreation Area, Donahue 242 (MSC).
Tate Co.: Senatobia, Gillis 2222 (MSC). union Co.: 6.3
miles SE of New Albany, Gillis 2242.(MSC). Wayne Co.: Beat
Four School, 7.6 miles west of Chickasawhay River, Gillis

6010 (MSC). New Jersgy. Atlantic Co.: Mays Landing, Fogg

 

2122 (GH); Pleasant Mills, Hermann 4221 (M0); Hammonton,
Gillis 2242 (MSC). Bergen Co.: without definite locality,
Oehler, 2,2, (NY). Burlington Co.: Riverton, Bassler 22
(KSC); Mt. Holly, Gillis 2222 (MSC). Camden Co.: Chew Road,
MacKenzie 2212 (NY); Crystal Lake, Westmont, 22222.222_(MO).
Cape May Co.: Cape May Court House, Gillis 2222,(MSC).
Cumberland Co.: Buena, Gillis 2242 (MSC). Hunterdon Co.:
Ellmaker Island, Raven Rock, near Stockton, collector unknown
(MIN); Warren Glen, Benner 2222_(GH). Mercer Co.: Princeton,
collector unknown (NY); between Princeton and Rocky Hill,

roadside, Gillis 5089 (MSC). Middlesex Co.: Middlebush,

199
old field, Gillis 2414 (MSC); East Millstone, Hutcheson
Memorial Forest, Gillis 2412 (MSC); Cheesequake State Park
near artificial brackish water pond, Gillis 2221_(MSC);
Perth Amboy, Gillis 2221 (MSC); New Brunswick, Stevens 11_
(ARIZ, GH, MSC, NY, RM, WIS). Monmouth Co.: Freehold,
MacKenzie 1241 (NY). Ocean Co.: Barnegat Pier, MacKenzie
2212_(GH); Island Beach State Park, dominant Shrub for 10
miles, Gillis 2122 (MSC); Toms River, MacKenzie 4222.(GH,
MO, US). Salem Co.: Penns Grove, 22222 2_22222 1222 (A, US);
Hancock's Bridge, 22222_2_22222_2221 (A, MO). Somerset Co.:
Watchung, Moldenke 1221 (NY); Hillsboro Twp., Sourland Mt.,
Neshanic-Zion Road, collector unknown (NDA); Mercury, Perr ,

2,2, (MO). Union Co.: Rahway, Gillis 5090 (MSC). Warren
Co.: Phillipsburg, Tidestrom 6009 (US). New York. Bronx

 

 

Co.: without definite locality, Wilson 12212 (NY); Bedford
Park, Edmondson 1422 (NY). Duchess Co.: Brush, Gillis 4222
(MSC). Greene Co.: without definite locality, Gillis 2222
(MSC). Manhattan Co.: Van Cortlandt Park, Hastings, 2,2,
(NY). Nassau Co.: Lawrence, collector unknown, (ND-G);
Woodmere, collector unknown (ND-G). Rennselaer Co.: Troy,
2211, 2,2, (F). Richmond Co.: Staten Island, Dowell 4241
(MIN). Suffolk Co.: Shelter Island,;22112, 2,2, (MSC);
St. John's Lake, Harbar C-l5648 (MINI; Northville, 4 miles
NE of Riverhead, Curtiss 1212 (G-DC); Grunport, 22212, 2,2.
(GH); Southold, Palmer 42212 (A); 3 Miles south of Montauk
Point, Gillis 4224.(MSC). Westchester Co.: Shrub Oak,
Gillis 4224, Earth Carolina. Alamance Co.: 4.6 miles

 

 

1“

 

 

 

 

200
north of Union Ridge, Ramseur 2,Hammond 2222 (FSU). Bruns-
wick Co.: Long Beach, 22y22 1424 (NY). Buncombe Co.:
Biltmore, Bartlett 2222 (ARIZ, TEX). Camden Co.: Elizabeth
City, Gillis 2222 (MSC). Catawba Co.: east of Conover,
Gillis 2222 (MSC). Currituck Co.: Currituck Sound, Randolph
2 Randolph 222_(GH). Dare Co.: Pea Island, Gillis 4241
(MSC). Davie Co.: Mocksville, Gillis 2222 (MSC); Farmington,
Gillis 2222 (MSC). Durham Co.: Duke Forest, Blomquist 122
(US); Nelson, Gillis 4222 (MSC). Forsyth Co.: Clemmons,
Gillis 2212 (MSC). Halifax Co.: Enfield, 22122.14211 (GH).
Hertford Co.: Potecasi Creek, SE of Mapleton, 22122 41211
(FSU). Hoke Co.: 0.2 mile NW of Hoke-Robeson Co. line,
22122 22224 (FSU). Macon Co.: Topten, Donahue 221 (MSC).
Orange Co.: University Lake, 22111_1211.(GH); Chapel Hill,
Gillis 4222 (MSC). Perquimans Co.: Hertford, Gillis 2222
(MSC). Polk Co.: Varyhans Gap, Peattie 1412 (F); Tryon,
Churchill, 2,2. (MO). Swain Co.: SW of Nantahala, Donahue
222 (MSC). Wake Co.: Raleigh, 2y222 2,2, (US). Locality

uncertain: Lillington, Pillsberry, 2,2. (US). Pennsylvania.

 

Adams Co.: East Berlin, Gillis 4222 (MSC); Bedford Co.:

New Baltimore, Gillis 2424 (MSC); 3.5 miles south of Center-
ville, Westerfeld 2222 (FSU). Bucks Co.: Bowman's Hill,
22222_2_Wherry 1212;(A); Bowman's Hill, Gillis 2242,(MSC);
Neshaming, McDowell 21_(GH); Washington Crossing, Gillis 2222,
(MSC). Dauphin Co.: Harrisburg,.22211,.2,2. (F). Franklin
Co.: Tuscarora Mountain, Gillis 2112 (MSC); Mercersburg,

Porter, 2,2, (M)). Lancaster Co.: without definite locality,

201
221224211 (US); Conewago, Heller, 2.2. (GH). Montgomery
Co.: Cold Point, _A_d_ax22 2 22222 1212 (A). Montour Co.:
Mausdale, 15.22212 5_9_4_2 (OSC). Northampton Co.: Northampton,
Churchill, 2.2. (GH, MO); Bethlehem, Molle, 2.2. (GH);

 

Johnsonville, Schaeffer 12221 (GH); Easton, Gillis 4224
(MSC). Perry Co.: Newport, 22222 2 222212 2222 (A, GH);
New Bloomfield, m 2 22_a_m_s 13_8_2 (A); Drumgold, 24222 2
Adm 2121 (A); Duncannon, 225122 2 22222 2222 (A). Phila-
delphia Co.: Mermaid, _A_d_2r_n_2 2 22222 18_82 (A); West Phila-
delphia, 42222 21_2 (GH); Cedar Grove, 222222 1_61 (GH).
Anyder Co.: Se'linsgrove, Moldenke 24_52 (NY); York Co.:
Woodbine, 22222 4291 (A); Dilsburg, Gillis 4222 (MSC).

Rhode Island. Bristol Co.: Barrington, Winslow, 2.2. (US).

 

 

Kent Co.: West Warwick, Gillis 4122 (MSC). Newport Co.:
Newport, Mearns 222,_222, 222 (A, GH, BM, ND-G, NY, US).
Providence Co.: Providence, Collins, 2.2. (US); Grant's
Mill, Palmer £422 (MO, NY). Washington 00.: Wickford,

Williams, 2.2. (GH). South Carolina. Aiken Co.: Granite-

 

 

 

ville, Eggert, 2.2. (MO); Opposite Augusta, Ga., Bartlett
2&2 (MICH, MSC). Anderson Co.: Whitner Park, 22212 1112
(MIN, OSC); Anderson, 221/12 216—8 (MIN, NY). Bamberg Co.:
Denmark, Palmer 22222 (A, MO, NY). Berkeley Co.: Pineville,
Godfrey 2 11122 212 (GH, NY, US). Florence Co.: Branch
Swamp, Bartlett 2212 (MICH). Lexington Co.: Batesburg,
MacGregor 22 (US). Oconee Co.: Clemson College, 20222 2l_02
(NY, US). Richland Co.: north of Columbia along highway

21, Gillis 5222 (MSC). Wood Co.: without definite locality,

 

 

 

 

202
Davis 2041 (US). York Co.: Ebenezer, Branch Swamp,
Bartlett 2373 (MICH, MSC). Tennessee. Sevier Co.: Great

 

 

Smoky Mountains National Park, % mile from Gatlinburg,
Gillis 9221 (MSC). Blount, Maryville, Godfrey, §,g, (GH);
Townsend, ggggg,'§ag. (MIN). Frankling Co.: Sherwood,
Eggert, 2,2, (MIN). Hardeman Co.: Bolivar, Palmer 11222_
(A, MO, US). Henderson Co.: Natchez Trace State Park,
Donahue 212 (MSC). Knox Co.: Knoxville, 3212, 212. (NY);
Island Home Park, Ainslie, 2,3, (KSC). :13222. Harris Co.:
Houston Memorial Park, Traverse 12 (F, GH, SMU); Lake
Houston, Traverse 1991.(F). Montgomery Co.: Willis,
Q1zgg,g2g_(F). Walker Co.: Stubblefield Lake Recreation
Area, Donahue 222 (MSC). Locality uncertain: Sunderland
Springs, Sargent, 3,2, (A). Vermont. Chittenden Co.:
Charlotte, Harsford, 2,2, (MIN). Windham Co. Westminster,
Gillis 3222 (MSC). ‘Windsor Co.: White River Jct., Gillis
flQ§2_(MSC). Virginia. Alexandria, without definite local-

 

ity, Steele, £12. (US). Arlington Co.: Fort Myer, Mearns
122 (US). Bedford Co.: without definite locality, Curtiss,
.222. (GH). Dinwiddie,Co.: Burgess Station, Fernald 2_1229
2212 (GH, PH). Fairfax Co.: Great Falls, Bartlett 11§§,
1112, 11§1_(MICH, MSC). Greene Co.: Bear Fence Mt.,
Walker 2221 (US). Hanover Co.: Mechanicsville, 22222 2
Whergy 2212 (A). Henrico Co.: Curles Neck Farm, Fernald,
1ggg,I2.§mg£1.§§22 (GH, NY). Isle of Wight Co.: Rushmere,
Fernald 2 1ggg_l3066 (GH, MIJL PH); Carrollton, Gillis 595§_
(MSC). Lancaster Co.: Lively, Allard 21222 (US).

203
Nansemond Co.: South Quay, Fernald 2:1229, _,3, (GH, PH).
Norfolk Co.: Ocean View, Kearney 1122 (US); Lambert's Point,
Kearney 1222 (US). Northampton Co.: Capeville, Gillis 2292
(MSC); Exmore, Gillis 2222 (MSC); Cape Charles, Tidestrom
11222_(GH); Coan, Tidestrom 2122 (A, US). Petersburg, Rehder,
_,2, (A, MO). Portsmouth, 22331,'§,Q. (E). Prince George
Co.: Desputanta, Fernald, 1ggg_2'§gggg,§§22 (GH, NY, PH,
US). Princess Anne Co.: Virginia Beach, Britton 2 22211,
.213. (F, NY, US); Gilmerton near Norfolk, Norton §1§.(US);
Little Creek, Fernald 2.1233, gag. (GH); Wolf Snare, Randolph
2 Randolph 232 (GH); Creeds, Fernald 2 Griscom 3322 (GH,
PH); Lynnhaven Inlet, Randolph 2 Randolph 222'(GH); False
Cape, Fernald 2:1ggg.fig1§.(A, GH); Cape Henry, 29121.40-190
(NY). Southampton Co.: Oak Grove School, Fernald 2 1223
§2§1_(GH, MO, PH). Surry Co.: Chippokes, Fernald 2 1323
12923 (GH, NY, PH, US); Claremont Wharf along James River,
Fernald 2 1229 2222 (GH, PH). Sussex Co.: Huske, Fernald
2 m 12131 (GH, PH, us); Waverly, 15531 121322 (smu);
Jarratt, Fernald 2_1gggl12221 (GH, PH, WIS); Stony Creek,
Fernald 2 1329,2222 (GH, MO, PH). Warwick Co.: Hampton,
Churchill, 2,3, (MIN, MO, WIS). Westmoreland Co.: Colonial
Beach, Tidestrom 2 Bartlett 2222 (A, MD, US). York Co.:
James River, Killip 29122 (US). Locality uncertain: Jericho
Run, Dismal Swamp, Boettcher £1.(RM, US). West Virginia.

 

Hampshire Co.: Capon Bridge, Gillis 5707 (MSC). Kanawha
Co.: Kanawha State Forest, Donahue 262 (MSC). Monongalia

Co.: Morgantown, MillSpaugh 146 (NY). Randolph Co.:

204
Huttonsville, Donahue 222 (MSC). Upshur Co.: Bucklin,
Pollack, 2.2. (US). Webster Co.: Hacker Valley, S_mi_t2
1232 (F). Locality uncertain: Little Falls, MillSpaugh,
2.21 (F).

*****

CANADA. Nova Scotia. Queens Co.: Port Mouton,

 

2222.2212 (A, F). Shelburne Co.: Harper Lake, Fernald‘2
1222.23222 (GH); East Jordan, Fernald 2 1222_22222 (GH);
Five-River (Morris) Lake, Fernald 2 1229 22222 (A, GH);
Welshtown (Birchtown) Lake, Fernald 2 1222 22223 (GH).
Yarmouth Co.: Gavelton, Fernald 2‘1222_22222 (GH); Argyle
Head, 1222 2 Linder 21122 (GH); Godfrey Lake, Fernald 2
1m 2_z1221((GH, MO); Yarmouth, £2.13 2 m 2173:: (GH);
Tusket,‘1222'2 Linder 21122 (GH).

* * * * *

BERMUDA. Paget East, Hughes, 2,2, (MSC). Tucker's
Town, 21222 222 (NY). Boaz Island, _Bgflvg 2 Britton 1222
(NY).

* * * * *

BAHAMA.ISLANDS. 112222. Marsh Harbor, Great Abaco,
21222.1222 (NY); Great Abaco, 3 miles south of Treasure Cay,
Gillis 2121 (MSC). :222122. North Andros, Staniard Creek,
22211.2 Carter 2222'(NY). 212121. South Bimini, without
definite locality, ngard 2 Howard 12221 (NY, SMU). :211:21y:
North Cat Cay, without definite locality, MillSpaugh 2222
(NY). Grand Bahama. Without definite locality. Britton'2

 

Millsgaugh 2446 (NY); Russell Town near Freeport, Gillis

205
(FTG). New Proviggnce. Delaport, Britton 3416 (NY); South-

 

west shore of Lake Cunningham, Gillis 2212 (MSC).

This subspecies probably represents the form of
poison-ivy which was first encountered by seventeenth
century explorers and settlers in the New World. Early
collections from North America, made in colonies close to
the Atlantic Seaboard, were either of this taxon or, if
found north of the 44th parallel, of Toxicodendron rydbergii.
The designation of a type for 2222 radicans L. has presented
a dilemma, because of the various Specimens known to have
been available to Linnaeus.

Linnaeus's first reference to poison-ivy was in
the Hortus Cliffortianus (1738): 5222_radicans foliis
ternatis: foliolis petiolatis ovatis 22212'intquerrimis."
Linnaeus based his early concepts of poison-ivy on the col-
lections other than at Uppsala, such as that of Clifford's.
The Clifford Specimens are probably the ones upon which
Linnaeus based most of his knowledge of poison-ivy. (These
specimens, now at BM, are a flowering representative of
1, gvdbergii, and a sterile branch of‘l. radicans subsp.
radicans.)‘ They undoubtedly represent Linnaeus's concept
of poison-ivy until after the publication of the Hortus
Cliffortianus. In 1751, however, Peter Kalm returned from
his eXploring and collecting adventures in North America,
bringing with him many specimens which were incorporated
into Linnaeus's personal herbarium. Among these was a form

of poison-ivy, presumably from the vicinity of Philadelphia

206

(Kalm, 1770), for which Linnaeus must have had some habit
data, for he edited his diagnosis of 2222 radicans from
the 2211, 22111, in Species Plantarum (1753) by the insertion
of the phrase {22212 radicante," thus suggesting that he
had learned of its habit as a liana. The pOpulation repre-
sented in the Clifford Herbarium by the flowering Specimen
(1, rydbergii) has no aerial roots and did not climb. Be-
cause both the Clifford specimens (already familiar to
Linnaeus), and the Kalm Specimen (representing a phase of
the plant new to Linnaeus), were all under consideration
by him at the time of writing of the Species Plantarum, the
Kalm Specimen has been confirmed as the lectotype. It is
the Specimen which demonstrates the climbing habit, thus
the added phrase in the 22, E1, The Clifford Specimen,
representing a non-climbing form of poison-ivy, must be
excluded as a candidate for the nomenclatural type. Fernald
alludes to the LINN Specimen as being the type by publish-
ing (1941) a photograph of it labeled "portion of type,"
but Brizicky (1963) formally designated it as the lectotype.
If Linnaeus knew of the non-climbing habit of the Clifford
Specimen and any other departures from the characteristics
represented in the Kalm Specimen, he chose not to separate
these plants nomenclaturally. This was finally done at the
end of the nineteenth century when Rydberg (1900) published
Small's name 2222 rydbergii.

The type material leaves much to be desired in the

way of diagnostic characteristics. After study of the

207
material in the Linnaean Herbarium, I have concluded that
the Kalm Specimen must represent the Atlantic Coastal pOpu-
lation, corroborating the hint in Kalm's records that he
collected it near Philadelphia. What has been most influ-
ential in making this decision is the characteristic tufts
of hairs in the main vein axils on the lower surface of the
leaflets, a characteristic only of Oriental and Atlantic
Coast forms. Had the Specimen been a fruiting one, the
pubescence of the exocarp would have solved the dilemma.
Populations from the eastern coastal regions of the United
States generally have entire leaflet margins, although not
uncommonly with a single lobe or notch each leaflet (two on
the terminal) as has the Linnaean specimen. The division
of the terminal leaflet into inequilateral halves is a more
subtle, although persistently accurate character, and is
found in this Specimen also. Complicating matters is the
additional Specimen in the Linnaean Herbarium (Sheet 387.15)
which has almost glabrous fruit. This Specimen is anomalous
as to the pOpulation which it can represent. It is there-
fore better to consider that Brizicky has chosen the appro-
priate Specimen for nomenclatural typification.

Curiously enough, most taxonomists, even those who
have worked extensively with the poison-ivy problem, have
ignored typification of this Species and the type subSpecieS.
Many subspecies, varieties, and forms have been described,
which all apply to the same taxon as the Linnaean Specimen.

This is due in part to the fact that the Kalm Specimen was

208
collected in flowering condition. Had it been a fruiting
Specimen with the fruit's characteristic trichomes on the
exocarp, the publishing of such epithets as microcarpon,
malacotrichocarpum, and littoralis (each used in several
combinations) might have been precluded.

The fruits of subSp. radicans are generally smaller
than the fruits of other subspecies of Toxicodendron radicans.
This is eSpecially true in pOpulationS which are found in
Florida and southern Georgia. The small-fruited forms seem
to demonstrate some geographical pattern, but otherwise do
not differ from other pOpulations of subSp. radicans, and
therefore do not seem to warrant nomenclatural segregation.
The various taxa using the epithet microcarpon or its ortho-
graphic derivatives are therefore treated in synonymy with
subSp. radicans.

(In regard to the type oflfi. littoralis Mearns, there
are two Sheets and a box of twigs and fruits comprising the
type at US, one of the Sheets containing a fruiting Speci-
men and one a portion of the stem with abundant aerial roots.
Presuming that these all represent a Single Specimen, thereby
fulfilling the requirement of Art. 9 of the Code, these all
represent the type Specimen. A Specimen labeled "isotype"
is deposited at GH, but collected 11 days before the type
(Mearns 222). Furthermore, US-403299 (Mearns 222) is prob-
ably an isotype. Mearns appears to have numbered each
Specimen with a separate collection number even though many

may be part of the same gathering.

209

Because the name 222§_radicans var. littoralis has
been applied (Deam, 1940) to pOpulations which do not apply
to this taxon, it seems apprOpriate to discuss the reason
that this application was made. Some pOpulations of poison-
ivy in Indiana, eSpecially from counties bordering the Ohio
River, have fruits which are pubescent to a greater or lesser
extent; they are, however, dissimilar in other reSpectS to
pOpulations from the coastal regions which are truly 1221-
codendron radicans subSp. radicans. This is also true for
some pOpulations in Oklahoma (which were named 1, aboriginum
by Greene, 1905). The epithet littoralis is misapplied;
the pOpulations of Indiana and Oklahoma Should be 1, radicans
subSp. negundo.

In describing several new varieties and forms,
Fernald (1941) described a number of taxa represented in
southeastern Virginia. He did not until later (1950) indi-
cate the reasons for his new taxa. He implied that forma
intercursa (= "running between"), with its supposed pubeSCent
leaf blades and glabrous petioles, was intermediate between
forma vulgaris, which he said had both leaf surfaces and
petioles glabrous, and f. negundo, which allegedly had both
leaf surfaces and petioles pubescent. These characters in
the pOpulationS in question, eSpecially petiole pubescence,
are highly variable.

We are able to typify Toxicodendron glabrum Mill.
by Miller's reference to Dillenius's Toxicodendron rectum

foliis minoribus glabris for which there is a Specimen in

210
the Dillenian Herbarium; this specimen has been designated
the lectotype. It was a cultivated Specimen, grown in 1727
by Sherard. It is not surprising, its being a cultivated
plant, that it is an atypical form of 1, radicans subsp.
radicans with very narrow leaflets and almost total lack
of the characteristic tufts of hairs in the vein axils on
the lowe:leaflet surface, conditions strongly suggestive
of the possibility that the specimen is a juvenile. By
reference to this Dillenian citation, 5222 radicans lucidum

Aiton and 2. humilis Salisburyl4 (as'fi. humile) can also

 

l4Dillenius (1732) intended that his Toxicodendron
amplexicaule, foliis minoribus glabra be different rom
Toxicodendron rectum, foliis minoribus glabris. Linnaeus
must have agree Slnce he quoted the former as synonymous
with his variety/A, and the latter with his variety 7‘ .
The descriptions by Dillenius indicate that he thought
Toxicodendron amplexicaule. . . . was a climber, and Toxi-
codendron rectum . . . was not. His illustration of Toxi-
codendron rectum . . . , however, appears to be of the
climber Toxicodendron radicans subSp. radicans as herein
treated. Moreover, the Specimen at the Sherard Herbarium
which Dillenius saw demonstrates Toxicodendron rectum . . .
to be 1. radicans subSp. radicans, desplte a written de-
scriptlon to the contrary. This becomes of importance when
one tries to determine what Salisbury meant by his Rhus
scandens and R. humile.

Rhus humilis must refer to 1. radicans subSp. radi-
cans also, although typified by another Specimen (the one at
Sherard), even though its description more nearly fits T.
rydbergii than 1. radicans. Because a Specimen can be found
atgleast one of _these polynomials of Dillenius, we can at-
tribute this one (1oxicodendron rectum . . .) to 1. radicans
subSp. radicans. Where we have a aescription, an —iIIUStra-
tion, and a Specimen, the Specimen must be granted prece-
dence over all else in typifying, even though it is not cor-
rectly defined by the illustration or description. Rhus
scandens, because of its reference to Rhus radicans L.,
must also be 1, radicans subSp. radicans.

211
thus be typified. Greene (1910) interpreted this plant as
near his 1. rufescens or R. littoralis Mearns. According
to Britten (1900), 1. glabrum Mill. "is made a synonym of
fihu§.glabra in Ind. Kew. The character 'foliis ternatis'
is sufficient to Show that this cannot be the case; Since
we have not Miller's Specimens, but the figure which he
quotes from Hort. Elth. represents 5. radicans." Toxico-
dendron glabrum (L.) Kuntze is most likely R. glabra L.
Since Kuntze virtually obliterated Rhgg by wholesale trans-
fers to Toxicodendron. In other words, Miller meant a
poison-ivy plant by his use of Toxicodendron glabrum; Kuntze
did not.

In naming Rhus blodgettii, Kearney (1894) described
clusters of flowers, but later in the paper admitted that
flowers were not seen. His specimen is fairly typical
Toxicodendron radicans subSp. radicans, except for the ab-
sence of pubescence on the fruits, but southern Florida
pOpulations often have glabrous fruits. Barkley annotated
the lectotype as follows:

Dr.’Kearney in the description of the type says,

'The Single specimen which I have seen was col-
lected by Blodgett at Key West, Florida. It is in
the Torrey Herbarium.‘ As this is the only Blodgett
Rhus in the Torrey Herbarium and it fits the descrip-
tion, it is probable that it represents the TYPE.

In regard to Toxicodendron goniocarpum Greene, two
Specimens at US fit Greene's description of the type. I have

chosen the fruiting specimen, McCulloch 4§_(Sheet 440443)
as the lectotype. McCulloch ޤ_(sheet 440444) is a probable

212
isotype (presumably McCulloch numbered Specimens rather
than collections). Greene probably had both Specimens under
consideration since he describes the "rooting freely" of the
sterile Specimen and the nature of the fruits of the second
Specimen.

Many Specimens of this subspecies are teratological
forms, in which the inflorescence proliferates primary and
secondary bracts in place of fertile flowers. As I noted
earlier (Gillis, 1960), "In these aborted flowers, the bracts
grow to several times their normal length (up to 1.4 cm)
and were arranged like the flowers, occasionally Opposite
and occasionally alternate." This monstrosity has been seen
expressed in most taxa of poison-ivy and the poison-oaks,
as well as in other Toxicodendron and.§hg§_species. 'Whether
this is a reSponse to a virus, insect damage, or some other
environmental factor is unknown. It is not genetic in that
it does not appear in the same pOpulation year after year.
It does, however, occur with greatest frequency in 1, Egg;-
gags subsp. radicans, most especially in the Middle Atlantic
and New England States.

This subSpecies is the most common of all poison-
ivy taxa, especially in New Jersey, southern New York, and
Long Island. Along a stretch of road between the Delaware
River and Princeton, New Jersey, poison-ivy forms a roadside
carpet for about 15 miles, covering every patch of ground
not occupied by some protruding plant, a house, or an inter-

section, and climbs every tree on both sides of the road!

213
At Island Beach State Park, New Jersey, in the zone pro-
tected from salt Spray behind the foredunes, poison-ivy
forms a dense, shrubby mat, seldom interrupted, between the
foredune and the roadway for a distance of 10 miles, about
50-75 meters in breadth. The subspecies becomes less common
toward the southern edge of its range, possibly being less
able to compete with the more vigorous vines and Shrubs in
warmer climes.

This subSpecies has greatest affinities with Asiatic
subsp. orientale and hispidum. All pOpulations have pubes-
cent fruits, entire leaflet margins (with, at most, only a
single notch on each abaxial edge), a characteristic tuft
of red-brown or hyaline hairs in the major vein axils on
the lower surface of the leaflets, and ovate or elliptic
leaflets with acute to acuminate apices. SubSpeciesugggi-
gang, however, generally has truncate or obtuse terminal
leaflet bases, the broadest portion of the ovate terminal
leaflet below center, the tufts of hairs in vein axils gen-
erally hyaline or straw-colored rather than red-brown, and
small, globose fruits. It is possible that these three
subspecies represent disjunction of a taxon that was once
contiguous through Eastern Asia and North America.

This subspecies appears to form intergrades with
other taxa where their ranges come into contact. Putative
hybrids, intermediate between subSp. radicans and 1, gig-
bergii, are known from the Berkshire Mountains of western

Massachusetts and elsewhere in southern New England. The

214
plant also appears to form a hybrid with subsp. pubens in
Louisiana, Arkansas, and Texas, and with subSp. negundo
along the Allegheny Ridge (especially Tuscarora Mountain)
in Pennsylvania and the Blue Ridge in Virginia, West Virginia,
and Maryland. (See discussion under HYBRIDS, CROSSES, AND
INTERGRADES; citation of presumed intermediate forms may
also be found under that heading.)

Toxicodendron radicans subsp. radicans has a broad
ecological amplitude. It is most abundant on sand dunes,
along fencerows, in floodplain forests, and old fields. It
manages to exist marginally as drainage becomes extreme,
either with high water table or excessive drainage, such
as borders of peat bogs or salt marshes, wet muck, on podzols,
or on Shallow lithosols. In Nova Scotia, however, one may
find it, in addition to sites mentioned above, on gravelly
barrens. In the Bahama Islands, it grows in pinelands or
hammocks on rocky limestone. In the Florida Everglades,
it inhabits leaf mold on potholed limestone.

This taxon has broad community relationships. In
its northern reaches, it is closely affiliated with the
beech-maple forest and what has evolved from the oak-chestnut
woods following the demise of the chestnut. Along the
Atlantic and Gulf Coastal Plains, it is often abundant in
loblolly pine woods, especially where there is little or
no oak present. Farther south, on the Coastal Plain, it
is found frequently in liveoak (Quercus virginiana) forests

and is reported as forming an association with Persea in

215
northern Florida (herbarium notes). Along the Gulf Coastal
Plain, it is common in southern hardwood forests of Liquid-
amba;.styraciflua,lfly§§a a uatica, Quercus vir iniana,
Fraxinus §p9,, and Age; rubrum.

On sand dunes it is abundant enough to form an asso-
ciation of its own behind the foredunes where it receives
only moderate salt Spray (I have observed that a good 2-day
northeaster in the Middle Atlantic States and New England
is sufficient to defoliate an entire pOpulation in a sand
dune association). On the dunes, its frequent associates
are Parthenocissus guinguefolia, Myrica pensylvanica, and
Smilacina racemosa.

In the Bahama Islands, its chief habitats are rela-
tively dry (no standing water) mangrove swamps associated
with Rhizoohora man le, Erithalis fruticosus, Blechnum
serrulatum, Achrostichum aureum, Bumelia salicifolia,
Chrysobalanus'igagg, and §aba1_palmetto. On some sites,
in relatively undisturbed communities, it favors the "black-
lands": woods of fairly dense Pings caribaea often associ-
ated with Bumelia, Pteridium aguilinum var. caudatum, 235?
gage re ens, 1£gmg_lamarckiana, and 11g; krugiana. On New
Providence, it is reported from the more Open pinelands
called "whitelands." At this extreme limit of its range,
it appears to compete poorly with other Shrubs and vines,
being absent from most thickets. I have searched for it
on some of the eastern islands in the Bahama Archipelago

(San Salvador, Rum Cay, Cat Island) to no avail.

216

Throughout its range, subSp. radicans, as well as
subSpp. negundo and pubens, are constantly associated with
Parthenocissus quinquefolia, ggmu§_§pp,, Fraxinus spp,,
and often with Clematis spp, and AmpelOpsis arborea.

4b. Toxicodendron radicans subsp. orientale (Greene)

Gillis, comb. nov. (Plate XXVII)

Toxicodendron orientale Greene, Leafl. Bot. Observ.
Crit. l: 127. 1905. Rhgg orientalis (Greene)
Schneider, Illust. Handb. Laubholzk. 2: 151.
1912. Type: Maximowicz, §,g, Hakodate, Hok-
kaido, Japan. 1862. (US-19948, not US-l9548
as cited by Greene, 1905, holotype; G-DC, iso-
type).

Bhg§_ambigua Lavallée, Arbor. Segrez., p. 54.

1877. Type: unknown, probably a living plant.

Rhg§_rishiriensis Nakai, Bot. Mag. (Tokyo) 36: 67.

1922. Rhug ambigua f. rishiriensis (Nakai)
Hara, Enum. Sperm. Jap., Part 3, p. 65. 1954.
Type: H, Koidzumi, §,Q, Japan, Rishiri To,
Hokkaido, Oshidomari Mts., growing with‘Abigg
sachalinensis. (TI, holotype).

Leaves bearing tufts of hairs in vein axils on lower
surface, generally light-colored to red-brown. Terminal
leaflet blade 1.2 - 1.7 times as long as broad, widest part
generally at the middle or above middle, 5 - 20 cm long,

3 - 12 cm wide; cuneate at base; petiolule 0.5 - 3.0 cm

long. Petiole 1.5 - 11.0 cm long. Fruits flattened,

217

PLATE XXVII. Specimen of Toxicodendron radicans
subsp. orientale. Type of Toxicoden-
dron orientale Greene.

 

 

218

TVPF SPECIMEN

 

U “EA/7*» n 1‘

r,fi’.c,,,.7, .1557} Ml~‘(‘(t 3],..-” .

B! If”. IOl'll 50!. PC 1 iv? ' - "" 3 3:"
Maximum-iv; Iler ll’d‘l ’1 . HERBARIUM.

I \III IV \I Hi « Ivl |~u:l\ll\l --I \ulm I I II In

I . W
Nu VIM/(Inn A'I‘JIQ‘J II"'f/,!ll ' ’, m/a

4

‘.u.?'\. ’: m.,.-z;
/

(13.-Mn

219
obliquely deltoid, 5 - 6.5 mm broad, Short hairy or papil-
lose. Distribution: Japan, U.S.S.R. (Kurile Islands,
Sakhalin). Common name: tsuta-urushi (Fig. 9).

Representative specimens. JAPAN: Hokkaido. Hidaka

 

 

Prov.: Shoya, Horoizumi-gun, Tokubuchi, §,g. (SAP, TI).
Iburi Prov.: Noboribetsu on road to Noboribetsu Hot Spring,

Gillis 4717 (MSC); Mt. Tarumai, Ito, §,n.

(SAP), Ishikari
Prov.: Mt. Teine, Yanagissawa, §,g, (SAP); Sapporo, 19kg:
bgghi, §,g, (GH); Sapporo, Suqimoto, §,Q, (MSC, TI); Lake
Shikotsu, Gillis 413§_(MSC); Usakuma National Forest, Gillis
41§Q_(MSC); Asahigawa City, Kamikawa, Miyabe, §.g. (SAP).
Kitami Prov.: Shiretoko Peninsula, 119 §1§ (SAP); Shaku-
betsu, collector unknown (TI); Shari, 5ggai,ugfln. (TI);
Rishiri Islands, Oshidomari, Suzuki, Komi a,l§ Shirahama,
I_flg. (SAP); Rishiri Islands, Oniwaki, Shirahama, §,Q, (SAP).
Oshima Prov.: Hakodate, Albrecht, §,g, (GH, MO, NY);
Maximowicz, §,g, (GH, NY, US); Yoshioka-mura, Miyabe g
Tashiro, §,Q. (SAP); Esashi, Sasayama, Mi abe, g Tashiro,
§,g, (SAP). Shiribeshi Prov. along theTomarikawa, Mt.
Shakotan, Shimamaki-bun, Misumi §,Igarashi, §,Q, (SAP);
Okuskiri Islands, collector unknown (SAP). Teshio Prov.:
Teshio, Tsujii g Kozuka, §,g, (MSC, SAP). Honshu. Aomori
Pref.: Mutsui Horosaki, Sakurai, gag. (A); Mt. Hakkoda,
collector unknown (TI). Chiba Pref.: Awa-Amatsu, 1.3 km.
SW of Kameyama, Gillis 461§_(MSC); Chiba, 53331,.gng. (TI).
Eshigo Pref.: Mt. of 5 Peaks, collector unknown (TI).

Fukui Pref.: Echizen, Tsuruga, Kitamura é Murata 293g (MSC).

220

Fig. 9. Distribution of Toxicodendron radicans
subsp. orientale.

 

 

 

 

 

 

 

 

 

 

 

222
Fukushima Pref. Kashi Hot Spring, 53%, 3.3. (TI); Jokogi,
Muramatu, 3.3. (T1); Mt. Bandai, Koyama, 3.3. (MSC). Gifu
Pref. Nakatsugawa, 33331;, 3.3. (TI); Kisogoe Pass, 53335;,
_.3. (TI); Mt. Ibuki, 3.313;, 3.3. (TI). Gumma Pref.: Mt.
Tanigawa, Takeushi, 22,- (TI); Mt. Hodaka, Kanto District,
I_<_a_n3_j._, 3.3. (TI); Oze, Nishida, 3.3. (T1); Usui Pass, 3%,
3.3. (TI). Hiroshima Pref.: Mt. Aza, m, 3.3. (T1).
Iwate Pref.: Moichi, 333:3, 3.3. (T1); Kanagawa Pref.:
Mt. Hakone, 33331, 3.3. (T1). Hata, Bisset M (E).
Miyagi Pref.: Matushima, Faurie £83 (A); Sendai, 935333,
3.3. (TI); Okunikawa, collector unknown (TI). Nagano Pref.:
Shimajimadam, Momose, 3.3. (MSC); Kuzu, Omachi, collector
unknown (TI); Nagawado, 53331, 3.3. (TI); Mt. Hachibuse,
collector unknown (TI); Nagasaki Pref.: Mt. Yatate, Tsu-
Shima Island, Yabe, 3.3. (TI); Irikan, 911131, 3.3. (T1).

 

Nara Pref.: Mt. Odaigahara, Ohwi, 3.3. (WIS). Niigata

 

Pref.: Mt. Azogi, Sado Island, I_(_a_n_a_i_, 3.3. (TI); Niigata,
Yamazaki, 3.3. (TI, US). Okayama Pref.: Okayama, w, 3.3.
(NY). Shiga Pref.: near Kyoto, collector unknown (TI);
Shiga Highland, Egan—a1, 3.3. (T1); Makino, 523131, 3.3. (TI).
Shizuoka Prefant. Amagi, Izu Peninsula, collector unknown
(TI); Mt. Daruma, Izu Peninsula, 133331, 3.3. (TI); Mt.
Ashitaka, 5333; _5_8£_7_ (TI); Mt. Senmai, collector unknown,
(TI); Gotemba, Hisanchi, 3.3. (MSC, TI). Tochigi Pref.:
Chuzenji, Juenberg, 3.3. (A); Nikko, 1(3n_a_i_, 4_o_39 (TI);
Nikko, Gillis 5188 (MSC); Chugiyi, Gillis 4_’78_l (MSC); Sampei-
toge Pass, Mishida, 3.3. (MSC). Tokyo Pref.: Mitake, 153331

223

'132 (TI); Motohachioji,'K3331,I313. (TI); Asakawa,'K3331
2221 (TI); Okutama, 53331_3§§§ (TI); university Botanical
Garden, Gillis 3323 (MSC). ‘Wakayama Pref.: Kii, Murata
13122 (MSC); Mt. Koya,‘g33,.3a3. (A). Yamagata Pref.:
Asahimachi, K3331, 3.3. (TI). Yamaguchi Pref.: collector
and exact locality unknown (TI); Yamanashi Pref.: Asagoduta,
Fuji, Hayata, 313. (TI); Mt. Kenashi, foot of Fuji'K3331,
3,3, (TI); Mt. Riugatake, foot of Fuji, 53331,I3.3. (TI);
Aokihara, Hayata, 3,3, (MSC). Other localities in Honshu,
Prefecture uncertain: Lake Chuging, Wilson 3213 (A, BM).
Ugosakai, Muramatsu, 3,3, (TI); foot of Mt. Nabekamudi,
collector unknown (TI). Mt. Azima, collector unknown (A,
TI). 533333. Fukuoka Pref.: Takosara, Takeda, 3,3, (E).
Kagoshima Pref.: Mt. Kirishima, Wilson, 313. (A). Kumamoto
Pref.: Mt. Asozu,.K3331,.3fi3. (TI); Tuabye Aso-san,
Maximowiqg,'_a3. (BM). Shikoku. Echime Pref.: Mt. Dogo,
Satonii, 3,3, (TI); Uagima, collector unknown (TI). Kochi
Pref.: Nanogawa, collector unknown (TI); Nanogawa, Watanabe,
3.3. (GH).

* * * *.*

”UNION OF SOVIET SOCIALIST REPUBLICS. Kurile Igland .
Kunashir, Rebansa, Shisahama g Takehashi, 3,3, (SAP); Kai-
karamur, Tatewaki, 3,3, (SAP); locality unknown, Vorobjov 22
(LE); Kinakai, Pobedimova, 3,3. (MSC). Shakotan, Mt. Shako-
tan,'Q331.§23 (SAP). Sikotan, Sikotan-yama, VOrobjov,.3fi3.
(MSC). Sakhalin: South Sakhalin Preserve, Nevelsky Region

Kuznepovo, Kravchenko 32 (LE); locality uncertain, Glehn, S.n.

(LE).

224

In naming a Maximowicz collection from Hakodate,
Greene (1905) expressed the idea that his Toxicodendron
orientale was sufficiently different from "any and all New
World Species that it forms a fair subSpecieS." The type
has both flowering and fruiting stems, but suggests two
inaccuracies in Greene's description. He indicated that
the Japanese poison-ivy differs from North American poison-
ivy by its petals not being "nervose and epicarp thinner . . ."
These characters are, however, not borne out on the type
Specimen.

The first poison-ivy from eastern Asia with which
one might be concerned nomenclaturally was described as 5333
ambigua Lavallée, a nomen nudum. Dippel (1892) supplied the
diagnosis. The name was used in connection with Ampeloosis
japonica, a supposed synonym. Presumably Lavallée studied
a living plant in the arboretum and therefore the name can-
not be typified. Because the description is so brief, the
name is frequently-~and erroneously--attributed to Dippel
(1892) whose description is more detailed. Masamune (1934)
erroneously equated this name to 5333 succedanea var. 13333-
133 Engl. in DC. For those who would recognize this taxon
as a Species of 2333, the combination 5333 ambigua may not
be used in any case since it was first published by Enger
(1849) for an Eocene fossil of Italy, a plant not similar
to the one under consideration here, but rather of the genus

Zanthoxxlon of later authors (Edwards and Wonnacutt, 1935).

225

Nakai (1916b) felt that the poison-ivy called 3333
toxicodendron by Miquel (1865, and often attributed to Miquel
as author of the binomial in Japanese literature) and 5333
toxicodendron var. vulgaris (sensu Engler, 1883) should be
.3. ambigua Lav. In his Opinion, 3. orientalis or Toxicoden-
3333 orientale Should be reserved for Chinese or Hokkaido
poison-ivy (Nakai, 1916b and Tanaka, 1916).13 Nakai (1313.)
went on to say that Japanese poison-ivy was not poisonous,
had larger flowers, more brown trichomes in the vein axils,
and more nearly entire leaflet margins in comparison to
American poison-ivy. His concept oflfi. toxicodendron was
essentially that of Linnaeus, and therefore he felt that the
binomial could not be used for any of the climbing forms in
Japan.

Hara (1939) compared Chinese and Japanese poison-ivy
and arrived at the conclusion that they are the same species.
He recognized that fruit pubescence is a critical character,
but pointed out that this feature is variable even among
specimens collected in the same area. He indicated that
Chinese Specimens are occasionally found to be nearly gla-
brous. It is likely that he was examining immature fruits

when he made this observation, for the mature fruits are

 

13Schneider (1901) suggested that Amgelopsis orien-
talis Planch. in DC. is synonymous with'g. orientalis, but
this plant is most likely Parthenocissus tricuSpidata.

14The common name of Japanese poison-ivy, tsuta-
urushi ("eliming urushi") attests to the fact that the plant
is a liana member of the lacquer tree group.

226
hiSpid in all Chinese poison-ivy specimens I have seen.
Although Hara did not believe that the narrowness of the
leaflets could be used to separate the two taxa, I feel
that the ratio of length-to-width of the terminal leaflet
is of taxonomic significance, and not narrowness (i.e.,
width) alone. This ratio is reasonably uniform within each
of the two pOpulations and therefore of importance in sepa-
rating the two. The terminal leaflet in the Japanese pOpu-
lations is always less than twice as long as broad (1.2 -
1.7), whereas the Chinese pOpulations are characterized by
terminal leaflets closely approaching a length twice the
breadth (1.9 - 2.0 times). The more subtle difference of
a cuneate leaflet base in Japanese forms as Opposed to an
attenuate base in Chinese forms is leSS useful, but never-
theless seems to be a consistent character.

Nakai (1922) described a new Species of poison-ivy
from northern Hokkaido and the Rishiri Islands which he
called 5333 rishiriensis (tachii-tsuta-urushi). Hara (1954)
reduced this Species to a form: '5333 ambigua f. rishiriensis.
Koidzumi (1924) described its habitat as being the conifer
forest of the Oshidomari Mountains of Rishiri To at 1050
meters. The taxon's only distinguishing character is that
it does not appear to be scandent. Whereas the habit is gen-
erally considered to be reliable in poison-ivy taxonomy, in
this case it is considered suspect. At the moment there
seems to be little to verify its restricted growth on genetic

grounds. The holotype (at T1) is indistinguishable from all

227
other Japanese poison-ivies. It is most likely not cognate
to Toxicodendron rydbergii of North America which, like its
Japanese counterpart, does not climb and curiously similarly,
does not grow south of the 44th parallel of latitude (Gillis,
1962). Forma rishiriensis appears impossible to characterize
outside of the field, has not been proven to be a non-climbing
plant at all times, and is inadequately collected. For the
present it must be regarded as an ecological form from the
windswept insular areas of northern Japan and far eastern
U.S.S.R., not to be recognized nomenclaturally.

Toxicodendron radicans subSp. orientale grows in
dense forests in Japan where it apparently Sprouted when the
forest was young, simply growing up with the forest. Repro-
duction within the forest is probably mostly restricted to
individuals with some foliage in good light situations.

From central Japan northward, it is often found in fir
forests. It is far less of a weedy Species than its North
American counterparts. Rather, it is a denizen of forest
communities of the hills and mountains. ‘Where there has
been disturbance, such as a road cut, it becomes more abun-
dant.

In northern Japan and the Kuriles, it is common in
the forests of 21333 jesoensis and 33133 sachalinensis along
with another liana, Actinidia kolomikta. In central Honshu
where Chosenia bracteosa forms large forests, it is also a
common woody climber (Tatewaki, 1950). Elsewhere in Japan,

it is found in the pine-fir-redwood woods (Pinus densiflora,

228
2, thunberqii, 33133 sachalinensis, nggtomeria japonica).
In addition, Tatewaki (1958) has reported it as a typical
inhabitant of the associations characterized by either
Aesculus turbinata, £3333 crenata, Pterocarya rhoifolia,
Cercidoghyllum japonicum,'3333H3333, and Qgercus crispula.
Its chief associates in most of these associations are also
vines: M1113.coignetiae, Hydrangea petiolaris, Actinidia
arguta, and SchiZOphragma hydrangeoides primarily, and
secondarily Celastrus orbiculatus var. papillosus, 3331-
sandra chinensis, and Euonymus fortunei. In the woods of
the B050 Peninsula highlands (Chiba Pref.) the most common
trees of the forests in which'1. radicans subsp. orientale
grows are Quercus serrata, Castanea crenata, Lithocarpus
edulis, 33133 11133, and 21333 thunberqii.
4c. Toxicodendron radicans subSp. hispidum (Engler in
Diels) Gillis, comb. nov. (Plate XXVIII.)
3333 toxicodendron var. hisgida Engler in DielS, Bot.
Jahrb. Supt. 29: 433. 1900. Type: 'Q.'3333'333
3, y, Rosthorn 874. "Bamm (Sic) 30', l" circ.,
Wald. Lung mO-ai. Nanch'uan, China (Szechuan),
11 Sept. 1891." (O, lectotype).
'3333 intermedia Hayata, J. Coll. Sci. Imp. Uhiv.
Tokyo 23 (art. 19): 73. 1908. Type: '1..§3333-
3131 £3 [_J. 31331 2923. Taiwan, Nantou County,
Mt. Morrison, 7500 feet elevation, (TI, lectotype)
Widest part of the terminal leaflet blade generally

at the middle or above the middle, nearly twice as long as

229

PLATE XXVIII. Specimen of Toxicodendron radicans subsp.
hispidum. Lectotype of Rhus toxicoden—
dron var. hispida.

 

 

 

     

4"» K
.1. W 73272.. .. Lu.

2.1144“- ‘4 MA- sr-l-I n.

TYPE COLLECT IO
4A1. 1'2».4~

    
 

17v.
Illa)”, ”v“. .27 .4”

I’n~\‘\. J“ , l A...

1.....4 ; .. .

 

llm- Wm hula]!!! um I hrbllunlrlhl‘ I'I.u./.~r
rlym: mm IIIIHH‘ ‘1'“ m: u .1), HM ”In .u/J. . m-

4 I Hm]: (I .I I lflr~(’/Hrll ”mum/mm!”

M... 7:..-:.erJv“ Z w
14" 1‘ 1’ ‘l‘Jz "“7‘

 

231

broad, attenuate at the base, 6 - 20 cm long, 4 - 9 cm
wide; petiole 3 - 11 cm long, terminal petiolule 0.3 - 3
cm long; tufts of hairs in vein axils below light-colored
to red-brown. Fruits flattened, obliquely deltoid, 5 -
6.5 mm brad, hiSpid with hairs nearly 1 mm long. Distribu-
tion: ‘Western and central China and Taiwan. Common name:
Taiwan tsuta-urushi. (Fig. 10.)

Representative Specimens. PEOPLES REPUBLIC OF CHINA:
333:3. Fang Hsien,'Wilson 233,(A, E, GH, US). Chanhyang,
Wilson 1123 (A, E, NY, US). Patung District, Ichang, 33313
3331_(K, NY), Locality unknown, 33311_333§ (A, E, BM, GH,
MSC, US). Sikang. Huping, Paohing, Wilson 3331 (A).
Szechuan. Nanchuan, 3333,3fl333 Rosthorn §13_(0); 1333.323.
(A, E, MSC); 33333.3§_(A). Mt. Wa, Wilson 2333 (A).
Tachienlu, Pratt, 3,3, (K). Cheng-Tu Plain,;333g,22333_(GH).

 

Yunnan. Liang-Shan I'cho,l1331 31233 (A). Localitx unknown.
“imam—5 00- was (A)-
* * * * *
REPUBLIC OF CHINA (TAIWAN). Chiayi Co.: Mt. Arisan
(also Spelled Alishan), collector unknOwn (TAI); Wilson
1Q§Qg.(US). Linsley-Gressitt 132 (BM, NY, U); Hayata (MSC);
113, 1133, 333.2321 (MSC). Ilan Co.: Mt. Oobi, collector

unknown (TAI). Mt. Bombon, Eizi, 3,n. (TI). Nantou Co.:

 

Saramao, Eizi 224 (TI); Mt. Taiheizan, Suzuki, 3,3, (UC).
Taichung Co.: Pai-koo-Shan, collector unknown (TAI); Sing
San, Liu, Liao, KuO, S.n. (MSC). Taipei Co.: Taikoku-syfi,

collector unknown (TAI); Taitung, Pa-Hsien-shan, collector

232

Fig. 10. Distribution of Toxicodendron radicans
subsp. hisgidum.

 

 

        

 

 

H. . ltipfifil
a f; x. .5... Eu...
.1 5 d. _ W M
_/ 28.3.. 3.8.31 9.4 :36
r.\.\/.\S J a; 335°
. 2422‘ 1., V
~ u. m.) \
‘<z_zoooz_ Tran!
.. . .\
. (.(\ Iv [Wu/“J!
. a)??\.
........ m IRA -- .\
...... ....qu x N315 ’-:-u-:vl..ryln-v 5.!
IV 1%. . .,\(( 2‘22... ..W .................
JCT. \ . \r\ . !!!!!
A )2. \ Ill
. I» IIIIII:
‘ W ~xoro_u;xrb \ /,

“.232fo 21.23... \...\ _ < I:

~ (III:

. . . ”(H . o z I/

r“.\ k m1) K:.w / J _ 1
L)+\.J.l

   

é . 3; 1. t\ J .7.\A.> s
. < i r u u N m . + /.I.\.u‘.\ lbx. A.) .r./
a .
\ . 4. (,
a»; :14w1uaaz . '02 z.m ./ . Jr
.1333. Vijw. ./. + x)
V ..- / If
\/.v 7 . ./
..a/ mi. 152., _ / ,1 f.
7% r . (...). . . .r
Sidw z<zox _mzuxm M/l/l l. l/,/ A //
. 2.” x z m m I L ... h.

255

 

234
unknown (TAI). Exact locality unknown: Niitaka (TAI).

A type was not designated by Engler for 2333_Toxico-
dendron var. hisgida, although several Specimens were cited
in the original description. These were all destroyed by
the 1943 bombing of the Berlin Herbarium. The only known
extant dUplicate of any of the cited Specimens is the one
at Oslo, hereby designated a lectotype.

Poison-ivy in Taiwan was namedlfi. intermedia by
Hayata (1908), who thought it was intermediate between sect.
Trichocarpae and sect. Venenata of the genus 5333 (sensu
Engler), due to the resemblance of the bristly trichomes of
the Taiwan poison-ivy fruits to those of Toxicodendron 331-
chocarpum (Miq.) Kuntze of Engler's sect. Trichocarpae.

Even those who would retain this taxon as a Species segre-
gate of 3333 cannot use 3333 intermedia since this combina-
tion has been antedated by 3333_intermedia Ettingshausen
(1888) for a Miocene fossil of Styria, Austria, a plant not
conSpecific with the taxon under consideration here.

In China poison-ivy is known from the central and
western mountains, bordering on the Yunnan and Kweichow
Plateaus and the foothills of the Tibetan Plateau in the
provinces of Szechuan, Hupeh, Sikang, Kweichow, and Yunnan.
There is a hiatus in the distribution of Chinese poison-ivy
between these western mountains and the highlands of Taiwan,
with none known in the intervening hills and plains. As Li
(1963) has noted, there are phytogeographic links between

the mountains of Taiwan and those of central and western

235

China. In Taiwan, the plant is found growing on the Paleo-
zoic slates of the central mountains.

The ripe drupes of Toxicodendron radicans subSp.
hispidum possess striking hiSpid hairs, often nearly 1 mm
in total length, the longest fruit hairs in the genus. 1331-
codendron radicans subSp. orientale has fruit with far Shorter
trichomes, Often no longer than those of fruits of eastern
North American taxa. On occasion, the fruits are papillose,
rather than pubescent. As mentioned elsewhere, the papillose
condition is interpreted as a form of incipient pubescence,
the papillae being evaginated or reduced hairs. More than
likely, mature fruits will be pubescent. The tufts of hairs
in the vein axils on the lower surface are red-brown in about
60% of the Specimens examined, hyaline or straw-colored in
the remainder.

The closeSt relative of this subSpecieS is Toxico-
dendron radicans subsp. orientale of the Japanese Archipelago,
which was probably derived from it. The three foregoing

taxa, 1, radicans subSpp. radicans, orientale, and hispidum,

 

15Attempts on the part of the writer to study the
poison-ivy complex in China have been frustrated by the
current political feud between the United States and the
PeOples Republic of China. CorreSpondence, both in English
and in Chinese, has been sent to the several botanical gar-
dens and herbaria in China listed in Index Herbariorum
(Lanjouw and Stafleu, 1964), all to no avail. A professor
of biology at Nanking University was willing, but unable,
to be of help. Even intercession on my behalf by journalist
'Edgar Snow with Premier Chou En-Lai yielded no Specimens nor
information. Because of these political difficulties, no
Specimens from any herbaria in the PeOples Republic of China
were examined.

236
are very Similar morphologically, almost enough to form a
separate cluster of taxa from all the other subSpecieS.
All these plants have pubescent fruits, entire leaflet
margins (with at most a Single notch on each abaxial edge),
the characteristic tufts of hairs that are red-brown or
hyaline in the major vein axils on the lower surface of the
leaflet, and ovate or elliptic leaflets with acute or acu-
minate apiceS.
They differ, however, by several reliable characters.

The North American taxon has truncate or obtuse terminal
leaflet bases; the broadest portion of the terminal leaflet
is below center making the leaflet ovate; the tufts of hairs
in the vein axils are generally hyaline or straw-colored
rather than red-brown; the fruits are globose, 3 - 5 mm
broad. The two Asiatic subSpecieS, on the other hand, have
attenuate or cuneate terminal leaflet bases; the broadest
portion of the elliptic or obovate) terminal leaflet is
above the mid-point; the flattened fruits are Often obliquely
deltoid as in the fruits of other Toxicodendron Species in
the Far East; the fruits are 5 - 6 mm broad; and the hairs
in the vein axils are generally red-brown. The leaflets are
generally smaller in subSp. radicans than those in the Ori-
ental subSpecieS, although there is considerable overlap in
Size.

4d. Toxicodendron radicans subsp. gubens (Engelm. ex

Watson) Gillis, comb. nov.

Rhus toxicodendron var. Rubens Engelm. ex Watson,

237
Bibliog. Index, p. 185. 1878. ‘3333 radicans
var. Pubens (Engelm. ex Watson) Fernald, Rhodora
43: 596. 1941. Type: 3, Lindheimer 231, High
in thickets, not climbing, fl. odorous. New
Braunfels, Comal Co., Texas. August 1846. (GH,
lectotype; M0-1773730, MO-l773731, US-19858,
isotypes).

3333 toxicodendron var. multifolia Vines, Trans.

Texas Acad. Sci. 26: 75. 1943. Type: Robert
3,'!1333 B-3092. Bray's Bayou, Houston, Harris
Co., Texas. 50 feet. July 1939. (SMU, holo-
type; MSC-190l69, isotype).

Vine or shrub with aerial roots. Leaflet blade
ovate, serrate, notched, or occasionally with one lobe on
each edge; scabrous or rarelnglabrous on the upper surface,
often with a line of minute, curly hairs on the midrib-
densely strigose, hirsute, or velutinous on the lower sur-
face, velvety to the touch; without tufts of hairs concen-
trated in the vein axils. Terminal leaflet 3 - 20 cm long,
3 - 11 cm wide, acuminate; petiole 2 - 13 cm long, hiSpid,
villous, or tomentose; terminal petiolule 0.3 - 5.0 cm long;
veins on lower surface light in color, usually prominent
and raised above the surface. Fruits glabrous, often glau-
cescent, rarely with incipient papillae or hairs, 3.0 -

4.5 mm broad. Distribution: Former Eocene Embayment region
of Lower Mississippi Valley west to East Texas, states of

Mississippi, Louisiana, Arkansas, Missouri, Texas,

238

southeastern Kansas. (Fig. 11.)

Representative Specimens. UNITED STATES: ArkansaS.
Arkansas Co.: Arkansas Post, Demaree 21333_(MO, NY, SMU);
St. Charles, White River National Wildlife Refuge, Demaree
32222 (MSC). Chicot Co.: along US-82 adjacent to Ditch
Bayou, Gillis 3232 (MSC). Crittenden Co.: West Memphis,
Demaree 11112 (GH). Drew Co.: Bayou Bartholomew, Tillar,
Demaree 13333 (MIL, MIN, NY, SMU). Faulkner Co.: Conway,
33333, 3,3, (MSC). Garland Co.: Hot Springs National Park,
Thousand Dripping Springs, Demaree 33333 (MSC). Hempstead
Co.: Fulton, Demaree 32121 (SMU). Henry Co.: ‘Windsor
Spring, Hitchcock, 3,3, (M0). Howard Co.: Ouachita Region,
Baker Springs near Umpire, Demaree 33123 (MSC). Lee Co.:

La Grange, Mariana-Helena Recreational Area, Demaree 31232
(MSC). Logan Co.: base of Magazine Mountain, near Blue
Mountain, Palmer 23312 (A). Marion Co.: Yellville, Demaree
31212 (MSC). Montgomery Co.: Mount Ida, Demaree 31233 (MSC);
Norman, Demaree 33333 (MSC). Poinsett Co.: Marked Tree,

edge of river, Demaree 32232. (MSC). Polk Co.: Big Fork,
‘33333,.313. (MSC). Pulaski Co.: Arkansas River bottom

below Natural Steps, Demaree 2332_(A, GH, MO, NY). Scott

Co.: Mill Creek Recreation Area, 5 miles east of Y City,
Donahue 222 (MSC). Searcy Co.: Marshall, Gillis 3133 (MSC).

Ken ck . Fayette Co., Lexington, Hille 2, (MSC). 13uisiana.

 

Avoyelles Parish: Bunkie, Thieret 3 Williams 17363 (MSC).
Bossier Parish.: Bossier City, Gillis 3773 (MSC). Caddo

Parish: Shreveport, Cotx, 3,3, (A). Catahoula Parish:

r‘

239

Fig. 11. Distribution of Toxicodendron radicans
subSpp. gubens and verrucosum.

 

 

240

 

 

2.0133 6.1... ucoo:=. .b 4
£3.02. to: .32...- 030030. .h C

    

\
r ,
I

A,

(

 

 

241
Sicily Island, Gillis 3233_(MSC). Iberia Parish: Jefferson
Island, Thieret 11323 (MSC). Lafayette Parish: Lafayette,
Thieret 11313 (MSC); Madison Parish; Tensas River terrace,
Tensas Gun Club Refuge, 3333.11313 (MO). Morehouse Parish:
4.8 miles north of Oak Ridge, Gillis 3233 (MSC). Natchitoches
Parish: small creeks near Clarence, Demaree 32332 (MSC).
Orleans Parish: New Orleans, Fendler, 3,3, (M0); New Orleans,
Drummond 13 (E). Plaquemines Parish: Pilottown, Delta
Wildlife Refuge,.3333‘11333_(SMU). Rapides Parish: Alex-
andria,'3311 322 (F, GH, MIN, MO, NY, US). Richland Parish:
north of Girard, Gillis 3232.(MSC). St. Landry Parish:
Bayou Carron, near Plaisance, Thieret'3 Williams 11333 (MSC).
St. Martin Parish: St. Martinsville, Langlois, 3,3, (MIN);
Lake Martin, Thieret 11322 (MSC). St. Mary Parish: Cote
Blanche Island, Thieret 12333 (MSC). Webster Parish: lo-
cality unknown, 31333 3332 (A). West Feliciana Parish:
Catalpa, Pennell 3212 (NY). MiS3iSSigg'. Adams Co.: 6
miles east of Natchez, Gillis 32§§ (MSC). Leflore Co.: 3.7
miles NE of Itta Bena, Gillis 322Z_(MSC). Quitman Co., north
of Marks, Gillis 3212 (MSC). Sunflower Co.: Moorhead,
Gillis 3232 (MSC); Shore of Sunflower River between Moorhead
and Indianola, Gillis 3233 (MSC). ‘Washington Co.: east of
Leland, Gillis 3233 (MSC). Missogri. Carter Co.: Big
Spring State Park, Gillis 3123_(MSC); west side of Current
River north of Big Spring State Park, 522E.§ Redfearn 333
(FSU). Henry Co.: Windsor Spring, Hitchcock, 3,3, (MO).
Iron Co.: Vulcan, 33113_331D(F). Jackson Co.: Courtney,

242
Bush 13892 (A, MIN, MO, ND, NY). St. Louis Co.: Kirkwood,
Woodson 738 (MIN). Oklahoma. Leflore Co. Page, Palmer

 

20948 (A, US). _Zgnnessee. Rutherford Co.: Rockvale,

 

limestone glades, Demaree 47568 (MSC). Shelby Co.: Cape-
ville, Gillis 5901 (MSC); south of Memphis, Gillis 5903
(MSC). Texas. Aransas Co.: 3 miles NE of Rockport. Jones

 

323 (SMU); Aransas National Wildlife Refuge, Traverse 3333
(F, SMU, US). Bexar Co.: San Antonio, 33532,'§flg. (GH, NY).
Brazoria Co.: Columbia, 3333.332 (A, MO, US). Brazos Co.:
College Station, 23££3_23233D(SMU); Cherokee Co.: Larissa,
3333 3332 (A). Colorado Co.: 5 miles NE of Columbus on
stream bank, Shinners, 14612 (SMU). Comal Co.: New Braun-
fels, Lindheimer 231 (GH, MO, US). Cooke Co.: Gainesville,
.3331,33323 (SMU). Hardin Co.: Sour Lake, Lundell 3 Lundell,
3,3, (NY). Harris Co.: Houston, Fisher 33331 (US); North
Shore Development, Lake Houston, Traverse 3331 (F, US);
Bray's Bayou, Houston, 23333H232, B-3086, B-3092 (SMU, MSC,
US). San Patricio Co.: Sinton, 33333 233 (SMU); Sinton,
Traverse 3233_(F, SMU, US). 'Wilson Co.: SE of Floresville,
MWWY).

The name gubens is attributed to Engelmann in Watson's
Bibliography (1878). It is accompanied by a reference to
Gray (1850) who supplied a description in Latin. No type
Specimen is cited. After Gray's discussion of Lindheimer's
collection of 3333 verrucosa (see treatment on that t0pic
under Toxicodendron radicans subsp. verrucosum), he discussed

a Specimen which he called only Rhus toxicodendron var.

243
A specimen at the Gray Herbarium fits his description and
furthermore was probably seen by Gray. It is labeled "3333
toxicodendron var. pubens" in Engelmann's handwriting, prob-
ably the source of Watson's attribution of the name to
Engelmann. It is the one specimen most likely seen by
Watson, the publishing author. Fernald (1941) felt that
this specimen was "either the type or isotype." I herewith
designate it as the lectotype.

The type comes from the western end of the range of
this subspecies where its range overlaps that of subsp.
verrucosum. The Specimen on the lower right of the type
sheet has been annotated as 3333 radicans var. verrucosum
(Sic) by Fernald, but it is too pubescent for that taxon.
It is more likely an intergrade between subSp. gubens and
subSp. verrucosum, bearing the pubescence of the former
and the lobing of the latter.

The namelfi. toxicodendron var. multifolia Vines is
not validly published (no Latin diagnosis), ungrammatically
named (Should be multifoliata), and later abandoned by its
author (Vines, 1960). Extra leaflets occur Sporadically
on individuals in the poison-ivy complex, but are not of
pOpulation significance, chiefly since plants with extra
leaflets generally possess the normal trifoliolate condition
as well. The type specimen indicated by the author comes
from a park in the city of Houston, Texas, although it was
not marked as the type. It was cited as being in Vines's

personal herbarium at the time of the publication of the

244
name, but this herbarium has subsequently been incorporated
into the SMU Herbarium.
Subspecies pubens is distinguished from subsp.
negundo by its pilose leaves which are broad, the terminal
leaflets often being as wide or wider than long. It occurs

in the western Gulf Coastal Plain where subSp. radicans is

 

also found. It may be distinguished from subSp. radicans
by its glabrous fruits and pubescent leaves. Furthermore,
the leaflets of subSp. gubens usually have coarse teeth on
the margins, whereas those of subsp. radicans are entire.
In Arkansas and Missouri, this subSpecies appears to form
hybrids over a large area with subSp. negundo and to some
extent with subSp. radicans. At least there is Significant
diversity over a large area. (See discussion under "Hybrids,
Crosses, and Intergrades.")

The subSpecies is most distinctive in the former
Eocene Gulf Embayment, particularly in what is locally called
the "delta" region of northwestern Mississippi, southeastern
Arkansas, and the eastern part of Louisiana, i.e., the
cotton-growing region. From the Mississippi River eastward,
the subspecies is replaced abruptly by subSp. radicans in
the pine region of southern Mississippi. The transition to
other taxa is more blurred to the north and west in Tennes-
see, Arkansas, Missouri, Louisiana, and Texas.

This plant inhabits the southeastern evergreen for-
est of the Mississippi alluvial plain. ItS tree associates

are most often Liquidambar styraciflua, Quercus virginiana,

245
and 33333 crassifolia. Because of the tendency of the plant
to be weedy, it often occupies a host of ruderal sites,
regardless of the plant associations involved. E. Lucy
Braun (1950) describes it as a shrubby component of the oak-
hickory forest of the Mississippi embayment section. Its
soil preferences are those of terraces of the Old Mississippi
Valley south to the present delta. In a few instances it
is known from highly calcareous lithosols.
4e. Toxicodendron radicans subsp. barkleyi Gillis, subSp.
nov. (Plate XXIX.) Vitis vel frutex radicellibus
aeribus. Folioli integri vel subintegri aut in-
cisuribus interdum, glabrati usque velutini pagina

SUpera folioli, strigosi, pilosi dense, vel tomen-

tose pagina inferna. Fructi glabri aut puberuli.

Differt ab subsp. divaricato foliis pubescentibus

et foliolis aliquantum latioribus, et ab subSp.

pubenti foliolis integribus vel subintegribus et

foliolo terminali elliptico seu lanceolato.

Vine or Shrub with aerial roots. Leaflets entire or
subentire or with occasional notches, ovate to elliptic or
elliptic-lanceolate, glabrate to velutinous on the upper
leaf surface, strigose, tomentose, or densely pilose on the
lower surface. Terminal leaflet blade 3 - 14 cm long, 10 -
10 cm broad, petiole 2 - 12 cm long; terminal petiolule 0.5 -
5 cm long. Fruits glabrous or puberulent, 3.0 - 5.0 mm

broad. Stem of the season pilose to wooly. (Fig. 16.)

246

PLATE XXIX. Holotype of Toxicodendron radicans
subsp. barkleyi Gillis.

 

 

 

MICHIGAN sun
1 2' 1 ‘ w.‘ l

UNIVERSITV HERBARIUM

 

 

247

 

hvzuch~ un- uu-u-‘Ivv _,n-‘IW.

I'LANT8 UF (Il' \TEHALA

Twcodondrm radium (L.) o. Kunt-

 

lnblnr an Huer’us and othar l'
{Tequont in a I011 urn. Callod
a "Hausa-111'.

‘Chanr'
Mm u. lie-mun No J11? 6 hum-t 1959

Ital-WI.- lat-r0—
IIK‘IIOAN "A“ Um

248

Fig. 12. Distribution of Toxicodendron radicans
subsp. negund .

 

249

 

 

 

 

 

 

250
Representative Specimens: MEXICO. Chiagas. 1 mile
north of Jitotol on road to Pueblo Nuevo Solistachuacén,

Breedlove, 333. (DS). 3istrito Federal. Chapultepec Park,

 

Ciudad Mexico, Bourgeau 23 (GH); Gillis 3726 (MSC). Durango.
North of Coyotes, Quebrado de San Juan, Maysilles 7194

(MICH, US). Guerrero. Ayahualtempam near Chilapa, Sessé‘3

 

Mocino 233 (MA). Omilteme, Nelson 1332 (US). Vallecitos,
Montes de Oca, Hinton 33333 (ARIZ, F, GH, NY). Mina, Hinton
33322 (A, MO, NY). Hidalgo. Rancho Viejo, Municipio de la
Misidn, Gonzélez Quintero 3332 (MICH, MSC). Tecozantla,
Salazar, 3.3, (US). Jalisco. Headwaters of Rio Mascota,

4 - 7 miles south of El Rinc6n, road to sawmill called LaCum-
bre, McVaugh 23323 (MICH). Summit of plateau and steep
mountainsides above Amacueca, road to Tapalpa, McVaugh
23333_(MICH). Summit of pass 7 - 8 road miles NW of Los
Volcanes, between Ayutla and Mascota, McVaugh 32323 (MICH).
Méxicg. Tenancingo, Salazar, 333. (TEX). Iztaccihwatl,
Purgus 232 (MO). Morelos. Cuernavaca, Trelease 333.(ILL).
Tetecala, Urbino, 333. (TEX). Valle del Tepeite, Lyonnet
231 (US). Na ri . Oak forest 9 miles north of Compostela,
McVaugh 33131 (MICH). Arroyo 3.5 miles SW of Jalisco, road
to El Malinal, McVaugh 33231 (MICH). 33339 Lefin. Alamar

 

about 15 miles SW of Galeana, Mueller 3.Mueller 133 (A, F,
TEX). 10 miles south of Monterrey, Hennen 31:23 (MSC).
Horsetail Fall, SW of El Cercado, 24 miles south of Monter-
rey, DeJong 3333 (MSC, MU). Cerro del Silla, near Monterrey,

Nelson 6688 (US). Oaxac . San Miquel Alborrados, Nelson

251
333 (US). Cuyamecales, Distrito de Cuicattan, Conzafta 3
3533325, _S_.3. (TEX). 334 Lg__,i__s Potosi. 1.5 miles east of E1
Plantanito, 20 miles east of Ciudad del Maiz, McVaugh 33332_
(MICH). Tamaulipas. Victoria, Palmer 223 (F, GH, MO, NY,

 

US). Veracruz. Orizaba, Botteri,.§33. (US). Colipa,
Liebmann, 3,3, (NY), Encinal, Matrata, Matuda 1368 (A, F).

 

 

Jalapa, Smith 1897 (GH). Other Mexigg. Valle de La Lagima.-

 

Martinez, 333. (TEX).

*****

GUATEMALA. Alta Vgrapaz. Coban. Turchheim 111784

 

(GH, NY, US). 33 Quich . Between San Francisco and Cotzal,

Sharp 45178 (F). ggatemala. San Juan Sacatepequez, Lewis,

 

3,3, (F). flgehuetenagg=. Paso de Boquer6n along Rio

 

Trapichillo below La Libertad, Steyermark, 33323 (F, TEX).
Los Pintos, SW of Huehuetenango, Steyermark 33333 (TEX).
Near ruins of Zaculeu on bank of Rio Zaculeu, Beaman 3332
(DUKE, ENCB, GH, MSC, NY, TEX, UC, US).

This subspecies differs from subsp. divaricatum by
its pubescent leaves and its somewhat broader leaflets, and
from subSp. gubens by its entire or subentire leaflets and
its elliptic or lanceolate terminal leaflet. The subspecies
is named for Dr. Fred A. Barkley, a long-time student of the
Anacardiaceae and one whose earlier (1937) monographic studies
in the poison-ivy complex have been a valuable source of‘
information to this study. Distribution: Mexico; moderate
elevations in Sierra Madre Oriental south to Sierra de los

Cuchumatanes in Guatemala.

252

Type: 3333 3, Beaman 3332. Guatemala, Dept. of
Huehuetenango. Near ruins of Zaculeu on high bank of Rio
Zaculeu, 1850 m elevation. Climbing on Quercus and other
trees; frequent in a local area. Called “Chang" and "Ma-
tazanal." 6 August 1959. (MSG-171021, holotype; DUKE,
ENCB, GH, NY, TEX, UC, US, isotypes).

3333 villosum Sessé and Mocino (as uillosum), Pl.
Nov. Hisp., p. 47. 1887, non 3333'villosa L.f., Suppl., p.
183. 1781, nec Hochet ex A. Rich., Tent. F1. Abyss. l: 145.
1847, nec E. Mey. ex Engl. in DC., Monogr. Phaner. 4: 427.
1883, nec Engler, Bot. Jahrb. Supt. 24: 501. 1898. Type:
33333 333 Mocino, Castillo et Maldonado, 1787-1795-1804,

.33. 333. "Habitat in Ayahualtempan prope Chilpam et in
capite Bonae Spei." (MA, lectotype).

This subSpecies iS found in the Sierra Madre Orien-
tal, the Sierra Madre del Sur, of Mexico and the Sierra de
los Cuchumatanes of western Guatemala. It forms intergrades
with subSp. divaricatum in the trans-Mexican volcanic belt
as far west as southern Jalisco. It is the eastern Mexican
counterpart of subSp. divaricatum, which occurs in western
Mexico although it was probably derived from an ancestral
pOpulation of subsp. gubens. The eastern subSpecies has
very pubescent leaves, whereas the western one is character-
ized by glabrous leaves. The leaflet shape and prominence
of veins are shared by both subSpecieS.

The Sessé and Mocino name 3333 uillosum is an enigma.

It has been thought to be a miSprint for villosum because of

253
the frequent interpolation of u's and v's in earlier writing.
This SUSpiCion is eSpecially cogent since the leaves are
often villous. The name is herein considered to be an ortho-
graphic variation of villosum.

This subspecies grows chiefly on soils high in cal-
cium, such as the limestones of the Cuchumatanes Mountains
and other limestone and calcareous outcrOppingS of Mexico.

It was found to be abundant on the edge of an orchard in
Galeana, Nuevo Léon, Mexico, where the soil calcium was
higher than in any other soil studied: Up to 5700 ppm.
4f. Toxicodendron radicans subSp. negundo (Greene)
Gillis, comb. nov. (Plate XXX).
Toxicodendron negundo Greene, Leafl. Bot. Observ.
Crit. l: 117. 1905. 3333 toxicodendron (subsp.)
negundo (Greene) Gates, Trans. Kansas Acad. Sci.
41: 106 (nomen nudum), 1938, republished legiti-
mately, Trans. Kansas Acad. Sci. 42: 13. 1939.
3333_radicans var. vulgaris f. negundo (Greene)
Fernald, Rhodora 43: 594. 1941. Type: 3,l3.
Norton 13. Low woods in Riley County, Kansas.
28 September 1895 (in fruit) and'1. 3. Norton 133,
same locality, 1896 (in flower). (US-352742,
US-352743, syntypes; GH, KSC, MO-l7737l7, MO-
1773718, NY, RM-18888, RM-l9505, isosyntypes).
Toxicodendron aboriginum Greene, Leafl. Bot. Observ.
Crit. l: 125. 1905. Type: '1. M, Bigelow, 3,3,,
Indian Territory (Oklahoma), Choctaw.Agency,

254

PLATE XXX. Specimen of Toxicodendron radicans subsp.
negundo. Type of Toxicodendron negundo
Greene.

 

 

255

’(LutAfl‘NQ‘O—dl‘d‘xhx uz_
~LL/Lf - 6,1929
.AJ"$_M€XK‘4‘A'

   

UNITLD STATES NATIONAL HERBARIUM
, ‘ ‘ . ‘ ‘ H

PLANTS OF KANSAS.
n- :"v' 4 EMA

.'V) ') '5: b ‘V‘J‘

v

911971-317
1211.112»...

TYPE SPECIMEN

 

 

256
1853. (US-19845, holotype).

Vine or shrub with aerial roots. Leaflet blade
ovate to elliptic, generally notched or serrate, often
lustrous in sun forms; glabrate to strigose on the lower
surface or with pubescence confined to the veins, not vel-
vety to the touch; no tufts of hairs concentrated in the
vein axils below; glabrous on the upper surface with a line
of minute curly hairs for several cm along the mid-vein
base. Terminal leaflet blade 3.0 - 20.0 cm long, 1.0 -
12.0 cm wide, petiole 2.0 - 16.0 cm long, glabrate, puberu-
lent, hiSpid, or villous; terminal petiolule 0.5 - 6.5 cm
long. Leaflet tip acute or acuminate; base obtuse or trun-
cate. Fruits glabrous or with a few scattered hairs, 2.5 -
5.5 mm broad, cream-colored to tan. Branches glabrate to
densely puberulent. Distribution: Canada: southern lobe
of Ontario: United States: Great Lakes States south to West
Virginia, Kentucky, Tennessee, Arkansas, Oklahoma, west to
eastern Kansas, Nebraska, Iowa. (Fig. 13.)

Representative Specimens: UNITED STATES. Arkgnsa .

 

Baxter Co.: Norfork, rocky bluffs above Norfork River,
Gillis 3132 (MSC). Benton Co.: War Eagle, Demaree 3132.
(GH, NY, TEX, US). Clark Co.: Okolona, Demaree 31233 (MIN,
MO, NY). Cleburne Co.: Heber Springs above Greer's Ferry
Dam, Demaree 33333 (SMU). Crittenden Co.: West Memphis,
Demaree 33333 (SMU). Garland Co.: Hot Springs, 33333 2323
(F); Hot Springs, Demaree 31333 (MO); Magnet Cove, Palmer
23331 (A). Independence Co.: Batesville, Demaree 31333

257

Fig. 13. Distribution of Toxicodendron radicans
subspp. eximium, divaricatum, and verru-
cosum.

 

 

 

258

 

 

 

32...... .52....- u:oo.no...—. 4
Baa-2...):- doaao ...-2.3.5 o
53......- doaau ...-0:... ..—. I

 

 

259
(MO, NY). Johnson Co.: Knoxville, Demaree 393:3 (SMU);
Hogarville, Demaree 23233 (MO). Perry Co.: Nimrod, Demaree
_3_9_333 (MSC). Phillips Co.: Helena, Palmer 23321 (A). Polk
Co.: base of Rich Mt., Hogkins 3 Nelson 323 (RM); Mena,
Demaree Q23 (GH, TEX, WIS). Pulaski Co.: Pulaski Heights,
Demaree 37_31_2 (MIN); Boyle Park, Little Rock, Gillis 3713
(MSC); Pinnacle, Palmer 23323 (A). Randolph Co.: 4 miles
SSW Maynard, Donahue 2_13 (MSC). 'St. Francis Co.: Forrest
City, Demaree 31332 (MSC). Sevier Co.: DeQueen, Fasset 8.
33333 32131 (MO); Bluffs over Little River, Brownstown Ben
Lomond, Demaree 3300—4 (MSC). Sharp Co.: Hardy, Demaree
23733 (RM, SMU, TEX). Illinois. Adams Co.: Quincy, Holton
313 (F). Bureau Co.: Ohio, 3333, 3.3. (MO). Calhoun Co.:
Kampsville, Turner 331 (F). Champaign Co.: Urbana, 331333
32333 (MIN, MO, ND, NY). Cook Co.: River Forest, Eifrig,
3.3. (F); Chicago, Nelson, 3.3. (MIN); Riverside, Pearsall,
_.3. (F). DuPage Co.: Naperville, Powell 31133 (WIS).
Hancock Co.: Hancock Twp., 3E3, _S_.3. (F). Johnson Co.:
Tunnel Hill, Palmer l_51_33 (A, MO). Kankakee Co.: Kankakee,
Schneider 6_7]_._ (F). Knox Co.: Abingdon, 3a_t_i_3, 3.3. (A).
Lake Co.: Lake Bluff, Hutchinson, 3.3. (F). LaSalle Co.:
Starved Rock State Park, Greenman, Lansigq, 3 M 33 (F,
GH, NY); Matthiessen State Park, 33333 33311 (TEX); Ottawa,
Herett, 3.3. (GH). Mason Co.: Bishop, 3333333321 (G).
Peoria Co.: Peoria Heights, 331333 2332 (MIN, TEX, WIS).
P0pe Co.: Golconda, Palmer 33323 (A, MO, US). Pulaski Co.:
Mounds, Palmer 33312 (A, MO, US). Randolph Co.: Red Bud,

260
Pammel, 3,3. (MO). Richland Co.: Olney, Ridgewgy 3333
(F, US); between Noble and Passport, Ridgeway 2322 (A).
Rock Island Co.: Park Bryan, Hagger 3'Ha33er, 3,3, (A).
St. Clair Co.: Cahokia, Eggert, 3,3, (MO). Sangamon Co.:
Shermfield, collector unknown (F). Stephenson, Freeport,
Johnson, 3,3, (MIN). Tazewell Co.: Kerfoot, 33333’33231’
(F); East Peoria, 33333 33222 (F). Vermilion Co.: between
Oakwood and Collison, 33333.33322 (MO,WIS); Joliet, Skeels
323 (F). Williamson Co.: Pittsburg, Douglass, 3,3, (MO).
Woodford Co.: south of Spring Bay, 33333 32233_(TEX).
Indiana. Crawford Co.: Milltown, 3333 33333 (A); 3 mile
east of West Fork, Kriebel 33333 (PUL). Dearborn Co.:
Aurora, 3333 33333 (A). Gibson Co.: Haubstadt, Schurmeier
.333 (WIS); Mr. Carmel, Tr on 2333 (F); Gordon Hills, 3333_
33332 (A). Harrison Co.: Elisabeth, 3333 33332 (A). Henry
Co.: Mt. Summit, Kriebel, 3,3, (SMU). Howard Co.: Kokomo,
33.31.(US). Huntington Co.: Huntington College campus,
Henderson, 3,3, (FSU). Jefferson Co.: Hanover, Coulter,
3,3, (F). Madison, 3333.33233,(A). Johnson Co.: Edinburgh,
3333 32312 (A). Lagrange Co.: 'Wolcottville, 3333, 3,2,
(SMU). Lawrence Co.: Rivervale, 3333 31332 (A). Madison
Co.: Ingalls, 33333 3332 (F, GH). Monroe Co.: north of
Bloomington, Hawkins, 3,3, (FLAS). Newton Co.: Kankakee
River State Park, Thieret 3223 (F). Orange Co.: West Baden,
3333 31333 (A). Owen Co.: McCormack Creek, 3333 31332 (A).
Perry Co.: Cannelton, 3333 33333 (A). Posey Co.: Shelton,

Heiser, 333. (MSC). Putnam, Co.: Green Castle, Grimes 531

261
(US). Ripley Co.: Versailles State Park, Donahue 232.
(MSC); Laugherty Creek, east of Versailles, 3333 33332 (A);
New Marion, 3333 33333 (A). St. Joseph Co.: South Bend,
Nieuwland, 3,3, (MO). Switzerland Co.: Cross Plains, 3333
33333 (A). Tippecanoe Co.: Lafeyette, 3333 31313 (A);
'West Point, Gillis 3213 (MSC). Vanderburg Co.: Evansville,
3333 33133 (A). Vermillion Co.: Blanford, Freisner 32333
(NY). 'Warrick Co.: Yankeetown, 3333 33133 (A). ‘Wells Co.:
Harrison Twp., Deam, 3,3. (GH, MIN, MO, US, WIS). ‘White Co.:

 

Furnessville, Johnson, 3,3. (NY). :3333. Clayton Co.:
(McGregor, Shimek, 3,3, (WIS). Decatur Co.: without definite
locality, Anderson, 3,3, (MIN, MO, RM). Fayette Co.:
Fayette, 3333, 3,3, (MIN). Johnson Co.: Iowa City, 33333-
I3333,I333, (MIN). Johnson Co.: Iowa City, Hitchcock, 3,3.
(KSC). Lee Co.: Keokuk, Jackson Twp., Hayden 2333 (GH,
MIN, MO, NY). Shelby Co.: without definite locality,
Fitzpatrick, 3,3, (F). Story Co.: Ames, Hitchcock, 3,3,
(MO). Kagsas. Allen Co.: without definite locality,
Hitchcock, 3,3. (KSC). Anderson Co.: without definite
locality, Hitchcock, 3,3, (KSC). Barber Co.: Hardtner,
McGregor 33322 (SMU). Barton Co.: without definite local-
ity, Hitchcock, 3,3. (KSC). Brown Co.: without definite
locality, Clothier 3 Whitford, 3,3, (KSC). Butler Co.:
White‘Water, Hitchcock, 3,3. (KSC). Chase Co.: without
definite locality, Hitchcock, 3,3, (KSC). Cherokee Co.:
Elm Branch, Jacobs, 3,3. (KSC); Baxter Springs, Hitchcock,

3,3, (KSC). Coffey Co.: LeRoy, Dunfield, 3,3, (KSC).

262

Concordia Co.: without definite locality, Hitchcock, 3.3.
(KSC). Cowley Co.: Arkansas City, Palmer .2233]; (A, US).
Crawford Co.: without definite locality, Clothier 3 W333-
mg, 3.3. (KSC). Douglas Co.: Lawrence, Gillis 6_4_8_33
(MSC). Elk Co.: without definite locality, Clothier 3
Whitford, 3.3. (KSC). Ellsworth Co.: Terra Cotta, McGregor

 

32333 (US); Ellsworth, Palmer 23232 (A). Franklin Co.:
without definite locality, Hitchcock, 3.3. (KSC). Geary Co.:
Alida, 3&3 23233 (KSC). Greenwood, without definite local-
ity, Clothier 3 Whitford, 3.3. (KSC). Jackson Co.: without
definite locality, Clothier 3Whitfor3_, 3.3. (KSC). Jefferson
Co.: Perry, Clothier, 3.3. (KSC). Johnson Co.: without
definite locality, Pinchomb, 3.3. (KSC). Kiowa Co.: with-
out definite locality, Hitchcock, 3._r_1_. (KSC). Labette Co.:
Montana, Rxdberg 3 .1333; 333 (KSC, NY). Lincoln Co.: with-
out definite locality, Hitchcock, 3.3. (KSC). Linn Co.:
Pleasanton, Hitchcock, 3.3. (KSC). Marion Co.: without
definite locality, Hitchcock, 3.3. (KSC). Mitchell Co.:
Beloit, Hitchcock, 3.3. (KSC). Montmorency Co.: without
definite locality, Hitchcock, 3.3. (KSC). Morris Co.:
Kelso, 3393 _22 (KSC). Nemaha Co.: Aenida, Hitchcock, 3.3.
(KSC). Neosho Co.: without definite locality, Hitchcock,
s_.r_1, (KSC). Osage Co.: Burlingame, Hitchcock, 3.3. (KSC).
Osborn Co.: Downs, bank of Solomon River, Palmer 23333 (A).
Ottawa Co.: without definite locality, Hitchcock, 3.3.
(KSC). Pottawatomi Co.: Floersh, Haymaker, s.n. (KSC);
Onaga, 1331, 3.3. (KSC). Republic Co.: T. 48., R. 3W.,

263
SW % Sec. 27, Morley 333 (SMU). Rice Co.: without definite
locality, Hitchcock, 3,3, (KSC). Riley Co.: without
definite locality, Norton 13 and 133, (GH, KSC, MO, NY, RM,
US); Manhattan, Norton, 3,3, (MIN); Wildcat, 33333.33333
(KSC). Rooks Co.: Rockport, Bartholemew, 3,3, (MO).
Saline Co.: Gypsum City, Palmer 23322 (A, US). Sedgwick,
Co.: without definite locality, Clothier 3 Whitford, 3,3,
(KSC). Shawnee Co.: TOpeka, Pogenoe, 3,3, (KSC). Smith
Co.: without definite locality, Hitchcock, 3,3. (KSC).
Sumner Co.: Belle Plaine, 33333.32212 (KSC). Wabaunsee,
33333.23331 (KSC). Washington Co.: Linn, 3333, 3,3, (KSC).
Wilson Co.: Neodesha, Palmer 23332 (A); ROper, Haller, 3,3,
(KSC). Wyandotte Co.: without definite locality, Clothier
‘3 Whitford, 3,3, (KSC). Woodson Co.: without definite
locality, Lathrog 323 (US). Kgnggckx. Ballard Co.: with-
out definite locality, Anderson, 3,3, (GH, MO). Boyd Co.:
Catlettsburg, 3333!,u3,3. (GH). Christian Co.: Pennyrile
Forest State Park, Donahue 233 (MSC). Fayette Co.: Lexing-
ton, 33333,'333, (MSC); 8 miles north of Kentucky River,
Gillis 3333_(MSC). Jefferson Co.: Louisville, N.G.R.R.,
Mueller 331 (HY). Kenton Co.: 2 miles south of Erlanger,
Gillis 3333_(MSC). Laurel Co.: 0.6 mile north of highway
30, Gillis 3333 (MSC). Nelson Co.: 4 miles south of Bards-
town, Donahue 273 (MSC). Rockcastle Co.: 0.6mile north of
Roundstone Creek, 7.9 miles north of Mt. Cernon, Gillis
3333 (MSC). Scott Co.: 1.5 miles north of Georgetown,

Gillis 6532 (MSC). Marxland. Montgomery Co.: Plummer's

r1

 

264
Island in Potomac River, Kearney, 3.3. (US). Without definite
locality: Foundling's Home and Eastward, Bartlett, 3.3.
(MICH, MSC). Micgigan. Arenac Co.: AuGres Twp., Gillis
3322 (MSC). Barry Co.: Maple Grove TWp., Gillis 3&3 (MSC);
Delton, Gillis 3332 (MSC); Wall Lake, Gillis 3333 (MSC).
Berrien Co.: Benton Harbor, D_odg3 363 (NY); Three Oaks Twp.,
Warren Woods, Gillis 3333 (MSC). Grand Beach, Nieuwland,
3.3. (ND, MSC). Branch Co.: Union TWp., H_oy_e_3 3 Hiltunen
1346 (WUD); Goldwater, Gillis 3 Wetmore 2323 (MSC). Calhoun

 

Co.: Athens, Gillis 3393 (MSC); Bedford Twp., Gillis 2393
(MSC). Cass Co.: Vandalia Twp., Gillis 330_9_ (MSC); Pokagon
Twp., Sgarks, 3.3. (MSC). Clare Co.: Clare, Flicke, 3.3,
(MSC). Clinton Co.: Rose Lake Bog, Howell 3 Roberts, 3.3.
(MSC); Bath, Yuncker _3_33 (US). Eaton Co.: Thornapple River
bank, Gillis 3273 (MSC). Genessee Co.: Montrose, Gillis
3333 (MSC). Gratiot Co.: Bethany Twp., Gillis 3331 (MSC);
3.5 miles south of St. Louis, Gillis 2233 (MSC). Hillsdale
Co.: intersection of Custer and Grace Roads, Gillis 31.313
(MSC). Huron Co.: Charity Island, Saginaw Bay, 23393, 3.3.
(GH). Ingham Co.: Haslett, Gillis 2333 (MSC). East Lansing,
bank of Red Cedar River, Gillis 231_7_ (MSC). Ionia Co.:
Morrison Lake, 343123, 3.3. (MSC). Isabella Co.: Mt. Pleasant,
Gillis 3313 (MSC). Jackson Co.: Jackson 2333 (MSC). Kala-
mazoo Co.: MSU-Kellogg Biological Station, shore of Gull
Lake, Gillis 2323 (MSC); adjacent to Gull Creek, Fischer,
3.3. (MSC). Kent Co.: Cascade TWp., along Thornapple River,

Hoffman, Swan, 3 Sarkissian 338 (MSC); Grand Rapids, Skeels 3

265
Shaddick, 3.3. (MSC). Lenawee Co.: Seneca Twp., Gillis
3548 (MSC); Hudson TWp., Gillis 3332 (MSC). Livingston Co.:
SW of Pinckney, Hiltunen M (WUD); Hamburg, Farwell 8343
(BLH); Island Lake Recreation Area, 3mg, _.3.~ (MSC);
Hamburg, Farwell _32_43 (BLH). Macomb Co.: Washington,
Cooley, 3.3. (MSC). Mason Co.: Scottville, Gillis 3333
(MSC). Mecosta Co.: Colfax Twp., Gillis 3323 (MSC). Mid-
land Co.: Midland, Erbisch 3 (MSC). Monroe Co.: Clark)
Island, McDonald, 3.3. (MSC). Montcalm Co.: Stanton, Gillis
_35_73 (MSC). Muskegon Co.: Casnovia Twp., Gillis 3332 (MSC).
North Muskegon, Gillis 3321. Newaygo Co.: Ashland Twp.,
Gillis 3333 (MSC). Oakland Co.: Walnut Lake, Bingham, 3.3.
(BLH); Proud Lake Recreation Area, Gillis 25_8_3 (MSC); Farm-
ington, Mfihling, 3.3. (MSC). Ogemaw Co.: West Branch,
Gillis 3333 (MSC). Osceola Co.: Sylvan Twp,, Gillis 3333
(MSC). St. Clair Co.: Port Huron, 33133 37_83 (SMU, BLH,
NY, TEX, E). St. Joseph Co.: Mendon, Gillis £33 (MSC).
Shiawassee Co.: Woodhill Twp., Gillis 3412 (MSC). Tuscola
Co.: Vassar, Gillis 3933 (MSC). Van Buren Co.: Bangor TWp.,
Gillis 3_52_l_ (MSC). Washtenaw Co.: Sylvan TWp., Donahue 313
(MSC); Chelsea, Gillis 2331 (MSC); Ypsilanti, Walgole 36_3a_
(BLH); Ann Arbor, Sudworth, 3.3. (US). Wayne Co.: Belle
Isle, Farwell 3322 (BLH); Bellville, Farwell 2333 (BLH);
Nankin, Farwell 13l_2 (BLH); Detroit, collector unknown, (E).
Minnesota. Houston Co.: Houston Village, Buttes 3 Rosindald
3333 (GH). Missouri. Adair Co.: Novinger, Palmer 23333
(A, MO). Atchison Co.: Watson, Palmer 33% (A, MO, US).

I1

 

266
Barry Co.: Roaring River State Park, Donahue 222 (MSC).
Benton Co.: Cole Camp, Palmer 33933 (A, MO). Carter Co.:
Big Spring State Park, Donahue 220 (MSC); Van Buren, Palmer
3231 (A, MO, US). Clark Co.: Dumas, Palmer 2332 (A).
C00per, without definite locality, 33133 33132 (A). Greene
Co.: Springfield, Standley 3321 (US). Henry Co.: Windsor,
Hitchcock, 3.3. (MO). Holt Co.: Forest City, Palmer 23332
(A, MO). Jackson Co.: Kansas City, .B_u3l*_1_ _13393 (MO). JaSper
Co.: LaRussel, Palmer 3331 (M0); Webb City, Palmer 2333
(A, MO); Jasper, Palmer 3333 (M0); Pr05perity, Palmer 3333
(A, MO); Joplin, Palmer 23333 (A, MO); Waco, Palmer 33333
(A, NY); Carthage, Palmer, 3.3. (A); Neck City, Palmer 2133
(A). Jackson Co.: Independence, 31333, 3333 (A, F, MO).
Jefferson Co.: Kemmswick, Wislizenus, 3.3. (MO). Johnson
Co.: Columbus, Palmer 33333 (A, MO). Marion Co.: Hannibal,
M33, 3.3. (A, MO). McDonald Co.: Noel, Ell—SEE (A, US).
Ozark Co.: Pontiac, Palmer 34192 (A, MO). Phelps Co.:
Jerome, Kellogg 333 (MIN, MO). Ralls Co.: Bear Creek, 33331,
_.r_1, (MO). St. Louis (city), 3m_i_g 231 (MO); St. Louis Co.:
Allenton, Letterman, 3.3. (MIN, MO, NY, US); Carsonville,
Fritchey, 3.3. (MIN). Shannon Co.: Montier, 3353333 (M).
Stone Co.: Galena, Palmer 2&3 (A). Taney Co.: Swan,
3333 3393 (A, MO); Branson, Palmer 4__7]fi (A, E, MO). Webster
Co.: Seymour, Donahue 223 (MSC). Nebraska. Cherry Co.:
Valentine, Petrides 28_43 (MSC). Lancaster, wooded banks of
Salt Creek, Turrell, .n. (ARIZ). Sarpy Co.: Bellevue,

Eastman, 3.3. (NY). Webster Co.: Red Cloud, Bates, 3.3.

267

(MIN). New York. Broome Co.: Mount Glenwood Ravine, Bing-

 

hamton, Millspaugh, 3,3, (F). Cayuga Co.: Montezuma, 33333,
'Wiegand,l3 Randolph 32333 (GH). Chemung Co.: without
definite locality, 3323,3333, 3333_(F, MIN, WIS, RM); East
Elmira, Johnson, 3,3, (NY). Franklin Co.: Brushton, Cutler,
_,g, (MSC). Jefferson Co.: Plessis, Gillis‘3313 (MSC);
Murray Co.: Price's Falls, Demaree 32333 (MIN); Seneca Co.:
Magee, Gillis 3393 (MSC). Tioga Co.: without definite local-
ity, Millspaugh, 3,3, (US). Tompkins, Co.: Ithaca, Casca-
dilla Creek, Palmer 768 (GH). 3333. Butler Co.: Oxford,
Overholts, 3.3. (MO). Coshocton Co.: Coshocton, Moldenke
32133 (SMU). Greene Co.: Yellow Springs, Demaree 33333

(A, GH, MIN, MO, NY, SMU,WIS). Hamilton Co.: Cincinnati,
33233, 3,3, (MIN). Lucas Co.: Sylvania, Wilson 3332 (WIS).
Sandusky Co.: Cedar Point, Jennings 3 Jennings, 3,3, (ND-G).
Shelby Co.: 2.9 miles north of McCartyville, Gillis 3333
(MSC). Summit Co.: Cuyahoga River, Akron, 33333,I3,3.

(MIN, US). Ok3§h9m§. Adair Co.: Watts, Wallis 3313 (SMU).
Caddo Co.: False Washita River between Fort Cobb and Fort
Arbuckle, Palmer 33 (NY, US). Canadian Co.: Yukon, Palmer
22323 (A). Cherokee Co.: limestone escarpments 8 miles
north of Talequah, Hogkins 3321,(TEX). Cleveland Co.:
Norman, Goodman, 3,3, (MSC). Commanche Co.: 'West Cache
Creek, Wichita Mts., 33333,.333. (US). Creek Co.: Sapulpa,
3333 31 (MO). Delaware Co.: Dripping Springs, Wallis 3339.
(SMU). Johnston Co.: Tishomingo, Palmer 3332.(A, MO, US).

Kay Co.: without definite locality, White, 333. (NY).

268
Leflore Co.: Choctaw Agency (Skullyville), Bigelow, $3.
(US); 6 miles east of Page, Donahue ggg (MSC); McCurtain
Co.: Boss, Houghton, gllfi (GH, MO, NY). Murray Co.: Platt
National Park, Merrill §._ Eager; Q23 (F); Price's Falls,
‘ Demaree l_2_8_<fl._ (GH, SMU); Davis near Arbuckle Mt., Palmer
M (MO, SMU); Sulphur, Merrill, s.n. (A). Ottawa Co.:
Commerce, 1332;; M (A, GH, MO, NY); Miami, Wallis _8_§_3§
(SMU). Payne Co.: 2 miles west of Stillwater, Cockrell fl
(TEX); 2 miles east of Perkins, Eggs _]._Q_I_3_ (TEX). Pontotoc
Co.: Ada, Robbins 2_‘5329 (SMU) Pushmataka Co.: Clayton, W
_l_2i (MO). Sequoyah Co.: Gore, Wallis §_£_1_§_0_ (SMU). Woods
Co.: without definite locality, m 3.3. (MIN, RM). Lo-
cality uncertain: Devils Canyon, Little §9_2_g (A). Pennsyl-
2233' Allegheny Co.: Mt. Lebanon Twp., along creek in
Kennedy's wood, Churchill, s.n. (MSC); Natrona, Shafer, s.n.
(MIL); Pittsburgh, Shafer, 3.3. (ND-G). Dauphin Co.: Harris-
burg, §_rfll_l, 3.3. (F, US). Fulton Co.: between Tuscarora
and Sideling Mountains, Gillis §_l_._l_§_ (MSC); Knobsville, Gillis
§l_l_€l (MSC); Hustontown, Gillis all (MSC). Indiana Co.:
Homer City, Wall]; _l_‘_7_2_5 (SMU). Mifflin Co.: Reedsville, Egg
_l_glgg (NY). Westmoreland Co.: without definite locality,
Pierron, 3.3. (F). MDakota: Clay Co.: river bottom,
Taxonomy C_l_a_s_§ 2;: lg (SDU). Tennessee. Davidson Co.:
Nashville, near confluence of Mill Creek and River, Bartlett
gag; (MICH, MSC). Egg. Bexar Co.: without definite local-
ity, 23353, 13529, I_3_-_E_>§ (MO, MSC). Cooke Co.: Gainesville,
Shinners £13236; (SMU). Dallas Co.: Dallas, Reverchon, s.n.

 

 

 

'
,.
,. , . r
.

,.

. . ,
,.
\

,

269
(GH). Fort Bend Co.: Richmond, B331 §_Q (US). Harris Co.:
Houston, Eggert, 3.2. (MO); San Jacinto, B993 ggggg (MO).
Hunt Co.: Commerce, Shinners l2_2_g§ (SMU). Jackson Co.:
Ganado, Palmer 2_2_3_§ (A, MO). Karnes Co.: Panna Maria,
Johnson l_3_OQ (SMU). Kaufman Co.: Terrell, T ler, 31.3.
(US). Red River Co.: Clarksville, Cory 56041 (US). Tar-
rant Co.: Lake Como, _R_u_1_:_h 91]; (GH). Taylor Co.: Abilene,
Talstead 19;; (GH). Tom Green Co.: Knickerbocker Ranch,
Dove Creek, Tweedy, §_L_r;. (US); San Angelo, Palmer _l_l_l_3_Z
(A, MO). Wood Co.: Hawkins, Gillis m (MSC). Locality
uncertain: Hortons, Reverchon, §.9_. (MO). Virginia.
Frederick Co.: Chambersville, Gillis 5_7Q§ (MSC). Giles
Co.: Pembroke, Eggg £12121 (GH, MIN, MO). Loudoun Co.: 1
mile down Limestone Branch from highway 15, Allard M
(US). Montgomery Co.: Blacksburg, ;A_d_a_m_s_ §_. Wherry _2_2_I_3_Q_
(A, GH, US); East Radford, m _8._ Wherry gggg (A); south
of Riner, Donahue _2_§_]_._ (MSC). Westmoreland Co.: Colonial
Beach, Tidestrom g Bartlett 22E (US). LL53 Virginia.
Cabell Co.: Milton, Williams 3312 (GH); Huntington, Gilbert
£9 (GH). Grant Co.: Mt. Storm, east flank, Gillis gzgg
(MSC). Jefferson Co.: Charles Town, Byrd roadside park,
Gillis m (MSC); Harper's Ferry, Gillis §_‘7_9_Q (MSC). Min-
eral Co.: Burlington along Patterson Creek, Gillis §11_4
(MSC). Preston Co.: Erwin on Wolf Creek, Gillis Q72};
(MSC). Ritchie Co.: Berea (Randolph _8_._ Randolph _l_._3_8_§ (GH).
Tucker Co.: Blackwater River, Mg _2_l_l_§ (GH). Upshur Co.:

Buckskin, Pollock, 3.3. (KSC). Wisconsin. Dane Co.:

270
Mazomanie, Seymour iiggg (WIS). Grant Co.: Rockville,
Wunderlin, §.g. (WIS). Mason Co.: Bath, Guther 1§;1
(WIS). Milwaukee Co.: without definite locality, Russell,
§,g, (GH). Racine Co.: Racine, Hale, §,g, (WIS). ‘Rock
Co.: Avon, Fassett ggilg (WIS); Milton, Ggggg, §,g, (WIS).
Sauk Co.: Smitg ggggg (MIL). Locality uncertain: Cedar
Twp., collector unknown (WIS).

* * * * *

CANADA. Ontario. Garleton Co.: March TWp., 223g
'g Breitunggglg (E). Grenville Co.: Prescott, Qggg'lgggg
(MSC). Middlesex Co.: POplar Hill, Gillis §§ZQ (MSC)..
Norfolk Co.: ‘Williams, §.g., Richardson, §,g. (A). ‘Water-
loo Co.: Galt, Montgomery g2; (GH); German Mills, Montgomery
953 (GH); Baden, Gillis §§§l (MSC).

Gates published the trinomiallghgg toxicodendron
negundo in a Species list in 1938, followed it a year later
with a formal description. He did not clearly designate a
rank for the epithet ne undo, but included this enigmatic
statement: "The trinomial in botany is usually referred to
as a variety, although the designation subSpecies would appear
to be more reasonable." Even though no rank is clearly
stated, it would appear that Gates intended the rank of sub-
Species. His 1939 publication should be considered the legit;
imate and valid publication because it refers to a previously
and effectively published name (Toxicodendron negundo Greene).

Greene described the fruits of his taxon, Toxicodendron

271
negundo, as "uncommonly small . . . the length distinctly
greater than the thickness," neither character being borne
out by the type material.

The type for Toxicodendron aboriginum was collected
at Choctaw Agency, Indian Territory by J. M. Bigelow on the
Whipple Expedition. Although there is no Choctaw Agency
today, Gaspare Signorelli (personal communication), philat-
elic history specialist, informs me that this locality was
a small settlement now known as Skullyville, Oklahoma. It
was "Choctaw Agency West," just across the border from
Arkansas Territory in what is now northern LeFlore County.
Greene's description "abruptly apiculate" for the leaflets
does not fit the type material which has acuminate leaflets.
The fruits have a scattered pubescence which will be dis-
cussed later.

It seems strange that this common plant of the mid-
western United States should have no formal subspecies or
variety name until now. Most manual writers presume that
the typical variety is this one and that the Atlantic Coast
subspecies must be one which was described later, rather
than the other way around. Such presumptions, of course,
fly in the face of the historical evidence that the earliest
collections in America were from the Coastal regions, and
not the interior.

It has been difficult to characterize this sub-
species, largely because of the distinctive characters it

does not have. It does not have the lobed margins of subSp.

272
eximium nor of subsp. verrucosum, nor the consistently entire
or subentire margins of subspp. radicans, orientale, hisgidum,
barkleyi, nor divaricatum, nor the pubescent fruits of
subspp. radicans, eximium, etc. Rather, it possesses a
variable and moderate degree of pubescence on the leaflets,
occasionally scattered-pubescent fruits, some notching on
the leaflets, which are clusters of characters found to some
degree in other subspecies.

Miller (Gard. Dict. Art. Toxicodendron, No. 6, 1768)
may have had reference to this taxon in his Toxicodendron
volubilis (sic). The description of "trifoliate leaves
whose lobes are oval, smooth, and cut into sinuses" does
not wholly apply, and could relate to other taxa, but also
might fit this one. Miller's reference to Dillenius's flggt.
filth, seems to provide a lead, but it is to Toxicodendron
amplexicaule, foliis minoribus glabris, which cannot be
typified. There is sufficient doubt as to whether Miller's
plant refers to subSp. radicans or to subSp. negundo that
the basionym volubile cannot be used. In all likelihood,
evidence suggests that, since most Specimens available to
eighteenth century England were coastal forms, Miller had
probably never seen this variety.

This subSpecies intergrades with Toxicodendron
rydbergii at its westernand northern limits and along Lake
Michigan. It intergrades with subSp. radicans in the
Appalachians, especially the Allegheny ridge and the Blue

Ridge Mountains. It seems to form large pOpulations of

273
intergrades with subSp. pubens in Arkansas and southern
Missouri where the infraSpecific boundaries become obscured.

Although subSp. negundo and subSp. radicans have
been confused because of their similarity, the closest
relationship of subSp. negundo seems to be with subSp. gubens.
These two subspecies are sympatric in Arkansas and Missouri.
SubSpecieS pubens is possibly a Specialized, pubescent form
from more generalized negundo stock. The differences are
mainly quantitative ones in degree of pubescence, length-
to-width ratio of leaflets, and degree of notching or ser-
ration of leaflets.

Some populations of subSp. negundo possess a Slightly
pubescent fruit, reminiscent of the fruits of subsp. radicans,
under which name Barkley (1937) included them. These popu-
lations appear in a few places in northwestern Indiana, the
Ohio River counties of Indiana (Deam, 1940), and a few
places in eastern Oklahoma. In no way, other than the
Slightly pubescent fruits, do these plants seem to diverge
from the more typical subSp. negundo, and they do not possess
the entire leaflets, nor tufts of hairs on the lower leaflet
surface characteristic of subSp. radicans.

Toxicodendron radicans subSp. negundo is frequent
in beech-maple, elm-ash, elm-maple, and weedy plant associa-
tions. In southeastern Wisconsin, at the northwestern extent
of its range, it has been described from the mesic and low-
land southern forests (Curtis, 1959). Farther to the south

it is a constant Species in the northern Appalachian highland

274
division of the hemlock-white pine-northern hardwoods of
the Alleghenies and the mixed meSOphytic forest and western
mes0phytic forest of the central states (Braun, 1950), but
it is not found in Similar forests in the northern reaches
of the country such as Upper Michigan, Wisconsin, New York,
and New England. There it is replaced by the less abundant

I. rydbergii.

 

Whereas its related taxon, subSp. pubens, is a com-
ponent of the oak-hickory woods, this subspecies and its
other relative, subsp. radicans, are seldom found in the
dry, undisturbed oak-hickory forest of the northeastern or
central United States. Although it is difficult to prove
negative associations, more than 58 oak-hickory stands have
been examined in the northeastern United States in only
four of which poison-ivy has been found. It was usually in
a marginal state of existence as a chance invader. Its
existence is also marginal in scrub-oak and pine woods of
the Coastal Plain or the jack-pine plains of Michigan. It
is generally replaced on Open dunes by Toxicodendron rygr
bergii, but may appear in the dune forests. Other than
these habitat preferences, it occupies habitats Similar to
those in which subsp. radicans occurs.

49. Toxicodendron radicans subsp. divaricatum (Greene)

Gillis, stat. nov. (Plate XXXI).

Toxicodendron divaricatum Greene, Leafl. Bot. Observ.

Crit. l: 122. 1905. Rhus divaricata (Greene)
McNair, Publ. Field Mus. Nat. Hist. 225, Vol. 4:

275

PLATE XXXI. Specimen of Toxicodendron radicans subsp.
divaricatum. Holotype of Toxicodendron
arizonicum Greene.

 

 

 

 

 

276

.._.,...,. u...
A w \T-‘nu‘m‘ u ‘04-» up
my» nr» 4 \-y..'H nun.

,4

   
  
  
   

IYI
A

>‘dvlwfinrml. LAU‘AL'KA‘ rm
Xhlgrmail n ‘s-
M‘ch .un Fuh-

Nsnnvo Numb". 3M

“(Mufi-Né ”1.4.1.6. ”A .Z‘,
J‘f‘" 2 b, '7 4' '1

 

C \
(nofli. HA .. {(A4}. o
«:w )TW,‘
5;? “"79
n . _ o
q .... .3.
[cgvy/ ,
HIZIII. [\ITIJI \TJTI'B DVI’AITIENT 0r llifllll Lfl‘ll.
‘ U
. “ , y / . ., r ; ‘
/
IA-nlm ; ‘
. , .
(”Huh-f ’ ‘
\1

277
69. 1925, non Eckl. and Zeyh., Enum. P1. Afr.,
p. 146. 1834. ‘Bhgg greenei McNair Qgig.),
insert after page 69. Toxicodendron radicans
var. divaricatum (Greene) Barkley (as divaricata),
Ann. Missouri Bot. Gard. 24: 433. 1937. Bhgg
radicans var. divaricata (Greene) Fernald, Rhoa
dora 43: 596. 1941. Type: Pur uS, §,g,, Mexico,
Calmalli, Baja California Norte, 700-9OO feet.
Jan. - Mar. 1898. (US-383431, lectotype of
Barkley, 1937).

Rhgg tridentatum Sessé and Mocino, P1. Nov. Hisp.,

p. 47. 1887. non Thunb ex L.f. Suppl. 184.
1781, nec Sond. ex Harv. and Sond., F1. Cap. 1:
511. 1859, nec Willd. ex Engler in DC., Monogr.
Phanerog. 4: 438. 1883. Bhgg saxatiliS DC.
Prodr. 71. 1825. Type: "Habitat in calidiS
Nouae Hispaniae regionibus, et capite Bonae
Spei. Floret uere." (MA, lectotype; F-l572884,
photograph of type).

Toxicodendron laetevirens Greene, Leafl. Bot. Observ.
Crit. l: 123. 1905. Bhgg radicans var. 13233:
virens (Greene) Fernald, Rhodora 43: 596. 1941.
Type: .l-.fl- Toumey, §,g, Arizona, Pima Co.,
Santa Catalina Mts., 5 May 1894. (US-441725,
holotype; A, GH, NY (3 specimens), US-619140,
isotypes; MO-144536, photograph of type).

Toxicodendron phaseoloides Greene, Leafl. Bot. Observ.

278
Crit. l: 123. 1905. Type: {9. y, Hartman 102.
Mexico, Sonora, Turicachi, 5600 feet altitude,
7 Dec. 1890. (US-306157, holotype; GH, NY,
isotypes).

Toxicodendron arizonicum Greene, Leafl. Bot. Observ.

Crit. l: 123. 1905. Type: Edward Palmer 4533,
Arizona, Cochise Co., Fort Huachuca, 26 April -
21 May 1890. (US-19847, holotype; GH, NY,
isotypes; MO-l44549, photograph of type).

Vine or Shrub. Leaflets entire, subentire, or occa-
sionally notched, elliptic, lanceolate, or ovate, glabrous,
rarely strigose on the lower surface or with Sparse pubes-
cence confined to the primary veins, not velvety to the
touch; upper leaflet surface glabrous, not lustrous, usually
decidedly darker than the lower surface; beins light-colored,
prominent on both surfaces. Leaflets usually 1.9 - 3 times
as long as broad with larger leaves southward in the range;
petiole 1.5 - 12.0 cm long, scattered to densely puberulent;
terminal blade 1.5 - 16.0 cm long, 0.6 - 9.0 cm wide; term-
inal petiolule 0.2 - 6.0 cm long; lateral leaflets often as
long or longer than terminal. Tip acuminate, usually long
tapering; base subcuneate; lower surface with occasional
tufts of strigose hairs in the major vein axils, disappear-
ing completely in the southern end of the range. Fruit glo-
bose, glabrous (glabrate in the northern end of the range),

3.0 - 6.0 mm broad, larger southward than in the north.

279
Flowering March to June. Distribution: Mexico and United
States: Sierra Madre Occidental, north to southeastern
corner of Arizona. (FIG. 12.)
Representative Specimens: UNITED STATES. Arizon .
Cochise Co.: Fort Huachuca, Palmer 453a (GH, M0, NY, US);

Bisbee, Goodding 4g (ARIZ); one mile above Paradise, Blumer

 

‘28; (US); Huachuca Game Reserve, Goodding 1104-49 (ARIZ);
Chiracahua National Monument, giggg,§ggg (ARIZ). Pima Co.:
Santa Catalina Mts., Toumey, §,g, (A, GH, MO, NY, US, ARIZ);
Santa Catalina Mts., Upper Bear Canyon, Gillis gggl (MSC);
Santa Catalina Mts., Sabino Canyon, Rehder £89 (A), Jgggg,
§,g, (RM), Toumey 132 (US); Fort Lowell, Lemmon, §,g, (MO);
Morista Canyon, Baboquivari Mts., 91g k Ilggg (GH); Florida
Canyon, Santa Rita Mts., Graham, §,g, (NY). Santa Cruz Co.:
Santa Rita MIS” Sycamore Canyon, Hggly, Mgggg, g Stokes 2251
(ARIZ); Stone Cabin Canyon, Thornber g Griffiths 1§2_(ARIZ);
Santa Rita Forest Preserve near Amado, Griffiths 4222_(M0);
vicinity of Nogales, fig g 112 3 (US); 2.95 miles SW of Pata-
gonia, Gillis g11g_(ARIZ, MSC).

* * * * *

MEXICO. Aguascalientes. Road to Calvillo, 21 miles

 

west of Aguascalientes, McVaugh‘g Koelz 158 (MICH). Baj

 

California. Calmalli, Purpus, §,g, (US). El Chorro, Moran

 

7307 (GH). LaLaguna, Moran 7431 (ARIZ, GH). Chiagas. Paraje
of Mahosik, Municipio of Tenejapa, Breedlove 14890 (DS, MSC);
Pokolom, paraje of Sibanikha, Municipio of Tenejapa, Breedlove

11736 (DS). Chihuahua. St. Diego, Hartman 589 (F, GH, NY,

 

280
US). Guadeloupe Canyon, Mearns 512_(US). Carretos, border
of Chihuahua and Sonora, Municipio de Janoes,'WQi§g|255;
(GH, MICH, TEX). Loreto near Rio Mayo, Gentry 2851 (A,
ARIZ, MO). Cajon Creek near US boundary line, Mearns 524
(GH, MO, US). Cgahuila. Sierra de la Gloria, southeast

 

of Monclova, Marsh 1973 (F, GH), Parras, Aguirre 5 Reko, gag.

 

(NY). Durango. Cerro Chupadero, 45 miles south of El Salto,
west of Pueblo Nuevo, Maysilles 8211 (MICH). Pueblo Nuevo,

Baker, 5.5, (MSC). Durango, Palmer 106 (F). Guanaiyato.

 

locality unknown, Qgg§§_fl§§ (GH, US). Hidalgo. Ixmiquilpan,
3555, Painer, §_Rggg.§25g (NY, US). Barranco de Toliman
below Mina Loma de Toro, Mggrg.2522_(GH). Jalisco. Sierra
del Nayarit, Diguet, 5,5, (NY). Zahotlan, Pringle, 5,5.
(TEX). Sierra del Halo, DeJong 2535_(MSC). Mazamitla,
Donahue 25 (MSC). Zapotlan, Salizar, 5.5. (TEX). Vicinity
of San Juan Cosala, north of Lake Chapala, ngg, 525. (MICH).
Barrancas of Rio Verde, 20 miles north of Tepatitlén, road
to Yahualica, MCVaugh 21425 (MICH). Ravine bottom in moun-
tains east of Mamantlan, 15 miles SSE of Autlan, Wilbur 5
Wilbur 2914 (MICH). Forested crest of ridge facing Pacific,
10 miles sough of Autlan, Wilbur 5 Wilbur 2522 (MICH). Sum-
mit of volcanic cliffs 8 miles SE of Jalostotitlan, road to
San Miquel el Alto, MCVaugh'LZ55§_(MICH). North lepeS
Mavado de Colima, near jct. of pack road to Zapotlan, McVaugh
23254 (MICH). Sierra del Digre, 3 miles south of Mazamitla,

McVaugh 13606 (MICH). México. Locality unknown, Sessélg

 

Mocino 930 (MA). Michoacan. Zitacuaro, Hinton 13151

 

 

281
(MICH, NY). Mt. Patamban, Nelson 5512 (GH, NY, US). Morelia,
Loma St. Maria, Arsene 2155, (A, GH, MO, US); Morelia,
Arséne 5511 (A, ILL, MO, NY, US). Apatzingan, Tancitaro
Region, LaS Barranquillas, Leavenworth 5 Hoagstrall 1511
(F, ILL). Patzcuaro, DeJong 5'Longpre 1551 (MSC). NE of
Puentacillas, north of Ario de Rosales, 11515, Koepgen,I5
11115 555 (WIS). Lava flow east of San Juan Nuevo, 8 miles

south of Uruapan, King 5 Soderstrom 4750 (MICH). Nuevo Le6n.

 

 

 

Hacienda Vista Hermosa, south of Monterrey, 55115 1555 AR12,
GH, MICH). Mountains 7 miles east of Iturbide, McVaugh

15515 (MICH). Oaxaca. Oaxaca Courts motel in Oaxaca,

Beaman 5555 (MSC). Tamazulapam, DeJong 1552 (MSC, MU).
Queretero. Tequisquiapan, Nelson 5515 (GH). San Luis Potosi.

2 km. SE Ahuacatlan, Xilitla, Rzedowski 10970 (MICH).

 

Tolentino, Meavry 1515 (TEX), between San Luis Potosi and

Rio Verde, DeJong 1511 (MSC, MU). Between Ciudad del Maiz
and El Naranjo, DeJong 1512 (MSC, MU). El Naranjo, Donahue
'25 (MSC). Sinaloa. Bufa de Surutato, Ocurahui, Gentry 5552
(ARIZ, GH, MO, NY, vs). East of Mazatlan, DeJong 1195 (MSC,
MU). 555511, Turicachi, Hartman 152 (GH, NY, US). Rancho
de la Nacha,'551£5_5225 (ARIZ, GH, MICH, US). Junction of
roads to Cananea and Remedios on Rio Magdalena, Wiggins 1522
(A, ARIZ, DS, MICH, US). Canyon de los Apaches, north of
Aribabi, 55155 2155_(GH, MICH). Region of Cordova, Borgeau
‘2251 (GH). El Rancho del Roble, northeast of El Tigre,

White 4238 (ARIZ, GH, MICH). Bacoachio near Esqueda, Wiggins
11115 (US). Alamo, Mt. Magdalena, Kennedy 1155 (US). Canyon

282
of Guadeloupe River, 55115, 515. (NY). Sabino Canyon,
Quirocoba, Rio Fuerte, Gentry 2255 (A, ARIZ, F, MO, TEX, US).
San José Mountains, Mearns 1552 (US). Santa Rosa Canyon,
near Bavispe, 55115.321 (ARIZ, MICH, GH), Canon Interna-
cional, 55115 5555 (GH, MICH). Sierra Madre Occidental,
Sierra Charuca, Canon de Tejas, Gentry 5552 (ARIZ, MICH).

Tamaulipas. Dulces Nombres and east into Tamaulipas, Meyer

 

 

5 Rogers 2552 (M0).

Barkley (1937) separated this taxon from others on
the basis of a constriction at the Side of the seed, a
feature often difficult to interpret if it does indeed ex-
ist. Even the divaricate nature of the inflorescence, which
gives the taxon its name, is difficult to see as unique to
this plant. When McNair transferred this taxon to 5555, he
neglected the Ecklon and Zeyher South African Species of
that name; so he had a correction Slip bound into his article
before distribution, changing the name to 5555 greenei.

In citing Specimens in connection with Toxicodendron
phaseoloides, Greene did not designate a Single Specimen
as type. Barkley (1515,) chose Hartman 102. Greene de-
scribes his material as having "abruptly acuminate" leaflets,
but they are not. Likewise, in his diagnosis of 1. 15555-
virens, Greene indicated that the foliage was light green.
The type has apparently no brighter green leaves than
Greene's twelve other poison-ivy Species. Presumably the
collections were already dried when Greene affixed names

to them, since he was not the collector; therefore, they

283

probably appear today much the same as they did to him.

According to Engler (in DeCandolle, 1883), 5555
saxatilis DC. (1825) is a doubtful Species and is probably
not anacardiaceous. No specimen is known to represent this
Species in the DeCandolle Herbarium. We would have no idea
of what DeCandolle had in mind except for a later reference
to the plant (DeCandolle, 1874). 5555 saxatilis was based
on one of the Sessé and Mocino paintings (cited as pl. 192
in the Calques des Dessins, DeCandolle, 1874). This is
the same plate that was cited as Ic. Fl. Mex. 196 under
5555 tridentatum by Sessé and Mocino (1887). Since there
is a type Specimen for the Sessé and Mocino binomial, we
can equate the dubious 5. saxatilis to a given Specimen
which is Toxicodendron radicans subsp. divaricatum. Accord-
ing to Dr. Rogers McVaugh (personal communication), the
type locality was in Morelos (Yacapixtla and Cuernavaca).

This taxon flowers from March to June, but chiefly
in late April and early May, as distinguished from subsp.
barkleyi, which flowers from May to August, chiefly in June
and early July. Although intergrades between these two
subSpecies are known, it is likely that their different
flowering periods act as a barrier to gene exchange in the
regions where their ranges overlap.

SubSpecieS divaricatum is found in southeastern
Arizona and western Mexico: the Sierra Madre Occidental
and the western portion of the trans-Mexico volcanic belt.

It is in this central region of Mexico wherein intergrades,

284
presumed hybrids, have been found between subsp. divaricatum
and barkleyi.

This taxon is found at moderate elevations, but
often is restricted to heavily wooded canyons or south-
facing SlOpeS intermixed with tussock grasses. 0n the one
occasion I have had to study it in the field, it was found
in moist canyons near creeks or Springs. The tree Species
making up the dominant vegetation in this locality in south-
eastern Arizona were 55112 laevigata, Fraxinus pennsylvanica
subsp. velutina, Cupressus arizonica, 21555 cembroides, and
Juglans mgigg, Other Species were Shrubs like 555§_115;
lobata, Rhamnus californica, Prunus emarginata, and Ceanothus
fendleri. The associated herbs were mostly weeds.

The taxon probably has its closest affinities with

subsp. verrucosum. The latter subspecies may have been

 

derived from subSp. divaricatum after it was isolated from

 

it aS a separate pOpulation.
4h. Toxicodendron radicans subsp. verrucosum (Scheele)

Gillis, comb. nov. (Plate XXXII).

5555 verrucosa Scheele, Linnaea 21: 592. 1848.
Toxicodendron verrucosum (Scheele) Greene,
Leafl. Bot. Observ. Crit. l: 124. 1905. 1521-
codendron radicans var. verrucosum (Scheele)
Barkley, Ann. Missouri Bot. Gard. 24: 435.
1937 (as verrucosa). 5555 radicans var. ygg-
rucosa (Scheele) Fernald, Rhodora 43: 596.
1941. Type: E, Lindheimer £51, In Hecken auf

285

PLATE XXXII. Specimen of Toxicodendron radicans subsp.
verrucosum. Type of Rhus rhomboidea
Small.

 

 

 

MAN'S 1H hOU'NtkN 70.145.
uun‘n r uuum
... . ...... . ... .

I Iluu . um.”

h .1411qu

 

 

 

 

 

 

 

287
felsigem Boden auf dem Missionshugel bei
Neubraunfels; Lindheimer. May 1845. (GH,
lectotype; MO-1774754, US-1814899, isotypes).

5555 rhomboidea Small, Fl. S.E.U.S., 727. 1903.

Toxicodendron rhomboidea (Small) Greene, Leafl.
Bot. Observ. Crit. l: 125. 1905. Type:

5. Arthur Heller 1515, (NY, holotype; E, F-
171075, MIN-149968, MO-l774757, MSC-53804, US-
213781, isotypes).

Vine or Shrub with aerial roots; leaflets rhomboid-
ovate, acute to acuminate, usually regularly and deeply
incised dentate or acute-angled lobed, glabrous (rarely
strigose) above, occasionally with a line of minute curly
hairs on the midrib above; glabrous to substrigose below,
with hairs confined to the primary and secondary veins,
rarely with tufts of hairs in the vein axils; leaflet blade
tip sharply acute, base cuneate or obtuse. Petiole glabrous
to pilose, rarely Short tomentose, 3.2 - 11.0 cm long;
terminal leaflet blade 3 - 13 cm long, 2 - 11 cm wide,
terminal petiolule 0.5 - 4.0 cm long. Fruit glabrous 2.5 -
5.5 mm broad. Distribution: Texas and Oklahoma: Edwards
Plateau and Arbuckle MtS. on limestone. (Fig. 11.)

Representative Specimens: UNITED STATES. Texas.

 

Bastrop Co.: Beuscher State Park, Gillis 3750 (MSC); BastrOp
State Park pine hills, Duval, 5,5, (TEX); Gillis 5161 (MSC);

 

Bexar Co., Leon Springs Clemons 5 Clemons 830 (RM). Comal
Co.: New Braunfels, Lindheimer num. glur. (GH, KSC, MIN,

288
MO, US). Coryell Co.: 5.6 miles east of Gatesville,
Shinners 25222 (FSU, SMU). Dallas Co.: Dallas, Northwest
Highway, Gillis 5132.(MSC). Gillespie Co.: locality un-
known, ggrgy,'5,g. (MO). Grayson Co.: Denison, Kohfeldt
55 (TEX). Harris Co.: Houston, collector unknown (MO).
Harrison Co.: Marshall, Gillis 5112 (MSC). ‘Hill Co.:
Hillsboro, Gillis 5155_(MSC). Kerr Co.: Kerrville,FE115£,
1§_7_o_ (ARIZ, E, 1:, GH, MIN, MO, NY, MSC, US); Turtle Creek,
5rgy_152 (TEX). Medina Co.: Castroville, Shore of Rio
Medina, Gillis 5155 (MSC). Menard Co.: Menard, Palmer
11525 (RM). Real Co.: locality unknown, Boyles, 5,5. (TEX).
Tarrant, Lake Como, 5555_2g1 (MSC, NY); ravine near Trinity
River, 5515 1555 (PH). Tom Green Co.: Dove Creek, Tweedy
155 (US). Travis Co.: Austin, Gillis 5512 (MSC); Bergstrom
Air Force Base, Gillis 515g_(MSC); Barton Springs, Krodel
'52 (TEX). Upshur Co.: Gladewater, Gillis 5155 (MSC).
Uvalde Co.: uvalde along wooded banks of Leana River,
Palmer 55555 (PH); UValde along Frio River, Gillis 5125
(MSC). Wood Co.: Hawkins, Gillis 5152 (MSC). Localities
uncertain: 18 miles SW of View on Edwards Plateau, Tolstead

7563 (TEX). Columbia, Bush 142 (A). Oklahoma. Love Co.:

 

Hickory Creek, Gillis 5148 (MSC). Murray Co.: Turner
Falls, Gillis 5142, 5137 (MSC). Woods Co.: locality un-
certain, White, 515. (RM).

*****

MEXICO. Tamauligas. Vicinity of San José, Mesa de

 

Tierra, Bartlett 10405 (MICH).

 

289

Because 5555 verrucosa was described by Scheele, a
Specimen was sought that he may have seen in order to typify
the Species. McKelvey (1955) indicated that plants col-
lected by Lindheimer in Texas in the middle decades of the
eighteen hundreds were intended for Engelmann and Gray; any
new species were to have been described by them. A.number
of Specimens, however, went with Roemer to Germany and
subsequently were described by Scheele in his series of
articles under the title "Beitrage zur Flora von Texas."
Unfortunately, no Specimen representing 5555 verrucosa has
been found at Heersum where Scheele was Pastor, nor at near-
by Hildesheim, nor has any other specimen been found that
could have been the one which Scheele examined.

Gray (1850) claimed that Lindheimer 5fi5_(Fascicle
iii) represented the Species that Scheele (1848) had in
mind when naming the plant. Scheele described the plant
from New Braunfels whence Lindheimer made several collec-
tions, each distributed in several herbaria. It appears
that all of the specimens in Table 4 represent Lindheimer's
poisonfiivy gatherings from New Braunfels.

If we can trust Lindheimer's number system, the 1847
and later collections must be excluded from considerations
of a type, Since, before the 1847 collections were made,
Roemer took the specimens to Germany--specimens which Scheele
described in 1848 (McKelvey, 1955). AS Gray (1850) and
Fernald (1941) decided, Lindheimer 555 probably represents

the collection which Scheele had under consideration, but

nu
Oz

2

 

 

 

 

omsqhsa-oz Hmma mcflpflaam Saganaacc:
omx omma GCHHoSOHw pmnmnsscc:
omsmhea-02 .nn.
.mmooov-sz Kama mcflpflaum Sam >H afloaommm
amfiahsa-oz .ru mama Sm: AaHaBOHo gamv mcapflaaw How
mmsvhsa-oz
.quVKuauoz .II.
.aamaama-m: .rw Gama mcfipaaao cam Haw aaaaammm
ameKSH-02 .ru Gama Hangapaam acapaaom cam
mosqhsa-os Gama paama< afiaoapm new
chom mHm£3
mwumnmom mpmo mmmpm Honssz cowpowaaoo
m<XmH .mqmmz:<mm gmz 20mm monHomaqoo mmEHmroqu--.v mqm<H

291
there are two collections so numbered. The collections given
the same number in 1847 (added in longhand on a printed label
which reads "fasc. iv. 1847") are probably intended to rep-
resent the same pOpulation, only a later collection, hence
the same number. Since Lindheimer 401 of the regularly-
numbered series is Lagenaria vulgaris, the handwritten labels
numbered 351 for a poison-ivy no doubt fall into the category
which Blankinship (1907) indicated were Specimens not col-
lected in quantity, nor even in numbered exsiccatae, but are
private Lindheimer numbers for the Engelmann Herbarium (now
at MO). In Gray's listing, Lindheimer 2450f the regular
series is Sedum gparsiflorum and 235.is miswritten as a typo-
graphical error: "555" (it appears in a consecutive list
between 255 and 251 and hence must be intended to be 255);
it iS Galium virgatum. Hence, 235 and 255 are again examples
of private numbers for the Engelmann Herbarium, not to be
confused with the series of numbered exsiccatae.

The problem remains as to which specimen should be
named lectotype. Lacking Scheele's specimen, a good candi-
date would be the one which was in Gray's possession at the
time, inasmuch as Gray wrote that Lindheimer 346 was 5555
verrucosa of Scheele. At the Gray Herbarium, there is a
pair of specimens which were obviously on the same sheet and
separated only in relatively recent years. The Sheet had
two Specimens and three labels on it originally. One Speci-
men was probably part of the May 1845 gathering (according
to one of the three labels), and the other a part of a

292
September 1846 gathering. The printed label (associating

the plants with Fasc. iii, Lindheimer 346) was superfluous

 

originally since neither plant was part of the gathering
which was distributed under the 535 number. The Specimen

to which the printed label is affixed is now highly Signifi-
cant because it has been annotated "Rhus verrucosa Scheele"
in Gray's handwriting. The Specimen which has now been
detached from Gray's annotation more nearly fits Scheele's
description than the Specimen which remains attached to it,
however, and hence has been chosen lectotype (see Plate
XXXIII.

The excluded specimen mentioned above (No. 255) has
been annotated 5555 radicans var. rydbergii by Fernald, but
is more likely a sparsely pubescent Toxicodendron radicans
subsp. pubens. Among other reasons for this choice is the
fact that 1, rydbergii does not occur in this part of Texas.
It is indicated in Gray (1850) by a symbol used to designate
Specimens which were not distributed. A Specimen at the
Engelmann Herbarium has the same printed label marked 555,
but also has an attached label in Lindheimer's nearly illeg-
ible handwriting giving it collection number551. The con-
dition of the drupes on both of the 351 Specimens suggests
that they were not fresh at the time of collection, but had
overwintered, thereby confirming their May date of gathering.
The fruits of all other 555_Specimens are fresher and were
no doubt produced during the season of collection. At least

one earlier worker at the Missouri Botanical Garden has

PLATE XXXIII.

293

Lectotype of Rhus verrocusa. Photo
was taken of overlapping sheets,
showing that they had once been joined.
Specimen on left is type of Rhus ver—

rucosa Scheele. Specimen on the right

is Toxicodendron radicans subSp. puben .

 

 

294

 

295
noted on two Specimens the idea that 5fi5_= 551.

It is possible that all poison-ivy specimens that
Lindheimer had on hand at the time of making his fascicle iii
were included in the distribution under the number 535,
whether they were collected in 1846 or before, especially
if they were destined for Gray or Engelmann, his employers.
Whether Scheele and Gray saw plants from the identical gather-
ing can never be ascertained without an extant Scheele plant.
What is assured, however, is that the May 1845 and the 5g5_
gatherings have the insect chewings on the leaves which pro-
duced exudations of poison, now hardened into black specks.
It was these which Gray maintained were the "verrucae magnae
subrotundae atr0purpureae lucidae" of Scheele's description,
exudations which may have led Scheele to his choice of epithet
"verrucosa." There is nothing else verrucose about any of
the plants he could have had under consideration. Further-
more, the handwritten labels attached to the Specimens which
Gray examined declare them to be "small Shrub on stony prair-
ies" and "high, in thickets and climbing, N. Braunfels" in
Engelmann's handwriting. These phrases agree with Scheele's
"In Hecken auf felsen Boden auf Missionshugel bei Neubraun-
fels: Lindheimer."

I believe that the May 1845 collections (Lindheimer
£51) distributed in part under exsiccatae number 515 with
the date 1846 represent what Asa Gray had before him. The
Engelmann Herbarium Specimen under the same number is there-

fore an isolectotype. All other Specimens distributed under

296
the number 5fl5_are considered to be t0potypes and therefore
of interest, but not of nomenclatural significance.

Greene transferred 5555 verrucosa and 5555_rhomboidea
to Toxicodendron. It is curious that he did not note the
similarity of these two, both of which came from the same gen-
eral region of Texas.

This subspecies is found on the Edward's Plateau of
central Texas and the Arbuckle Mountains of Oklahoma, reaching
its northern limit at Turner Falls, Murray County, Oklahoma.
In northeastern and south central Texas, the subSpecies is
found off the Plateau, but more or less maintains its distinct
glabrous and sharply lobed leaflets. There is little apparent
intergrading with other taxa, although some is suspected.

Collections of subsp. verrucosum have occasionally
been confused with subsp. eximium, probably because both taxa
have lobed leaflets. The latter has pubescent fruits and
leaves, whereas the former is glabrous. Furthermore, leaf-
lets of subSp. verrucosum have acutely—angled lobes, whereas
the lobes of leaflets of subSp. eximium are rounded.

This taxon grows in the variety of habitats like those
of most of its related subSpecies but it is most frequently
found in limey soils: lithosols, talus, limestone gravels,
limestone bluffs, etc. It is also a roadside weed, found on
alluvial banks, forest borders, and even in oak-pine woods
on fairly excessively drained acid soils. It has also been
reported from thin soil over chalk in Dallas County, Texas.

Its community relationships have not been extensively explored.

"t -> :55. mW

 

297
4n. Toxicodendron radicans subsp. eximium (Greene) Gillis,

comb. nov. (Plate XXXIV.)

Toxicodendron eximium Greene, Leafl. Bot. Observ.
Crit. l: 123. 1905. ‘5555 eximia (Greene)
Standley, Contr. U.S. Natl. Herb. 23. 668. 1923.
5555 toxicodendron var. eximia (Greene) McNair,
Publ. Field Mus. Nat. Hist. 225 Vol. 4: 68. 1925.
Toxicodendron radicans var. eximium (Greene)
Barkley, Ann. Mo. Bot. Gard. 24: 435. 1937 (as
eximia). Type: 51, Edward Palmer 155, Mexico,
State of Durango, vicinity of the city of Durango.
April - November, 1896. (US-305009, holotype;

A, F, GH, MO-1774807, NY, isotypes).

Toxicodendron biternatum Greene, Leafl. Bot. Observ.

 

Crit. l: 124. 1905. Type: y, Havard, 515.,

Eagle Nest on Rio Grande River (Presidio), Pre-

sidio County, Texas. 1881. (US-156164, holotype).

Vine or Shrub with aerial roots; leaflets broadly

ovate, deeply-cut, rounded, incised-dentate, sinuate-lobed,
sometimes tending to be trilobate; scabreScent to velutinous
above, relutinous to tomentose below, veins hirtellous. Term-
inal leaflet blade 1.1 - 13.0 cm long, 0.7 - 11.5 cm wide;
petiole hispid, tomentose, or villous; terminal petiolule
0.1 - 4.5 cm long. Fruits 3 - 5 mm broad, glabrate, setose,_
or hispidulous, yellow-cream. Blade tip acute to acuminate,
base cuneate. Distribution: Uhited States: southwestern

Texas; Mexico: Coahuila, Durango, Tamaulipas. (Fig. 12.)

 

 

 

PLATE XXXIV .

298

Specimen of Toxicodendron radicans subsp.

eximium. Note variation in leaf morphology

in a single individual. The smaller
leaves are from a shaded branch of the

same vine as the one from which the larger,
sun leaves were collected.

M‘CHIVLAN STATE

LINN. ( WS'TV HERDARIUH

 

299

  
  
  

   

PLANTS 0" l". n:

[511500. 51705 ”I ”In; (L, 1 Aunll. war.
5“ (Or-om) lurk

urn-nor Co., In; Band hallo-ml Pun An
I

n
f at DIIVAIIDI

 

all|r loll.
In!
llllknl T. 01111. P .a11r-L.r 111~
u57~5

Snail-r lonv-I nru fro- IL-do (fro.

Inhlu-Ihl Ill".-
Ilflflflll IIAII UIIVIIIIII

300

Representative Specimens: UNITED STATES. Texas.

 

Brewster Co.: Big Bend National Park, Chisos Mts., Warnock
1212 (ARIZ, GH, TEX); Willow Creek Canyon, Sperry 251 (US);
Palmer 53515 (A, MO, NY); Goode Place, 1555 , 515. (GH, TEX);
Oak Canyon, 553y_1515, 15552 (GH); Mueller 5555 (F, TEX);

Oak Canyon, base of Ward Mt., Gillis 5155 (MSC). Maverick
Co.: Eagle Pass, Havard, 5,5, (GH). Presidio Co.: Camp
Eagle Nest (Presidio), Havard, 5.5, (US); Fresno Canyon,
tributary of Mexican Canyon, Hinckley 5525 (SMU, U); Marfa,
Hinckley 2515 (ARIZ, GH, NY); Shafter, Cibolo Canyon, Hinckley

 

.2555 (A, ARIZ, NY, SRSC). Valverde Co.: Nueces River near
Rio Grande, Palmer 55355 (A, MO, NY, US); High bridge of
Pecos River, Pilsbr , 515. (PH); Devils River, Palmer 15555
(A, MO, WIS); Devils River, Traverse 21g5_(SMU).

*****

MEXICO. Coahuila. Muzquiz, Marsh 1166 (GH); Sierra

 

 

Hechiceros, Stewart 158 (GH). Durang . Near Durango (Nombre
de Dios), Palmer 106 (A, GH, MO, NY, US). Tamaulipas. Vic-

 

 

toria, Palmer 154 (US, A, GH, NY). 'El Rosario, vicinity of
Marmolejo, Bartlett 10881 (MICH). Nuevo Le6n. San Augustin,

 

 

Monterrey, A55gg_2fl5_(A, F, MO, US).

The type was collected a long distance from any other
known stations for this subspecies. Greene (1905) indicates
that it was collected at Nombre de Dios, 40 miles southeast
of Durango. The type for 1, biternatum was collected at Eatle
Nest on the Rio Grande River. Barkley (1937) mapped this in
Maverick County, Texas, but he probably confused this locality

301
with Eagle Pass. The actual locality was Camp Eagle Nest,
where Havard was post surgeon from August 1880 until June
1881 (McVaugh, 1957). The camp has long Since disappeared,
but it occupied the present site of the city of Presidio in
Presidio County. This flowering Specimen must have been
collected in the Spring of 1881, although the type specimen
indicates only "188_,"

This subspecies is one of the more unusual and rare
forms of poison-ivy. It is found rarely along the Rio Grande
River itself, but is more likely to be found in the washes
and tributaries to that river. It apparently had not been
collected between 1937 and 1964, at which time Dr. Barton
Warnock helped me find a station that was known to him years
earlier in Oak Canyon of the Chisos Mountains of Big Bend
National Park. This plant is restricted to the canyon bottom
where it is abundant, but no trace of it is found anywhere
on the walls of the canyon, nor on the ridges above.

The subSpecies is distinctive by its peculiarly

lobed leaflets which often tend to be trilobate to the extent
that they resemble the "club" on a deck of playing cards.
The pubescence of the leaves and fruits suggests a relation-
ship of Toxicodendron toxicarium, but the climbing habit and
aerial roots require its placement in 1, radicans. As is
the case in other taxa, shade leaves are much smaller than
sun leaves (sic), as Shown in Plate XXXIII.

Superficially this taxon resembles Toxicodendron

radicans subsp. verrucosum and has frequently been

302
misidentified as such. Earlier maps of its distribution
have been in error when they indicated its being found in
central Texas, probably because of confusion with subsp.
verrucosum, a more widely distributed taxon. (See discus-
sion under 1, radicans subSp. verrucosum.)

It is not easy to characterize the niche which is
occupied by this plant for there have been few studies of
its community associations. The only time that I have stud-
ied subSp. eximium in the field was in the one canyon bottom
mentioned above. There it had the following assemblage of
associates: Quercus gravesii, Garrya 23313, Juniperus 1155-
5155, 55 5 virens, 11115 arizonica, Arbutus texana, and

Ungnadia Speciosa. The soils in which it grew were shallow

 

volcanic lithosols containing calcium concentrations of
1099 ppm.
5. Toxicodendron rydbergii (Small ex Rydberg) Greene.
(Plate XXXV.)
5555 rydbergii Small ex Rydberg, Mem. New York Bot.
Gard. l: 268. 1900. Toxicodendron rydbergii
(Small ex Rydberg) Greene, Leafl. Bot. Observ.
Crit. l: 117. 1905. 5555 toxicodendron var.
rydbergii (Small ex Rydberg) Garrett, Spring Flora
Watsatch Range, 3rd. ed., 68. 1917.. 2555 radicans
var. rydbergii (Small ex Rydberg) Rehder, J.
Arnold Arbor. 20: 416. 1939. Toxicodendron 15517
5555 var. rydbergii (Small ex Rydberg) Erskine,

Pl. Prince Edward Island, 190. 1961. Syntypes:

303

PLATE XXXV. Specimen of Toxicodendron rydbergii.
Type of Toxicodendron fothergilloides
Lunell.

 

504

  
   
   

..rn- 1....
A nouns-mar mm 0' ”(1" AND VT!
WA?! nun ”u Wm “all"
Iflflfllflfli [3. mg. 3.1 "rm!“

'5 = Wm 91mm 110143;
n in". 11“. r: t. 2: lid. 191.2.)
a

{d 4 2/3/31

MONT” DAKOTA FLORA

H.170». Egg: n, / 7

Cow-nub, D- ) Ludl

305
15. 5. Williams 291, Montana, Cascade Co., Great
Falls, 1885; 5. 5. Anderson, 5,5,, northern

 

Montana, Helena, Lewis and Clark Co. (NY, syntypes).
Rhus toxicodendron var. vulgare Michx., Fl. Bor. Am.,
183. 1803. Rhus toxicodendron vulgaris a. radicans

Engler in DC., Monogr. Phaner. 4: 394. 183. Rhus

 

radicans var. vulgaris Roemer et Schultes, Sys.

Veg. Vol. 6. 1820, non Toxicodendron vulgare Milka;
Gard. Dict., Art. Toxicodendron, No. l. 1768.

Type: "Hortus parisiensis," lectotype (P).

Toxicodendron macrocarpum Greene, Leafl. Bot. Observ.
Crit. l: 117. 1905. Type: 5, 5,‘Thompson 152,
Kansas, Hamilton Co., 11 July 1893 (US-265734,
holotype).

Toxicodendron longipes Greene, Leafl. Bot. Observ.
Crit. l: 118. 1905. Type: Lester E, 5535 212,
Canyon south of Glenwood, Sevier Co., Utah, alti-
tude 6300 feet, 12 June 1875 (US-153641, holotype).

Toxicodendron hesperium Greene, Leafl. Bot. Observ.
Crit. l: 118. 1905. Type: .5155 Whited 241,
Washington, Chelan Co., Wenatchee, 6 August 1893
(1896?) (US-268197, holotype).

Toxicodendron pumilum Greene, Leafl. Bot. Observ. Crit.
l: 124. 1905, non 1, pumilum Kuntze, Rev. Gen.

Pl. pt. 1, 154. 1891. Type: 5, 1, MacDougal 25,
Arizona, Coconino Co., vicinity of Flagstaff,

altitude 7000 feet, 28 June 1898 (US-334125,

 

 

306
holotype; F-69784, GH, NDA, NY, RM-3l903, iso-
types).
Toxicodendron punctatum Greene, Leafl. Bot. Observ.
Crit. l: 125. 1905. Type: 5, 5, Metcalfe 1555,
New Mexico, Sierra Co., Black Range, near Kingston,
9 July 1904 (US-498281, holotype; F-187484, GH,
MIN-294394, MO-l773740, NY, US-890258, isotypes).
Toxicodendron desertorum Lunell, Amer. Midl. Naturalist
2: 185. 1912. Type: .1. Lunell, 5,5,, North
Dakota, McHenry Co., gravelly soil sand hills,
13 July 1899 (MIN-245797, holotype).
Toxicodendron fothergilloides Lunell, Amer. Midl.
Naturalist 2: 186. 1912. Type: '1. Lunell 155,
Ramsay Co., North Dakota, thickets at Devil's
Lake, 29 June 1902 (MIN-245796, holotype).
Sub-shrub often less than 1 meter high to shrub up to
3 meters. Stem simple or Sparsely upright-branched, arising
from much branched subterranean stolons; aerial roots absent,
plant never climbing. Leaves approximate near summit of
stem, often appearing falsely verticillate. Terminal leaflet
blade broadly ovate, rhomboid or suborbicular and usually
acuminate, 3.0 - 15.0 cm long, 2.0 - 11.0 cm wide; petiole
glabrous (-glabrate), 1.5 - 25.0 cm long; terminal petiolule
0.6 - 5.5 cm long. Leaflet margins dentate or undulate,
subtruncate or rounded at the base, glabrous above, often
with a line of minute, curly hairs along the midrib, glabrous

or appressed strigose below; autumn coloration generally

307

yellow to orange, less often bronze or red. Flowers in small
clusters. Fruit globose, cream to yellow, (3-) 4 - 7 mm
broad, glabrous, smooth.

Distribution: Canada: southern provinces north to
520 N. Lat. United States: central Washington, Oregon, Utah
south to northeastern Arizona and western Texas, east to
central Kansas, Nebraska, Iowa, southern Wisconsin, northern
Michigan, New York, New England; few mountain tops, Appalachian
Mountains. Common name: Rydberg's poison-ivy, poison-oak
(erroneously). (Fig. 14.)

Representative Specimens: UNITED STATES. Arizona.
Apache Co.: Bog Creek, 8 miles east of McNary, Deaver 5521
(ARIZ). Cochise Co.: Big Emigrant Canyon, Blumer, 5,5, (ND-G,
NY); 2 miles sw of Portal; Matthews _8._ Matthews 383 (MSC);
Chaperon Canyon, Blumer 1211_(E, ND-G, NY); also Blumer 1525
(ARIZ, E, F, GH, KSC, MIN, MO, ND-G, NY, RM, US); Cabeyas Mts.,
Simmon 155 (GH); Fort Huachuca, Wilcox 215 (US); Chiricahua
National Forest, Burrall 2522 (US). Coconino Co.: bottom,
West Fork Canyon, Oak Creek Canyon, near Sedonia, Demaree
41255 (ARIZ); Bright Angel Canyon, Nelson125 (US); Flagstaff,
McDougal, 5,5, (F, GH, NDA, NY, RM, US); Navaho Reservation,

 

Vorhies 52 (A, ARIZ, GH, MIN, NY); Gila Co.: Parker Creek
Branch Station, Little 4551 (ND). Navajo Co.: Betatakin
ruins, Howell 24552 (ARIZ); Fort Apache, Mayerhoff 41 (F).
Pima Co.: Coronado National Forest, 51532 4552 (SMU). Santa
Cruz Co.: Santa Rita Forest Reserve, Griffiths 4222 (US).

Yavapai Co.: Montezuma Well near Camp Verde, Gillis 6602

308

Fig. 14. Distribution of Toxicodendron rydbergii.

 

 

509

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

310
(MSC, US); Fort Whipple (2 miles east of Prescott), 55555‘5
Palmer 455 (MO). Locality not located as to county: Willow
Spring, (possibly Coconino Co.) Rothrock 254 (F); Love Lake,
Brinkerhoff, 515. (ARIZ); Colorado. Boulder Co.: Boulder,

 

 

 

Tweedy, 5,5. (NY, RM); Gregory Canyon near Boulder, Robbins
1607 (RM). Denver Co.: Denver, Smith, 515. (KSC). El Paso

Co.: Pikes Peak, Ball, 515. (US); Williams Canyon, Harper 5

 

Harper, 5,5, (A). Larimer Co.: Fort Collins, 55555 25, 125_
(GH, US); Canyon of the Thompson near Forks, Osterhout 5551
(RM). Las Animas Co.: Dixon Canyon, collector unknown (NY);
Region of the Mesa de Maya, Troy, Rogers 5525 (TEX, US); Mesa
Co.: Grand Junction, Eastwood 1211 (A). San Miquel Co.:
Norwood Hill, Walker 425 (GH, MIN, NY, RM, US). Montezuma
Co.: Navajo Canyon, Mesa Verde, 551y_151.(US). weld Co.:
Windsor near river, Osterhout 1125 (ND-G, RM). 15555. Bonner
Co.: Big Potlatch River, Christ 4215 (NY); HOpe, Sandburg
255_(GH). Bannock Co.: Inkom, 5552 1142 (A). Bonner Co.;
Lake Pen D'Oreille, Christ 11181 (NY); Sunnyside, Christ 2555
(NY); Boundary Co.: Bonner's Ferry, Epling 15455_(MO). Idaho
Co.: Cottonwood, Christ 15552 (NY). Kootenai Co.: Coeur
d'Alene, Christ 12125 (NY). Latah Co.: Julietta, Christ
15151 (NY). Lemhi Co.: Indianola Ranger Station, Hitchcock
5_Muh1ick 14524 (NY). 111igg1=. Cook Co.: Glencoe, Shore

 

of Lake Michigan, Churchill, 5,5, (MIN); Stoney Island, Smith
6003 (F, GH, MO). Du Page Co.: Wheaton, Moffatt 594 (WIS).
Lake Co.: locality unknown, Gates 2506, 2805 (F); Waukegan,

Steyermark 5 Standley 28186 (F). Winnebago Co.: locality

311

unknown, Bebb, 5.5. (F). Indian . Benton Co.: Pine, Merner

 

14251 (WIS). Lake Co.: Gary, Thieret 1555 (F); Whiting,
Lansing 5E (F). Laporte Co.: wooded dunes 2 miles west of
Michigan City, Deam 8854 (IND). Marshall Co., Lake Maxin-
kuckee, Evermann 833 (US). Porter Co.: Chesterton, M
15551 (A); Dunes State Park, 1r_y_o_p 1511 (F). _I___cm__5. Cerro
Gordo Co.: Mason City, Shimek, 5.5. (NY). Clayton Co.:
Giard, Grether 8136 (WIS). Decatur Co.: locality unknown,
Anderson, 5.5. (MO). Dickinson Co.: Spirit Lake, Evermann,
5.5. (F). Johnson Co.: Lake Macbride State Park, 5.5.5.
54155 (FSU). Shelby Co.: locality unknown, Fitzpatrick,
5.5. (MSC). Kansas. Cloud Co.: locality unknown, Fraser
155 (KSC). Commanche Co.: locality unknown, Hitchcock, 5.5.
(KSC). Cove Co.: locality unknown Hitchcock, 5.5. (KSC).
Cowley Co.: locality unknown, Clothier 5 Whitford, 5.5.
(KSC). Decatur Co.: locality unknown, Hitchcock, 5.5.
(KSC). Dickinson Co.: Westgate, _s_.5. (KSC). Douglass Co.:
locality unknown, Hitchcock, 5.5. (KSC). Ellsworth Co.:
Ellsworth, Palmer 21252 (US). Hamilton Co.: Coolidge,
Hitchcock, 5.5. (KSC); Syracuse, Thompson _l_02 (GH, KSC, MO,
US). Hodgeman Co.: locality unknown, Hitchcock, 5.5. (KSC).
Johnson, Olathe, Hitchcock, _s_._r_1_. (KSC). Kearney Co.: local-
ity unknown, Hitchcock, 5.5. (KSC). Kiowa Co.: without
definite locality, Hitchcock, 5.5. (KSC). Leavenworth Co.:
without definite locality, Hitchcock, 5.5. (KSC); Logan Co.:
without definite locality, Hitchcock, 5.5. (KSC). Marshall
Co.: without definite locality, Clothier 5 Whitford, 5.5.

312
(KSC). McPherson Co.: Lindsberg, Bodin, s.n. (MIN). Meade
Co.: without definite locality, Hitchcock, 5,5, (KSC).
Morton Co.: without definite locality, Hitchcock, 5,5. (KSC).
Ness Co.: without definite locality, Hitchcock, 5,5. (KSC).
Osborne Co.: Osborne City, §D§é£.222 (RM, US). Pawnee Co.:
without definite locality, Hitchcock, 5,5, (KSC). Pottawat-
omie Co.: St. George, Kellerman, 5,5, (MO). Pratt Co.:
without definite locality, 15p_Vranken 155 (KSC). Reno Co.:
without definite locality, Hitchcock, 5,5. (KSC). Republic
Co.: without definite locality, 55115, 5,5, (KSC). Rush
Co.: without definite locality, Hitchcock, 515. (KSC).
Russell Co.: without definite locality, Hutchinson 215 (KSC).
Saline Co.: Salina, Hancin 2425 (KSC). Sheridan Co.: with-
out definite locality, Hitchcock, 5,5, (KSC). Stanton Co.:
without definite locality, Hitchcock, 515. (KSC). Sumner Co.:
without definite locality, Clothier 5 Whitford, 5,5, (KSC).
Thomas Co.: without definite locality, Hitchcock, 515. (KSC).
Trego Co.: without definite locality, Hitchcock, 5,5, (KSC).
Wallace Co.: without definite locality, Hitchcock, 515. (KSC).
Webster Rock Co.: without definite locality, Griebel, 5,5,
(KSC). Wichita Co.: without definite locality, Hitchcock,
5,5. (KSC).. M5155. Aroostook Co.: Fort Fairfield, Fernald

 

1225 (GH); Fort Kent, 55555 1154 (A). Kennebec Co.: WinthrOp,
Sturtevatii, 515. (MO). Lincoln Co.: Boothbay, Fassett

52525 (SMU); White Island, Churchill, 5.9, (Mo). Somerset
Co.: Fairfield, Fernald 5 155g 14511 (GH). Washington Co.:
OxCove, Fernald 1225 (GH). County unknown: Bristol,

313

 

Chamberlain, 515. (ND-G). 555sachusetts. Berkshire Co.:

Berkshire, Churchill, 5,p1_(MO). Michigan. Alcona Co.:

 

Crooked Lake, Gillis 4155 (MSC). Alger Co.: Au Train Lake,
Gillis 5545 (MSC); Devil's Slide, Gillis 5542_(MSC). Allegan
Co.: Old Baldy Dune near Saugatuck, Gillis 5525 (MSC).
Alpena Co.: Long Rapids Twp., Gillis 4125 (MSC). Antrim
Co.: Eastport, Gillis 5522_(MSC). Baraga Co.: Sturgeon
River, Gillis 5114_(MSC). Benzie Co.: Crystal Lake, Gillis
5154 (MSC); Point Betsie, Gillis 5511 (MSC). Charlevoix Co.:
Charlevoix on dunes, Gillis 2245_(MSC); Beaver Island, 5155,
‘5.5, (BLH); Hat Island, 55155 22552 (KSC). Cheboygan Co.:
university of Michigan Biological Station, Gillis 2245 (MSC);
south shore of Douglas Lake, 55555.525 (RM); Cheboygan,
Shinners 15155 (MIL, SMU); Wolverine, 55555 11525 (KSC).
Chippewa Co.: De Tour, Gillis 4111 (MSC); Drummond Island,
Hiltunen 1552 (MSC, WUD); Sugar Island, Hiltunen 251 (WUD).
Crawford Co.: Lake Margrethe, Gillis 5514 (MSC); Wakeley
Bridge, Gillis 5515 (MSC); Delta Co.: Isabella, Gillis 2255
(MSC); base of Burnt Bluff, Gillis 5545,(MSC). Dickinson
Co.: Quinnesec, Gillis 5552 (MSC); Norway, collector unknown,
(MSC); Menominee River, Gillis 5552 (MSC). Emmet Co.: Mack-
inac City, Gillis 5515_(MSC). Gogebic Co.: Gogebic Lake,
Darlington, 5,5, (MSC); vicinity of 8-Mile Creek, Darlington,
5,5, (MSC). Grand Traverse Co.: East Bay near Acme, Gillis
2552_(MSC); south shore of East Arm of Grand Traverse Bay,
Gillis 5525 (MSC); mixed woods below Brown Bridge Dam on

bank of Boardman River, Dieterle 1722 (MICH). Houghton Co.:

314
Bear Lake, Richards 5225 (MICH, MIN, NY). Iosco Co.: Nissik-
one Camp for Boys, Gillis 5552 (MSC); East Tawas, McVaugh
11451_(BLH, MICH). Iron Co.: Horserace Rapids, Gillis 5541.
(MSC). Kent Co.: Ross, 55555,.515. (MO). Keewenau Co.:
Anich, Farwell, 5,5. (MSC); Lac LaBelle, Farwell 2212 (BLH,
MSC); Eagle River, Bourdo 5145 (MSC). Lake Co.: Pine River,
Dover Twp,, Gillis 5525 (MSC). Leelanau Co.: Sleeping Bear
Dune, base of pinnacle, Gillis 5452 (MSC); Glen Haven, Herman
2225 (F); Good Harbor Bay, Grigsby, 5,5, (MSC); South Manitou
Island, morainic bluff, 1555 2515_(MICH). Luce Co.: Man-
istique Lake shore, Gillis 5552 (MSC). Mackinac Co.: St.
Ignace, Gillis 5555 (MSC); Thuja forest, Mackinac Island,
Potzger, 7308 (ND). Manistee Co.: Portage Park dunes,
Umbach 5451 (MICH). Marquette Co.: Tankowa Lake, Gillis
5415 (MSC). Mason Co.: Ludington, dunes, 55555 12555 (ILL,
NY). Menominee Co.: Menominee, Grassl 2515 (MICH, NY);
shore of Lake Michigan along highway, Waterman 512 (MSC);
evergreen woods near beach at Fox, Grassl 1514 (MICH). Mis-
saukee Co.: HOpkins Creek, Gillis 5555 (MSC); Goose Lake,
Gillis 5551 (MSC). Montmorency Co.: West Twin Lake, Gillis
5511 (MSC). Muskegon Co.: near Coast Guard Station near
North Muskegon, Gillis 5525 (MSC). Oceana Co.: Pentwater
dunes,'55555 12115 (ILL); Little Sable Lighthouse, Gillis
5551_(MSC). Ogemaw Co.: Piper Lake, Gillis 5554 (MSC).
Ontonagon Co.: Escarpment overlooking Lake-of-the-Clouds,
Gillis 5155 (MSC). Oscoda Co.: Mio,"GilliS 411519 (Msc). Ot-
sego Co.: Otsego Lake, Gillis 5551 (MSC). Ottawa Co.:

 

 

F
.
.
< h n '
t Q
s r
\
.
O
0 I
O
5 .

 

,315
Ferrysburg dunes, Gillis 5525 (MSC). Presque Isle Co.:
Rogers City, McVaugh 2225 (BLH, MO, MICH). Roscommon Co.:
Roscommon, Gillis 5555 (MSC). Schoolcraft Co.: Hiawatha
National Forest, Shore of Clear Lake, Waterman 551 (MSC).
Wexford Co.: Greenwood TWp., above Manistee River on High
Roll Way, Gillis 5525 (MSC). Insert Bay Co.: 1 mile north
of Waterworks Park, Bay City, Dreisbach 1541_(MICH). 51555-

sota. Becker Co.: Detroit, Woladron, 5.5. (NDA). Carlton

 

Co.: Thompson, Sandberg54l, (KSC, MISC). Carver Co.:
without definite locality, Anthony 5 Waconia, 5,5, (WIS).

Clay Co.: Buffalo State Park, Stevens, 5,5. (NDA). Clear-
water Co.: Schoolcraft Island in Lake Itasca, Coffin 5
15555_15_(ARIZ). Dakota Co.: without definite locality,
55555 5 _M_c_>_o_r_e_ 15214 (BRY). Hennepin Co.: St. Anthony Park,
Schuette 5,5, (GH); Itasca Co.: Ingersol, Johnson 1521_(GH,
MIN, US); Bena, Gillis 5115 (MSC); Sand Lake, Johnson 2122,
(M0). Jackson Co.: Kilen Woods State Park, 55555 5 5511
12541_(SMU). Lake of the Woods Co.: without definite local-
ity, Mggg5_5.55555 12155 (NDA). Marshall Co.: Warren, Gillis
5111 (MSC). Norman Co.: Frenchman's Bluff, Stevens, 555.
(NDA). Ramsey Co.: St. Paul, St. Anthony Park, Lugger, 5,5,
(NDA). St. Louis Co.: Duluth, Lakela, 515, (F, NY, US);

Ely, Vandrey, 515. (MSC). Stearns Co.: Paynesville Twp.,
‘55555 5 Phinney 12552 (A). Winona Co.: Winona, Holzinger,
5,5, (GH). Montana. Cascade Co.: Great Falls, Williams 221
(NY, US). Dawson Co.: Glendive, 5555, 5,5, (US). Flathead
Co.: Big Fork, Butler 5552 (NY); Flathead Lake, Clemens,.5,5.

316
(F). Gallatin Co.: Bozeman, Blankinship 106 (F, MO, US).

 

Lake Co.: Swan Lake, Jack 2377 (A). Lewis and Clark Co.:
Helena, Anderson, 5.5. (NY). Madison Co.: Upper Madison

River, Lenz, 5,5. (MO, ND). Park Co.: Livingston, Booth

 

3802 (NDA). Stillwater Co.: Absarokee, Hawkins, 515. (WIS).
Nebraska. Brown Co.: Long Pine, Rutter, 515. (US); Columbus,

 

Palmer 36060 (MO, NY). Dixon Co.: Newcastle, Clements 2555
(NY, US). Franklin Co.: Franklin, Laybourn 41 (M). Gage Co.:
without definite locality, Knight, 515. (RM). Kearney Co.:
Minden, Hapeman, 515. (ARIZ, FLAS, MIN, MO, RM, SMU). Otoe
Co.: Syracuse, Osborn 1525 (M0). Saline Co.: Crete, 51551,
‘5,5. (BRY). Sheridan Co. Metcalf Public Hunting Grounds,

13 miles north of Hay Springs, pine forest,'51555 55 (RM).
Sioux Co.: Monroe Canyon, 5555;, 5,5, (MO). Stanton Co.:
Stanton, Lindbo, 5,5. (US). Thomas Co.: Natick, Rydberg
1415 (US); Helsey sand hills, Breneckle, 5,5. (NDA). 555

 

Hampshire. Carroll Co.: Intervale, between Conway and Bart-

 

 

lett, Rehder 221 (A, US). Coos Co.: Shelburne, Mpg;5_4111
(GH). Grafton Co.: Hanover, Hitchcock, 5,5, (MIN, US);
Thornton along railroad, Gillis 4525 (MSC);West Thornton,
ridge west of Pemigewasset River, Gillis 4524 (MSC); Warren,
Williams, 5,5. (GH). Hillsboro Co.: Fox State Forest,

Baldwin, 5,5, (MSC); Manchester, Smith, 5,5, (A). 555 Mexico.

 

Bernalillo Co.: Capelin Canyon, Sandia Mts., Ellis 245
(MO, NY, US). Catron Co.: Pleasanton, Anderson 1507 (MSC);
Gila River west fork, Metcalfe 339 (GH). Colfax Co.: Ute

 

Park, Standley 14036 (US). Lincoln Co.: Gray, Earle 5 Earle

317
455 (NY); White Mts., Woolton, 515. (US); Ruidosa Creek on
Mescalero Reservation, 7 miles west of Ruidosa, Hinckley155
(ARIZ, F, GH, NY, SRSC, TEX). Sandoval Co.: Tijeras Canyon,
Sandia Mt. north end, Stoutamire 5212 (MSC). San Miquel Co.:

 

las Vegas, Tuttle 41 (NY); Pecos River Division, Santa Fe
National Forest, Standley 4011 (GH, NY, RM, US). Sierra Co.:
Kingston, Metcalfe 1088 (F, GH, MIN, MO, NY, US). Socorro

 

Co.: Gila River, Mogollon Mts., Metcalfe 339 (ARIZ, E, MIN,

 

MO, ND-G, NY, RM, US); Water Canyon, Magdalena Mts., Herrick
‘5 Herrick 15 (F). Locality uncertain: Orate Creek, Wisli-

zenus 502 (M0). New York. Essex Co.: Port Henry, Britton,

 

‘5.5, (NY). Monroe Co.: Inspiration Pt,, Matthews 2151 (WIS).
Niagara Co.: Niagara Falls at base of Bridal Veil Falls,
Gillis 5515 (MSC). Onondaga Co.: near Syracuse, Underwood,
5,5, (E). Oswego Co.: Selkirk, Fernland, 55555, 5 Wiegand
14514 (GH). St. Lawrence Co.: Lisbon, Phelps 542 (GH, NY).
Tompkins Co.: without definite locality: Wiegand 12451 (GH).
Ulster Co.: Lake Mohonk, Gillis 4551 (MSC). Warren Co.:

The Glen Pack Forest, Gillis 5515 (MSC). Washington Co.:

Pilot Knob Lake George, House 28972 (TEX). North Carolina.

 

 

Cumberland Co.: Fayetteville, Miles, 555, (US). North Dakota.

 

Barnes Co.: Valley City, Bergman 525 (NDA); Little Yellow-
stone Creek, collector unknown (NDA). Benson Co.: Butte,
Lunell, 515. (NY); Ft. Totten, Bergman 1525 (MIN, MO, NDA);
Minnewauken, Lunell, 5,5, (NY); Pleasant Lake, Lunell, 515.
(MIN). Billings Co.: Medora in ravine, Bergman, 555. (NDA).
Bottineau Co.: Lake Metigoshe, Stevens, 515. (NDA).

318
Burleigh Co.: Bismarck, Apple Creek, Stevens, 5,5, (NDA).
Cass Co.: Fargo along Red River, Bergman, 515. (NDA).
Cavalier Co.: Tongue River, Stevens, 5,5, (NDA). Dunn Co.:
Killdeer Mts., Stevens 5_Mg15,.515. (NDA). Eddy Co.: Hamar,
Stevens, 5,5. (NDA). Emmons Co.: Linton, Stevens, 5,5,
(NDA). Golden Valley Co.: Sentinel Butte, Bergman, 5,5.
(NDA). Grand Forks Co.: Turtle River State Park, Stevens,
5,n. (NDA); Grand Forks, Egebretson 145 (F). Grant Co.:

Wade, Bell, 5,5. (NDA). Kidder Co.: Dawson State Game Ref-

 

uge sand flat, Stevens, 5,5. (NDA). Logan Co.: Napoleon,
Stevens, 515. (NDA). LaMoure Co.: Kulm, Brenckler, 555.
(NDA). McHenry Co.: Sand Hills, Lunell, 515. (MIN); Towner,
Lunell 154, (A, MIN). McLear Co.: T. 48, R. 91 woods,
Heidenreich, 5,5, (NDA). Morton Co., Mandan, Bergman, 5,5,
(NDA). Oliver Co.: Fort Clark, Hayden, 5,5. (MO). Pembina
Co.: Walhalla stream bank, Bergman 1255 (NDA, OSC). Pierce
Co.: Lake Girard, lower edge of prairie coulee, Stevens,
5.5, (NDA). Ramsay Co.: Devil's Lake, Lunell 155_(MIN, RM,
WIS). Ransom Co.: McLeod, Stevens, 515. (NDA). Richland
Co.: Hankinson, lake Shore, Bergman 554 (NDA); Kindred,
sandy prairie, Stevens 2455 (NDA); Wahpeton, Bergman, 5,5.
(RM, WIS). Rolette Co.: Duneith, Stevens, 5,5. (NDA).
Sargent Co.: Milnor, Stevens, 5,5, (NDA). ShOpe Co.: Black
Butte in coulee, Stevens & Moir, S.n. (NDA). Stark Co.:

Hebron, Side of butte, Stevens 5 Moir, s.n. (NDA). Ward Co.:

Minot, Lakela 389 (MIN). Wells Co.: NE slope of Coteau,
Stevens, 5,5. (NDA). Williams Co.: Williston, Bell 191 (NDA).

 

319
Ohio. Ashtabula Co.: Geneva-on-the-Lake, Rood 2484 (MIN).

 

Oklahoma. Beaver Co.: Knowles, Stevens 516 (MO). Oregon.

 

 

Baker Co.: Wallowa National Forest, Peterson 255_(OSC).
Bonner Co.: Kootenay, 5y511, 5,5. (GH). Deschutes Co.:
Redmond, Whited 122 (ND). Grant Co.: Canyon City, 2555
15114 (NY). Jefferson Co.: west of Culver on bank of
Crooked River, Gillis 5215 (MSC); Warm Springs, Gillis 4215
(MSC). Sherman Co.: Rufus, Gillis 4221 (MSC). Omatillo
Co.: Pendleton, Henderson 22 (GH). Wallowa Co.: Snake
River Canyon, Eureka Creek, collector unknown, (WTU); T.4N,
R. 49E, Sec. 35, Ferguson 255 (OSC); College Creek Ranger
Station, Ingram 2245 (ND). Wasco Co.: Maupin, Gillis 4211

(MSC). Pennsylvania. Lancaster Co.: Lancaster, Ely, 5,5.

 

(MIN). South Dakota. Brookings Co.: Brookings, Moore 52

 

 

(MIN). Clay Co.: floodplain of Brule Creek, 12 miles east
of Vermilion, 5y5£ 1525_(SDU). Custer Co.: Hermosa, Rydberg
524 (NY, US); Jewel Cave National Monument, Stoutamire 4554_
(MSC). Davison Co.: Mitchell, Small, 555. (F). Fall River

 

Co.: Edgemont, Visher 2552 (F, RM, SDU); Hot Springs,
Thatcher 122 (MIN). Hanson, Co.: Alexandria, 55555,.515.
(FLAS). Harding Co.: Cave Hills, 55555_1525 (MIN); Ashcroft,
Visher 224 (RM); Little Missouri River, Visher 244 (F, SDU).
Kingsbury Co.: Spirit Lake, Thornber, 5,5, (ARIZ). Lawrence
Co.: Deadwood,.5555 55 (F, MIN, MO, NY, SDU, US, WIS). Meade
Co.: Ft. Meade, Forwood 55 (US); Piedmont, 25511 5115 (F,
MIN, RM, WIS). Minnehaha Co.: Sand Hills, Duncan, 5,5, (MIN);
Palisades State Park, Messerli 152_(SDU). Pennington Co.:

320
Rockerville, 55155,'515. (MO); Wall, Palmer 51254 (MO, NY,
US); Quinn Table, Lindstrom 512 (SDU); Dark Canyon, 7 miles
west of Rapid City, McIntosh 515 (SDU); hillsides near Rapid
City, 555 515 (RM, SDU). Shannon Co.: White River Valley,
Visher 2154 (NY). Spink Co.: Mellette, Brenckle 41:25
(FLAS, MO, NY); James River woods, 4 miles east of Mellette,
Brenckle, 5,5, (SDU). Stanley Co.: Cedar Pass, 5y5£ 5551
(SDU). 15555. Armstrong Co.: Gamble's Ranch, Palmer 15225
(GH, MO). Brewster Co.: Alpine, Warnock 5452 (SMU); Pine
Peak, Davis Mts., Hinckley 1151_(ARIZ, F). Briscoe Co.:
Lubeck, Demaree 7629 (M0). Culberson Co.: McKittrick Canyon,
Hinckley 4454 (SRSC). Floyd Co.: between Floydada and Emma,
Coville 1512 (US). Jeff Davis Co.: Musquiz Canyon, 13 miles
north of Alpine, Warnock 5452_(SRSC, TEX); 10 miles SE of
Fort Davis, 55£y 55555 (SMU, US); Elbow Canyon near Pine
Peak, Hinckley1151 (ARIZ, GH, NY, TEX); Little Aguja Canyon,
Palmer 54521 (A, MO, NY). 5555. Cache Co.: Logan, Cotten,
5,5, (MIN). Daggett Co.: Grouse Creek Canyon, Graham 5115
(GH, MO). Davis Co.: Farmington, Clemens, 515. (A, GH).
Juab Co.: without definite locality, Maguire 5 Holmgren,
5,5, (GH, NY). Salt Lake Co.: City Creek Canyon, Salt Lake
City, Palmer 55552 (M); Salt Lake City, Leonard, 5,5, (GH,
KSC, ND). Sevier Co.: Glenwood,‘fl555 212 (US). Summit Co.:
Wahsatch, Watson 215 (US). Uintah Co.: Ashley Creek near
mouth of Dry Fork, Graham 5155_(MO); Green River, Island Park,
Graham 2151 (US); Split Mt. Gorge, Dinosaur National Monument,
55155 551 (BRY). Washington Co.: Springdale, 19555 5_242

321
(MIN, MSC, MO, US); Zion National Park, springline on Cathe-
dral Mt., Gillis 5552 (MSC, US). Vermont. Addison Co.:
Leicester, Dutton, 5,5, (FLAS); Middlebury, Brainerd, 515.
(GH). Caldeonia Co.: Barnet, Blanchard, 515. (MIN, MO, US).
Chittenden Co.: Charlotte, Hosford, 5,5, (F). Lamoille
Co.: Cambridge Junction, Donahue 272 (MSC). Orange Co.:
Bradford, Gillis 4522 (MSC). Orleans Co.: Willoughby Lake,
Churchill, 515. (GH, MIN, MO). Rutland Co.: Proctor, Banker

 

247 (NY). Windham Co.: Westminster, Gillis 4081 (MSC).
Locality not certain: Mt. Pisgah, 5. 5,'5.,‘5,5. (US);

Middleton Springs, Carpenter, 5,5, (MIN). Virginia. Greene

 

Co.: summit of Rocky Mount near Skyline Drive, Ewan 17205
(TEX). Page Co.: Luray, Steele 5 Steele 151 (E, GH, MIN,

MO, NY, US). Wgshiggton. Chelan Co.: Chelan, Muenscher 5

 

Muenscher 11224 (A); Wenatchee, Whited 241 (US). Grant Co.:
Dry Falls, Coulee City, Thompson 2115 (GH, MO, ND, NY); Grand
Coulee, Soap Lake, Eyerdam 525 (A, F, MO). Klickitat Co.:
near John Day Dam on bluffs overlooking Columbia River, Gillis
5245 (MSC). Okanogan Co.: Epsom Lake, Kruger Mt. Oroville,
St. John, Courtney, 5 Parker 5415 (WTU). Spokane Co.: Hang-
man Creek,:Sudeorf 254 (GH); Spokane, Kraeger 555 (GH, MIN,
NY, US). Stevens Co.: Northport, on Columbia River, Rogers
545 (MO, NY, SMU). Walla Walla Co.: Waitsburg, Horner 122
(GH, US). Whitman Co.: Revere, Eastwood 15211 (A); Wawawai,
51555 115 (A, ILL, MIN, MO, US, NY); Yakima Co.: N. Yakima,

Mosier, 5,5. (US). West Virginig. Mineral Co.: Allegheny

 

Front Mt., Gillis 5717 (MSC). Upshur Co.: Buckhannon,

 

 

322

Pollock, 5,5. (MO). Wisconsin. Barron Co.: Poskin Lake,

 

 

Cheney, 515. (WIS); Barron, Goessl 1555 (MIL). Bayfield Co.:
Tomahawk Lake, Fassett 9814 (WIS); Port Wing, Cheney 1155
(WIS); south of Cornuc0pia, Fassett 5 Knowlton 25514 (WIS).
Brown Co.: Green Bay, Brannery Hill Creek, Kruschke, 5,5,
(MIN). Buffalo Co.: near Fountain City, 55115_Z1fl1 (MIL,
WIS). Burnett Co.: Viola Lake, Fassett 2521 (WIS). Clark
Co.: Neillsville, Goessl 1425 (MIL). Columbia Co.: near
Prairie du Sac Bridge, 55155 5545 (A, MIL, WIS); Poynette,
Russell, 5,5. (WIS); southeast of Cambria, Waterworth, 5,5,
(WIS). Dane Co.: Belleville, Fassett 25112 (WIS); Belle
Isle, Madison, Fassett 22152 (WIS); Beeche's Woods between
Sauk City and Mazomanie, Fassett 21542 (MIN, WIS). Door Co.:
Washington Island, Fuller 1422 (MIL); Sturgeon Bay, Goessl
5111 (MIL); near Ellison Bay, Palmer 25112 (A); Garrett Bay,
Fassett 15255 (GH, Mo, WIS); Newport, 59551;, 5.5. (WIS).
Douglas Co.: Upper St. Clair Lake, Museum Expedition, 5,5.
(MIL); Superior, Cheney 1552 (WIS). Dunn Co.: Menomonie,
Goessl 2225_(MIL); near Fall City, Fassett, Chavannes, 5
Thompson 25551 (WIS). Eau Claire Co.: EauClaire, Goessl
1115 (MIL); Florence Co.: bank of Menominee River near
Menominee, Michigan, Gillis 5551 (MSC); Fond du Lac, Moon
Lake, 55155 5 Fuller 1515 (MIL). Grant Co.: Bascobel,
Sylvester, 5,5. (MIL). Iowa Co.: Arena, Bullamore, 5,5,
(WIS); Barneveld, Kieckhefner, Lindfors, 5 Thompson, 515.
(WIS). Jackson Co.: Black River Falls, 55115 5524 (MIL,
WIS); Hatfield, Grether 5524 (WIS); Proy, Grether 5151_(WIS).

323
LaCrosse Co.: Mormon Coulee, Harmon 551_(WIS); near Coon
Valley, M5555,'5\5. (WIS). Lafayette Co.: near Belmont,
§5515 1151_(A, MIL, WIS). Longlade Co.: Hollister, Seypour
15115 (WIS); Lincoln Co.: Pine River Twp., Sey5our 14555
(WIS); Merrill, Seymour 5_Schlising, 5,5, (SMU, WIS). Mara-
thon Co.: Wassau, Goessl 2545 (MIL). Marinette Co.: east
of Amberg, 55555 225 (WIS). Milwaukee Co.: Milwaukee,
Kruschke K-4l-24 (MIL, ND); Wauwatosa, Finger, 515. (MIL).
Oconto Co.: Pensaukee, Goessl 4511 (MIL). Oneida Co.:
Rainbow Rapids, Cheney 1551 (MIL, WIS). Ozaukee Co.: Sauk-
ville TWp,, Kruschke K-60-4O (MIL). Pierce Co.: Maidens

Rock, collector unknown (MIN); Prescott, Museum Expedition,

 

5,5, (MIL). Price Co.: Fifield, Graenicher, 515. (MIL).
Racine Co.: Horlicksville, Wadmond 455_(MIN); Burlington,
collector unknown, (MIL). Richland Co.: Richland, Fassett
22425 (GH, MO, WIS). Sauk Co.: Ableman, Palmer 21551 (A,
MO); Baraboo, collector unknown (WIS). Shawano Co.: Shawano,
Goessl 4155 (MIL). Sheboygan Co.: Sheboygan, Goessl, 5,5.
(WIS); Elkhart Lake, Schuette, 5,5. (F); Oostburg, Grittinger,
_,5, (WIS). Vernon Co.: Genoa, Museum Expedition, 5,5.
(MIL); Viroqua, Hansen 145 (TEX, WIS). Washburn Co.: Noble-
ton, Goessl 1415 (MIL); Minong, near Bass Lake, Fassett 5525
(A, GH, WIS, US); Waushara Co.: Plainfield, Fassett 21551
(MO, WIS); W. Bloomfield, 5515 5 Christensen, 5,5, (WIS).

Wood Co.: Remington, Fassett, Catenhusen, Grange, 5 Oehmeke
25521 (WIS). Wyo5ing. Albany Co.: Pole Creek Canyon,

Nelson 154 (GH,MIN, ND-G). Carbon Co.: Freezeout Hills,

324
Nelson 4551 (M0, NY, RM); Seminoe Dam, N. Platte River,
Porter 5 Porter 5525 (RM). Crook Co.: Hulett, Ownbey 555
(RM). Fremont Co.: 10 miles west of Lander, middle fork
of POpO Agrie Creek, Porter 5552 (RM). Hot Springs Co.:
Wind River Canyon, south of ThermOpolis, 55555_25545 (MSC).
Laramie Co.: Pole Creek near Table Mt,, Nelson 154 (RM).
Natrona Co.: CaSper Mt. are, Garden Creek Falls, Joswik 551
(m). Platte Co.: Hartville, Nelson 551 (GH, ND, RM, us);
15 miles west of Wheatland, Palmer Canyon, Porter 4415 (RM,
SMU, TEX). Sheridan Co.: Big Horn, Tweedy 2252 (NY); Upper
Goose Creek Ditch, Vie Willits 155 (RM); Sheridan, Rollins
.555 (GH, MO, NY). Uinta Co.: Evanston, Pannel 5 Blackwood
5555 (M0). Washakie Co.: north end of Tensleep Canyon,
Big Horn Range, Porter 5_Porter 5241 (RM). Locality uncertain:
Big Woods River Mts., Johnson, 5,5, (NY).
* * * * *
CANADA. Alberta. Fort Saskatchewan, Turner, 5,5,
(ALTA). Medicine Hat, 5555 11255_(MSC). Rosedale, Moodie 1215
(F, NY, US). Linnet Lake, Breitung 11112 (ALTA, F, NY).

5gitiah Columbia. International boundary between Kettle and

 

Columbia Rivers, Macoun 55155_(ND-G), between Bridesville and
Grand Forks, Macoun 55155 (NY). Fairmont Hot Springs and
Radium Hot Springs, Sansom 1551 (NY). HOpe, Newcombe 545
(F). Lytton, Kra'ina, 5,51 (SMU). Spence's Bridge, Macoun,
5,5, (NY). Vancouver, Davidson 12 (SMU). Fort Vancouver,

Hooker, 5,5. (K). Manitoba. Grand Rapids, Scoggan 4751 (MIN).

 

Arveme, Griddle, 5,5, (A). Fort Garry, Winnipeg, Boivin,

325
15y5, 5 55y5_1255_(MSC). Lorette, Chainaie, St. Boniface,
Boivin 5454 (MSC). Dulta, 55155,‘515. (A). Riverton,
Krivda 5:54 (NY). Sundown, collector unknown (WIN). Brandon
woods, collector unknown (WIN). Einnepegosis, collector

unknown (WIN). Brokenhead, collector unknown (WIN). Big

 

Grass Marsh, southern Lake Winnipeg, collector unknown (WIN).

New Brunswick. Carleton Co.: Woodstock, Dore 5 Gorham

 

 

45894 (US). Victoria Co.: Aroostook River, Williams, 5,5,

(GH). Westmoreland Co.: Cap Brulé, Knowlton, 5,5, (GH).

 

Nova Scotia. Halifax Co.: Shubenacadie Grand Lake, Fernald

 

 

5 Bissell 21787 (GH). Hants Co.: Five Mile River, Pease 5

 

1555 21152 (GH). Lunenburg Co.: Bridgewater, Fernald 5

555g 24525 (A, GH). Pictou Co.: Ferrona Junction, Robinson
.522 (NY). Victoria Co.: Port Bevis, Fernald 5 555g 21122
(GH). Ontario. Algoma Co.: Batchawana Falls, Taylor 55.51.
214 (GH). Bruce Co.: Sydney Bay, DeJong 1555 (MSC); Stokes
Bay, Krotkov 2125 (M0, NY). Carleton Co.: Ottawa, Marshall,
5,5, (MSC). Frontenac Co.: Westbrook, Gillis 5515 (MSC).
Lambton, Pinery Provincial Park near Grand Bend, Gillis 5552
(MSC); Pt. Edward, Cooper, 5,5, (MIN). Lenox Co.: Newburgh,
.55555,‘5¢5. (F). Manitoulin Co.: Manitoulin Island, Grassl
5145 (NY). Renfrew Co.: Cobden, Gillis 5442 (MSC), Stormont
Co.: Longue Sault Rapids, 5555_5 555y 15551 (MSC). Parry
Co.: Island in French River, 5555y'52 (US). Thunder Bay Co.:
Sibley Peninsula, Boivin 2514 (MIN). Wentworth Co.: % mile
from Lincoln Co. line, 2 miles from Winoma, Gillis 5511 (MSC).

Timiskaming Co.: Dawson Point, Lake Timiskaming, Morton 5

326
LePage 11687 (US). Locality uncertain: Chinook Island,
McDonald 165 (US). Prince Edward Island: Malpeque, Fowler

 

 

ggig4_(w18). Quebec. Argenteuil Co.: Chatham Twp. Cushing,
Jenkins gggg (MSC). Bellechasse Co.: Beaumont, Chabot g
Verret A-2204 (QUE). Berthier Co.: Berthier en Bas, Rousseau
22911 (GH). Chambly Co.: Longueuil, Marie-Victorin, gag.

(ND). Charlevoix Co.: Baie-des-Rochers, Cayouette g Brassard

5847 (QUE); De-la-Rive, Boivin 1123 (M0); Port-aux-Quilles

 

(part of town of Cap-aux-corbeaux), Cayouette 4 Brassard 5846
(QUE); St. Simeon, Cayouette g Brassard 5525 (QUE). Chicou-
timi Co.: Anse St. Jean, Brassard, 4,3. (QUE); Canton Labrosse,

Cayouette g Brisson, §,g. (QUE); Petit Saguenay, Cayouette g

 

Brassard 5511 (QUE); Sainte Rose-du-Nord, Brassard 60-ll9A
(QUE); Deux Montagnes Co.: Oka-sur-le-lac, Cartier, §.g.
(QUE); La Trappe, Piniere Sablonneuse du lac, Louis-Marie,
§.Q. (OSC). Gaspé Co.: Riviere Petit-Pabos, Marie-Victorin,

Jacgues, g Rolland-Germain 44919 (GH). Gatineau Co.: Grand
Remous, Louis-Marie g Lamarre 294 (GR). Hefford Co.: Granby,

 

Fabius §1§ (NY). Hull Co.: Hull, Macoun 421_(GH). Iber-
ville Co.: Iberville, Cinq-Mars 4 Samoisette, §,Q, (QUE).

Lac Saint-Jean Co.: Saint Gédéon, Cayouette g Brisson Q442
(QUE); Val-Jalbert, Cayouette 58-254 (QUE). Laval Co.: Saint
Leonard, Ricard g Boivin 1929 (GH, MO); Ile aux Praises, near
Montreal, Rouleau gg42_(TEX); Mount Royal along railroad
tracks, Gillis 8444 (MSC). Matapedia Co.: Matapedia River,
Rousseau 82411 (NY); Routhierville, Cayouette §_Cayouette
gggg (QUE). Montgomery Co.: Isle of Orleans, 443k, §,Q, (A).

327
Montmorency Co.: Montmorency, Marie—Victorin, Rolland-
Germain, Raymond, 8 Champagne 55882 (MSC). Napierville Co.:
Napierville, 1555,8885 (MSC). Quebec Co.: Cap-Rouge, Penault

 

4585 (MSC); Plaines d'Abraham, Cinq-Mars 8 Barabé, 5.5. (QUE).
Roberval Co.: Pointe-Bleue, Lalonde 8 Cayouette 5581 (QUE).
Saguenay Co.: Albert, Cayouette 845 (QUE); Cap 5 1a Boule,
Brisson, Brassard, 8 Dramville 60-370 (QUE); Petites-
Bergeronnes, Cayouette 8_Brassard 5858 (QUE); Ruisseau Gagnon,
Brassard 8 Brissom 60-331 (QUE). St. Jean Co.: St. Jean,

CinqrMars 8 Samoisette, 5.5. (QUE). Saskatchewan. South

 

 

ridge of Qu'Appelle River, Shumovich, Selleck, 8 Gelleta 51
(RM). Assiniboia, Boivin 8 Breitung 5552 (MSC). Nipawin,
Breitung 1488 (MSC). Moon Lake, near Saskatoon, Boivin,
Russel, 8_Breitung 5152_(MSC).

Toxicodendron rydbergii is the most northerly ranging
species in the Anacardiaceae. It is a taxon which is easily
separated from related taxa by a number of constant characters,
yet forms morphological intermediates on the fringes of its
range where it apparently intergrades with neighboring taxa.
At least eight new Species have been described by writers
who apparently did not know its range of variation; at least
three other taxa have been described which are probably inter-
mediates between this Species and other taxa, and represent
pOpulations from the edges of its range.

Prior to 1900 any specimens of this taxon were called
by the names of other taxa--their collectors not being aware

of any distinctions between them. It is possible that

328
Linnaeus had such differences in mind in the Species Plan-
tarum since he described three forms of poison-ivy, but his
references are to the older literature and refer to names
which cannot be typified.

The species is distinct from all other poison-ivy
taxa in not producing aerial roots in the wild, thus always
remaining a sub-shrub or, at best, a shrub of dimensions not
more than several meters high. Its fruits tend to be larger
and lighter in color than those of other taxa, its petioles
always glabrous and usually quite long, and its leaflets
very broad, often suborbicular. Because of the size and
shape of the leaflets, they are difficult to press intact,
and therefore will often be folded along the midrib in her-
barium specimens. In the field, this character is expressed
by a spoon shape to the leaflets. The leaves are usually
clustered near the t0p of the aerial stem, borne on petioles
that are sometimes short, but more often than not, are more
than one and a half times as long as the blade. Because the
petiole length seems to be somewhat related to the degree of
sunlight received by the plant, the length of petiole alone
is not a definitive character; its glabrous nature is. The
short pedicels bearing fruits larger than other members of
the complex make the fruiting inflorescence much more compact
and tightly clustered than those of other taxa, hence the
ability of the infructescence to stand erect.

In describing Toxicodendron hesperium, Greene distin-

guished the taxon by its wrinkled fruits, apparently unaware

329
that unripe fruits of all members of this group become wrin-
kled upon drying. In regard to I. Qumilum Greene, a differ-
ent situation exists. Toxicodendron pumilum Kuntze (1891)
is meant to refer to 8555 gumila Michx. (= 8. michauxii
Sargent) which is a homonym for 8. gumila Meerb. Toxicodendron
Qumilum Greene therefore is also a homonym, having been pub-
lished 14 years after Kuntze's name. Greene felt that this
species was characterized by nodding inflorescences, but con-
fused the issue by his statement describing the type: "In
higher mountains of northern Arizona, . . . labeled R. 5;-
versiloba, though the plant, despite its inclining panicles,
is strictly of the Atlantic type of the genus." The type is
a collection from early in the growing season with new leaves
barely Opened and fruits dating from the preceding season.

In regard to Toxicodendron punctatum Greene, isotypes
at NY and at US were annotated by Greene himself (after
publication of this species) as 8555 rydbergii! Greene's
description mentioned "tufts of hairs in and near axils of
midveins" as a definitive character, but this trait does not
characterize the type.

J. Lunell collected two forms in North Dakota, naming
them Toxicodendron fothergilloides from the same locality
which, even though given the same collection number, were
collected two years apart. Barkley (1937) chose "type no. 1"
as the type inasmuch as the Code provides only for a single
specimen to serve as the nomenclatural type. It is a flower-

ing Specimen with a separate Sprig of fruits attached to the

sheet.

330

The range of this species extends throughout western
United States roughly from the 100th meridian to the Great
Basin and the Cascade Mountains. It is also the sole rep-
resentative of the complex in North America found north of
the 44th parallel of latitude. It iS the taxon which inhab-
its the dunes of the Great Lakes and so follows the shorelines
of Lakes Michigan, Ontario, and Erie which extend south of
the 44th parallel. A disjunct populartion is found on moun-
tain crests in the central Appalachians from Pennsylvania,
West Virginia, and Virginia. Except for the southern lobe
of Ontario which bounds Lakes Huron, Erie, and Ontario, and
for Nova Scotia, Toxicodendron rydbergii is the sole taxon
of poison-ivy found in Canada from the Gaspé Peninsula to the
east flank of the Cascade Mountains.

In garden plots at Michigan State University, where
poison-ivy populations were grown experimentally, plants of
Toxicodendron rydbergii maintained their distinctiveness
especially in regard to the non-climbing habit. A specimen
from the Davis Mountains of southwestern Texas, however, demon-
strated a different behavior in a growth chamber. After sev-
eral months at high relative humidity, aerial roots were pro-
duced. It appears that a response to moisture may induce
aerial root formation even in this taxon, as a reversion to
the stock from which it was derived. In the wild, it is this
ability to produce roots which presumably permits the stolons
to root upon stimulus by soil moisture, thus expanding a

clone Spatially.

331

Because of the apparent range limits at the 44th
parallel in eastern United States, a study area was sought
along a north-south river valley which would likely be
hOSpitable for the climbing forms to the south (Toxicoden-
5355 radicans subspp. radicans and negundo), yet a river
valley which would cross the 44th parallel. The Connecticut
River Valley, both in New Hampshire and Vermont, was explored.
As I searched southward through New Hampshire, the first
place where climbing poison-ivy was encountered was on a
stream bank at the West Thornton-Woodstock city line in
Grafton County at 440 N. This collection appears to be an
intergrade between I. rydbergii and I. radicans subsp. £55;-

cans.

 

There is a cognate form of poison-ivy in Japan, also
restricted north of the 44th parallel, and reported to be a
sub-shrub. This taxon, named 8555 rishiriensis Nakai, is
discussed under Toxicodendron radicans subsp. orientale, and
treated herein as an ecotype without being recognized nomen-
claturally.

Plants of Toxicodendron rydbergii grow on sandy or
rocky lakeshores, floodplains, river terraces, talus, prec-
ipices, railroad rights of way, and other ruderal sties, sand
dunes, and cutover woods. It is frequently found with 235557
5155 aquilinum whereas most other poison-ivy taxa do not occur
with bracken fern. It is seldom found in deep woods or old
forests, having been shaded out, if it ever did occur there.

When it is found on sand dunes, it is missing from the

332
adjacent dune forest, being replaced, within their geographic
ranges, by I. Radicans subsp. radicans and negundo. In Nova
Scotia and the Gaspé region it inhabits gypsum cliffs, Slaty
ledges, and the cobbly upper beach. It produces stolons that
intertwine the chinks and crevices in stone walls of northern
New England farms. Although often in wet places like rills,
ravines, and near waterfalls, it is reported from sub-alpine
rocks on Pike's Peak (Colorado). It is also reported as
forming an "association" with Shepherdia in Dinosaur National
Monument, Utah. In growth form, it is usually a sub-shrub,
generally achieving no more stature than a meter. When it
is undisturbed for several decades and the Site conditions
are favorable, it may reach two or more meters in height
and be a genuine shrub. I have collected it under such con-
ditions, from moist washes along the Columbia River in Wash-
ington and Oregon, and at the outlet of Montezuma Well in
Yavapai County, Arizona.

Habitats for this taxon are many and varied. Because
the taxon reproduces by rhizomes, it is difficult to assess
the true abundance and ease of establishment of this species.
The many stems appearing on a single Site may all be of a
single individual, or of several. It is not usually abun-
dant, but may find its niche in the more mesic and better
illuminated portions of the hemlock-white-pine-northern hard-
woods association, the Spruce-fir forest formation, the maple-
basswood formation, or in a disturbed woods of pine, cherry,

and red maple. In southern Ontario, Toxicodendron rydbergii

333

inhabits woods of elm, oak, and ironwood (Ostrya). It inhab-
its some of the more mesic habitats in the prairies and moist
valleys in the Rocky Mountains. The ruderal nature of this
plant permits invasion along roads, railroads, lakeshores,
floodplains, fencerows, etc., where there are many weedy
associates. It is probably safe to say that the plant's eco-
logical amplitude is such that it will grow in situations too
varied to characterize. Rather, the need for moisture and
sunlight likely determines in large measure where it occurs.

6. Toxicodendron toxicarium (Salisb) Gillis, comb. nov.

(Plate XXXVI.)

Rhus toxicodendron L., Sp. Pl. 1: 266. 1753. 8555
toxicarium Salisb. Prodr. 170. 1796. Toxico-
dendron magnum Bert. (Sic) ex Steudel, Nom. Bot.
2: 694. 1841. (nomen nudum). '8555_toxicodendra
St. Lager, Ann. Soc. Bot. Lyon 7: 133. 1880.
Toxicodendron toxicodendron (L.) Britton in
Britton and Brown, Ill. Fl. U.S. 2: 484. 1913.
Type: "Habitat in Virginia, Canada." (LINN
378.16, lectotype, specimen onleft).

8555 toxicodendron var. guercifolium Michx., Fl. Bor.-
Am., 183. 1803. 8555 quercifolia (Michx.) Steudel
(in synon.), Nom. Bot. l: 689. 1821, non 8. 5555-
cifolia Goeppert, Die Tertiére von Schossnitz in
Schlesien, 37. 1855. Toxicodendron quercifolium
(Michx.) Greene, Leafl. Bot. Observ. Crit. l: 127.

1905. Type: "Virginie et Carolines." (P,

334

PLATE XXXVI. Specimen of Toxicodendron toxicarium.
Holotype of I. monticola Greene.

 

555

     
   

   

\n... “tar-l \ux W(::)
.3 if't",‘9)q‘ll

(ahvxnxfizu‘u
If; 12233211};

 

UNHLD SVAHS NAV’ONAL MHkAmuM
Au.» 1
. , . ...
K! > :1 >' \ - v y
y_va. ‘MI.H.AL
umu‘: v x-.__;l.£.l:li

 

- um. um.

4......-
m ...
mmun

HDRA UF [Hf SUUIHEHN APMABHWI REGION.

4 num- A

A
w .n..) A
lasagna-'1; m4; mm.) Groom

/ w
1' »>V’/1/
i! 5 L A V «'1
lTYPE 29‘.‘;,ge..4,'.g a .3951; .n.:n.
LA‘... I.) . r 1.1; '..c. 12"..
a] . .">""-,l 2,5,51

 

 

 

 

 

branc

336
lectotype).

Rhus acutiloba Turcz., Bull. Soc. Imp. Naturalistes
Moscou 36: 612. 1863. Type: Berlandier 8885,
Mexico (actually Texas), (KW, lectotype, G-DC,
GH, MO-1773755, isotypes).

Toxicodendron monticola Greene, Leafl. Bot. Observ.
Crit. 1: 126. 1905. Type: Albert BEED.§§Q9
Georgia, Lookout Mt., July 1898, (US-345540,
holotype; ND-G, MO-l773749, NY, presumed iso-
types).

Toxicodendron compactum Greene, Leafl. Bot. Observ.
Crit. 1: 126. 1905. Type: William Hunter, 5.5.,
Virginia, Carroll Co., Woodlawn, August 1899,
(US-364942, holotype; GH, isotype).

8585 toxicodendron f. elobata Fernald, Rhodora 43:
598. 1941. Type: 8, 8, MEE.E£££:.§mfl-: Bridge-
ton, Cumberland Co., New Jersey, 1% miles s.e.
by N.J. Central R.R., 20 July 1909. (GH, holo-
type).

Rhus toxicodendron f. leiocarpa Fernald, Rhodora 43:
599. 1941. Type: ‘M. 5, Fernald 8 Bayard 5555
8841, dry open sandy pine and oak thickets near
the county line, south of Jarratt, Sussex Co.,
Virginia, June 8, 1938, (GH, holotype, NY, US,
isotypes).

small shrub or subshrub with Slender, pubescent

branches and branching subterranean stolons. Leaflets often

337
subcoriaceous, 3 (-5), ovate to oblong or oblong-obovate,
lobate-dentate or lyrate, sinuate-pinnatifid with 3 - 7
rounded, blunt or rarely subacute lobes, occasionally just
undulate; often more deeply lobed in male than in female
clones; apex rounded, subacute or acuminate, base obtuse or
cuneate. Petiole hiSpid or villous, 1.4 - 12.5 cm long,
terminal leaflet blade 0.7 - 3.2 cm long. Leaves pilose,
strigose, hirsute, or velutinous above, strigose, velutinous,
or woolly below, often drying brown. Lateral leaflets in-
equilateral, subsessile or with petiolules up to 2.5 cm long;
leaflets attached in palmate fashion when trifoliolate, pin-
nate when more than 3 leaflets. Inflorescence a lateral
paniculate thyrse, up to 1 dm long; bracts deltoid to lanceo-
late, O.6 - 10 mm long, 0.3 - 3 mm broad at the widest point,
glabrate, ciliate, deciduous; sepals 5, deltoid-ovate, 1 - 2
mm long, 0.5 - 1.5 mm broad, glabrate; petals 5, cream to
yellow-green, dark-veined, Oblanceolate, glabrous, 2 - 4 mm
long, 2 - 3 mm broad, reflexed in male, recurved in femals,
larger in male flowers than in female. Anthers lanceolate,
l - 1.5 mm long, 0.5 - 1.1 mm broad, filaments linear-subu-
late, l.O - 1.8 mm long. Drupe hard, yellow-brown to tan,
globose-reniform, sometimes depressed-blobose, pubescent
(rarely glabrate), exocarp becoming papery upon ripening
and readily separating from the mesocarp; mesocarp striate
in white waxy matrix; endocarp dun-colored, bony. Bark gray-
brown to red-brown on new wood, becoming gray on older wood;

leaf scars V- or U-shaped with several bundle scars;

338

lenticels usually conSpicuous; young branches villous. Buds
stalked, naked, mucronate, brown, covered with a brown woolly
pubescence. Autumn coloration red-bronze or brown. Distribu-
tion: southeastern United States, New Jersey to Florida,
west to Texas, north to southeastern Kansas. Common name:
Eastern poison-oak. (Fig. 15.)

Representative Specimens: United States. Alabama.
Autauga Co.: Booth, Gillis 4588 (MSC). Baldwin Co.: Loxley
Research Plots, Stapleton, Grelen 5159-46 (FLAS). Blount Co.:
Eggert, 5.5. (MIN, MO, NY). Choctaw Co.: Bladonsprings, M555,
5.5. (US). Clark Co.: Jackson, Gillis 5881.(MSC). Cleburne
Co.: north of Edwardsville, Donahue 848 (MSC); waterfall,
Che-a-ha Mt., Mohr, s.n. (ALU). Cullman Co.: St. Bernard,

 

Wolf, 5,5. (MO). Dekalb Co.: Ft. Payne, Sherman 298 (SMU).

 

Elmore Co., Wetumpka, Gillis 4584 (MSC). Escambia Co.,
Atmore, 45155.1818 (ILL),‘ Hale Co.: Payne Lake Recreation
Area, Talledega National Forest, Donahue 845 _MSC). Lee Co.:
University, Gillis 4584 (MSC). Macon Co.: Tuskegee, Gillis
4588 (MSC). Marion Co., north of Hamilton, Gillis 5848_(MSC).
Tuscaloosa Co.: 3 miles south of Fosters, Gillis 5888 (MSC);
Tuscaloosa, M553‘8815 (ALU). Washington Co.: 1 mile east

of state line, Gillis 5818 (MSC). Arkansas. Clark Co.:
Arkadelphia,Demaree 11884_(A, MO, SMY); Graysonia, Demaree
81848 (A, ND-N, NY, SMU). Cleburne Co., Heber Springs, rim
of Round Mt., Demaree 55814 (MSC). Conway Co., Petit Jean
Mt., Morrilton, M5555 488 (SMU). Drew Co.: Ladelle, Demaree

339

Fig. 15. Distribution of Toxicodendron toxicarium.

 

 

FMS x7”
\ x, 2

 

    
 

 

 

341
88881 (A, M0); Monticello, Demaree 85488 (SMU). Faulkner Co.:
Cove Creek, Demaree 18 (M0); Greenbrier, Gillis.8111 (MSC).
Franklin Co.: north slope of Altus Hill, Donahue 225 (MSC).
Garland Co.: Hot Springs, Demaree 11855 (M0, NY, SMU); Hot
Springs, Palmer 84588 (A). Grand Prairie Co.: Hazen,
Wheeler 18 (F). Grant Co.: Poyen, Demaree 11858_(SMU).
Hempstead Co.: Fulton, §E§2.§§ZZ (A, F, PH, M0). Hot Springs
Co.: Magnet Cove, Cook Mt., Demaree 18581 (MO, NY, SMU);
Bismarck, Demaree 17405 (MO, WIS); Butterfield, Demaree 14518

 

(M0, NY); Malvern, Palmer 88851 (A). Lawrence Co.: Straw-
berry, Demaree 85818 (MU, TEX). Logan Co.: West Mt., Palmer
84588 (MO); east Side of Magazine Mt., Donahue 881 (MSC).
Lonoke Co.: Carlisle, Demaree 11518 (A, GH, MIN, M0, NY,
SMU). Miller Co.: Texarkana, Demaree 84415 (MO, NY, SMU).
Ouachita Co.: Camden, Demaree 15188 (A, M0, NY). Prairier
Co.: Hazen, Palmer 85858 (A, ML). Pulaski Co.: Little
Rock, Palmer 22943 (A). Saline Co.: Bauxite, M5515 480198
(SMU). Scott Co.: 3 miles east of Y City, Donahue 888_(MSC).
Sebastian Co.: Mansfield, Demaree 18148 (SMU); Arkansas
National Forest, Palmer 88815 (ND). Seiver Co., DeQueen,
Fassett 8.85515 18181 (GH, NY, WIS); Mineral, Brinkley 818_
(F, TEX). Yell Co.: Dardanelle, Demaree 88158, 88855 (MIN,

M0, NY, SMU). Delaware. Sussex Co.: Laurel, Commons, 5.5.

 

 

(GH, MO, NY). District of Columbia. Fort Totten, Holm, 5,5,

 

 

(M0). Woods near Sligo Creek, Moffatt, 5,5, (ILL). Terra
Cotta, Holm, 5,5, (ILL, LCU, M0). Woodley Park, Steele, 5,5,

 

(A, F). Without locality, Hitchcock 12962 (US). Florida.

342
Alachua Co.: Gainesville, Gillis 4815 (FLAS, MSC); oak-
pine-hickory forest, University of Florida campus, Wiggins 8
Wiggins 19855 (FLAS); Wacahoota West, 5,5, (FLAS). Clay Co.:

 

Goldhead Branch State Park, West, 5,5, (FLAS). Dixie Co.;

 

5.5 miles north of Oldtown, Pasture Survey, 585. (FLAS).
Gadsden Co.: Aspalaga, collector unknown (M0). Gilchrist
Co.: Fort White, Cooley 8 81555_8188 (SMU); Wilcox, Martin,
DeVall, 8 Arnold, 5,5, (FLAS). Jackson Co.: west side of
Apalachicola River, north of Victory Bridge, Exploration Eggiy
1881 (FLAS); Marianna, Godfrey 58111_(MSC, SMU). Jefferson
Co.: Wacissa and Capps, Godfrey 58548.(FSU, MSC, SMU).
Lafayette Co.: north of Mayo, 8555‘8 Arnold, 555. (FLAS).

Leon Co.: Tallahassee, Godfrey 53105 (FSU, GH, NY, SMU, VDB);

Hammock, Small, DeWinkeler, 8.Mosier, 5,5. (NY). Levy Co.:

 

locality unknown, Hitchcock 885 (F). Liberty Co.: pineland
Bristol, Godfrey 8 Henderson 58815 (FLAS, FSU, MSC); Marion
Co.: 2 miles south of Silver Springs Gillis 4481, 5851 (MSC).
Okaloosa Co.: Laurel Hill, Godfrey 51815 (FSU, MSC). Suwanee
Co.: locality unknown, Hitchcock 888 (F, MO); pineland 8
miles west of O'Brien,'8555 8 Arnold, 5.5. (FLAS). Wakulla
Co.: Wakulla, Godfrey 58188 (MSC, NY, SMU). Georgi . Bartow
Co., Cartersville, Rhoades, 5,5, (WIS). Bulloch Co.: local-
ity unknown, Harper 848 (F, GH, MO). Clarendon Co.: Pax-
ville, Godfrey 8 1£y55_1811 (GH). Clarke Co.: Athens,.1§ggg,
_,5, (MSC). Cobb Co.: Kenesaw Mt., Rhoades, 5.5. (WIS).

Dade Co.: Lookout Mt. Ruth 356 (MO, ND-G, US). Decatur Co.:

 

343
Bainbridge, Gillis 4588 (MSC). Dekalb Co.: Stone Mt.,
Miller, Egpgy, M ers,‘8‘85y51_548 (GH); Lithonia,_85££y 8
My5£5_881 (GH). Emanuel Co.: Swainsboro, Gillis 4855 (MSC).
Floyd Co.: Rome above Cossa River, Demaree 8 11pp5_58184,
(MSC). Gwinnett Co.: Thompsons Mill, Allard 146 (US).

 

Habersham Co.: Currahee Mt,, Small 5,5, (MSC, NY). Haralson

 

Co.: Tallapoosa, Donahue 251 (MSC). Jefferson Co.: local-

 

ity unknown, Hopkins, 5,5, (NY). Lowndes Co.: 4 miles
north of Clyattville, Gillis 6583 (MSC). McDuffie Co.: near

 

Sweetwater and Thomson, Bartlett 1702, 1703, 1704 (MICH, MSC).

 

Peach Co.: near Byron, 1.5 miles north of jct. I-75 and
Ga-49, Gillis 6551 (MSC). Randolph Co.: Shellman, Sandlin,
5.5, (BLH). Richmond Co.: Augusta, Sargent, 5,5, (A);

Augusta, Gillis 4958 (MSC). Walton Co.: Loganville, Wiegand
8 Manning 1862 (CH). Ware Co.: Waycross, Gillis 4975 (MSC).

 

Whitfield Co.: Taylor's Ridge, Wilson 155 (E, F, GH).
Kansas. Chautaqua Co.: locality unknown, Hitchcock, 5,5,

(KSC). Louisiana. Bienville Parish: Castor, Moore 6407

 

 

(GH). Jackson Parish: Indian Village, M55£5_8 Marman 5818

(GH). Natchitoches Parish: ChOpin 1885 (A, F, M0). Rapides
Parish: Valentine Lake, Thieret 11888_(MSC). Vernon Parish:
LaCamp, Donahue 841 (MSC); 4 miles SE Ft. Polk,l8£51 8 Ricks
15188 (FLAS). Mgry1apg. Ann Arundel Co.: Severn Run, 8555.
w (FLAS, FSU, MO, NY, US). Baltimore, Forman, 8.8. (NY).
Montgomery Co.: north end of Rock Creek Park, Bartlett 8848
(MICH, MSC). Prince Georges Co.: Clinton, 8518, S.n. (GH,

KY, LCU). Wicomico Co.: Salisbury, Chickering, 5,5. (F, NY).

344

Worcester Co.: Snow Hill, Norton, 5,5, (MO). Mississippi.

 

 

Attala Co.: locality unknown, McDougall 1615 (US). Coving-
ton Co.: Collins, Gillis 6005 (MSC). George Co.: Lucedale,

 

Cooley, Pease, 8 Demaree 3374 (GH). Harrison Co.: DeSoto

 

National Forest, Diener 885 (ILL). Jefferson Davis Co.:
between Lucas and Prentiss, Gillis 5888 (MSC). Jones Co.:
Laurel, 155; 54 (SMU, WIS). Lauderdale Co.: Meridian,
Rhoades, 5,5. (GH, RM). Smith Co.: Marathon Lake, Donahue
848 (MSC). Stone Co.: McHenry, Demaree 88818 (FSU, GH, SMU,
VDB). Wayne Co.: Beat Four School along US-84, Gillis 5811
(MSC). Missouri. Mississippi Co.: Charleston, Palmer 8
Steyermark 41516 (A, MO). Ozark Co.: Tecumseh, Palmer 88885
(A, M0); Gainesville, Bald Jesse, Palmer 48881 (M0, NY);

Palmer 33068 (A, MIN, NY). New Jersgy. Camden Co.:, C00per

 

Creek, 155g 18858 (GH); Somerdale, Meredith, 5,5, (GH).

Cape May Co.: Goshen, Gillis 5581 (MSC); Locust Grove, West
Medford Branch of railroad, 155g,.§1p. (MICH). Cumberland
Co.: Bridgeton, Van Pelt, 515. (GH); Millville, Gillis 5584
(MSC). Gloucester Co.: Woodbury, 155g,,§,5. (GH); Malaga,
shore of Lake Malaga, Gillis 4844 (MSC). North garolina.

 

Bladen Co.: Tackee1,.45155 88151 (GH). Buncombe: Asheville,
Schneck, 5,5, (ILL). Carteret Co.: Longleaf, Gillis 8888
(MSC). Cumberland Co.: FayetteVille, M1155,I§flg. (US);

Cedar Creek, 85155 8 Hammond 84441 (NY). Dare Co.: Manteo
near old fort, Roanoke Island, Churchill, 5,5. (MSC). Durham
Co.: Nelson, Gillis flQ§Q.(MSC)- Franklin Co.: Nash, Ahles
15514 (SMU). Lexington Co.: Gaston, Gillis 4888_(MSC).

345
Moore Co.: Carthage, 858, Godfrey 8 Campana 1881 (FLAS, GH,
SMU). Onslow Co.: Gum Branch, E2122.§ Moreland 548 (GH).
Pender Co.: 4 miles south of Wallace, 5222.1992$.(F5U)-
Polk Co.: Tryon, Millsbaugh 4060 (F). Richmond Co.: local-

 

ity unknown, Wiegand 8 Manning 1861 (GH). Oklahoma. Atoka

 

 

Co.: 12 miles south of Atoka, HOpkins 8 Nelson 1188 (RM).

Cleveland Co.: Norman, Woodward 88 (MO). Creek Co.: Bristow,
Gillis 5188 (MSC). Garvin Co.: Davis, Palmer 51 (NY). Logan
Co.: Guthrie, Carleton, 5.5. (ILL). Love Co.: Hickory Creek
north of Marietta, Gillis 5144 (MSC). Oklahoma Co.: Oklahoma

City, Campbell, 5,5, (MIN). Payne Co.: Stillwater, Waugh

 

124 (F, KSC). Pittsburg Co.: McAlester, Palmer 6402 (A, F,
MIN). Pottawatomie Co.: St. Louis, Faulkner 55_(GH). South
Qarolina. Aiken Co.: Mt. Pleasant, Gillis 4614 (MSC); north

 

of Augusta, Ga., Gillis 4888_(MSC); Graniteville, Eggert, 5,5.
(MIN, MO); Aiken, Harbison 1158 (A, E). Anderson Co.:
Anderson, 23212.1122 (US). Barnwell Co.: Williston, Palmer
88818 (A, MO). Cherokee Co.: Draytonville Mt., ESE of
Gaffney, 45155 85888 (FLAS). Clarendon Co.: Paxville, Godfrey
‘8 Ipypp 1811 (NY), Darlington Co.: Darlington, Rehder 855

(A), Greenville Co.: near Greenville, Small, 5,5, (NY).

 

Horry Co.: locality unknown, Weatherby 8 Griscom 15588 (GH).
Kershaw Co.: Camden, Palmer 48884 (A, M0, NY). Lexington

Co.: Gaston, Gillis 4888 (MSC); Pleasant Hill, Gillis 4511
(MSC). Oconee Co.: Tomassee, 85555_8851 (US). Orangeburg
Co.: Eutawville, Godfrey 8_1£y55_888 (G, GH, MO, NY). Pickens
Co.: Clemson, 855§5_8854 (NY). Richland Co.: Columbia,

346
Bartram, 5.5, (NY); north of Columbia along highway 21,
Gillis 4588 (MSC); Fort Jackson, Gillis 4814 (MSC). Tennes-
55; Grainger Co.: Rutledge, Norton, 5,5, (ND-G). Hamilton
Co.: Chattanooga, Churchill, 5.5. (GH); Lookout Mt., 85555
8 Hesler 1858 (SMU). Knox Co.: Knoxville, Lamson-Scribner,
‘5.5. (US). Sevier Co.: Sevierville, Palmer 48588 (A, M0).
1588:. Bastrop Co.: BastrOp, Lundell 8 Lundell 18884 (F,
NY, SMU); BastrOp State Park, Gillis 3694 (MSC). Bexar Co.:

near Rio Medina, Berlandier 2035 (KW, G-DC, GH, M0). Bowie

 

Co.: 8.8 miles NNE of Sulphur River, Donahue 881 (MSC).
Brazos Co.: 12 miles NE of Bryan, Donahue 884:4;(MSC).
Burleson Co.: Somerville, Shinners 14694 (SMU). Caldwell
Co.: locality unknown, McBryde, 5,5, (TEX). Dallas Co.:
Dallas, Reverchon 154, (A, F, GH, M0). Grayson Co.: Denison,

 

Reverchon, 5,5, (MIN, M0). Hardin Co.: Kountze, Demaree
55415 (MSC). Harris Co.: Hickley, Thuron, 5,5. (US).
Harrison Co.: Marshall, Palmer 5888 (A, MIN). Henderson
Co.: NNE of Frankston, Donahue 888 (MSC). Houston Co.:

WNW of Ratcliff, Davy Crockett National Forest, Donahue 888
(MSC). JaSper Co.: Bouton Lake Recreation Area, Angelina
National Forest, Donahue 848 (MSC). Lamar Co.: Arthur City,
Demaree 18885 (A). Montague Co.: St. Joe, Shinners 13272
(SMU). Montgomery Co.: Conroe, Palmer 88881 (A, M0, NY).
Polk Co.: Livingston, Palmer 5888 (A, F, MIN); NE of Camden,
Thar e, Turner 8 Johnston 54188 (TEX). Rusk Co.: Henderson,
Shinners 14885 (SMU); Tatum, 85£y 55454 (SMU). Smith Co.:
western Tyler, north of Pine Bluff Rd. .853y'55181 (NDA, F,

347
SMU). Tarrant Co.: locality unknown, 5233.229 (F, US).
Upshur Co.: Big Sandy, Shinners 14888 (SMU). Walker Co.:
Huntsville, Palmer 18848 (A, M0). Waller Co.: Hempstead,
.flflll.l§ (F, ND-G, NY, MO). Wilson Co.: Ted Cover Road,
'855y 55485 (SMU). Wood Co.: Mineola, Shinners 14858_(SMU);

county unknown: between Bejar and Trinidad Rivers, collector

 

unknown (G-DC). Virginia. Arlington Co.: Oakcrest, Hotch-

 

kiss 5861 (ND- , NY); Naucks, Steele, 5,5, (A). Augusta Co.:
locality unknown, Caro 466 (GH). Caroline Co.: Milford,
Fernald8 ong 7518 (GH, PH). Carroll Co.: Woodlawn, Hunter,

 

515. (GH, US). Culpepper Co.: Buzzard Mt., Allard 20851
(US). Dinwiddie Co.: Carson, Fernald, Long, 8 Smart 5834
(GH, PH); Petersburg, Ferna1d8 Long 8346 (GH, MO, NY, PH).

Fairfax Co.: Great Falls, Bartlett 1784 (MICH, MSC, SMU).

 

Fauquier Co.: western slope of Bull Run Mt., Allard 18885_
(GH, US). Greensville Co.: Orion, Fernald 8 L229.l§§2§

(GH, PH, US). Nansemond Co.: South Quay, Fernald 8 155g
15:22 (GH, PH, us). page Co.: Kimball, Miller, 5.5. (US).
Prince Edward Co: Farmville, Fosberg 15541_(GH). Princess
Anne Co.: Cape Henry, 8g15£,40-l5l (NY). Richmond, University
of Richmond Campus, Clarke 885 (GH); Churchill, 5,5, (GH).
Roanoke Co.: Hanging R00k:.fl22§.§92£ (GH). Shenandoah Co.:
Massanutten Mt., Allard 5518 (F, MO, NY, US); South of Eliza-
beth Furnace Forest Camp, Chrysler 8881 (CHRB). Spotsylvania
Co.: Fredericksburg, 1111§_888 (SMU). Sussex Co.: Jarrett,
Fernald 8 155g 8841 (GH, NY, US). Williamsburg, Grimes 8551_

(GH). West Virginia. Kanawha Co.: Pine T0p Ridge, Loudon

 

348
District, Fletcher, 5,p, (MSC).

 

The Tournefort name Toxicodendron, which has been
used for this generic segregate of 8585, was chosen by Lin-
naeus as the Specific epithet for Eastern poison-oak.
Tournefort (1700), Barrelier (1714), and Gronovius (1743)
may have had this plant in mind when they described Toxico-
dendron triphyllum £5115 sinuato pubescente, for its leaves
are characteristically sinuate or lobed and pubescent.
Linnaeus included this polynomial phrase in synonymy under
his 8585 toxicodendron. However, no Specimens representing
this Species which were probably seen by these early authors
appear to be extant to confirm their identity. Furthermore,
in the Michaux Herbarium, there is a specimen labeled with
this same polynomial phrase that is 1, rydbergii. It is
with Michaux that much of the nomenclatural confusion begins,
as mentioned earlier. The point to be made herein is that
authors before and after Linnaeus have had different taxa in
mind in their use of 8585 toxicodendron as a name, or poly-
nomial phrases associated with it. Typification of the
Linnaean name is therefore exceedingly important.

The lectotype sheet at LINN (sheet 378.16 Plate
XXXVII) bears two plants; the one on the left represents the
taxon here recognized; the other is aromatic sumac [8585
aromatica Aiton = Lobadium aromaticum (Ait.) Raf. ex Steudel].
The Specimen on the left, then, is chosen as the lectotype
0f.§fl2§ toxicodendron. This species name has been used in-

discriminately for poison-oak (both Toxicodendron lobatum

'—

PLATE XXXVII.

 

349

Lectotype of Rhus toxicodendron L.

 

Photo taken at the Linnaean Herbarium.
The plant on the upper left with flower
buds present is the type. The other,
larger plant is Rhus aromatica

(= Lobadium aromaticum). The small
letter K beneath the latter specimen
indicates that it was collected by
Peter Kalm.

 

 

550

 

 

W

m

G.
)IIH‘u Nur‘hrL.

351
and 1. toxicarium and poison-ivy (1, radicans and 1, gyg-
bergii). Of the infraSpecific names prOposed for various
members of the complex using 8585_toxicodendron as a base,
one has been applied to Western poison-oak, three to Eastern
poison-oak, and twelve to poison-ivy!

Many persons have used the name 8555 toxicodendron
in preference to 8, radicans because it is perhaps more
euphonious and more suggestive of the toxic nature of the
plant. A rough count of herbarium Specimens examined shows
that it has been used more frequently (and carelessly) than
any other binomial with reference to either poison-ivy or
poison-oak. It was not even spelled in full by Linnaeus
until the publication of Systema Naturae, ed. 11 (1760). A
good discussion of the orthography is given in Fernald (1941,
pp. 597-599).

Michaux published the epithet quercifolium without
rank indicated, but variety implied. It was Eaton (1818)
who designated the rank as‘8555 toxicodendron var. guerci-
folium. Steudel (1821, p. 689) made the combination'8.
guercifolia, but lists it in synonymy and cannot, according
to the Code (Lanjouw.58.51., 1966) be cited as author of the
combination. Robinson and Fernald (1908) made the combina-
tion legitimately, although they attribute it to Steudel.

In the meantime, however, this combination was validly pub-
1ished as a fossil Species, 8555_quercifolia Goeppert from

the Tertiary of Bohemia (Goeppert, 1855).

352
To give some idea of the nomenclatural "mess" (to
quote Fernald's well-chosen word regarding this complex)
involving the epithets toxicodendron and radicans, some com-
binations are listed below. Where their true identity can
be correlated with a known species, the names are listed in
the apprOpriate synonymy in the taxonomic treatment; where
they are not, such names are listed in a group of doubtful
Species.
Rhus toxicodendron L. (1753, p. 266)
8. radicans var. toxicodendron Persoon (1805, p. 325)
Rhus-Toxicodendron radicans Marshall (1785, p. 131)
8585-Toxicodendron toxicodendron Marshall (1785, p. 131)
8. toxicodendron var. radicans Eaton (1818, p. 400)
8. toxicodendron forma radicans McNair (1925, p. 68)
8. toxicodendron subsp. radicans Clausen (1949, p. 8)
Toxicodendron toxicodendron (L.) Britton (1913, p. 484)
All of these have been cited by other authors. Also,
Rhus toxicodendron radicans (without rank specified) has been
attributed in the literature to Torrey, Dippel, Miquel, and
to "Farr." (sic) (Dodge, 1911), and 8. toxicodendron attrib-
uted both to Linnaeus and to Small by McNair (1925). McNair
(155.|515.) also mentions a var. typica which has no nomen-
clatural status according to the Code.
With respect to Toxicodendron vulgare Miller (Gard.
Dict., ed. 8, Art. Toxicodendron, No. 1, 1768), confusion of
a different sort pertains. It was not Miller's custom to

cite Linnaean binomials when transferring species to other

353
genera, or when providing new epithets for them, but to
give a reference at the beginning of the account which
established a connection with the earlier literature. He
used Linnaean epithets where apprOpriate; in other cases,
he used pre-Linnaean names, citing the source of the orig-
inal publication. A synonym which Miller (and others)
frequently cited as "Toxicodendron triphyllum 15115 sinuato
pubescente" (Tournefort, Inst. Rei, Herb. 611) which linked
the name 1, vulgare Mill. with 8585 toxicodendron L., being
a polynomial phrase shared in the two descriptions. The
lectotype of the Linnaean name accordingly typifies the
name substituted for it by Miller. This is the species
which herein is called 1, toxicarium.

Should not the epithet vulgare of Miller (1768) have
priority over toxicarium first applied as a species epithet
by Salisbury (1796)? It probably would if Miller's Latin
and English diagnoses accurately described Eastern poison-
oak, but they do not. He described the "foliolis" as:
"obcordatis, glabris, integerrimis," and adds further, "caule
radicante." Still further, he continues, "with roundish,
heart-shaped, smooth, entire, trifoliate (sic) leaves, and
a stalk putting out roots." Eastern poison-oak has leaflets
which are not obcordate, not glabrous, and are lobed, not
entire. The stem is not climbing; the stalk never produces
aerial roots. Inasmuch as the species which Miller named as
Toxicodendron vulgare is too inaccurately described to be

Eastern poison-oak, even though indirectly linked with that

354
Species in the literature (thus satisfying Art. 32 of the
Code), this binomial must be rejected under Art. 69 of the
Code.

A Specimen at Kiew representing 8585 acutiloba
Turcz, was collected in Texas by Berlandier. At the time
of Berlandier's explorations in Texas (1827 - 1830), this
region was part of Mexico, hence the word "Mexico" on the
label. This taxon is not found in present-day Mexico. It
was presumably collected in Southeastern Texas (McKelvey,
1955). An isotype at the Gray Herbarium completes the story:
the Specimen was collected near the Medina River in July
1829. According to McKelvey (155, p11,),this would have
been 15 - 16 July, just south of present-day San Antonio
in Bexar County. Another Specimen cited by Turczaninov (at
KW) is Mathes 158 from Taylor County, Texas. This is a
mixed collection; the two large Specimens on the sheet are
Toxicodendron radicans subsp. verrucosum, but the small
Sprig in the upper right corner is 1. toxicarium.

There are many pOpulations of Eastern poison-oak in
which leaflets of the male clones differ from the leaflets
of female clones. Most often, the leaflets on male plants
are more deeply lobed than those of female clones. In fact,
some of the leaflets in female plants may be elobate or with
undulating leaflet margins. Fernald named the form 8885
toxicodendron f. elobata, separating it simply on the basis
of sexually dimorphic characters, apparently unaware that

this was the case.

355

In collecting material which later became the type
for Toxicodendron monticola, Ruth apparently numbered Speci-
mens, rather than gatherings. Hence, the specimens at ND-G,
NY, and MO are presumed isotypes. Greene did not clearly
cite a type specimen, but indicated,8515,855 as representa-
tive. This Specimen is therefore designated lectotype. Be-
cause of the uncertainty with which Greene deals with Eggpy
Wilson 155 from Taylor's Ridge, Georgia, this specimen is
not considered a syntype. Greene's diagnosis described the
leaves as bright green, yet the leaves of the type have
dried brown, and he probably saw it after the plant was
dried.

Often used in connection with this species is the
epithet pubescens. Philip Miller originally published the
name Toxicodendron pubescens (Gard. Dict., ed. 8, Art.I1581-
codendron, No. 2, 1768). His description fits 1, toxicarium
which has pubescent leaves and fruits, but not to the exclu-
sion of all other taxa. Apparently no specimen upon which
this name could be based is extant. Thunberg (1794) pub-
lished the epithet again, this time as 8885 pubescens for a
plant from the Cape of Good HOpe, not a member of the poison-
ivy complex, but a 8885 sect. Thezera. 8585 pubescens
Engelmann ex Engler in DC. (1883) was a new name (and a
homonym) published for a plant that is likely to be 1, 5551-
5555 subsp. radicans, which has pubescent fruits. Type
Specimens--designated or implied--are known for none of these.

In 1900, Farwell published another Rhus pubescens,

356
indicating that he was transferring the name from Miller,
although, judging from the region covered by his work
(Michigan), not to represent the same Species. As was his
custom, he named a "type specimen" (sic) which is at BLH

with a duplicate at ND-G. Farwell's "type" is 1, radicans

 

subsp. negundo.

The formal epithet of 8. toxicodendron f. leiocarpa
means "smooth-fruited," but the holotype has pubescent
fruits (although the sheet bears the notation "glabr. fr.");
furthermore, the fruits are immature and shriveled, not
readily demonstrating any papillose condition. The glabrate
or glabrous nature of the fruit might be due in part to
their immaturity (their having been collected early in June)
and therefore should not be considered diagnostic. The
holotype has some aborted ovary tissue such as that described
for New Jersey poison-ivy by Gillis (1960). Isotypes at NY
and US were annotated by Fernald as a form of 8585 radicans
(sic) in error.

Toxicodendron toxicarium clearly is the most mis-
understood species in the poison-ivy complex. The common
name, poison-oak, is related to the typical leaflet shape,
which resembles the leaf of a Species of the white oak group.
This common name has been used in different senses by physi-
cians, Boy Scouts, extension agents, botanists, and others
until it has lost its meaning. Most often it is used to

represent 1, rydbergii or a shrubby or arborescent form of

357

‘1. radicans. The plant is seldom encountered in the wild
by non-botanists because of its habitat requirements. It
is practically restricted to low-nutrient sands of the
Atlantic and Gulf Coastal Plains, generally in association
with scrub-oak and pine woods (sandhills vegetation) that
have little economic value. It is seldom abundant. Although
it produces aerial stems from its stolons, it does so at
far greater intervals than its counterparts 1. radicans or
‘1. lobatum. In habit, it resembles the sub-shrub 1, nge
bergii more than any other member of the complex.

This plant is found most often in scrub-oak and
pine woodland savannah that has an understory of ericaceous
shrubs and bunch grasses, chiefly Aristida, 81155, and
AndrOpogon, characteristically found in the sandhills of
the Carolinas. In Oklahoma, where the same general vegeta-
tion type, minus the pine, may be found, Toxicodendron 1581-
carium occurs also. The soils in which it grows are most
often coarse sands, generally low in calcium, magnesium, and
potassium. In fact, the Lakewood and Norfolk sands in which
it has been found most often are the soils with the lowest
nutrient values of any along the Atlantic Seaboard. For this
reason, its requirements are such that it seldom occurs in
the same association with any other taxa of the complex, all
of which generally prefer richer soils.

In a very few instances, putative hybrids between
Toxicodendron toxicarium and 1. radicans subsp. have been

found. The former is contained within the combined ranges

358
of 1. radicans subsp. radicans, pubens, and verrucosum. It
might be expected, given the environment intermediate be-
tween that restricted one required by 1, toxicarium and the
rather broad one permitted by subspecies of 1, radicans,
that some gene interchange might take place. In the field,
although such situations are comparatively rare, several
have been found. One from BastrOp County, Texas, suggests
crossing between 1. radicans subsp. verrucosum and 1, 1581-
carium in a region where both are abundant. Several collec-
tions have been made in New Jersey, Georgia, and South
Carolina which seem to be intermediate between 1, toxicarium
and 1, radicans subsp. radicans. Other putative hybrids
have been seen among herbarium Specimens. A more complete
discussion of putative hybrids is discussed in the section
on "Hybrids, Crosses, and Intergrades."

The relative rarity of intermediate forms throughout
the southeastern states where both of these species are
frequent, if not common, has been a guiding factor in recog-
nizing them aS distinct species. Although hybridizing can
apparently occur, it is not an important factor in the gen-
etics of these pOpulations.

Part of the difficulty in recognizing the species
Toxicodendron toxicarium in the field is the degree of loba-
tion of the leaflets. Leaflet margins may be undulate to
almost entire varying through degrees of lobing to crenate
or deeply lobed. Often, as mentioned before, one sex may

have leaflets more distinctly lobed than the other, all

 

359

within the same pOpulation. It is usually the male which
has more deeply lobed leaflets than the female, but the
converse may also be true. The most significant morpho-
logical features, then, are the pubescent nature of the
leaves and fruits, the non-climbing habit, and the rather
impoverished habitat.

In published keys, one finds that some previous
workers have considered thickness of the leaves to be a
major factor separating 1, toxicarium from related taxa.
From anatomical sections, I have measured a number of leaves
of various of these species, and found no consistent differ-
ence. The leaves of 1, toxicarium are also described as
leathery. They perhaps have a slightly thicker cuticle, a
possible evolutionary result of the plant's being found on
drier sites than is poison-ivy, but otherwise, this char-
acter is variable. The thickness or leathery nature of the
leaflets is hardly conducive to field determination in such
a dermatitis-producing plant!

7. Toxicodendron magnifolium (MacGinitie) Gillis, comb.
nov. (Plate XI). 8585_magnifolia MacGinitie, Publ.
Carnegie Inst. Wash. 465: 144. 1937. Type: 85;;y
8, MacGinitie 5,5,, California, Trinity County,
Oligocene beds at Redding Creek; Weaverville flora.
(University of California, Dept. of Paleontology,
holotype; several isotypes).

This fossil plant Shares with modern poison-ivy

(eSpecially the group including Toxicodendron radicans subsp.

360

radicans, hispidum, and orientale) several morphological

 

characteristics. Dr. MacGinitie noted the close resem-
blance of the fossil with leaves of Japanese poison-ivy (1,
radicans subsp. orientale), a resemblance that is closer
than with North American taxa today. The fossil suggests
a link between Asiatic poison-ivy and North American pOpu-
lationS. Aside from the possibility of long-distance dis-
persal as a mechanism of accounting for poison-ivies in
Asia and North America, some ancient population may have
existed continuous across the Alaskan-Siberian land bridge
during the Paleocene-Eocene interval. There are unfortun-
ately no fossils from the northern Orient, the Aleutians,
Alaska, nor western Canada to sustain the hypothesis that
the plants really had this distribution.

The fossil associates of Toxicodendron magnifolium
(1158, 8y555, Juglans, and Taxodium) are similar to poison-
ivy's present-day associates in the southeastern United
States, and it may actually represent an organism whose
ecological requirements are little changed now. The fossil
specimen, however, was found in a region not presently in-
habited by poison-ivy, but rather by western poison-oak,
1, lobatum. Because of the great separation in time and
distance between the fossil and modern plants, I choose to
recognize the taxon named by MacGinitie as distinct from
the modern plants, but nevertheless to transfer the species

from Rhus to Toxicodendron to place it with the toxic sumacs.

HYBRIDS, CROSSES, AND INTERGRADES

A study of naturally-occurring suspected hybrids
found in the field was made, and an attempt was made to pro-
duce hybrids artificially. Difficulty in obtaining germina-
tion precluded the latter's being wholly successful, but
nevertheless some data of significance were accumulated.
SUSpeCted natural hybrids have been found and are herein
described along with ecological and geographical considera;

tions.
Field Studies

Because several subspecies of Toxicodendron radicans
and 1. toxicarium overlap in range, a search was made for
natural hybrids between them. AS will be discussed later,
these two Species rarely exist in the same vegetation associ-
ation or soil type. Even on the few occasions where one
finds them occurring together, there is a paucity of inter-
mediate forms. Several of these associations were investi-
gated in detail.

In the San Felasco area of Alachua County, Florida,
there is an extensive tract (T. 98., R. 19E., Sec. 21) of
mixed community of hammock and sandhills vegetation (Monk,

1960), where both Toxicodendron radicans subsp. radicans and

361

362
.1. toxicarium grow abundantly side by side. Despite inten-
sive searching, I was unable to find any intermediate forms
between these two readily distinguishable species. Both
flower at the same time (early April) and are pollinated

by honey bees, although few female flowers of 1, radicans

 

were found in the population.

In a study area north of Columbia, Richland County,
South Carolina, in a narrow band of overlap between poison-
ivy and Eastern poison-oak (c. 10 meters), there was only
one individual with intermediate characters was found (Gillis
4588). This individual has some of the leaflet shape,
notching and color of Toxicodendron radicans subsp. radicans,
but the pubescence and growth form (sub-shrub) of 1, 1581-
carium. The cited Specimen is a female in flower.

In another study area in BastrOp County, Texas, be-
tween Beuscher and BastrOp State Parks, one finds pockets
of Toxicodendron radicans subsp. verrucosum and others of
,1. toxicarium, the latter growing on the poorer soils. Only
one suspected hybrid was discovered, a sterile specimen
which possessed the leaf pubescence of'1. toxicarium but
the climbing habit of 1, radicans subsp. verrucosum (Gillis
5151). The lobing is intermediate between the rounded lobes
of 1, toxicarium and Spreading, sharp lobes of 1, radicans
subsp. verrucosum.

A suspected hybrid clone (between Toxicodendron 1581-
carium and 1, radicans subSp. radicans) was found in Jeffer-

son County, Georgia, in a loblolly pine and sassafras woods

363
(Gillis 4855). It shared with poison-ivy the climbing
habit and the tendency to have entire leaflet margins, but
had the pubescence of Eastern poison-oak.
Artificial crosses were made between Toxicodendron

radicans subsp. radicans and 1. toxicarium in Alachua County,

 

Florida, and in Richland County, South Carolina, in the
areas mentioned above during the spring of 1961. Because

of a dearth of female flowering poison-ivy plants, reciprocal
crosses were not possible in the Florida experimental plots.
Crosses which were made in South Carolina were each repli-
cated in three inflorescences and bagged to prevent con-
tamination by insects. Pollen was transferred from anther
to stigma by means of a loblolly pine needle. The popula-
tion of Eastern poison-oak had sexually dimorphic leaves,

so that the male clones could be distinguished from the
females even when only unopened flowers were present. Flow-
ers which had not yet Opened, but for which flowering ap-
peared imminent, were chosen exclusively. Results were
considered successful if fruits were produced.

Where Toxicodendron radicans was the female parent,
the fruits were small (3.5 - 4.0 mm broad), whereas the
fruits were more typically the size of 1, toxicarium fruits
(5.5 - 6.5 mm broad) where this plant was the female parent.
The hybrid seeds were indistinguishable from normal seeds
whose parentage represents but a single Species. It is to
be noted that no fruits at all develOped from unpollinated
controls, thus making it unlikely (but not impossible) that

364
apomixis might be taking place. Germination of the hybrid
seeds was attempted, but no seedlings resulted. But then,
I have succeeded in germinating only two seeds of 1, 1581-
carium in trials numbering more than 200 seeds of pure
parentage.

Results, as shown in Table 5, suggest that there is
no apparent difference in percentage of successful crosses
between Open-pollinated flowers and artificial hybrids, at
least not in the South Carolina series. In the Florida
trials, however, the number of interspecific crosses produc—
ing fruits was half that of Open-pollinated poison-ivy
flowers.

In the gorge of the Columbia River between Oregon
and Washington, there is overlap in range of Toxicodendron
lobatum from the west and 1, rydbergii from the east. In
studies made in late summer and early autumn of 1960 and
1962, the pOpulations of these two were examined and samples
collected from along the two Sides of the gorge. The zone
of overlap produced an apparent introgressive series between
these two taxa. The zone of apparent hybridization extends
in Washington from the confluence of the Klickitat and the
Columbia Rivers eastward 43 miles to the vicinity of the
John Day Dam, and in Oregon from the Hood River eastward
23 miles to the vicinity of The Dalles. There is evidence
that this zone of overlap is restricted to the gorge because
pOpulations immediately south of Hood River appear to be

genetically pure 1, lobatum (Gillis 5231 from 2 miles south

365

.COHpmHSQOQ mEmm Eon mm; cmHHOQ mo monsomm
.mCOflpmasaoo pcmhmmmwp Scam mm; cmaaoa mo mousowH

 

 

 

 

 

 

 

 

 

 

 

 

o o NV 0

o o o um I. m
o 0 mm asaaaaaxoe .H empacaaaoaea < o
00 ON vm o

5v om OH no .I m
mm mm «a Eswumoflxow .H pmmefiHHOducmdo < v
on ma ON 0

mm <0 mm av I l m
vv 0H vm Eswumofixov .H mamowUmH .H < m
<0 mm on o

no on OH ON I .l m
on Ca ma Esfiumowxop .H NEDHHmonov .H < N
on ma mm 0

mm on b OH _I .I m
SN 6 mm Ezwumoflxop .H HESHHmofixov .H < H

mm vm an mo I I m
ow ad on mcmofipmn .H mamofipmu .H < N
cm 0 ON 0

av om 0H av .I .l m
mm mm mm mcmoflpma .H EnfiumbMMOp .H < >
mommoao mpflsHm pmvaHHHom III (I

pch Hod Hammmmoosm mafiuasmmm mumzoaw pcmymm pcmnmm coapmo 966832
Cmmz mmmpcmonmm Mo Monasz mo 906832 mHmEmw mam: ufiHowm mmoMD

 

 

IHXOH

mam; mHADmmm

mmHHm :HOm H< Hood mmeMHmmm >qm<m ZH Qmmmmh<w
.AUmeQEDCV Hood ><2 NH 20 <ZHAOm<U mhzom .>HZDOO DZ<AIOHm 2H 024
Anwwmppmav Hooa AHmQ< o 20 .<QHmOAm .>HZDOO <D$U<A< 2H mQ<E mmma mmmmomu

.EDHm<U

.H 02¢ mZ<UHQ<m .mmmbm mZ<UHQ<m zomDZmDOOHXOH meBHmm mmmmomo UHmHUmmmmmHZHll.m mqm<H

366
of Hood River city limits). Furthermore, neither taxon
grows above 1275 meters in the Cascades which rise to the
north and south of the gorge.

In a roadside park at the confluence of the Klickitat
and Columbia Rivers there is a large shrub of Toxicodendron
lobatum surrounded by a clone of suSpected hybrids. These
pOpulations were seen late enough in the season to demon-
strate their autumn coloration. The difference between the
two taxa was therefore striking. From this location, each
pOpulation of the taxa was studied as I collected eastward
through the gorge. Each population in the zone of overlap
was sampled until pOpulations appeared to be fairly constant
as typical 1, rydbergii. It is in the same region that
flowering and fruiting material were collected by Suksdorf
in the late 1800's. The latter collections were the basis
for two new Species names by Greene (1905). Presumably
these collections represent hybrid pOpulations.

Eastward through the gorge from Hood River in Oregon
or from the Klickitat River in Washington, fruits of the
Toxicodendron Species become smaller, less pubescent, more
nearly globose, and on shorter pedicels. The inflorescence
becomes more compact and more erect than pendant. The leaves
progressively become less crowded at the summits of branches,
lose the pubescence of the petioles, assume a more yellow-
orange (less red-brown) autumn coloration, lose the tendency
for the peculiar mottling of the dried leaves of 1, lobatum,

and tend to have the broadest part of the leaflet below the

367

midpoint, becoming ovate rather than elliptic. (Fig. 16.)
Herbarium Studies

In the course of examination of populations in the
field plus examination of herbarium Specimens, I have deter-
mined a number of these to be intergrades between taxa. In
several cases, the situation indicates that these intergrades
are very likely hybrids. In others, the nature of the inter-
grades is less certain, possibly examples of incomplete
divergence among taxa. Toxicodendron lobatum and 1, rydbergii,
for example, occupy completely separate ranges except for a
zone of overlap in the Columbia River gorge, as mentioned
above. It seems justified to interpret the intergrades in
the gorge as putative hybrids. The rather blurred separation
between 1. radicans subsp. negundo and subsp. radicans in the
Allegheny Mountains may be interpreted as hybridization, but,
on the other hand, may also be looked upon as intergrading
due to current Speciation° The Alleghenies are not a com-
pletely effective barrier to gene exchange between subSp.
radicans to the east and subSp. negundo to the west. They
are treated here as hybrids, but the case of incipient Speci-
ation cannot be ruled out.

Toxicodendron lobatum (Hook.) Gillis X 1, rydbergii
(Small ex Rydberg) Greene.
Toxicodendron lobadioides Greene, Leaf. Bot. Observ.

Crit. l: 119. 1905. Type: Washington, Western

Klickitat County along the Columbia River, rocky

368

 

Fig. 16. Distribution of Toxicodendron lobatum and
1. rydbergii in the Columbia River gorge
of Oregon and Washington and location of
intergrades between them.

 

 

_— ”MA

 

“or“

569

 

 

 

 

 

 

\

 

.28>-O'

./

.x)
-

...)

(x

\
=20“>s .h
0.0.20-3. 00:00
It'll! .h .8...

   

.05.. .008
1 «I

g

30...... 0:33

  

 

  

370
places. ‘fla_8. Suksdorf, 5,5,, July 1885. (US-
47202, holotype; F-155984, F-255875, isotypes in
part).

Toxicodendron coriaceum Greene, Leafl. Bot. Observ.
Crit. l: 120. 1905. Type: Washington Territory,
1v. _N. Suksdorf, 5.3., 6 May 1885. (US-19803, holo-
type; DS-855ll, F-l55984 (in part), F-255875 (in
part), MO-l7737l9, NY, isotypes).

Representative Specimens: UNITED STATES. 8£§gon.
Hood River Co.: Hood River, near mouth of river on flood-
plain, Gillis 5881 (MSC); same, Whited 1117 (ND); same,
rocky bluffs, Henderson 885 (MO). Wasco Co.: The Dalles,
Howell, 5,5. (GH); Gillis 4222 (MSC); Seufert, Gillis 4888
(MSC). Sherman Co.: east of Rufus, Gillis 4888 (MSC).

Washington. Klickitat Co.: western part of county, Suksdorf,

 

 

5.5. (DS, F, GH, MO, NY, US); roadside park near confluence
of Klickitat and Columbia Rivers, Gillis 5848 (MSC); 11.2
miles east of Klickitat River confluence with Columbia,
Gillis 5241 (MSC); 11.3 miles east of confluence of Klickitat
and Columbia Rivers, Gillis 5848 (MSC); Murdock, east of
bridge, Gillis 5848 (MSC); savanna valley, T. 2N., R. l5E.,
SW’% Sec. 11, Gillis 5844 (MSC).

Presumably these two Greenean species are hybrids
between Toxicodendron rydbergii and 1. lobatum, for they
were collected in the region where the two species overlap
in range. They are probably part of the same pOpulation,

from the basaltic washes in the hills near the confluence

371
of the Columbia and Klickitat Rivers in western Klickitat
County, Washington.

Data on an isotype of Toxicodendron coriaceum and
of 1, lobadioides at F suggests that one pOpulation which
was sampled by Suksdorf on two different occasiona furnished
the type material for two of Greene's Species! This sheet
has both a fruiting and a flowering specimen and bears a
printed label indicating the date of collection as "May 6,
July 1885." The holotype of 1, coriaceum at US is a match
for the fruiting representative except that it bears a
handwritten label. The type of'1. lobadioides at US is a
match for the flowering representative on the Sheet at F,
but bears c0pies of both the handwritten and printed labels.
Also, the isotypes of 1, coriaceum at NY and at M0 have the
printed label with the May date crossed out.

Presumably the flowering specimen dates from May,
and the fruiting one from July, both from the same general
area, representative of the same or closely related pOpula-
tions. The two type Specimens mentioned above appear to be
intermediate in morphology between Toxicodendron lobatum
and 1. rydbergii. Whereas 1. lobatum has densely Short-
pubescent fruits and 1. rydbergii has glabrous fruits, these
intermediates have scattered hairs, or in one instance, a
crown of hairs at the base of the fruit. Table 6 shows the
diagnostic characters used in analyzing these intergrades.

Because of the already burdensome collection of names

that the poison-ivy complex has acquired, I have preferred to

372

 

mmOQOHm mpHSHm

mmoQOHO mHHDHm

emcee ea m.m - 0.4 aeaaem

mmeHEDom
Ho mvsom maHH pmHmmmH

onmb mmomHHpm pmmmmhd
udm Ho msoHQmHO mpwmemH

HmHsoHnHonsm
ow mHm>o >Hpmonn meHmmmH
mHmHspc: mCHOHmE HmHmmmH
OHBH So OH Ham 0 I

v mm>mmH

mspomHm wHoHpmm

mmoQOHmnzm
mmOHHHQmQ Ho mmOHspmm mszam

meQOHObsm
.eacaeeaHe >HeemaHa meaaem

emcee ea m.a - m measam

meCHESom
>HeemaHa oe apnea mafia eaHmmaa

30Hmb
mmomHHpm op mpmHanO pmHmmoq
mpm>o >Hpmohn mvammmH

CHOHmE OCHpmHDOc:
59H; Ho .mHHpcm mpmHmmmH

OHSH Eo OH Hmd OH I v mo>m6H

mHmHQmHO mHOHumm

mmopmw >Hmmcov mHHsuw

eaoeHaa
>HaseaHao .aaeeeeeam mpasem

Umonn SE n n O mpHSHm

mmspbo mQHp pmemmH

mHHm: vamemom 36m cpHs
no .BOHmb msoubmHO mpmemmH
HmHsoHQHOQSm Ho .mpm>ono
.aaeeaHHa >Heaoaa meeHLaaa
OmQOH Ho

.mpmHspcs .mHmHHmm mpmHmmmH

OHSH Eu OH Hod OH I s mm>mmH

vcmHDHmnsa Ho wvaanO wHOHpmm

 

 

aaaaaaaxa .e

eHea>z e>Heaesa

 

enemaoH .H

 

 

 

mmwmmmmObImmmm>r

m>HH<HDm mHm$H Qz< .HHDmmmo>m .H Dz< EDH<mOH zomDZmQOUHXOH .mHme<m mo mmMHU<m<IUII.© mHm<H

373

use a formula to designate this hybrid, rather than use
either of the Greene names, or some other at the subspecies
level. Moreover, no one population possesses the same clus-
ter of intermediate characters nor the whole array of them,
so that a hybrid taxon would be difficult to circumscribe.
The formula seems to be a better means to indicate a Speci-
men or pOpulation that demonstrates intermediate characters
anywhere along the Spectrum between one parental form and
the other.

Toxicodendron toxicarium X 1, radicans subsp. radicans.

Suspected hybrids between these two Species have
been found along the American Atlantic Seaboard where their
ranges overlap. Because their habitat requirements are gen-
erally quite different, it is not often that a pollinating
vector will visit one parent plant while carrying pollen
from the other. Criteria which were employed for determina-
tion of these forms as hybrids are given in Table 7.
Representative Specimens: UNITED STATES. Alabama.

Covington Co.: edge of titigum-magnolia bay, 17 miles SSW

of Andalusia, Hardin 8 Duncan 15010 (GA). Arkansas. Conway

 

Co.: Morrilton, Demaree 88888_(GH, MIN, MO, NY, SMU. Logan
Co.; Blue Mountain, Palmer 84818 (MO). Geo i . Bibb Co.:
savanna woodland, Gillis 5515_(MSC). Walker Co.: rocky
area below Harrisburg, 10.3 miles SW of Lafayette, Hardin,

Humphrey, 8 Duncan 13872 (GA). North Carolina. Forsyth Co.:

 

Kernersville, Ahles 8 Britt 40711 (NY). Stanley Co.: Yadkin
River, southeast of Badin, Ahles 8_Radford 11936 (SMU).

374

 

mHme CHm> CH mHHmL
mo mpmsp :HHS HmHmmmH Hmzou

meHHHo Ho: mCHOHmE pmHmmmH
pcmommbsa mszem

mHmHQmHO Ho

msoHQmHm momeSm pmHmme Hosea
mvaQmHm mommasm HmemmH Hedda

oHpQHHHm Ho mpm>o mpmHmmmH

szopoc >HmHmH Ho
.mcHOHmE mHHHcm :HHB pmHmmmH

mocmommbsa >HmmHQO mmHoHpmm

mCH> Ho bannm

mHHXm ch> OH mHHm:
mo mvmsp 39H; HmemmH Hoaoq

meHHHo mcHOHmE HmHmmmH

mHmH mHHSHm

mDQCHvsHm>
mommysm pmHmmmH Hmaoq

Hcmommbso >HOCH
uHQO oommusm HmHmmmH Homo:

oHHdHHHm Ho mpm>o
>Hpmogb ow OHooEOLH mHmHmmmH

mCHOHmE mecmHo
no mHmHHmm >Hmmumoo :HHS
Ho .mOQOH Qnmcm anz pmHmmmH

mmomHHpm Ho mpsmHHL mmHoHpmm

OCH> no bannm

QHHOHE

mcon HmmH>mon mgmsnma
go OopanHpme >Heuoch:
mocmommnsa pmHmmmH Hmzoq

mHmHHHo mCHOHmE HmemmH

mmOHHHQmQ
cmpmo Ocm pcmomwbno pHsuw

msoCHpsHm> Ho mpzmHHc
momMHsm HOHmmmH HmsoH

OHQmHL Ho msoybmom
momwuzm pmHmmmH Homo:

mpm>onoumcoHno Ho
OCOHQo ov mpm>o mpmHmmmH

mchHmE mchmHzpcs
Ho meOH 59H; pmHmmmH

mmovcmEOH
Ho mpsmHHn mmHOHHmm

aeeea-aam

 

 

 

mCmOHOmH .dmbzm mamoHpmH .H

aaea>m a>aeaeae

 

EOHHmonop coupcmpoonOH

 

 

 

 

UszammmO DHmm>I m>HH<HDm QZ<

 

 

.mZ<OHQ<m .mmmDm mz<UHQ<m .H Qz< EDHm<UHXOH zomQZmQOOHXOH .mHme<m m0 mmmHO<m<$Unu.b mqm<H

375
8gy_t Qarolina. Pickens Co.: Clemson College (MO). Rich-

.—

 

 

 

land Co.: north of Columbia on highway 21, pine woods,
Gillis 4588 (MSC). 1::g1p15. Fauquier Co.: Bull Run Mts.,
Allard 8841 (GH).

Toxicodendron rydbergii Xl1. radicans subsp. radicans.

Toxicodendron rydbergii and 1, radicans subSp. 3551-

5555 have ranges which meet in the northern United States
and bordering Canada east of the Appalachian Mountains and
the Laurentian Plateau. Where their ranges overlap, there
appears to be some introgression. A number of Specimens
have been seen from Quebec, Nova Scotia, the Berkshire Moun-
tains, and other places in southern New England which appear
to be intergrades between these two taxa. (Fig. 17.) The
characters upon which their hybrid nature has been suspected
are listed in Table 8.

Representative Specimens: UNITED STATES. Connecti-

 

581. Litchfield Co.: 2.5 miles SE of Washington, Donahue
851 (MSC); near Massachusetts line, Fernald, 515. (GH).
851:5. Knox Co.: Rockland, Fernald 1884 (GH). Berkshire
Co.: Huosac Valley, Lanesboro, Churchill 558 (MIN). ‘85555-
chusetts. Essex Co.: Bradford, part of Haverhill, Churchil1,

 

 

(M0). Suffolk Co.: Dorchester, section of Boston,

 

Churchill 552 (M0). New Hampshire. Strafford Co.: NE of

 

Durham, Donahue 268 (M80) Grafton Co.: West Thornton, Gillis
4097 (MSC). Merrimack Co.: Concord, Gillis 4100 (MSC).

Bennsylvania. Mifflin Co.: without definite locality,

 

Rothrock, 5,5, (F). Rh 8: 1,la :pd. Providence Co.: Diamond

u———_- _r—e-u—
———

376

Fig. 17. Distribution of 3 taxa of poison-ivy in the
United States and Canada and location of
intergrades between them.

 

577

 

 

I ..

... z, 8. a r:
. .1 .

JIgI (IIIIII1II.

.—"!..h I 0803‘: do... 0000.....h I 1 I
.
It... in... ogfi I 0803': ...... 0:00;: .5. O

is... do... clan—‘05.... I Evtfi ‘ 1
(I,

  

29>UU311 "134‘

              

..1 n~
,,,. . u. I 0000‘: 6’.- 2.08.... ..p '
1x . ill... in... taco-i! .... I 2
. 3211.5“ :
I
1
0n.

 

 

 

 

.1 ,1. . 1.. L

meta—m QwEZD

 

378

 

mmOH
IHHde Ho HCmommnsa mdumooxm

HmHmEme
CH so «.0 OCHOmHm>m mHHSHm

mHHHCm mCHOHmE HmHHmmH

onHanO oomwnsm HmHmmmH Rodd:
mHme CHm>

CH mHHmC mo COHHmHHCmocoo
CHHB mommusm HmHmme szoH

oHHQHHHo Ho mHm>o mHmHHmmH

HCmH
unambza no meHCmHO mmHOHHmm

Eo OH
CmCH mmmH OCHOmHm>m mmHOHHmm

mCH> OCHQEHHo Ho CDHCm

msoHQmHm Ho .mHHmC
OmeHHmom :HHB wsom “mmOH
IHHQma >HHC¢HQH6CH Qumooxm

“myoEmHO CH Eo mv.O OCHOmHm>m
Han .mNHm CH mHCmHHm> mHHDHm

mHHHCm
Ho mHmHspCD mCHOHmE HOHmmmH

mspo
anO momwnsm HwHHme Honda

mHme CHm> CH mHHmC 3mm
CHHB mommHSm HmHmme HoBOH
mHm>o >Hpmomn mHmHmmmH
>HHmHmHO HCmHOHmosa OCH
quomb Ho msouanO mmHOHpmm
Eo NH OCHOme>m meoHHmm

mHooH HmHHmm >2 OCH
nbEHHo mHCmHQ 36m “QCHCmunsm

mspomHm dnmooxm

HmHmEmHO
CH Eo m.O OCHOmHm>m mHHDHm

meHCOCC
Ho mHmHHmm mCHOHmE HmHHme

mzoun
umHO mommhsm HmHmmmH Head:

mHme CHm> CH
mHHmC Ho COHHmHHCmoCoo 0C

CHHS mommusm HOHHmmH Hmaoq
HmHsoHnHonsw

6H mpm>o >Hpmonn mHmHHmmH

mOoHCmHm mmHoHHmm

Eo OH CmCH

mHoE OCHOmHm>m mmHoHHmm

anaca-e:m

 

 

 

mCmoHpmH .ansm mCmOHOmH .H

eHaa>m m>Heaeam

 

HHOHwCONM coHpCmOoonOH

 

 

 

 

I oszmmmmO QHmm>I m>MH<HDm QZ<

 

.mz<oHo<m .ammam mz<oHo<m .H oz< HHommmo>m zomazmoooHXOH .mHme<a mo mmmHo<m<:o-I.m mam<H

379

 

Hill, Palmer 46188 (A). Tennesseg. Blount Co.: Walland,

 

Hudson, 5,5. (RM). 'Vermong. Caledonia Co.: Barnet,
Blanchard, 5,5. (MIN, MO, NY, US); Lamoille Co.: without
definite locality, Grout, 5,5, (F); Windsor Co.: Ascutney,

Gillis 4088 (MSC). Virginia. 1 mile north of HOpewell

 

Gap, western slope of Bull Run Mts., Allard 462 (GH, US).

West Virginia. Hampshire Co.: Slanesville, Donahue 265

 

(MSC).

*****

CANADA. Nova Scotia. Halifax Co.: Fall River,

 

Jack 3317 (M0). Lunenberg Co.: East Bridgewater, Macoun
81289 (F). Quebec. Brome Co.: Glen Sutton along the Mis-

——-.._-_

 

siquoi River, Marie-Victorin, 51.51, 55888 (GH, F, MO, NY).
Toxicodendron toxicarium X 1, radicans subsp. verrucosum.
This hybrid is likely to be a limited one because
the two parental types are sympatric for only a small por-

tion of their total range. Moreover, Toxicodendron toxi-

 

carium tends to be found in rather acidic soils and 1, 5551-
5555 subsp. verrucosum is generally foundin more alkaline
soils or in shallow lithosols derived from limestone. Sev-
eral pOpulations in Texas, however, have produced suspected
hybrids, plants which are intermediate in morphology between
the two taxa. (See Table 9.)

Representative Specimens: Texas: Aransas Co.:
.922!.flEQlé (GH); BastrOp Co.: climbing post oak in BastrOp
State Park, Gillis 5151 (MSC); Bexar Co.: San Antonio, gggmy
885 (GH); McLennon Co.: Robinson,‘85£y 55118 (SMU); Red

380

 

mmHHHomd OCm moCmo

mzoabmHO mHHCHm ummbso OmHmHHmom CHHB mHHDHm
mHmHOmHO no monbmHO mmOHHQ
mmommunm HmHHmmH HoBOH mmommysm HmHmmmH Hmaoq
HmHsmCm mmCOH HmHmmmH mmzvbo monoH HmHmmmH
HCmmmHQ mHooa HmHHm< HCmmmHQ mHOOH HmHHo<
mCH> Ho psanm mCH> Ho CCHCm HHmEm

mmOH
nHHQmQ Ho HCwommnsa mHHsem

msoCHHsHm> op mmOHHQ
mmomwnsm HmemmH HmBOH

OmOCOOH meoH HmHmmmH
HCmmnm wHoou HmHHm<

anaca-nzm

 

 

 

Eamoosuhm> .mesw mCmoHpmH .H UHHQ>$ m>HHmpsm

 

EOHHmoHXOH CogpcwnooHXOH

 

 

wszmmmmO QHmm>I m>HH<HDm

mHmIH Dz< EDmOUDmmm> .mmmbm mZ<UHQ<m .H QZ< EDHm<UHXOH zomDZmQOOHXOH mo mmmho¢m<$0|l.o mqm<H

381
River Co., Clarckville along Red River, Cory 56041 (SMU);
White Oak Bayou, collector and county unknown (MIN).

Toxicodendron radicans subsp. divaricatum X 1, rydbergii.

 

These two taxa are relatively uncommon throughout
their respective ranges, but eSpecially so in the arid and
semi-arid regions of southern Arizona and northern Sonora.
SuSpected intermediates have been found in only three places:
in the Chiricahua Mountains and the Santa Rita Mountains of
Arizona and the San José Mountains of Sonora. The character-
istics which appear to intergrade are listed in Table 10.

Representative Specimens: Mexico: Sonora, San José
Mountains, Mearns 1052 (US); United States: Arizona; Santa
Cruz Co., Florida Canyon, Santa Rita Mountains, 4000 feet,
Graham, 5,5, (NY). Cochise Co., above Portal, Chiricahua
Mountains, 5000 feet elevation. Stoutamire 8885 (BLH, MSC).

Toxicodendron rydbergii (Small ex Rydberg) Greene X
Toxicodendron radicans subsp. negundo (Greene) Gillis.
Synonym: Toxicodendron rufescens Greene, Leafl. Bot.

Observ. Crit. 2: 46. 1910. Type: Indiana, Marshall

County, vicinity of Lake Maxinkuckee, 8, Walton

“81558 8, 14 September 1906 (US-645163, holotype).

Representative specimens: UNITED STATES. Illinois.

 

Will Co.: Romeoville (formerly Romeo), umbach, 5,5. (F, WIS).

Locality uncertain: Gates 745 (F). Indiana. Marshall Co.:

 

Lake Maxinkuckee, Clark 8 (ND-G, US). Michigan. Gladwin Co.:

 

Sage Twp., shore of Wiggins Lake, Gillis 3602 (MSC). Lapeer

382

 

msoubmHm m>m3Hm mHHCHm

mHmHDOCD
Ho mHmHCOO mCHOHmE HmHmme

OmOHn mm OCOH mm
mmEHH m.H u H.H .HmHonnHonsm
0H me>o >HOmoHC mHmHHmmH

HmemmH HmCHEHmH
CmCH HmOCoH Lose >HHm3m:
“OCOH >Hm> >HHmsm: wHOHHmm
msouanO m>m3Hm mHOHHmm

OCHQEHHU Hm>oc
HCmHQ “HCwmbm mHooH HmHHm<

moCmommbsa
pmHmHHmom CHHB mHHCHm

mHmHCmO mCHmHmE HOHHmmH
OmOHn mm mcoH

mm mmEHH O.H n m.H HmHHmmH
umHHmmH

HmCHEHmH CmCH HmOCOH Ho
CHOCmH msmm mCH Hsonm mHOHHmm

HCmHOHmbsd Ho mspomHm mHoHHmm

mHmHmE m 0H OCHQEHHU
HCmHQ “HCmmmHQ mpoon HmHHm¢

HmCONHH<v mmCmH mo Hymn
CHOCHHOC CH mHmHanO mHHsam

mHHHCm mCHmHmE HmHHmmH

pmonb mm mcoH
mm mmEHH m u O.H meHmmmH

HOHmme
HmCHEHmH CmCH HOHHQCm mHoHHom

HCmHCHmnsd
>HxaHee oe >HeHee eHOHeaa

OCHQEHHo CmHHo
HCmHQ “HCmmmHQ mHoOH HmHHm<

 

 

aHmeaaawa .H

eHaa>E a>Hpeeam

 

 

EOHmOHHm>H
O

O
.amnsm mCmoHpmH H

 

 

 

 

. Oszmmme QHmm>I m>HH<HDm mHmIH QZ<
Hmeme>m .H Dz< EDH¢UHm<>HQ .mmmDm mZ<UHD<m ZomDZmDOUHXOH .mHme<m mo mmmho<m<IUIn.oH mHm<H

383
Co.: Imlay City, Farwell 6693 (BLH). Oscoda Co.: south

 

of Oscoda, Hagenah 2365 (BLH). St. Clair Co.: Port Huron,

 

Dodge, 5.5. (MIN). Minnegpta. Spring Grove, Rosendahl 280
(GH). New York. Chemung Co.: without definite locality,

 

Lucy 1694 (F, RM, WIS). Lewis Co.: Harrisville, Gillis
5377 (MSC). Oklahoma. Beaver Co.: Knowles, Stevens 516

 

 

(GH, MIN, MO, NY). South Dakgtg. Union Co.: Ravine, T.

92N., R. 49W., Eslick, 5,5. (SDU). West Virginia. Kanawha

 

Co.: Quick, Reed, 5,5, (CHRB). Wisconsin. Sauk Co.:

 

 

m 1954 (A, MIL, WIS).
* * * * *

CANADA. Ontario. Waterloo Co.: Hamburg, river
banks. Merner 15881 (WIS); Quebec. Laval Co.: St.-Laurent,
Adrien 1888 (A, GH).

Greene (1900) placed his Species Toxicodendron ruf-
escens as near 1, glabrum Mill. or 8555 littoralis Mearns

(= 1. radicans subsp. radicans). It appears rather to be

 

an intermediate between 1. radicans subsp. negundo and 1,
rydbergii, with fruits and petiole length in common with the
former and leaflet, shape, habit, and absence of aerial
rooting of the latter. Both taxa are known from this gen-
eral region of Indiana.

The two parental taxa are very different from one
another when they Show the most definitive characters, but
not all specimens seen in the field, or worse yet in the
herbarium, demonstrate the most important diagnostic char-

acters. A fruiting Specimen with tight, erect inflorescences,

384
fruits greater than 0.5 cm broad, very long, glabrous
petioles, and suborbicular leaflets is unquestionably 1581-
codendron rydbergii. A flowering male Specimen with puber-
ulent petioles and broadly ovate leaflets is not so easily
pigeonholed. It has been mentioned earlier that subsp.
negundo is difficult to characterize because it has much in
common with several other taxa and is characterized more be-
cause of what striking characters it does 551 possess, than
because of those it does. Along the zone of overlap in the
ranges of 1, rydbergii and 1. radicans subsp. negundo, the
characters which distinguish these taxa become blurred. A
number of populations in this tension zone look suspiciously
like intergrades between the two parental forms.

The regions where the two ranges meet are: the
southern lobe of Ontario, the central part of the Lower
Peninsula of Michigan, areas around the southern half of
Lake Michigan, and an irregular line cutting southwest across
Illinois, Iowa, Nebraska, and Kansas, and through parts of
Oklahoma. Suspected intermediates have been examined in
the herbarium and a few in the field from this region. (Fig.
17.) The suSpected hybrids Show their intermediate nature
as indicated in Table 11.

Toxicodendron radicans (L.) Kuntze subsp. radicans X

 

1, radicans subsp. negundo (Greene) Gillis.
Rhus radicans forma hypomalaca Fernald, Rhodora 43: 592.

1941. Type: West Virginia, Ritchie Co., Berea,

385

 

HmHmEmHO CH So m.O OCm
q.O Cowapmb OCHOCmH mHHCHm

mCOvaHHmm .Ho mmnopoc mo
aeHaa m I H EpHa agaHmaaa
HmHspoHobsm meHHmmH

OCoH Eo OH I m mmHOHHmm

msoHQmHm mmHoHHmm

anacaIaam

HmHmEmHO
CH Eu 0.0 OCHOme>m mHqum
mHmHmmmH C0 Cpmmp N u H
mpm>o >vaoun mHmHmmmH

OCOH Eo mH
vsonm OCHOmHm>m mmHoHme

>HHmC pmHmHHmom Ho

.HCOHCHOOCQ .msoHQmHm mmHOHHmm

HCmmmHo on

yo: >mE Ho >mE wHooH HmHHmm
“OCH> OCHQEHHQ Ho QCHCmubsm

HmHmEme CH Eo mv.O
CmCH mmmH OCHOmHm>m mHHCHm
meHooo >Hmmemoo mHmHmmmH

mHm>o meHmmoH

OCOH Eo OH CmCH
mmmH OCHOmHo>m mmHOHHmm

mmomHHHm
Ho HCmHsHmbsd mmHOHHmm

mHQOH
HmHme CHHS mCH> Ho nannm

 

 

Hamuwbpmu .H

aHaa>z a>Hpeesa

 

opcsmmc .mesm mCmoHOmH .H

 

 

 

OszmmmmO 0Hmm>I m>HH<HDa

mHmzh 02¢ .HHommm0>m .H 02< 0020002 .mmmDm m2<UHO<m .H .mHme<0 mo mmMHU<m<IUII.HH MHm<H

386

 

dry Upland woods. 1,18. and 8, 8. Randolph 1385,
23 August 1922. (GH, holotype).
Representative Specimens: UNITED STATES. Arkansas.

 

Drew Co.: Monticello, Demaree 13692 (A, NY). Garland Co.:

 

Starlight Bay, Hot Springs National Park, Demaree 54056 (MSC).
Johnson Co.: Knoxville, Demaree 19943 (MIN). Perry Co.:

 

Ouachita National Forest, Demaree 37842 (SMU). Pulaski Co.:
Galloway near Little Rock, Demaree 53848 (MSC). District

 

 

 

51 Columbia. Washington, Hunting Creek,‘85;8, 5.5. (US);

 

 

Giesboro Road, Steele, 5,5, (US). Kentucky. Breathitt Co.:

 

 

Robinson Forest, Smith 1717 (KY). Oldham Co.: 3 miles SE

of Westport, Duncan 22750 (KY). Logisiana. Madison Co.:

 

Tensas Gun Club Refuge,'8855 11518 (M0). Orleans Parish:
New Orleans, collector unknown (E). Plaquemines Parish:
Pillottown, Eflifl.llfl§é (MO). St. Mary Parish: Cote Blanche
Island, Thieret 11448 (MSC). Massachusetts. Berkshire Co.:

 

without definite locality. Milligan, 5,5. (ARIZ). Bristol
Co.: Nunquit, part of Dartmouth, Sturtevant, 515. (MIN).
Essex Co.: Bradford, part of Haverhill, Churchill, 515.
(MIN). Corfolk Co.: West Quincy, Churchill, 5,5, (MIN).

Mississippi. DeSoto Co.: 3 miles north of Nesbit, Gillis

 

5906 (MSC). Ngrth Carolina. Buncombe Co.: Asheville,

 

Donahue 259, (MSC). Swain Co.: Alarka, Ahles 8 Bell 14210

(FLAS). Penngylvania. Bedford Co.: without definite local-

 

ity, Westerfield 6568 (SMU). Chester Co.: Berwyn, Darby
Creek, Stone, 5,5. (RM). Huntingdon Co.: Raystown Branch,

1 mile NE of Entriken Bridge, Wahl, Handwerk, 8_Westlake 5627

 

c
a. Q
, n
. - a
v v o
. . . -. , .
. g .
u
\
1 .
. 3 .
.9 a \
a - I. L o
O V a
A o
. 1
A o v
o .
o c .
_ . n
\ \
J a 5 0
\
~

 

387
(FLAS). Juniata Co.: Port Royal, Fogg 16858 (NY). Adams
Co.: South Piney Mt., Gillis 5112 (MSC). Franklin Co.:

east slope of Tuscarora Mt., Gillis 5110 (MSC); near summit,

 

east slope of Tuscarora Mt.,GGillis 5114 (MSC). Tennessee.

 

Franklin Co.: without definite locality, Eggert, 5,5. (MO).
nggg. Bexar Co.: Rio Medina, Berlandier 2475, 1045 (E,

 

GH); Smith Co.: Tyler, Moore 580 (GH). Virginia. Bedford

 

Co.: without definite locality, Curtiss, 5,5, (GH). Fred-
erick Co.: Winchester, Gillis 5188 (MSC). Floyd Co.:

State highway 8 near Blue Ridge Parkway, Donahue 858 (MSC).
James City Co.: Toano, Menzel, 5,5, (GH). Montgomery Co.:
East Bradford, 45555 8 Wherry 2263 (GH). Smyth Co.: Marion,

Britton 8 Wail, S.n. (F, NY). West Virginia. Greenbrier

 

 

Co.: Blue Bend Recreation Area, Donahue 858_(MSC). Morgan
Co.: near B. 8 0. Railroad and Potomac River, Donahue 855
(MSC). Taylor Co.: Thornton, Gillis 5188 (MSC).

Whereas 1, radicans subsp. radicans iS primarily an
Atlantic Seaboard taxon and subsp. negundo is a land-locked
taxon of the Midwest and the Ohio River Valley, their ranges
come into contact in the Piedmont, the lower Appalachians
and west in Arkansas. Characters separating these taxa are
more subtle than those separating many others; it is there-
fore not always easy to determine with assurance what may
be an extreme form of one of the parents, or what may actu-
ally be a hybrid. It is on the basis of certain key char-
acters, however, that determinations of suspected hybrid

origin have been made. These characters are listed in Tablelfih

388

 

mmOHHHQmQ >HHC¢HQH0CH Ho
msoHanm mHHsnm mHHmC Omumppmom CHHB mHHsnm HCmomobsa mHHCHm

mHHXm CHm>
mHHXm CHw> CH wHHmC mo mumsp CH UmHmHHCwoCoo mHHmC 36H mHme CHm> CH mHHmC mo mHHDH
HDOCHHB mommusm HmHmmmH ngoq m CHHB mommuzm HmHmmmH HmSoH CHHB mommusm HmHmme 96300

opmHHmm Ho mHmHCUCC
Ho .meopoc HCmCHEOHQ
meHHmm Ho OmCoHOC mHmHmmmH CHHB CHOHmE HmHHmmH mHHHCm mHmHmmmH

wmomHHHm >HmmHQO
mmomHHHm mmHOHme 0H HcmHsmmbsa mmHoHHmm HCwommbsq >HmmHQO mmHOHHmm

 

 

I mCmoHpmH .ansm
opcsmmc .mesm mCmoHOmH .H OHHC>I m>HHmpsm mCmoHpmH CoHOCmOooHXOH

 

 

 

 

 

I oozawmz .emmsm
mz<oHo<m .H oz< mz<oHo<m .mmmam mz<oHo<m zomCZmooonoH mHzmm<a Co mmMHo<m<moII.mH mam<H

 

389

Although Fernald intended to differentiate forma
hypomaloca from others by the pilose lower leaf surfaces,
the type Specimen actually represents an intermediate be-
tween Coastal poison-ivy (subsp. radicans) and the midwestern
taxon (subsp. negundo). The pilosity of the lower leaf sur-
faces is interpreted as a local variation. It is perhaps
the result of the prOper "dose" of additive genes. The fruits
have a few scattered hairs and papillae, suggestive of 1,

radicans subSp. radicans, but the leaflets are notched and

 

acuminate in the fashion of subsp. negundo. Since the speci-

men comes from the mountains of West Virginia which form

part of the barrier between these two taxa, it is not un-

likely that intermediate forms Should be found. (Fig. 17.)
Toxicodendron radicans subsp. divaricatum X‘1. radicans
subsp. barkleyi.

The two Mexican subSpecies of poison-ivy are not
abundant in comparison to taxa found farther north. It is
not surprising therefore that apparent intergrades between
the two taxa are scarce. Only a few pOpulations from the
trans-Mexico volcanic belt or an extension of this mountain
region seem to Show traits intermediate between the two par-
ental types. Such pOpulations have been found in Jalisco,
Guerrero, the Distrito Federal, Michoacan, and Oaxaca. (Fig.
18) Characters which have been used to demonstrate the sus-
pected hybrid nature of the pOpulations are few. They are

listed in Table 13.

390

Fig. 18. Distribution of 2 taxa of poison-ivy in
Mexico and location of intergrades between
them.

 

 

No. ”2

MEXICO

ll,

 

‘

)0.

 

 

 

c’OI

10'

tulle-us subsp. div-tional»

Hg!

0
Q
C
b
O
h
C
on
B
-

§
I
:

 

 

 

 

 

392

 

6:09bmHO m>63H6 mHHs9m

HC6o
Im6bza 6660936 H6HH66H 96am:

msoCHH3H6> 0H
660HHQ 6omm9sm H6HH66H 96300

mso9bmHm
>H966C 9o 6oC6ow6bsa
O696HH666 CHHB mHHs9m

m9HmC
6696HH666 CHHS 9o 6Hm9n6HO
666M936 H6HH66H 96am:

m9HmC 660HHQ

6H69Hmo9a >HH6HHC6666
“6oC6om6bsq O696HH666

CHHS 6omm9zm H6HH66H 9630H

6OC69 mo 6C6
C96CH9OC CH >HC0 HC6om6bso
.mzo9bmHO >HH636: mHHC9m

mso9bmHm
666M956 H6Hmm6H 96am:

m6EOCoH9H 6moOH9H6 H6696
O696HHmom CHHB 9o mzo9anO
666H9sm H6HH66H 9630H

 

 

 

HN6Hx9mC .mesm mCmoHpm9 .H

eHae>m a>Heaeaa

 

EOHmoH9m>HO .mezm
mCmoHpm9 C09OC6poonOH

 

 

 

 

 

DZHmmwmmO 0Hmm>I m>HH<H00 mHmIH LO 02< .H>qum<m .mmmbm

m2<UHO<m .H 02< EDH<UHm<>H0 .mmmbm mZ<UHO<m zom0Zm0OUHXOH H20m<0 mo mmwho<m<IUII. mH mqm<H

393

It may be noted from Fig. 18 that these two sub-
Species appear to be sympatric in several regions. The
concept of sympatric subspecies is an anathema to some tax-
onomists or evolutionists, because, by definition, subspe-
cies have distinct geographical ranges. The Mexican pOpu-
lations of poison-ivy, however, bring to the fore some
possible explanations for this phenomenon. One possibility
is that the two subspecies evolved in separate regions, sep-
arated from one another by any of the traditional barriers
to gene flow. Then, in relatively recent times, perhaps
assisted by the activities of man, the two have expanded
their ranges to overlap. Thus, only recently have they be-
come sympatric and share genes. Another possibility is that
they are not strictly sympatric in that they do not occupy
the same communities as do, for example, 855; nigrum and

8. saccharum in eastern North America. There has been no

 

location found where they are growing close to one another.
It is, then, their gross ranges which overlap, but not the
refined patters too small to Show on a standard political
map; that is to say, they are ecologically different, OCCUpy-
ing the same overall range, but not the same communities.

A third possibility, not likely to be valid here,
is that the two subspecies are just now Speciating and exist
relatively close together only as relicts of pOpulations
which existed prior to the erection of a barrier to gene
flow. As interpreted herein, the two subSpecies under con-

sideration here were not derived from one another, but both

394
have their closest relatives farther to the north of Mexico.
A fourth possibility is also unlikely in this case, i.e.,
that either or both of the subspecies is apomictic and un-
able to share genetic information, yet have come to occupy
portions of the same range.

Representative Specimens: Mexico: Guerrero: Bar-
ranca, Mina, Hinton 8488 (MICH, NY); Jalisco; Nevado de
Colima, Beaman 2388 (MSC); Sierra del Halo, lumber road
leaving Colima Highway, 7 miles SSW of Tecalitlan toward

San Isidro, 2000 m., McVaugh 16229 (MICH); Michoacan: east

 

of Morelia between Rio del Salto and La Polvilla,'815g 8
Soderstrom 5888 (MICH, SMU); Nayarit: Mountains 9 miles
south of Compostela, heavily forested stream valley in oak
zone, 1000-1200 m., McVaugh 16484 (MICH); Oaxaca: Reyes,
Nelson 1188 (CH, US).
* * * * *
Toxicodendron radicans subsp. radicans X subsp. pubens
and 1, radicans subSp. negundo X subSp. pubens.
Considerable taxonomic difficulties are encountered

in the region of overlap where Toxicodendron radicans subsp.
pubens, subSp. radicans, and subSp. negundo intergrade. The
number of Specimens which combine characters of these taxa
is numerous, but not enough specimens are available from the
most Significant area, i.e., the second and third tiers of
counties west of the Mississippi River in northern Louisiana,
southern Missouri, and the intervening region of Arkansas.

Perhaps it is due to active divergence among the taxa. More

395
likely this large-scale variation is attributable to three-
way hybridization among the taxa.

This region of the foothills of the Ozarks repre-
sents the western edge of the range of subSp. radicans, the
northern edge of the range of subSp. pubens, and the south-
ern edge of the range of subsp. negundo. No conclusions
other than the probability of hybridization among these sub-
Species seem to be warranted at this time. Further field
work in this area is desirable for resolution of the true
nature of these pOpulations. It is possible that the region
in question is the center of diversity of a rapidly evolving
Species, but more likely a region in which the three subspe-
cies have been invading for a considerable period of time,
perhaps since the Pleistocene.

For the time being, Specimens from this region have
been classified by the combination of several key characters.
Pubescent fruits and tufts of hairs in the vein axils on
the lower surface of the leaves have been interpreted as
indicative of the presence of genes of subSp. radicans. The
presence of dense, erect hairs in the interveinal area on
the lower surface of the leaves is interpreted as indicative
of the presence of genes of subsp. pubens. The presence of
appressed hairs on the lower leaflet surface is indicative
of the presence of genes of subsp. negundo. These clusters
of characters and the taxa they represent are Shown in

Table 14.

 

6
O/
3

6H6>o
>Hpmo9b 6H6Hm66H

6H6996m
9o 66C6H6C
HH6 mH6Hmm6H

H6696
6C6 666669aam
66xHE 66C66m6bsm

66:66
666m936 H66H
9636H Co 69H6$
HC6mn6

9:2 .8 BEH

636966HO mHH39m

6Hm>o
>H6669b 6H6Hmm6H

69HHC6nsm

96 69HHC6
6866 “66:6H6C
6H6Hmm6H 656m

H6696
6C8 666669556
66xHE 66C66m663m

66C66
666M936 H66H
9636H C6 m9Hm:

HC6m69Q
69H6C Ho mHHCH

HC6666QCQ
HmCa6Eom mHHz9m

6H6>o
>Hpmo9o 6H6Hmm6H

6H6996m
9o 66C6H6C
HH6 6H6Hm66H

H6696 66C6om6bsm
66C66

6660936 mm6H
9636H Co 69Hmm
HC6666

meaee Ho aHHaH

mso9bmHO mHHz9m

6H6>o 6H6Hmm6w

66C6H6C
>HHmoE 6H6HH66H

666669566
66C6666nsm
6696HH666
666H9sm H66H

9636H C6 m9H6$

HC6mnm

aaHee Ho mHHaH.

6:6966HO mHHs9m

6H6H666C6H
6H 6H6>o 6H6HH66H

69HHC6
>HHmoE 6H6HH66H

666669556
66C6666nsm
6696HHm6m

666M936 mm6H

9636H Co 69H6$
HC6m69a mHme CH6>

ea meHae Ho aHLeH

HC6666CCQ mHHs9m

 

66C3m6C x
6C6bsm .mesm

6C6nsm x
mC66H669 .anzm

 

6C6nsm .ansm

66C5m6C .mesw

 

mCm6H669 .ansm

 

 

 

ZOHDmm 2<H¥m<NO mmH 2H

mz<UHO<m 20m0Zm0OUHXOH mo mZOHH<H0000 mDOHm<> 2H 00Hzmmmmmmm mmMHmDHU mMHU<m<$UII.vH mHm<H

397

1. radicans subSp. radicans X,1. radicans subsp. pubens

 

Representative Specimens: UNITED STATES. Arkansas.

 

Conway Co.: Morrillton, Petit Jean State Park, Demaree
22801, 36876 (GH, MO, NY, SMU). Craighead Co.: Lake City,
Demaree 7218 (GH, SMU, TEX, WIS). Garland Co.: Hot Springs,

 

Demaree 22056 (A, MIN, MO, NY); Hot Springs, Palmer 23098

 

(A, MIN). Hempstead Co.: Fulton, §2§2.§§&Z.(A1 US). .Mont-
homery Co.: Norman, Demaree 8558 (GH, MO, SMU). Pope co.:
Hector, Merrill 5_2_ (A, MIN, NY); Nogo, Merrill 13.2. (MIN, Mo).
Pulaski Co.: Little Rock, Gillis 8115 (MSC). Stone Co.:
Ozark National Forest, rocky creek above Blanchard Springs,

Allison, Demaree 52504 (MSC). Louisiana. New Orleans,

 

Cocks 51 (A). Tangipahoa Parish, 3 miles north of Maurepas,

Seibert 703 (MO); Hammond, 13 miles east of city, Rose-

 

Innes 8 Warnock 721 (GH, TEX). Mississippi. DeSoto Co.:

 

3 miles north of Nesbitt, Gillis 5906 (MSC); Jackson Co.:
Ocean Springs, Demaree 28090 (SMU). Oklahoma. LeFlore Co.:

 

 

Stapp, creek banks, Demaree 15732 (MO, NY, SMU).

1, radicans subsp. pubens X‘1. radicans subsp. negundo
Representative Specimens: UNITED STATES. Arkansas.

 

Benton Co.: Twin Mountain, Demaree 4518 (F, MO, WIS).

Fulton Co.: Salem, Demaree 85885 (SMU). Washington Co.:
Lake Weddington Recreation Area, Donahue 884 (MSC). 1585.
Pottawattamie Co.: Council Bluffs, Demaree 88555 (MO, SMU).
Kansas. Pottawatomie Co.: St. George, Kellerman, 515. (KSC).

Sumner Co.: Geuda, Gassler, 5,5, (MIN). Missouri. Oregon

—

398
Co.: Pine, on county road J., Donahue 219 (MSC). Ozark
Co.: Tecumseh, Palmer 32912 (A, MO, NY).

Uncertain or Excluded Species

Rhus-Toxicodendron radicans Marshall, Arbust. Am.
p. 131. 1785. This name probably refers to Toxicodendron

radicans, but there is neither citation of previous published

 

names nor of any Specimens. It is therefore a nomen nudum.

Rhus-Toxicodendron toxicodendrum Marshall, Arbust.
Am. p. 131. 1785. This name probably refers to Toxicoden-
dron toxicarium, but there is neither citation of previous
published names nor of any Specimens. It is therefore a
nomen nudum.

Toxicodendron vulgare Mill. Gard. Dict. ed 8, Art.
Toxicodendron, No. l. 1768. The description can apply to
a number of taxa. No Specimen to represent this name can
be found.

Toxicodendron pubescens Mill. Gard. Dict. ed. 8,
Art. Toxicodendron, No. 2. 1768. The description can apply
to a number of taxa. No Specimen to represent this name can
be found.

Toxicodendron crenatum Mill. Gard. Dict. ed 8, Art.
Toxicodendron, No. 5. 1768. The description can apply to
a number of taxa. No Specimen to represent this name can
be found.

Toxicodendron volubile (as "volubilis") Mill. Gard.

Dict. ed 8, Art. Toxicodendron, No. 6. 1768. Rhus radicans

399
var. volubile (Mill.) DC. Prodr. p. 69. 1825. .52!§.£2£l‘
codendron subsp. vulgaris var. volubilis Engler in DC. Monog.
Phaner. 4: 394. 1883. 1, radicans var. volubile Farwell,
Amer. Midl. Naturalist 12: 125. 1930. All uses of the epi-
thet volubilis refer to Miller or DC, the former of whom
equates his species to Toxicodendron amplexicaule . . . of
Dillenius. None of these plants can be typified. They prob-
ably refer to some kind of 1, radicans, but which subspecies
is uncertain. Engler cites "85, Toxicodendron L. Spec. 381
sensu stricto" but describes it as "caule scandente, vix
radicante" which does not accurately describe 8585 Toxicoden-
5555 sensu stricto. He further cites a large number of Speci-
mens, some at Kew and some presumably then at Berlin, but
these Specimens represent almost every taxon in the poison-
ivy complex.

Toxicodendron serratum Mill. Gard. Dict. ed 8, Art.
Toxicodendron, No. 7. 1768. No Specimen to represent this
name can be found.

Toxicodendron arborescens Mill. Gard. Dict. ed 8,
Art. Toxicodendron, No. 9. 1768. The description can fit
most taxa in the poison-ivy complex. No Specimen to repre-
sent this name can be found.

Rhus pgmilum Pursh, F1. No. Am. 1: 205. 1814.

(= 8. michauxii Sarg.).
Rhus variifolium DC. Prodr. 2: 73. 1825. Nomen

nudum.

400
Rhus bahamensis D. Don. 1832. A General System of
Gardening and Botany, p. 72. London. 1832. Nomen nudum.

This is possibly Toxicodendron radicans subsp. radicans,

 

the only member of the Rhus complex known from the Bahamas.

Rhus quercifolia Goeppert, Tert. F1. Schossnitz.

1855 (= Monopleura in Aceraceae).

ECOLOGICAL LIFE HISTORY

The bulk of scattered information about the natural
history of Toxicodendron sect. Toxicodendron is here summar-
ized as an ecological life history. The pattern followed
is essentially a synthesis of the pattern of the Biological
Flora of the British Isles format as published in the Journal
of Ecology and the outline suggested by Pelton (1951). In-
asmuch as most of the material relating to anatomy, morph-
ology, hybridization, and taxonomy is included elsewhere,
such material is not repeated; only additional tOpics will
be discussed. Excluded from discussion is sect. Simplici-
15115 which includes only 1, borneense about which virtually

no data are available.
Morphological Limits

Although the total length of time to elapse from
the time of seed germination until the time a young plant
can produce flowers is not known for all taxa, plantation
studies of Toxicodendron radicans subsp. negundo have shown
that three years were required for plants to progress from
seed to flowering stage. The longevity of the species is
unknown. However, trunks of 1, radicans subsp. negundo are
known up to 12 cm across, bearing 38 growth rings. The
widely distributed collection of George Cooley (No. 8488)

401

402

of 1, radicans subsp. radicans from Sanibel Island, Florida,

 

represents a tree of some five meters in height. This same
individual tree is reported to have had a trunk diameter of
9 cm dbh, and one of its branches 13 dm long bearing 6130

flowers! Trunks of even larger 1, lobatum individuals have

been reported in California.
Altitudinal Limits

Data on elevation extremes for each species and sub-
Species are summarized in Table 15. The data come primarily
from herbarium information. They are only approximate. Ex-
cept for relatively high elevations, most collectors do not
often record elevations with their collection data. A few
additional remarks on altitudinal distribution from my own
observations are perhaps in order.

In the Great Smoky Mountains Toxicodendron radicans
subsp. radicans occurs at lower elevations but dr0ps out of
communities abruptly at 1070 m. Because of its abundance
at elevations below 1070 m, the apparent scarcity above that
elevation prompted a search for the plant at higher eleva-
tions, a search that proved fruitless. Toxicodendron radicans
subSpp. negundo, pubens, and verrucosum might grow at higher
elevations if the requirements of soil, moisture, and com-
munity preference were satisfied there also. That is to say,
altitude alone probably does not limit their distribution.
Indeed, any upper altitudinal limit for these taxa probably

is not reached anywhere within their natural ranges.

 

oaoeoe oomH oe Aev Ho>oH

66m

66Hx62 362 CH m96H6E OOmN

6H C6O69O CH 696H6E omHH 6H mOmCmO CH H6>6H
mCOHH6>6H6
696H6E OOmH 6H H6>6H
696H6E OOCm I
696H6E OOOH 6H H6>6H

403

66m
360
66m

omHH

66m

696H6E OOON 6H mC6HHm>6H6 6Hm96662

m96H6E OOmN 6H mCOHHm>6H6 6Hm96UoE

oaopoe ooeH I

m96H6E 05m 6H mCoHH6>6H6
mCOHHm>6H6

A.mm ..HE 696966639v 696H6E who 6H mCOHH6>6H6

H.6HE >onmv m96H6E OBOH 6H H6>6H

00m
360
360
660

66m

 

eoonaHr

 

6H6HC6H9o

 

HsoHeaoo

 

ECH66H96>H6

 

EDHdew

 

ESwOUDHHw>

6C6nsm
66C5m6C

 

ow6H669

 

 

 

 

 

 

 

 

 

 

 

 

 

EsmOOOC .H
HHoeooo99 .H
ECH966onH .H
sapwooHH.H

.ombsm 6C66H669 IH
.anzm 6C66H669 .H
.ansm mCm6H669 .H
.an56 6C66H669 aH
.mesm mCm6H669 .H
.mesm 6C66H669 .H
.an36 mCm6H669 .H
.ansm mCm6H669 .H
.mesm 6C66H669 .H

 

 

 

 

20m02000OHXOH .Homm 20m0zw0OUHXOH

mo mmHommm mow 2<0m 20HH<>mHmII.mH mHm<H

404

Toxicodendron radicans subsp. hiSpidum appears to
be a true montane plant, having been collected nowhere at
elevations less than 1150 meters. In Taiwan, where its
distribution is better known than in China proper, it is
restricted to the central mountains, and only at the higher
elevations. In mainland China, it has been found at eleva-
tions estimated at 3450 meters at a latitude of about 250 N.
It is perhaps these higher elevations which make the growing
conditions more temperate at the lower latitudes.

Toxicodendron radicans subsp. orientale is found at
sea level in Hokkaido, but is rare at low elevations farther
south, as in the Tokyo valley for example. In southern
Honshu, it is more likely to be found in the mountains at
moderate elevations. This taxon, however, does appear to
have an altitudinal limit. In the Hakone and Nikko regions
(35 - 370 N.) of central Honshu, where the plant has been
collected most frequently, it is abundant at elevations of
1200 - 1300 m., but totally absent at 1800 m. In other
words, for these taxa, there appears to be a relationship
between latitude and elevation maximum such that the plants
will reach higher elevations farther south in their respec-
tive ranges.

The relatively low Allegheny and Blue Ridge Moun-
tains appear to act to some degree as a quasi-barrier to
gene exchange, since Toxicodendron radicans subsp. negundo
occurs west of these mountains whereas subsp. radicans grows

east of them with some putative introgression in the

405
mountains. Where 1, rydbergii occurs in the Appalachians,
it is found at the summits of mountains. Whether it is a
relict there or is the result of recent invasion is diffi-
cult to assess. Where it has been found most frequently
at higher elevations has been along road cuts, and it may
well be expanding its range couthward as habitats Open up
which satisfy the correct latitude-altitude coefficient.

Toxicodendron rydbergii probably reaches a true ele-
vation limit that is caused by a complex of factors such as
length of growing season, frequency of late or early frost,
etc., factors which are associated with altitude. It is
found no higher than 1150 meters in Oregon and Washington,
but grows at elevations up to 2500 meters in New Mexico.

An attempt was made to calculate the coefficient of
latitude-elevation maximum for Toxicodendron rydbergii, the
only member of the Species complex for which sufficient data
are available to make such calculation. The determinations
were based on highest known elevations for this Species in
Oregon and New Mexico, and checked by collections at high
elevations from intermediate localities. The relationship
between latitude and maximum elevation is a linear one whose
formula is:

Kf = L (Ef + 12,000)

where Kf = coefficient of latitude and maximum elevation

when elevation is measured in feet.

L latitude in degrees north

I]

Ef elevation in feet. Then, substituting for

406
latitude and elevations values in Oregon and New Mexico
and elsewhere:
Kf = 7.2 x 105
Where elevation is measured in meters, the following holds:
Km = coefficient of latitude and maximum elevation
when the elevation is measured in meters.
Em = elevation in meters
Then, again by substitution
Km = 2.18 x 105.
Daubenmire (1954) made a graph of altitude of timber-
lines in North America plotted against latitude and found a
curve which was essentially linear for that part of the
world between 35 - 700. At more southerly latitudes, the
curve changed markedly. For the range of latitudes involved
in the distribution of Toxicodendron rydbergii (32 - 520 N.),
the computing of a constant for maximum elevation should
be more meaningful than for a spread from O - 700 N. Lat.
It should follow a straight-line relationship, because the
latitude Span is all temperate. Furthermore, one is here
concerned with the distribution of only one Species, there-
fore involving a Single niche, rather than that of an entire
life form, i.e., forest trees. Daubenmire (1515,) concludes
that
. . . along the main axis of the North American Cordil-
lera, the elevation of alpine timerline rises steadily

at a rate of about 110 m per degree of latitude between
60 and 30°, then declines very gradually to the equator.

407

Considering that his data include a variety of
tree Species and deal with highest elevations of species
that live far higher than poison-ivy, it is interesting
that the altitudinal maximum for Toxicodendron rydbergii
also rises steadily at a rate of about 109 m per degree!

This close approximation to Daubenmire's figures
suggests a biological value of some significance. The
altitudinal limit of Species distributions and its corre-
lation with latitude might be explored for other plants
and also for animals to see if the change of lO9-llOm/degree

is approached.
Fire

The plants of this complex generally occur in dis-
turbed habitats which are likely to be disturbed repeatedly.
The activities of man are important in creating this dis-
turbance, but fire is also a major source of disturbance,
especially for the forests in which Toxicodendron toxicarium
is found. Toxicodendron radicans, 1. lobatum, and 1, rygr
bergii are plants which respond favorably to disturbance
and are likely to become more abundant following fire, but
I have no data to substantiate this supposition. For 1.

toxicarium, on the other hand, there are good data to sup-

 

port this claim. In a stand of oak-pine in Lee County,
Alabama, Auburn University Forestry Department maintains
control-burn plots in which 1, toxicarium is native. A set

of ten meter square random quadrats were examined in each

408
of two stands--both burned and unburned. The number of
shoots per square meter was recorded for each. For ten
quadrats in the unburned area, the average was eight
shoots/m2; for ten quadrats in the burned area, the aver-
age was 26 soots/m2. At another site in Peach County,
Georgia, where the most abundant Eastern poison-oak was
found Of any Site investigated, there was ample evidence
of burning a few years before (charred stumps, charcoal in
the upper soil horizons, etc.). Fire, then, appears to en-

hance the abundance of 1, toxicarium.
Biological Interactions

There are a number of interactions of poison-ivy
and the poison-oaks with other plants and with animals. A
number of these have been observed first-hand in the field,
but most of the following references to such interactions
have been derived from the literature. Where references
have been consulted, attempts have been made to establish
the taxon Of Toxicodendron involved. Precision is not pos-
sible in most instances because the writers have failed to
file a voucher Specimen of the plant in question in any
museum or herbarium, and/or the scientific name which may
have been used has been a source of confusion, i.e., 8585
toxicodendron.

Among the parasitic fungi, two were found by Rhoads
(1918) as examples of wound parasitism: Polyporus gilvus

and Irpex lacteus. A rust genus, Pileolaria, is found

409
primarily on Toxicodendron Species according to Dr. Joe
Hennen (personal communication), having been found infecting
1. lobatum and a number of subspecies of 1, radicans in
herbarium material. Arthur (1934) claimed that Pileolaria

l6
toxicodendri infects 1, radicans throughout its range.

 

Other fungal infestations are summarized in Table 16.
Insects and other arthrOpods work on Toxicodendron
in a variety of ways. One of the chief pollinators of the
poison-ivies in Michigan, the Carolinas, New Jersey, and
the prairie states is the honey-bee (4515 mellifera), which
makes a non-toxic honey from its nectar. Presumably the
honey-bee is a pollinating vector elsewhere, but has not
been Specifically recorded as such. Other visitors to the
flowers that appear to transport pollen included other
HymenOptera and several ColeOptera. A number of insect
larvae have been observed eating the leaves, and indeed
several have been reared to adult stages in the laboratory
on leaves solely of this genus. Mr. Murray Hanna reared a
nymph of ClastrOptera obtusa on 1, radicans subSp. negundo
leaves in Okemos, Michigan. Dr. Roland L. Fischer reared
larvae of Argg _p. to adult sawflies on the same taxon from
Gull Lake, Michigan. From field observation, he noted dam-
age to individual plants by this larva. Stripped leaflets

on which only midribs remain are deposited as vouchers in

 

16= 8, brevipes according to Hennen (personal com-
munication).

410

TABLE 16.--FUNGI PARASITIC ON TOXICODENDRON SPP.

 

 

‘~V~

 

Polyporaceae Polyporus gilvus (Schw.) Fr. Rhoads, 1918

 

Irpex lacteus Fr. Rhoads, 1918
*Pileolaria brevipes Berk. Arthur, 1934 and
and Rav. Dr. Joe Hennen,

personal
communication

leindrOSporium toxicodendri Schwarze, 1917
(Curtis)Dearness

 

*Known to parasitize various North American sub-
Species of 1. radicans, 1.10batum, and the Asiatic sub-
species of 1. radicans, subSp. orientale and subsp.
hispidum.

 

 

 

411
the Michigan State University Herbarium (MSC). LeConte
and Horn (1883) raised a dermestid beetle of the genus
Aspectus from a tumor on a stem of poison-ivy. Mites were

found in leaf galls on specimens of living 1. radicans

 

subSp. divaricatum brought from Mexico City, and determined
by Mr. Robert Fouts of the Plant Quarantine Division of the
USDA to be Aculus toxic0phagus (Ewing).

Criddle (1927) observed that Epipaschia zelleri
larvae made the leaves curl lengthwise to form tunnels. He
claimed this larva was the most destructive of all insects
to poison-ivy. Larvae of Lithocelletis guttifinitella tun-
nel into the petioles and leaves, making white circular
cysts, notable eSpecially as autumn coloration occurs. The
pupae then overwinter in the leaves. From Specimens depos-
ited in the Michigan State University Entomology Museum,
Mr. Julian Donahue has kindly furnished data to the effect
that larvae of all members of the subfamily Eutelliinae
(family Noctuidae) are known to feed on poison-ivy. Table
17 summarizes the known effects of insects on poison-ivy
species.

Some birds eat the fruits of Toxicodendron radicans
occasionally, and others eat them preferentially (Martin,

17

.g§.a1., 1951). It is noted that various flickers, which

 

17Although the species of Toxicodendron are not enu-
merated for each bird in the list, one can assume that refer-
ences to the West Coast must be to 1. lobatum, to Arizona,
1. rydbergii, etc. There are certain references which cannot
be pinned down precisely, e.g., New England.

 

 

412

mama .yamm
COHPmUfiCDEEOo

Hmcomemo .pmmcmwmho meHm

coevmofic36600
Hmcomnmo .Ummzmfimho meHm

coma .mmnaom

coma .mmngom
oooa .mmnuom

mmoa .mxofi:
mama .meA:

mama .mxoflm
emoa .manaano

coma .Hflmnefix “mmoa .cmmH02
«mod .mmnnom

«mod .mmnuom
moofi .HHmQEH¥ .amoH .mmnnom

 

mHocH mHDmCMme

 

. m mHmPQOfimmAm

 

A.>:mmmv wmeHmESL mmH<H

 

mHmnEmno maawflHOMflocH mflhmafiommw

 

mamEmHo maampugamnppsm maymaamoozpflq
.csmam maamnHOMHOLH Hanofiyamz

 

 

meama maflmmoumem mwomoomu
A.Emaov magmpsmhm mseflmdmma

 

 

mflampmumqwm .m
.wHO HHmHHwN mficommofimmm

 

 

Ammosqv mzcowcp mswmmmo mmvopmmq

 

meamz mflammmn mmwNfipmHmEH
mpohu mEHHHmcoasm mflamuzm
Awmcwsuv xwupmazoo mmpommma

 

 

mmufiH>EopflomU
mmpfiH>EopHomo

<mMHmHD

mmpwmh<

¢mMHmOZME>I

mmpwwumaflomum
mmnfinHSUAPsz
wmpfloHHpHOH
mmpwamh>m

mmpwcmmo>q

mmpwspooz

<mmHmOQHmmq

 

 

 

.mam zomoszOUHXOH ZO DZHDmmm mQOmomIHm<nl.bH wqm<H

413

 

.cflnu

 

.nflnfl

ofloa .mcmu new >manopmam

ovoa

Hmcomymo

 

.afln

ca

 

.n

-H

n

0H

 

.efinn

 

.pflnfi

.>mmmm2 new Hmmm

soapmoHCJEEoo

.Ummcmwmuo meHm
mpmv Enema: >moaoEOpcm 3m:

mmmfi .cno: new mpcoomu .mmoa .mxofi:

 

.mana
.pflnfi

Hmoa .Hmcopflm pcm .cmpao: .cmpaoz

 

.omq mHHHEwmcoo manocpzoo>pwm

 

.Qo: ancmpooflxop mDEwcmnpoo>I

 

A>mmv mapmaspc: mzaamnmpmz

 

.ecoflm macammm .x

 

.xooz magmomo mayonma>x

 

.omq mzwmomsw manocwchopo>uo

 

mcmowmoam mswwvapomq

 

A>mmv masome mammoa>pm<

 

mafimsm mwonh msHonpnmoNpflm

 

A.Lmam2v mcmpflup w>£omhn>£omm

 

swam: mspflomwc mzpomom<

 

mpcoomq mpmpcmpwo mfimeonwo
Hmccmaa>w mafiuumamm> mowmmm

 

 

spasm “mean mmmcaoaasca

mmefip>floom
mmcupsaoom
mmpficoaasonso
mmnwv>aoom
mmufip>aoom
mmprOfiazonso
mmpwo>nEmHmo

mmpflo>oEmHmo

omvflQH
mmpfiHmEom>H£o

mmpflpmmenma

mmuficpcoaoamz

memeQmHmuw

<mMHmOmAOO

 

 

 

cascapcoo--.ea mqm<e

414

.mvoon Hmfinmm co maamwm

.vaHQ >>Hucomfioo may on wommcfl m>wpozupmmp pmoe 0:9 mo 09 mEOm >3 anwpwmcoom

.mw>mwa co mm>HmAH

 

cowpmoHCDEEoo
Hmcomnmd .mpsom pnmnom

mNOH .pommHH
cowpmoficsesoo

Hmcomuwa .mcch >mHHDE
mama .HfimZQE

cowpmowc:EEoo
Hmcomnwo .pmmcmflmno meHm

 

 

 

 

 

 

 

 

Amcflamv mammnooofixop.MMflmmfla mmpfl>5Q0HHm
mmHHE
pommfle macs“ mflmcfiaoumo mmuflnfica<
mmsuno mnmpqonpmmao mmpfinmvoonvmmao
.mm mayonaamw< mmafinflammfia
<mmhmozom
amsmpam: mpmaamoo mmnmno mmmom

 

 

 

smzcfipcou--.ea mgm<e

415
are fond of ants despite their formic acid content, are
partial to fruits of Toxicodendron, its fruits comprising
up to 25% of their diet. Barrows (1895) suggested that
dissemination of the species was accomplished largely by
crows. Certainly one may see birds of many kinds as being
responsible for the seeds being dropped at the base of
trees, telephone poles, and fencerows. There is some indi-
cation from anomalous germination tests that I obtained

with disseminules of Toxicodendron Species, that passage

 

through an animal's digestive tract might be needed to
break down the waxy and bony layers over the seed to facil-
itate germination.

Of mammals and their eating of Toxicodendrons, little
is known. Although bears, rabbits, and mule deer have been
observed browsing the plants (Martin, g1_a1.,.1gg.‘91§.),
only the pocket mouse in California has been seen eating
seed enough to comprise as much as 25% of the total diet.
More than once I have heard of cases of dermatitis in human
beings attributed to the Skinning of rabbits or squirrels
whose stomach contents were largely that of poison-ivy leaves
and seeds. No experiments have yet been carried out to test
this hypothesis, however. On the other hand, members of
the poison-ivy complex are used by small animals in a way
other than as food. In localities where ground cover iS
dense with Toxicodendron radicans, 1. rydbergii, or 1. 19b;-
‘Egm, the cover and protection they afford are important

contributors to the life of rodents which appear to be

416

unaffected by the toxicity of the plants.
Periodicity

The flowering times, length of ripening periods,
and earliest date of ripened fruits differ among Toxicoden-
dron rydbergii, 1. lobatum, 1. toxicarium, and the subspecies
of 1. radicans. Among the latter subspecies, there is some
correlation of these periods with latitude. Dates discussed
here are mean dates of maximum flowering and do not account
completely for some precocious and some serotinous flowering.
Whether late flowers are a response to photOperiod is not
known. Early flowering never preceeds the Opening of new
leaves.

Time of flowering for Toxicodendron radicans subsp.
radicans begins at the southern tip of Florida at some time
between the end of January and the middle of February. The
flowering date appears progressively later toward the north,
arriving in late March and early April in northern Florida,
mid-May in the Carolinas, early June in the middle Atlantic
States, and the first week in July at the northern limit of
its range in southern Maine. Flowering periods for 1. gyd-
bergii begin the second week in May at the southern end of
its range, and progress to the end of June and early July
at its northern limit. For 1. radicans subSp. negundo, the
Span is from mid-April to the latter part of June. Figs.

19 and 20 Show the flowering date progression for these taxa.

417

Fig. 19. Flowering periodicity of Toxicodendron
rydbergii. Numerals refer to the week
of the month in question.

 

 

 

 

M
w
t i
«y ) Q
. F.. ‘
) g I
‘ - o a ‘ i a” :' .
‘ : f H“
t 4 4
l ‘ l
‘ t w 1‘ l , , V - m“.
‘ J t "w" '
\
l 7 J 7 7 r " . \ u
‘i W 7 ,7 W, Li“; ~ , )7 l t - ,, ;
:cALz ‘ ‘ 1&4
a
no no no loo noon nu." “ t
\
0 no so: “1 .00 AM 095 I” llLfll‘l'C‘S ‘
l t
)
LAMIY AllWYNhL [QUAL'AIIA PM‘CTION ‘ ‘
o l
.... ,7 1 l
,, , 1 L | g 1
120 no

  
   

  
    

y—l
................ ,._,. H.u;x‘(_ ,

' (*1. ‘
May— 1 Jun
w \/ 'r

   

 

 

 

 

at 9V LO um wot

 

m BASE MP SERIES
nan-nu o: m
M mus-v" or arm
I‘m :- Lyn-9 Imo-

 

Mum—VAL...“
.4...

w. v.. ..-...

 

 

 

 

 

 

 

419

Fig. 20. Flowering periodicity of Toxicodendron
radicans subsp. negundo. Numerals refer
to the week of the month in question.

 

 

 

NORTH AMERICA

No. |02

 

 

 

r

 

” ,1 lg! "_lvn’. U .L‘

Ail; l ‘Cvl'fflfi

 

 

 

 

 

 

 

GOODS BASE MAP SERIES
meARYAINY ’3‘ (406-£va
Yul 1.1.1.193!" -_\{ at if!)
.‘NOV n LENA-O SOHO-

421

In Japan, Toxicodendron radicans subsp. orientale
flowers in late May in Kyushu and Shikoku, between the
second week in June in southern Honshu to the first week
in July for northern Honshu. In Hokkaido, the plant flowers
during the last two weeks in June except for the Rishiri
Islands, where it flowers early in July. In the Kuriles
and Sakhalin, it flowers between the last week in June and

the first week in August. Subspecies hispidum appears to

 

flower in Taiwan at the end of April or in early May, and
on the mainland of China during the last two weeks of May.
Within Toxicodendron radicans subsp. divaricatum
there appear to be two periods of flowering maxima, one in
April and May, and one in early to mid-August or even early
September. There seems to be no geographical nor photo-
periodic pattern to the flowering cycle. Even within a
given area (e.g., Michoacan)flowering has been reported in
early April, early May, early June, and late August from
different pOpulationS. Further study might reveal correla1
tions with rainfall or some other ecological factor, but
for now, these differences are attributed only to local
pOpulation variation. Subspecies barkleyi Shows a similar
Spread of flowering dates, except that any secondary flower-
ing period is earlier than that for subSp. divaricatum
(early July rather than August or September). One Jalisco
population, however, is recorded as flowering abundantly in

November.

422

Flowering periods for Toxicodendron lobatum shift
northward with the season, beginning in late March in Baja.
California, commencing later in northern California and
Oregon, and finishing the last week of May to the second
week in June in Washington and British Columbia (see Fig.
21). For 1. toxicarium, on the other hand, a simple geo-
graphic pattern can be established. Fig. 21 also shows the
flowering dates for this species.

Length of time required for fruits to ripen also

differs with latitude for several taxa. For others, there
are too few data to determine these figures precisely.
Fruits are judged to be ripe when the exocarp is no longer
green, no longer shrivels in the drying of herbarium Speci-
mens, and becomes brittle, separating from the mesocarp.
At the southern tip of Florida, fruits ripen in Toxicoden-
dron radicans subsp. radicans in three months, but in north-
ern Florida, this ripening extends to four months. The
ripening then shortens with increase in latitude to only
two months in southern New England at the northern end of
the range. Between central Florida and Maine, this average
ripening difference is Shortened by roughly four days per
degree of latitude (see Fig. 22).

The Span of latitude for subsp. negundo is smaller

than that of subsp. radicans, so there is not so noticeable

 

a difference in ripening time with changes in latitude.
Throughout the range, except for upper New York (two months)

and northern Texas (four months), the ripening time is about

423

Fig. 21. Flowering periodicity of Toxicodendron
lobatum and 1. toxicarium. Numerals refer
to the week of the month in question.

 

 

NORTH AMERICA

 

 

I"
I

    

 

Is. I
I I”)
I 4MCIy—2Jun -- f,
.7. 3May—IJun
2Apf- 3MOY N “)3:
3Mar—2Apr. LN
”W’

.....

I [lobatum

MALI

no 400 I!) no man an.“

o no oov can too not: at» not: ”manna

LAMIT AINUTNM. (mu-nu PNICTION

fly I

HO

 

o

 

    

 

'
$14. fi‘flflh‘ufi 4 AP"

 

1.“? 2— Qgig'hh‘\ IApr

I‘\‘f" B
\ ( ‘ VAL)
T. toxicarium -"‘ ,é “Nu?-

 

30 var Lam-mun:

 

”R

I
\«1

 

 

GOODS BASE MAP SERIES
mum-nu or “W
“may a: Lulu no
A panama:

Mann—Aw
wan. -..” ...

Fig. 22.

425

Flowering and fruiting periodicity of Toxi-
codendron radicans subsp. radicans. Numerals
refer to the week of the month in question.
Shading refers to the length of time required
for fruit ripening.

 

 

 

NORTH AMERICA

 

 
    
  
   
 
  
   
    
  

..._, 7,, '

   
 
    
  

,/
4Jun—1JUIV

— 3 Jun
4MCI)’—IJun

  

I |.

v
4
\,
. »_, .4 ‘l
‘~__‘fr "v
I
_ .

:5

’3—4Moy

2--3 May

I—2May
2—3Apr

3 Mar —IApr

4 Dec «- 4 Feb
I ,x/fZTM‘rM

6’30 3:) MIR/kg

 

IIIIIIIIII 2 ms T0 mm

m 2.5 MONTHS TO arm

% 4 mourns TO RIPEN

712.5 ' t .
3 mourns TO RIPEN
SCALE
"1 62—2 H’s”; ‘Il'll V-.t"l
I I O i v 4 I'.“.U(I.(9$ J I
LAMBERY AZIMUTHAL EQUAL-AREA PROJECTION /

 

 

 

GOODE BASE MAP SERIES human,”
O‘DADthq' 0‘ 5;;345..~, y k“ 1”” '
m5 jH'vinng 5‘ 5“, A60

*"“" U IEWABC, spa).

427
three months. Although extensive data are lacking for subsp.
orientale, it appears that two months are required for rip-
ening in Honshu, but only one and a half months in Hokkaido
and north. No fruiting Specimens from Kyushu or Shikoku
have been seen.

Ripening of fruits of Toxicodendron radicans subsp.
divaricatum requires about three months. The ripening time
lengthens to four months in Baja California, and shrinks to
two months in Sonora. Because of the divergence of flower-
ing times, precise data for more than two or three locali-
ties were not obtainable. With what little data could be
trusted from which to draw inferences about ripening time
in subSp. barkleyi, it appears to require the longest period
of all--up to four and a half to five months. Again from
sparse data, one may estimate ripening time for subsp.
eximium at one month, for subSp. pubens at three months,

and for subSp. hispidum at two months.

 

Toxicodendron toxicarium requires from Six to nine
weeks for ripening of fruit, with no particular geographical
pattern to the variations. Toxicodendron lbbatum requires
eight to ten weeks for ripening with the shorter period in
northern California north to Canada and the longer period
in central California south to Baja California.

Toxicodendron nodosum has been collected more times
in flowering stage than in fruiting condition. Flowering
Specimens have been collected from Java in July, August,

and September, from Malaya in September (Larut) and December

428
(Perak), and from Sarawak in June. Reliable ripening dates
are not available, but what little data there are suggest
that it takes about two months for the seeds to ripen.

In summary, one may say that the fruits of most
members of the poison-ivy complex are ripe by the end of
August or first part of September with three exceptions:
(l) the early flowering (December to February) pOpulations
will be ripe by early summer; (2) the serotinous flowering
pOpulationS will not produce ripe fruit until late in the
calendar year if at all; and (3) Toxicodendron nodosum has
different flowering and fruiting periods in different re-

gions.
Reproductive Biology

Most of the reproduction of poison-ivy and its
allies is vegetative, enabling a single disseminule to grow
to cover up to an acre of land. New pOpulations, of course,
are begun by seed. Cuttings do not fare well and are very
susceptible to damping off. Only three cuttings grown in
the laboratory (out of Sixty-three attempted) were still
alive three months later. None survived a winter out of
doors in experimental plots in southern Michigan. These
cuttings represented seven taxa (Toxicodendron lobatum, 1.
rydbergii, 1. toxicarium, 1. radicans subspp. radicans,
negundo, divaricatum, and verrucosum).

Germination is epigeal. The cotyledons are still

borne within the tough endocarp and testa until after the

429
hypocotyl emerges. A few never fully emerge, and die. The
root hairs are brown. In those taxa wherein it was tested,
viability is generally good to at least six yearS--no longer
tests were run--but seeds require Special treatment for
germination.

A series of tests were set up to determine germina-
tion requirements. Seeds were gathered from 35 sources
representing eight taxa. One hundred seeds of each pOpula-
tion were placed in petri dishes on wet filter paper, repli-
cated four times for each treatment. One treatment tested
immediate germination without Special treatment; another
tested germination after six weeks of moist storage at l0 C.
Results were tabulated up to four weeks after the start of
the germination at room temperature, at which time the experi-
ment was terminated.

Only seeds from pOpulationS of Toxicodendron radicans
subSp. radicans in Gainesville, Florida, and in Bermuda germ-
inated without after-ripening chill. No germination was
recorded at all from seeds of 1. toxicarium, 1. radicans
subSp. barkleyi; little germination occurred with seeds of

1. rydbergii, 1. radicans subspp. verrucosum and 1. lobatum.

 

Germination results ranging from 10.75% to 62% were obtained
for 1. radicans subSp. negundo, and between 3.5% to 81.75%

for T. radicans subSp. radicans (mean was 40%). Seeds pro-

 

duced during four seasons were gathered from one pOpulation

of 1. radicans subsp. radicans at Island Beach State Park,

 

New Jersey (1955, 1956, 1958, and 1959). The average

430
germination remained essentially the same, but appeared to
increase very slightly with age. Table 18 gives the germ-
ination results. In order for conclusive growth studies to
be possible in future experiments, the problem of germina-
tion must be solved, especially for the taxa for which no
germination was effected. Scarification with acid, sand-
paper, or fire, or removal of the mesocarp before testing
were not tried.

Seeds of a number of pOpulations were weighed to
determine if differences existed between taxa in seed weight.
Fifty disseminules were weighed, each being an intact seed
with endocarp and mesocarp attached. The total weights are
tabulated in Table 19. The heaviest propagules are those of

Toxicodendron lobatum. Those of 1. toxicarium and of 1.

 

rydbergii are next heaviest. Those of 1. radicans subsp.

radicans and subSp. verrucosum are the lightest. Putative

 

hybrid "seeds" of 1. lobatum Xl1. rydbergii in the Columbia
River Valley of Oregon were heavier than those of any pure

Species measured, averaging 2.38 g/5O seeds.

431

TABLE 18.-~GERMINATION OE TOXICODENDRON SEEDS

recorded in per cent germination

Trials followed Six weeks of moist storage at 10 C.
Results are based on 100 seeds per trial and are

 

 

 

 

 

 

 

 

 

 

 

 

 

Replicate Mean
1 2 3 4
1. radicans subsp. radicans
Cape Hatteras, N. C. 27 43 54 46 42.5
Gainesville, Fla. 70 65 65 51 60.25
Chapel Hill, N. C. 17 10 19 21 16.75
Sapelo Island, Ga. 25 26 21 28 25.0
Bermuda 64 59 54 49 56.5
Williamstown, Mass. 13 29 21 16 19.75
Cape Charles, Va. 66 72 86 66 72.5
Middlebush, N. J. 95 72 77 83 81.75
Columbia, S. C. 4 6 3 l 3.5
Island Beach, N. J.
1955 59 42 44 43 47.0
1956 21 29 35 39 31.0
1958 34 34 37 40 36.25
1959 33 36 34 29 33.0
1. radicans subsp. negundo
East Lansing, Mich. 60 67 62 59 62.0
Haslett, Mich. 57 49 61 54 55.25
Norman, Okla. 10 14 10 9 10.75
1 radicans subsp. verrucosum
Austin, Texas 1 2 6 3 3.0
1. radicans subSp. orientale
Nikko, Japan 0 0 0 0 0
1 radicans subSp. barkleyi
Galeana, Mexico 0 0 0 0 0
Huehuetenango, Guatemala 0 0 0 0 0

432

TABLE 18.--Continued

 

 

 

Replicate

2

3

Mean

 

ha

I'—I

I~—I

. rydbergii

Beulah, Mich.
DeTour, Mich.

Kah-Nee-Ta Hot Spring, Ore.

Saugatuck, Mich.
Lake Itasca, Minn.

. lobatum

Berkeley, Calif.
Sacramento, Calif.
Placerville, Calif.
Trail, Ore.
Corvallis, Ore.

The Dalles, Ore.
Kalama, Wash.

. toxicarium

Morehead City, N. C.
Gaston, S. C.
Gainesville, Fla.
Bastr0p, Texas

OOOI—‘OCOO Ol—‘OJOO

OOOO

OOOOOI—‘OO OOUIOH

OOOO

OOI—‘l—‘OOH Oh-I-‘OH

OOOO

OOOOONOO I—‘I—‘OOOO

OOOO

OI—‘OOOO
I001
0"

OOI—‘OOMI—J

OOOO

U'IOxlO
UIU'I

 

433

 

 

 

 

 

 

 

 

 

oe.a-o¢.a mo.m-ee.o m¢.H-oo.o H».H-mm.o mm.H-4H.H mmcmm
om.H mmb.a NH.H 00.0 Hom.0 nom.a CmmE
mv.a 0
No.0 Nm.a 0
No.0 ma.a h
00.H m>.0 NN.H 0
vh.a 00.0 om.a m
vo.a mv.a 00.0 0N.H v
hm.a vm.a 00.0 00.0 vm.a m
0n.H mo.m AH.H H>.H mm.H N
OV.H wh.a no.0 mm.o N.H VH.H H
mmmmmmm. Ezmoosgum> mamofipmn
Eswnmofixou _amwmmmw .meSm .anzm .ansm Hfimhmnpwu
.H aH mamowpmh .H mamowpmu .H. mamofipmn .H. aH coapmasoom

 

 

coavmasooo cm>H0 m Eonm
mummm Emmosou>HEopCmH >vmflm How mEmHm ca mam pmppoomu mosam>

 

mmHUmmm zomDZwDOUHXOH mo mHIwHMS Qmmmu|.0H mqm<H

434
Community Specificity for
Toxicodendron Toxicarium

The range of community types in which the various
taxa of the poison-ivy complex grow is generally broad
except for 1. toxicarium. Other than for a very few unusual
cases, its community preference is so marked that it can be
readily spotted from a moving automobile. Its greatest abun-
dance is in a sandhills type of vegetation typical of the
Carolina fall-line sandhills, but common elsewhere on the
Coastal Plain. This habitat may be described Simply as a
fairly Open, dry, savanna woodland of oak and pine (more
oak than pine). In Oklahoma and parts of Texas beyond the
range of pine, most other components of the community are
nonetheless present. The woods are usually disturbed, often
by fire. They are never very old, generally less than fifty
years. The wide Spacing of the trees allows for a great
deal of light to reach the herb layer where 1. toxicarium
grows. The soil is excessively drained; often the degree
of dryness is emphasized by the presence of lichens such as
Cladonia cristatella on the surface. The herb layer always
has bunch and wire grasses: AndrOpogon, Stipa, or Aristida,
and a variety of legumes: Clitoria mariana, Petalostemon,

Stylosanthes, Tephrosia, and Lespedeza to mention a few of

 

the more important. The acid litter of oak and pine have
formed acid soils in which Vaccinium and occasionally Gay-
lussacia grow well. Crataegus uniflora and Cnidoscolus

stimulosus are additional faithful components of this

435
association. The general sandhills type of vegetation of
which this association is a part has been described in
detail by Wells and Shunk (1931) and Duke (1961). The
Florida component of this formation is detailed by Laessle
(1958) and Monk (1960). The chief associates of 1. 1951-
carium, from my collections, are listed in Table 20.

Since the environments were Toxicodendron radicans
is generally found are those which do not support 1. £251-
carium, an attempt was made to examine intensively the com-
munity preference for the latter to see what was Special-
ized about its niche. A search was made for a habitat on
the Atlantic Coastal Plain where both Species grew close
together and, if possible, grew in the same community. Two
such communities were found that fit the objective well,
plus a marginal third. These three are discussed herein
and in the section on soils analysis.

One of the two major communities examined is in
northern Florida in an area described as the San Felasco
region about 12 miles NW of Gainesville in Alachua County
(T. 98., R. 19E., Sec. 21, 1.7 miles off Devil's MillhOOper
Road). It is best described as intermediate between hammock
and sandhills vegetation and has been studied by Monk (1960)
as he explored the transition between sandhills and hammock
over a gradient of soil and vegetational change. An area
of somewhat more than an acre was found near the site stud-
ied by Dr. Monk, in which both Toxicodendron toxicarium and

1. radicans subSp. radicans grew together abundantly, i.e.,

436

C.

X

XXXXX XXCXX

XXXXX

X

X

Escwnam mmnmmmmmm
mmomvmm mvfiofiasowvm
mCmQCH .<
mflaomfipmsmcm mafieflm<
machHmHfi> mfiaocmmE

mampfinoam manofiHwOHm

mamoaumsm maamcom>aom

EamopcmEOp Escomownm

AmppmhmmHmE .Hm> .HQCHV mpmaampm .0

mama: .0

moflpCmaHHmE .
mpmn>a
mw>mma
mamoca

mpmoamm .

00000

mnam msoumzo
mupmam .0
mmopcwEOP .0
mHEHomfivuoo .0
mpwaama m>Hm0

muwwwnmo mofiH>2
mcmflcwmnw> mSHmQficsh
mvmmp .m

mmeMEUm .0
mflupmsama mscwm

 

>mmwmh 362

mmxmh

mmcHHOHm0

msuuoam

mfimnomo

 

 

ZOHH<Hm0m> mAAHIQZ<m 2H EDHm<0HXOH zomQZMDOOHXOH mo me<H00mm<nu.ON mqm<H

437

CCC

X

333

3x

33333

XXXX

XX

XXXXX

X

msEmcowH>mHm couono
mzmoazswpm msaoomopwco
mHOHMHUCmnm mamm>aom
mcmH:fime> mflmoundmh
whoaman mepCmmoa>pm

EamopcmEop .Hm> mfiaomwoflHQEwm mfimogoc>cm
ESHmDCHm msaommmcm

Esvmccaa coEwpmoampmm

msmoaafl> mzcwdsq

mpmwnvm .4

mpnflc muwpmommq
mcmkuoam mfipomamo
memommCmo ESHUoEme
mflaomwpcsvou mflHmHOpoao
mcmwhms mHHOpfiH0

Enamficamnfi> mammmOchmo
mampwpoflc .0
mpmasoflommm mammmo
anaemosma mwmflpomm
mHHmH>fiHp .m

mafiaomamcso mansm
mflHOMHPSQHm mficoH¢
mpmaawnas .m
mcmwcflaoumo msczum
whoamflcs mammmpmho

 

>mmnmh 3oz

mmxmh

mmcflaoumU

mpwhoam

mfimhowo

 

 

P'l.’

Umscwpc00|l.om mqm<H

 

438

33

33

33

XX

XXXXX

X

mcwnfi>HmQEmm sawsmmamw
mnemownmsm m::p:m8m0
mcwnESooua EDEwHQ>Hom
mcmaawom> .>

Esaamcmp .>

Eamone>uoo .>
EsmcaEmpw .>
Esmnonnm Enficwoom>
mmoazv .0

mmOUCOHm mfiommmza>m0

mamasome mHHchEHc0

Espwmnoc>m .Hm> Esfiaomfioos> Sawmcwum
flflcmaawp mflamxo

mmoaaa> .4

mefipflnme monomq

Enamwmnomm EDEmcymeam:
mwvwoownmq>n EsoHHmQ>I
HafizommCAH .Hm> mflmcmempmze .>
mflaomwpnoo mfipfl>

momqo meH

mafiaamooo mscm
mcmoflHwEm manpocmm0
moupm>a>m muscuaausm
mpmaaouoo mwouocazm
mspmuwomo .0

 

>mmnm0 362

mmxmh

mmCHHonmo

mUHHon

mwmuomo

 

 

 

nmscupcoo--.om mum<e

60
0,8

3

m

.SHHOH

m0

3
.ecwH
we

KICF

 

 

 

 

 

 

 

 

 

 

439

333

33

3 3><

XX

X

X

flfipoon .m

maa>cq>nomup 00mpwaom
Espflmoasoo Ezfinoaflm
flflaamEm owomcmm
mnemficaHOHmo madamdocnn>m

meadows mumwux
mammmOHnmow madamaocwe>m
Ezflaomwscwp Enficwamz
EswaomHQOmm>£ EDHHOHmosm
mamOpcwEOp m:00#¢mcqwam

mflHOMHCHEmnm wflwaom>nn0
mmOpcmEop mHmMUCmHHmm
mmHmp mapoaQ

Eamoawo Eswamw
mflmcmficaaonmo mflaawsm

mamflcwHOHmo mcmnnm>

mfiamupmsm coEwpmcmm

E:Eowocofip mEmpmocoflHH

mpmu>a mH>Hmm

mmoazpcmam .Hm> mHHOMHH06pcw mwumaamusom

mcmoflHmEm mammoflaamo
mampfiHoam .m

mflam>ac xoacm
mpmaafioflvnw> ¢
mmonmbsv mmflamaom<

 

>mmnm0 sz

mmme

mmCAHOHm0

mUHHon

mwmuomu

 

 

 

prCchoonn.ON wqm<H

440

.mmme mo pmmm pflEHH cumpmmz myfl mmsomou mmfiomam I

I
3

.>wmnmh 362 mo EHDOm pflefla camnpuoc mpfi monommh mmwomqw I c

.002 pm pmwflmOQmp Hmcoso> .Uwgowaaoo mwflomam I x

 

XCXX

X3X

3.

XX

XXXX

XXXX

a>mum mmnpcmnflom
Eamonnfim .m

EncapCmem ESHLUCAH>mHm
E:pmp:€&oo Esvmcom>aom
mcmfimnomm mafiaoz
msucsowm msoc:h

mosmam xmaflsm
mfiHmH3>o .0
wzpomHmOHumH mahma>0
mcmomh .m

mHHnmpse .0
mmcmoHHOpoa Esoflcmm

mmofluwm mficocpcma
mSAHmcuwp .<
msfiumdoom comooonpc<
mvowupm .<

mmOCmH mpfivman<

mamxmp .>

mfiaomwpmsmcm mwcocum>
mmpwocpCmHHmn momnpocommupmh
mflaomfipuop .m

whopo .m

 

>mmhm0 362

mmxmh

mmcwaonm0

msuuoui

mfimyomo

 

 

 

Umscfipcoouu.om mqm<H

441
an extensive tract of essentially transitional vegetation.
The soils of the transition zone have a higher percentage
of finer particles than the soils in which 1. toxicarium
generally grows, and are higher in calcium content than
those where this species is usually common. On the other
hand, it is probably the fine particles and increased cal-
cium and other nutrients, especially in the upper soil
horizons where the roots are concentrated, which enables
1. radicans to survive. AS Monk (ibid.) notes, this transi-
tion zone has species of both sandhills and hammock communi-
ties, although some Species which prefer extremes of each
type are absent.

A clue to the ecological requirements of Toxicoden-
dron toxicarium comes from a gradient study north of Colum-
bia, Richland County, South Carolina (Plate XXXVIII). A
llO-meter transect was run along a 5% slope from a heavy

textured soil where 1. radicans subsp. radicans was fre-

 

quent, to a sandhills community on sand where 1. toxicarium
was plentiful. Only in the middle of the transect did both
Species occur together, and there they occurred abundantly.
Although the two plants flowered at the same time and had
the same pollinating vector (honey-bee), there was evidence
of only one individual which could be classed as a putative
hybrid.

The vegetation analysis was made by selecting quad-

rats of ten meters on a Side along the transect line,

PLATE XXXVIII.

442

Gradient through forest in South
Carolina.

a. View into mesic hardwood—pine
forest on loamy sand. Poison—
ivy is abundant.

b. Middle of gradient where poison-
ivy and poison—oak overlap.

c. Dry end of gradient on coarse,
shallow sand. Sandhills type
of vegetation. Poison-oak is
abundant.

p.— “_W—fi.“ M ,——.~—‘-___.~u—‘ ...—....— #5.“.m‘

 

 

$3.41.. .
r r . .
...: ..

. r ._

... .n.

 

. .. .. ‘a
. . . . .
V V . ... . .. 1
.._, 1 ... . vi“.
. .. . .r ,. es
,. .. a;

444
alternating Sides of the base line. In these quadrats,
all tree Species were recorded in terms of diameter at
breast height (dbh), and the measurements totaled for the

2 quadrat

quadrat. Two quadrats within the larger 100 m
were selected in which stems of all Shrubs were counted.

Four smaller quadrats were selected, nested within the

100 m2 quadrats in which per cent cover of the plants in

the herb layer was estimated. The tables which follow (21 -
23) Show a gradual Shift to the north from a hardwood forest
to a sandhills oak-pine scrub, with lessening density of
trees. Calcium content of the upper soil horizons drOpS

and the soil texture becomes lighter; there is more leaching
in the A horizon. Further, the species numbers become fewer.
Loblolly pine occurs in each quadrat, but becomes less abun-
dant (as do all tree Species) upward along the gradient.

In short, as measured Up the gradient, the vegetation changes
to one more characteristic of the typical vegetation in

which Toxicodendron toxicarium may be found. Due to the
plants' abundance, there is no lack of seed source for each
extreme of the gradient to receive disseminules from the
other end of the gradient. The differential in distribution

of the taxa must involve factors in the soil and/or associ-

ated vegetation.

445

 

00.0
00.0

00.0

00.0 00.0 00.0

00.0

HH.0
0H.0

00.5
00.0
VH.0

00.0

00.0

00.0
0H.0

00H.0
00.0 00.0

00.0 00.0

00H.0
VH.0
H.0

NH.0
No.0 00.0H
00.0
00.0
00.0

m600CmHHHmE msons0
memflcfimaw> mwaocmmz
mpmaampm msonmso
mcmHCfiaonmo .0
mcwpoumm macsn0

mama: muonmso
mcwHHmQOo mzcm
mamoflHmEm mschCmm0
mwmoamm msonmso
momao meH

Esnnsn Hmo<
mvmmv mscfi0
mvfihoam mscuoo

msawwomh>pm HmpEmvaUAA

mmopcmEov m>Hm0

 

0H 0 0

.0

0 v

mpmunmnd

mmflomqm

 

 

coaumpmmm> mHHMLUCmm cw ma 0H pmunmsc

“pmmuow voozvnm: ca ma H amnnmso

.EU .Um CH 0mmmwu0xm

mum yam pmunmzc mnp Canvas mmwup Ham How .nnv CH 0mampov mum 0cm
.mpmunmsc kume OH x 0H cw 0mausooo >000 ma 0m0HoomH mum mmwom0m mmnh

$2 22 NH £7.qu50 558 .i238 02355 $5238

.H20H0<00 mmthHOE AHom 02H0<00000 < 0204< IHmOZIIH300 0230 H0002<0H
.<ZH000<0 Ihbom .>HHZDEE00 mquI02<m OH H00000 000300<$ 2000 H0002<0H 20 02000 0000HII.HN mqm<H

446

0

Ha
0H

00

mama: mzoamso
mmaonhm mfim00am0e<

weaponwm .0
mamHCfiHonmo macsa0

mmOpcmEop m>nm0
mosmam xmafiem

EswCHEmym Eswcfloom>
mamaCfi0Hfi> mfiaocmmé

Esnnsn Hmo<
whoamwc: mammmpmn0

mconw>900cmm EzflEmmeu
mfiaomHUCSpOH meH50

mfiaomwvhoo mflpfi>
mcmoflUmH .0mnzm
mcmowvmn coavcmnooHXOH

 

00

am

00

am mm mm mm 9H mH
mumyvmsa

mmwomam

 

 

mpcmEmnsmmmE mmflomam mme How #30 vama
mpmuvmsv 8 OH x 0H mUHmCA vammc .mpmavmsd

E 0 x 0 m ca mamum mo mHmnEDC >0 nmuammme mum mmflom0m naucm

Head ><2 NH .<quom<o zeaom

.>ez:oo az<qonm .<Hm2:uoo .HZmHa<mo mmbemHoz
qum osz<mmoma < ozou< memoz-:e:om mzam Howmz<me

.<ZH400<0

IHbom .>HHZDEE00 quHIQZ<0 0H Hmmmom 000300<I 2000 Hommz<mh 20 02300 0000:0II.NN mqm<H

 

 

Eli]

447

00

r-IC’)

o—Iv—I r—IN

mQAUCmeHmE mzoumsc
mDHMflomH>pm HmnEmvfisvfiA

Esmnonum Esficfloom>
mvmmp maca0

mampwuofic mammmo
mmUHooflHm0>£ 85009003:

mpmuamm msoH030
mafiaamQOo mscm

mflaomeUCHsc mammfloocmcpnm0
mpmaampm msoumso

mflamw>anp manna
EsfiumoHXOp councmnoonOH

mpmH>H msonso
mnfiHon mscno0

 

m0

00

00

av mv 00 mm 00 mm
mvmnvmsd

QHI mH
mmwom0m

 

 

 

UwDCHpc00un.NN mqm<H

 

 

 

, moo va «00 m Imc Ill/III
ILllLlflflflI mHH 00H H mpmnbmso 0 0 03 mm 1mm

wmwomqw

 

Iii-Ell,

bmzcfiwcoonu .NN mamasm

'1'
III"

448

 

0H

Ha

0H

0H

Ha

(\I
v—IWC")

r—ILO

moHUCmHHHmE msonmd
msamwomh>pm Manamvfiscfiq
Eswnonnm EDflcwoom>
mvmmp mzcfi0

mampfipofic mfiwmm0
mmvfiooHHm0>£ EsoflHmQ>m
mpmoamm wonmSO

mcwaamQOQ mscm

mfiaowmzvcflsu mammfioocwcvum0
mpmaampm msonDO

mwamw>flnp manna

Enfiemofixov convcmvoofixOH
mpmu>a msonSG

mvfieoam washoo

mama: muonmso
mmnonuw mfim00am0e<
weaponmm .0
mamflcwaoamo macsh0
mmoucmsop m>Hm0
mosmam meH80
EsmCHEmpm Esflcfioom>

mcmficwmufi> mflaocmmz
EDHQSH Hwo<
mHoHMfic: mammmpmu0
mcoufi>H00cmm ESflEmmHmw
mflaomfiUCSpon xmawem
mfiaomwunoo mflvfi>
mamoflvmn .0mnsm
mamownmh couvchOQMXOH

 

mHflh

mHH

00H

mOH

£0

mo
mummvmmd

00

00

De we mm

mmH6000

 

 

 

uwscwvc00un.mm mqm<H

 

ErUO>KCCW£rC~'((4

 

$3 ><2 NH 23050 :58
.3238 02326: £82208 £25920 0020302
dem 02355.3 2. 020.2 5002-258 2600 0222 06022.00 £23005
:58 £222,228 3.552% B 0828 0820?: 26mm Emmzfit 26 02:60 mmmmr--.mm 33.0

 

 

N v Esaamcmp Esflccwom>
0 0 0 Banana Hmo<

p N mwaommsvcfisv mamwwoocmnunm0
a N 0 a p p p p mCQOH maamzovfiz

0 mamwcwmufi> mfiaocmmz
p p .00 000

p p mfiamfl>flup manna
a m mwne

N H Esfihmofixop .H
a a 0 v 0 mamofl0mn .00030
mamofivma 00H000000onh

449

00

OH

San

Va

00

Ha

00

00 00 on 00

0 N 0 H

00

00

00

Hmppflq
EspfimOQEOU ESflLQHH0

p mcmufi>HwQEmm Sufismmamo
0H 00 00 H0 00 NH 0 0 0H 00 00 mwaomfiunoo mfipfl>

 

omI 00, mm (mm, 00 am, mNI IbH, 0H 9H 0H
09090000 mmfiow0m

#000 900 a 00:9 wmma .00099 n p .mpcmemnswmme 0000000
mmup How 050 Ufima mpmuvmsv E OH x 0H 00H00H 09090030
E 0.0 x 0.0 Ca Hm>oo p000 H00 20 00930008 0H0 www000m mum:

0002 ><2 NH <2H000<0 zeaom
.3208 023202 {32300 €208,200 002.082
qum 02Hm¢mmomo < 0200< 20m02-2e:00 2020 02:0 Hommz<me .<2qum<0
xeaom .>HH202200 000H102<m 00 000000 000200<2 2000 Hommz<me 20 02:00 0000:--.mm mqm<e

450

000009080 035900000
mcmflcwm9fi> 00009000H
00fl9oam 030900

000000000 630008000
000929 0309030

630008090 E3fiCfloom>
mvmaao9oo 000905030

0090090 0H000H0QE<
mcflpo90m 030390

Magmaaflu mHmeo
mpmoamm 0309030

099“: 0N000000q
09H000003po9 mefiE0

0H90003E 5300000
HwHHmEm 0000000

.00 x0900
0030H0 xmafiem

 

00

00

.00

00 £0 00
00090030

0H

00.

0H, 0H
0000000

 

 

 

003prc00un.0w wqm<H

451

m o 00 m N N N N m 003000 x0008m
83000000 80000000>

839039 9000

N m 0 0 0 000000300030 03000000000900

0 000009 000000002

000000090> 00000002

.00 000

0 v 0 00000>090 03030

N v 0 m0 0 009B

mm 00 0 m0 8309000x00 .H
0 m 00 ”mm 00 0 m 0 0 mam00000 .00030

00000009 0090000000x00
00 00 00 N0 00 00 on 00 cm om 900000

8000000800 8300000m
0 m 0 m N 0 0 N m m 00090>900800 800800000

00 00 00 mm mm 00 00 00 00 00 0000000000 0900>

 

0m 00 ‘00 0m 00 0m 00 0v 00 00
00090000 000000m

'Illtlll'l- -‘l““lll‘|.l“l.ll“‘lu

 

I'I-Il-ll.‘|'|l ll IIIII“II

003000000uu.mN w0m<H

 

452

00

0o

000

00

0m 00
00090030

‘1"
[HJ

00

 

 

00

£0

 

 

000009080 030000000

000000090> 000090000
0009000 030900
000000000 830008000
0009>0 0309000
830008000 80000000>
000000900 000900000
0090090 000000008<
00000900 030090
00000000 0000x0
0000000 0309030

00900 000000000

N 000000003009 x0008m

00000038 8000000
0000080 000000m
.00 x0900

00
0000000

 

 

 

003:00co0--.mm 000<0

453

00000900 030090

GHNOHmm mSOHmDO

 

 

 

0 0 0 00900 000000000
0 0000080 000000m
N m 0 003000 x00080
m 000000300030 03000000000900
.00 000
N 0 v N 00000>090 03030
0 m 00 N m m 0090
N0 0 0 0 N0 N m0 8009000x00 .0
0 0 m 00000000 .00000
00000009 0090000000x00
00 00 mm 00 00 00 om mm 0N 0N mN m0 900000
8300000800 8000000m
0 m 0 v m 0 N 0 N 0 N 0 00090>900800 830800000
00 0N 0 mm 00 ON no 0 on 00 00 00 0000000900 0000>
00 00 00 00 0m 00 ‘00 on m0 00 x00 00
00090030 000000m

 

 

 

 

 

00::00c00--.00 000<0

04 .0 +J -0 0+

454

030900900 000000900<
00000000908 0309030
000000090> 000090000
830008000 80000000>
00000000900 030390
0000900 000000000
8000000 830000
8300038800 83000000000
000000000 830008000
000900 .0

09000 0309000

00000 03000
000009080 000000000

000000900 00090000m

0900000 000000000>0m

 

 

00090030

000000w

 

 

0030000001I.mN m0m<0

455

00 N0 mm on 00 mm

N

no

00 om

0N

cm on

On

00

00000900 030390
0000000 0309030

00900 000000000

0000080 000000m

003000 x0008m

000000300030 03000000000900
.00 000

00000>090 03030

0090

8309000x00 .0

00000009 .00030
00000009 0090000000x00

900000
8300000800 8300000m

00090>900800 830800000

0000000000 0000>

 

000, 000‘ 000, 000‘ .000 000
00090030

000 x00

00

0000000

fl

 

003000000un.mN w0m<H

 

456

mafihmcump comoaohnc<

ma moaogmafiume muonmzd
mameflmHfi> mfimOHchH

p EswcwEmvm Eswcwoom>
N mamficflaonmo mscznm
mmowamm masonpcmo

Esmoafia Eswamo

EspmpDEEoo Esvmcom>aom

mcmomemo Eswvosmmo

mpmn>a .0

mnmfic msoumso

mvmmp mscwm

mamoHHmEm manpOCmmo

mpmHHonoo mannocasm

H p mHoamwn mmLHCmmoa>um

 

mpmuvmsd

no 00
mmwomaw

fl

 

Umscwvcoonu.mw mqm<H

457
Soil Studies

A number of soil and site analyses were made to
determine if members of the poison-ivy complex demonstrate
any soil requirement specificity. A total of 38 soil pits
was dug and samples of each horizon were brought back to
the laboratory for analysis. These samples represented the
soils in which nine taxa grew, of which four seemed to show
some very Specific requirements.

Each horizon was analyzed for texture, pH, and major
nutrients available in parts per million. Soil texture was
analyzed by the hydrometer method (Bouyoucos, 1936) and
verified by touch. pH tests were run on a Beckman pH meter
against a known phOSphate buffer. The concentration of P,
K, Ca, and Mg available was determined by the Soil Testing
Laboratory of Michigan State University. From these data,
soil profile information was synthesized and the profiles
identified as well as could be done with available for the
Asiatic taxa, except that it is known from herbarium label
data that Toxicodendron nodosum has been found only on shal-
low limey soils. No data are available for the soils in

which I. radicans subsp. divaricatum grows.

 

Although soil profiles were studied from a variety
of sites where Toxicodendron lobatum and the various poison-
ivies grew, most of the common and widespread taxa occupy
such a variety of habitats that the samples tell us little.

The profiles which were examined demonstrate only a segment

 

 

   

458
of the range of soil types in which these taxa grow: I, ryd-

bergii, I. lobatum, I. radicans subsp. radicans, I, radicans

 

subsp. negundo, and I. radicans subSp. pubens. Even digging

 

of soil pits would not be necessary to indicate that these
plants will grow in a wide range of soils.‘ They seem to be
absent from podzols and heavily leached lateritic soils,
however, and only rarely make incursions into peat. One

taxon which seems to show narrow ranges of soil preference

is Toxicodendron toxicarium; hence, this plant's soil require-
ments were examined more extensively.

Soil samples are of particular value from regions
wherein several taxa form putative hybrids. If the parent
taxa generally inhabit different kinds of soils, then a
sample of the soil in which a suspected intermediate is
growing may be used to complement morphological data to
confirm the hybrid nature of the pOpulation. That is to
say that the probability that the individual is a hybrid
is increased if the individual appears to be growing in a
soil that is also intermediate between the extremes pre-
ferred by its possible parents. Such samples have been used
in corroborating the hybrid nature of pOpulations of Toxico-
dendron toxicarium and I. radicans subSp. radicans, as will
be shown later.

As one may note from collection record data, Toxico-
dendron lobatum, I. rydbergii, and several subSpecies of
I. radicans generally grow well--and, in fact, reach Optimum

growth conditions--on floodplains. In those regions where

459

the soils have been sampled also, individuals were climbing
nearly every tree and formed a ground cover approaching 100%
in the undifferentiated river alluvium. Such soils were
determined to be an Underwood stony loam near Hood River,
Oregon where I. lobatum and the hybrid between it and I.
rydbergii grew. The floodplain soil was a Genessee silt
loam in East Lansing and Saginaw, Michigan, where I. gagi-
gggs subsp. negundo was abundant.

The soil preferences for Toxicodendron radicans sub-
Sp. negundo were examined in a small study within Baker
Woodlot on the Michigan State University campus for which a
detailed soil map was available. The pOpulations of poison-
ivy were mapped for correlation with the soil patterns. To
be sure, soils could not be adjudged to be the only factor
Operating to determine distribution patterns of the plant,
but the conditions suggest that they are involved to a large
measure.

Baker Woodlot (Ingham Co., East Lansing, Michigan.
T. 4N., R. lW., SW % Sec. 19) is probably a fragment of the
forest that was cut over about one hundred years ago inas-
much as the oldest trees are about a century old. It is pre-
dominantly a beech-maple woods, but there are pockets of
other tree assemblages that appear to be related to soil sub-
strate and soil moisture. Toxicodendron radicans subSp.
negundo was expected to be common at the edge of the woods
where light conditions would be more suitable than those in-

side the dense canOpy, but it will be seen that some factor

 

460
other than light must be involved in producing somewhat
contrary data.

The "edge effect" is manifest on the east side of
the woodlot where poison-ivy is found commonly. The soils
in which the plant is growing have a medium to high natural
fertility and moderate to high available moisture. On the
south side of the woods there is little poison-ivy despite
an improved light regime from that inside the woods. The
south side is clearly disturbed. It serves as a railroad
right-of-way, was clear-cut long ago and planted with pine
and locust. For the most part, the soils which occur on
the south side are--with the exception of Miami loam, one
of the better agricultural soils Of the state--poorly
drained with low natural fertility and with low or variable
moisture available for uptake. Such soils are Coloma loamy
sand, Hillsdale sandy loam, Locke sandy loam, Berrien sandy
loam, Granby loamy sand, Carlisle muck, and the Washtenaw
soil. Poison-ivy is likewise uncommon along the west and
north edges of the woods. In the northwest corner is a
patch of Spinks loamy fine sand which supports a tangle of
brambles, gooseberries, and trees, but no poison-ivy. The
area is lower than much of the woods and supports a vegeta-
tion that is fairly Open to light. Presumably it is the
low fertility which has precluded its supporting any pOpu-
lation of poison-ivy.

Perhaps this analysis of a single woods, its soils,

and its poison-ivy pOpulation can form a generalization

 

461

about Toxicodendron radicans subSp. negundo and its soil
requirements. Contrary to pOpular Opinion, poison-ivy is
not ubiquitous, but rather requires fairly rich soils. It
is all but missing from soils which are low in fertility,
or in available moisture, and does not generally gorw in
standing water. Whereas the plant generally was found to
be uncommon on the poorly drained Wauseon soil (composed %[
of a two-storied parent material: a fine sandy loam to
loamy fine sand on tOp of a clay loam to silty clay loam

till lower story), it was most abundant on this soil sur- g j

 

rounding a muck patch in the northern center of the woods.
This abundance is presumably attributable to edge effect
(border of an open Cephalanthus community), to increased
fertility, and to the greater drainage afforded by the in-
crease in lepe. Poison-ivy thus appears to grow best in
soils of high nutrient value and good, but not excessive,
drainage.

From the soils sampled for Toxicodendron rydbergii
pOpulations, it appears that this species generally inhabits
slightly more alkaline soils than does I. radicans subsp.
negundo. Toxicodendron radicans subsp. radicans, from
similar types Of soils studies, appears to inhabit slightly
more acid soils than either of the taxa mentioned above.
Both of these generalizations are admittedly oversimplifi-
cations, eSpecially in view of the fact that only six pro-
files were examined for I, rydbergii and five for,I. radicans

subSp. radicans.

462

When one considers the broad geographical sources
of these soil samples (New England to the Pacific Northwest
and to Florida), the data assume a greater significance
despite their apparent skimpiness. In examining the soil
profiles for those sites in which Toxicodendron radicans
subSpp. or I. lobatum were found, the A horizons of the
soil--and to a lesser extent, the upper B--seem to be the
most significant in determining growth conditions for the
plant, for it is in the A where maximum root develOpment
takes place. Only occasionally do roots penetrate to the
B horizon, and then only in the upper B. It is perhaps for

this reason that I. radicans is virtually absent from well-

 

develOped podzols or laterites wherein the A horizon is
highly leached of minerals, especially Ca and Mg.

A study was made to determine the mineral nutrient
requirements for poison-ivy. Seeds of various taxa were
germinated on filter paper in petri dishes and then trans-
ferred to a series of hydroponic solutions. Seedlings were
grown in complete nutrient solution as a control, and in a
series Of solutions lacking reSpectively Ca, Mg, P, K, Fe,
N, and S. Four replicates, involving different taxa each,
were carried out in the greenhouse. After two months, the
experiment was concluded and evaluated.

Seedlings grown in complete nutrient solutions pro-
gressed well. Those grown in Mg-poor solutions initially
performed poorly and drOpped all leaves, but axillary buds
develOped and the plants were relatively healthy at the end.

463
Furthermore, plants grown without Mg develOped the longest
roots. Except for Ca, no one mineral nutrient seemed en-
tirely crucial. Absence of Fe produced mottling on the
leaves and some chlorosis; absence Of N03 resulted in
chlorosis, etc. In Ca-poor solutions, however, the most
marked effect of all resulted. The seedlings did not de-

velOp beyond the cotyledon stage before they died. Even

..y
3;... .
V
l-
|.

though no tests were run to determine at what minimal con-

centration calcium becomes critical, it appears safe to say

 

 

that calcium is the most important of the elements tested 3 ’j
in the soil for the growth of taxa in the poison-ivy com- l?"'
plex.

It was noted earlier that Toxicodendron rydbergii
has generally been found growing in soils with fairly high
concentrations of calcium. In agricultural sites, these
soils would be described as not requiring additional lime
to support good growth of most common crOps. Toxicodendron
lobatum has been found in soils with similar Ca concentra-
tions-~none analyzed yielded less than 230 ppm of Ca. For
I. radicans subSp. verrucosum, only one soil was found in
which the significant concentrations of calcium were below
350 ppm in those horizons in which roots were found. At this'
site (Harrison County, Texas, east Of Marshall), soil calcium
in the A and B horizons ranged from 98 - 231 ppm. Here,
poison-ivy was rare, found only with difficulty after a
long search, probably because it was at the lower limit Of

its calcium requirement. Furthermore, I. radicans subSp.

464
verrucosum has its main distribution on the limestone and
dolomite Edwards Plateau and Arbuckle Mountains, where soils
are naturally high in calcium.

Only once was Toxicodendron radicans subsp. eximium
studied in the field, in the Chisos Mountains Of the Big
Bend region Of Texas. It was growing in a lithosol on ba-
salt. The calcium content Of the soil was determined to be

1099 ppm. The soil in which one pOpulation of I, radicans

 

subsp. barkleyi grew at the edge of an orchard in Galeana,

 

Nuevo Leon, Mexico, was sampled and examined critically in
the laboratory. Here the soil calcium was higher than that
for any other soil studied, namely, Up to 5700 ppm.

For Toxicodendron toxicarium, the soil regime is
very different. Unlike any other taxon in this complex, it
grows in podzols at its northern limit (Lakewood sands in
southern New Jersey), but otherwise is mostly confined to
well-develOped lateritic soils of the red-yellow podzolic
great soil groups, both types Of soil having a strongly
leached A horizon and a well-develOped layer of illuviation
in the B. The soils are chiefly light textured, at least
in the A and B horizons, where maximum root development takes
place. Theyeme moderately to strongly acidic, and uniformly
have very low concentrations of Ca. Of the nine soil samples
from I. toxicarium habitats, sampled from seven states
throughout the range Of the plant, the highest calcium con-
tent was found in the B1 of a loamy sand in BastrOp State

Park, Texas (91 ppm). The parent material of this soil

 

465
contained far higher Ca concentrations, but was out of reach
of normal root develOpment. Interestingly, there was no
1. radicans found there. Most soils which were tested in
which I. toxicarium grew had far lower Ca, even down to 9
ppm. To some extent, K concentrations were also impoverished.

All of the soils in which Toxicodendron toxicarium
is indigenous have develOped under acid-forming litter, thus
leading to heavy leaching. Scrub oaks are dominant with
pine a usual constituent in this kind of woods except in
the western end of the range of I. toxicarium. Drainage
of the soil is high to excessive. Such soils are the Lake-
wood sand, a podzol, and the Lakeland sand, Norfolk sand,
and Stephenville fine sandy loams, lateritic soils. The dif-
ferences in these soils are in parent material, extent of
horizon develOpment, and thickness of sand over heavier sedi-
ments. The Lakeland, Lakewood, and Norfolk sands have devel-
Oped on unconsolidated acid sands of the Coastal Plain; the
Stephenville series has develOped over parent materials de-
rived from reddish sandstone or interbedded sandstone and
sandy shale. The latter was the one soil sampled for a poison-
oak habitat that was off the Atlantic or Gulf Coastal Plains
(in Oklahoma).

The Lakeland soils show little profile develOpment,
chiefly because of the low clay content of the parent ma-
terial, lacking the finer textured subsoils of the Norfolk
and Lakewood series. The Norfolk, Lakeland, and Lakewood

series differ in that there is a greater thickness Of sand

 

466
over heavier sediments in the Lakewood and Lakeland than in
the Norfolk. There are generally heavier textured upper
soil horizons in the Norfolk than in either other series.
Suffice it to say that all of these soil series have much
in common, despite minor differences in profile development.
The physical and chemical attributes of the root environment
are appreciably similar. No collections were made from any
soils determined to be the Orangeburg, a major soil Of the
Carolina sandhills according to Duke (1961).

Although poison-ivy (Toxicodendron radicans subspp.)
may inhabit sand dunes and the coastal plain sands, it is~
not indigenous to the poor soils of the sand scrub as is
Eastern poison-oak (I. toxicarium). Toxicodendron radicans
appears to grow in soils that have a higher Ca and usually
higher Mg content than the ones in which I. toxicarium grows.

Furthermore, I. radicans generally grows in heavier textured

 

soils than I. toxicarium, with less leaching in the A horizon.
Several areas of transition between soil types and

corresponding vegetation have been studied where both Toxi-

 

codendron radicans subsp. radicans and I, toxicarium were

 

growing intermixed. One area is a gradient along a 5% slope
between a Norfolk loamy sand (thin solum phase) and a shallow,
incipient Lakeland sand north of Columbia, South Carolina.
There is abundant poison-ivy at the lower end Of the lepe

in a heavy-textured soil, poison-oak at the upper end in a
shallow coarse sand, and both in the middle of the zone de-

scribed in greater detail in the section of this paper on

467

Community Specificity for I. toxicarium. A second plot

 

that was studied is in an extensive area of essentially
transitional zone between a mesic hammock community and
a sandhills community in the San Felasco region of Alachua
County, Florida. A third area where poison-ivy was found
in association with Eastern poison-oak was in an oak-pine
woods in Lincoln County, Mississippi (T. 7N., R. 7E.).

The soil in the South Carolina site where poison-
ivy and poison-oak both occur lies between a sandy clay
and loamy coarse sand and has been determined to be a sandy
loam. The lower (poison-ivy) end of the gradient has finer
textured soil than the upper (poison-oak) end and a higher
Ca content than either the middle or upper end (44 ppm vs.
21 ppm in the middle vs. 16 ppm in the tOp for the Al).
The pH is slightly higher in the middle of the gradient
than at either end. As the finer particles become less
prevalent, the soil becomes more conducive to the growth
of poison-oak and less to that of poison-ivy. The same is
true as the calcium content drops, especially in the root
zone of the A and upper B horizons (see Tables 24 and 25,
and Fig. 23).

The Florida site is mapped in the Lakeland series
(being only a few miles from the type location where the
series was established) as a sandy loam. The texture is
finer than that of the adjacent woods where only poison-
oak grows, the soil of which is mapped as a Lakeland sand.

The pH in the transition zone is higher and has a higher Ca

 

468

 

 

 

 

 

v.4 ¢.H m NH om.v asmm v
o q.H m mm oo.v ncmm >EmoH m
m A.H m 6H mm.¢ nemm >EmoH m ..
m.m m.H 6 mm mm.v ecmm >88H H 0064 asHumonop .H m
m.mH m.mH NH mm Hm.o scam o
m 6.0 6 OH 50.0 EmOH >acmm mm
o 0.0 .v mH qH.m EmoH >ucmm Hm
m.m m.H m 0H No.0 smOH seamm m<
v.6 6.H a Hm mm.q EmoH >ncmm H< I
m H.H 0 HH mm.m amOH >ecmm o< kHoq asHumonop .9
some memoHan .H_ m
H H mHH mm.q ucmm o
o H mH mk.e namm EmoH mm
0.0 mo om.¢ >mHo >UCmm Hm CH name
6 H mm mo.¢ >mHu >ncmw mm
H m.mHH 00.0 >mHu >acmw o< I
a m n.4q Aq.q >mHu >ncmm H< oHoa mcmoHcmu .H H
HOOESZ
mo IO waspxmh CONHHoI CONHHoI coawowaaoo come OCON
mHHHHo

 

 

<ZHHOm<U IHDOw .<HmEDAOO .mQOOg mZHmux<O mo mHm>q<z< qumoma AHOmII.VN mqm<H

469

 

 

 

 

 

 

 

 

 

0.0 m 0 0H0 00.0 0000
. Emoa
0.0 0.0 00 00 00.0 00000 0
Emoa
0.0 m 0H 00 00.0 00000 0< .n
Emoa Esfinmofixov .H
Emoa >pcmm IIII. 4 I
0.0 0 0H 00 00.0 00000 H< 0cmmemH 0 v mcmoHnma .H 00HH0H0
0.0H 0.0H 0H 00 H0.0 0:00 o
Emoa
m 0.0 0 0H 00.0 >0000 00
EmoH
0 0.0 0 0H 0H.0 00:00 H0
Emoa 00H
0.0 0.H 0 0H 00.0 >0Cmm 0<
Emoa
0.0 0.H 0 H0 00.0 00:00 H< linl. :II. I
0000 0H00 ssHumoH 0» .0
EmoH >EmoH I
0 H.H 0 HH 00.0 >0Cmm o< 0cmmemH 0000 mcmoHan .H mcHHOHmO .0
0000 eaHumUonp .H
EmoH I
0.0H 0.0 0.0 000 0.0 emoH < 0HH0 0050 0000 mchHnmu .H H00H00H00H2
HmOEDZ
x a 02 00 Id manpxmh CONHHO: 00>H Hflom cowpomaaou coxmh >vaamooq
C0380 mHHHHo

 

zomu EDHm<UHXOH

 

 

 

 

.H DZ< mZ<UHD<m

.H mhom mmmrz m<mm< mo mUHHMHmmHU<m<IU mAHmomm AHOmIn.mN wqm<H

 

 

 

470

Fig. 23. Road cut in Bastrop County, Texas showing
profile of soil in which I. toxicarium is
abundant.

 

471

 

472
content in the upper soil horizons than in the Lakeland
sand which supports poison-oak but no poison-ivy.

The Mississippi site was determined to be on a Bude
silt loam, poorly drained, low in fertility, and slightly
eroded (Milbrandt, gt 31., 1963). Its Ca content was moder--
ate--273 ppm--at the lower end of the presumed tolerance

range for Toxicodendron radicans subSp. radicans, but higher

 

 

than that recorded for most poison-oak sites. This soil
may be looked upon as an intermediate habitat between the
usual soil types in which either poison-ivy or poison-oak
are normally found, but within the tolerance ranges of both.
Although poison-ivy will usually live in the range of par-
ticle size included in a silt loam, the low fertility of
this profile makes it a marginal site for poison-ivy.

In summary, one may conclude that Toxicodendron ragi-
[gags subsp. radicans inhabits soils which are higher in pH,
higher in nutrient concentrations (eSpecially calcium and
magnesium), and heavier textured than those in which 13.32li'
carium will grow. There is no clear-cut break between the
requirements of one and the other, at least from the data
presented herein, but rather an overlap of habitat require-
ments leading to the occasional ecotone wherein both Species
may be found intermixed. These ecotones are characterized
by an intermediate type of soil, accompanied by an intermedi-
ate type of vegetation between a mesic woods and a sandhills

vegetation--a woodland savanna.

LITERATURE CITED

Abbott, F., Jr. 1886. Notes on a recent case of poisoning
caused by the exhalation of Bhus radicans (Toxico-
dendron) at the Botanical Gardens, Hobart. Pap. &
Proc. Roy. Soc. Tasmania. 24: 182.

Adanson, M. 1757. Histoire naturelle du Sénégal. Coquil-
lages. Avec la relation abrégée d'un voyage fair
en ce pays, pendant leS années 1749, 50, 51, 52 et
53. Paris.

. 1763. Familles deS plantes, Vol. 1. Vincent CO.
Paris.

 

Anonymous. 1908. J. & Proc. Roy. Soc. New South Wales 42:
xlviii.

. 1949. English grow poison ivy. Sci. News-Lett.
56: 70.

 

. 1961. Occupational poison oak dermatitis in
California. State of California, Dept. of Public
Health.

 

Arthur, J. C. 1934. Manual of the rusts in the Uhited
States and Canada. Purdue Research Foundation.
West Lafayette, Ind.

Backer, C. A. and R. C. Bakhuizen Van den Brink, Jr. 1965.
Flora of Java, Vol. II (in English). N.V.P. Noord-
hoff, Groningen.

Balls, E. K. 1965. Early uses of California plants. Univ.
Of Calif. Press. Berkeley.

Barkley, F. A. 1936. "Five-leaved" poison ivy. Biologist
17: 122.

. 1937. A.monographic study of Rhus and its immedi-
ate allies in North and Central America, including
the West Indies. Ann. Missouri Bot. Card. 24: 265 -
498.

 

473

le F
Bark Y, (A

Beissen, L
La:

Blankinshi;
An]

Boerhaave,

Bouyoucos,

an;

Sci
Brandegee ,
Zoe

Braun, E. I
Ame

Britten. J.

Britton. N.
of

Nev

--~‘~‘9rc

e

216

 

474

Barkley, F. A. 1957. Generic key to the sumac family
(Anacardiaceae). Lloydia 20: 255 - 265.

. 1963. A criticism of the traditional concept
of the genus Rhus. Prospects of Iraq Biology 3:
52 " 580

 

, and Elizabeth Ducker Barkley. 1938. A Short
history of Rhus to the time of Linnaeus. Amer.
Midl. Naturalist 19: 265 — 333.

 

Beissen, L., E. Schelle, H. Zabel. 1903. Handbuch der
Laubholz-Bennennung. Berlin.

Blankinship, J. W. 1907. Plantae Lindheimerinae, Part 3.
Annual Rep. Missouri Bot. Gard. 18: 123-- 223.

Boerhaave, H. 1727. Historia Plantarum. Ludg. Batavia.

Bouyoucos, George J. 1936. Directions for making mechanical
analyses of soils by the hydrometer method. Soil
Science 42: 225-228.

Brandegee, T. S. 1890. The plants of Santa Catalina Island.
Zoe 1: 110.

Braun, E. L. 1950. Deciduous forests of eastern North
America. Blakiston Co., Philadelphia.

Britten, J. 1900. Notes on Rhus. J. Bot. 38: 315 - 317.

Britton, N. L. 1881. A preliminary catalogue of the flora
of New Jersey. Geological Survey of New Jersey.
New Brunswick.

. 1894. List of PteridOphyta and SpermatOphyta
growing without cultivation in northeastern North

America: Anacardiaceae. Mem. Torrey Bot. Club 5:
216.

 

. 1901. Manual Of the flora of the northern states
and Canada. New York.

 

, and J. A. Shafer. 1908. North American trees.
New York.

 

, and Addison Brown. 1913. An illustrated flora
Of the northern United States, etc. Vol. 2. New
York.

 

475

Brizicky, G. K. 1962. The genera of Anacardiaceae in the
southeastern United States. J. Arnold Arbor. 43:
359 - 375.

. 1963. Taxonomic and nomencaltural notes on the
genus Rhus (Anacardiaceae). J. Arnold Arbor. 44:
6O - 80.

 

Burkhill, I. H. 1935. A dictionary of the economic prod-
ucts of the Malay Peninsula. 2 vols. London.

Burse, C. F. H. 1811. Dissertatio inauguralis medica de
Rhoe toxicodendro et radicante. Thesis, University
Of Berlin.

Candolle, A. P. de. 1825. Prodromus systematis naturalis
regni vegetabilis. Part III. Paris.

Candolle, A. de. 1874. Calques des dessins de la flore du
Méxique de Mocino et Sessé, qui ont servi de types
d'eSpeces dans la Systema ou le Prodromus. Geneve.

, et C. de Candolle. 1883. Monographiae Phanerog-
amarum. Burseraceae and Anacardiaceae by Adolph
Engler. Vol. 4. Paris.

 

Chaney, R. and H. L. Mason. 1933. Studies of the Pleisto-
cene paleobotany of California iii. A Pleistocene
flora from the asphalt deposits at Carpinteria,
California. Publ. Carnegie Inst. Wash. 415: 45 - 79
plus plates.

ChOpra, R. N., R. L. Badhwar, and S. Ghosh. 1949. Poisonous
plants of India (ed 1). Indian Council of Agricul-
tural Research. New Delhi.

Clavigero, F. J. 1786. The history of (Lower) California
(translated by Sara E. Lake and A. A. Gray), Stanford
University Press. Stanford. 1937.

Cleland, C. E. 1966. The prehistoric animal ecology and
ethnozoology of the Upper Great Lakes region. An-
thrOp. Pap. Mus. AnthrOp. Univ. Michigan NO. 29.

Connor, H. E. 1951. The poisonous plants in New Zealand.
Bull. 99, D.S.I.R.O. Wellington.

Cooke, C. W. 1945. Geology of Florida. Bull. 29. Fla.
Dept. Conserv. 29.

476

COpeland, H. F. and B. E. Doyel. 1940. Some features of
the structure of Toxicodendron diversiloba. Amer.
J. Bot. 27: 932 - 939.

Corner, E. J. H. 1939. Notes on the systematy and distribu-
tion of Malayan phanerogams iii. Gard. Bull. Straits
Settlem. 10: 239 - 328.

 

1940. Wayside trees of Malaya I. Singapore.
Cornut, J. 1635. Canadensium plantarum. Historia. Paris.

Criddle, N. 1927. LepidOptera reared in Manitoba from
poison ivy. Canad. Entomol. 59: 99 - lOl.

Crum, H. A. 1951. The Appalachian-Ozarkian element in the
moss flora of Mexico with a check-list of all known
Mexican mosses. Univ. Mich. Doct. Diss. Ser. Publ.
3486.

Curtis, J. C. 1959. The vegetation of Wisconsin--an ordina-
tion of plant communities. Univ. Wisc. Press.
Madison.

Daubenmire, R. 1954. Alpine timberlines in the Americas
and their interpretation. Butler Univ. Bot. Stud.
11: 119 - 136.

Dawson, C. R. 1954. The toxic principle of poison ivy and
related plants. Rec. Chem. Prog. 15: 38 - 53.

. 1956. Chemistry of poison ivy. Trans. New York
Acad. Sci. 18: 427 - 443.

 

Deam, C. C. 1924. Shrubs of Indiana. Publ. of Dept. of
Conserv. State of Indiana. Indianapolis.

. 1932. Shrubs of Indiana, ed. 2. Division of
Forestry, State of Indiana. Indianapolis.

 

. 1940. Flora of Indiana. Dept. of Conservation,
Div. of Forestry. State of Indiana. Indianapolis.

 

Delestinne, E. 1956. Contribution a la pathogénésie de
Rhus Toxicodendron. Cahiers d'HoméOpathie et de
Therapeutique comparée. 1956: 313 - 324.

Dillenius, J. J. 1732. Hortus Elthamensis, Vol. 2. London.

Dippel, L. 1892. Handbuch der Laubholzkunde, Vol. 2.
Berlin.

Dodge, C.
t
i

Domin, K.
1

Don, G.
V

Dressler,
N

Duke, J,
l

Dyer, H.
E

Eardley ,
E
Eaton. A.
U

EdWa rd 8 ,
V
ElliOtt.

In nl

Engler,

lnflfifix

477

Dodge, C. K. 1911. Results Of the Mershon Expedition to
the Charity Islands, Lake Huron. Annual Rep. Mich-
igan Acad. Sci. 13: 173 - 190.

Domin, K. 1930a. Jedovate Amerike Skumpy. Veda Prir. 11:
121 - 125 (in Czech).

. 1930b. Jedovate Amerike Skumpy. Veda Prir. 11:
203 - 209.

 

Don, G. 1832. A general system Of gardening and botany,
Vol. 2. Calcyiflorae. London.

Dressler, R. L. 1954. Some floristic relationships between
Mexico and the United States. Rhodora 56: 81 - 96.

Duke, J. A. 1961. The psammophytes of the Carolina fall-
1ine sandhills. J. Elisha Mitchell Sci. Soc. 77:
3 - 25.

Dyar, H. G. 1904. Poison ivy caterpillars. Jour. N.Y.
Entomol. Soc. 12: 249-250.

Eardley, A. J. 1962. Structural geology of North America,
ed. 2. Harper and Row. New York.

Eaton, A. 1818. A manual of botany for the northern and
middle states Of America. Albany.

Ecklon, C. F. and C. Zeyher. 1834 - 1837. Enum. p1. Africae
australis extratIOpicae. Hamburg.

Edwards, W. N. and F. M. Wonnacott. 1935. Fossilium Cat-
alogus II: Plantae, Part 20: Anacardiaceae.
W. Junk's Gravenhage.

Elliott, S. 1821. Sketch of the botany of South-Carolina
and Georgia, Vol. 1. J. R. Schenck. Charleston,
S.C.

Engler, A. 1881. Ueber die Morphologische Verhaltnisse und
die geographische Verbeitung der Gattung Rhus und
der mit ihr verwandten Anacardiaceae. Bot. Jahr.
Supt. 1, Vol. 4: 365 - 426.

. 1898. Beitrage zur Flora von Afrika. xv.
Anacardiaceae africanae ii. Bot. Jahr. Supt. 24:
493 - 502.

 

. 1883. Burseraceae und Anacardiaceae, in De
Candolle. Monographiae Phanerogamarum., Vol. 4.
Paris.

 

Engler, A.

A

C

Be

Epstein, I

an,
Erdtnx V

 

1:aegri,

FaSSet.

FthSC

478

Engler, A., und K. Prantl. 1892. Die natfirlichen Pflanzen-
familien 3: 138 - 178.

1897. Die natfirlichen Pflanzenfamilien 5.

 

Berlin.

Epstein, W. L. 1958. Rhus dermatitis: fact and fiction.
Bull. Kaiser Found. Med. 6: 197 - 204.

Erdtman, G. 1952. Pollen morphology and plant taxonomy,
Vol. 1. Chronica Botanica Co., Waltham, Mass.

. 1954. An introduction to pollen analysis.

 

Chronica Botanica Co., Waltham, Mass.

. 1963. "Palynology" in R. D. Preston (ed.) Advan-
ces in botanical research. Academic Press. New
York.

 

, and Vishnu-Mittre. 1958. On terminology in pol-
len and spore morphology. Grana Palyno. l: 6 - 9.

 

, and H. Straka. 1961. CormOphyte Spore classifi-
cation--an outline based on the apertures (tremata).
Forh. Geol. Foren. Stockholm 83: 65 - 78.

 

Ettingshausen, C. von. 1888. Die fossile Flora von Leoben
in Steiermark. II. Akad. Wiss. Wien, Math.-Natur-
wiss. K1., Denkachr. 54: 319 - 384.

Faegri, K.,and J. Iversen. 1950. Text-book of modern
pollen analysis. COpenhagen.

Farwell, O. 1900. Flora of the park system, and throughout
the city of Detroit. Ann. Rept. Commissioner Parks,
Boulevards Detroit 11: 73.

. 1930. Botanical gleanings in Michigan vi. Amer.
Midl. Naturalist 12: 113 - 134.

 

Fassett, N. C. 1940. Preliminary reports on the flora of
Wisconsin xxix. Anacardiaceae. Trans. Wisconsin
Acad. Sci. 32: 103 - 106.

. 1941. Mass collections; Rubus Odoratus andR
parviflorus. Ann. Missouri Bot. Gard. 28: 299— -
374.

 

 

Fedtschenko, B. A. 1914. Flora Asiatskoi Rossii, Part 5.
Sapindales. Colonization Bureau, Ministry of Agri-
culture. Moscow (in Russian).

479

Fernald, M. L. 1941. Another century Of additions to the
flora of Virginia: some varieties and forms of
Rhus radicans and Rhus Toxicodendron. Rhodora 43:
589 " 5990

 

. 1950. Gray's manual of botany, ed. 8. American
Book Co., New York.

 

Forbes, F. B. and W. B. Hemsley. 1886 - 1888. An enumera-
tion of all the plants known from China prOper,
Formosa, Hainan, Corea, and Luchu Archipelago, and
the Island of Hongkong. J. Linn. Soc. Bot. 23: 1 -
521.

Forbes, W. T. M. 1920. LepidOptera Of New York and neigh-
boring states. Part I - Primitive forms, Microlepi-
doptera, Pyraloids, Bombuces. Cornell Univ. Agric.
Exp. Sta. Mem. 68, Ithaca.

. 1948. LepidOptera of New York and neighboring
states, Part II. Geometridae, Sphingidae, Moto-
dontidae, Lymantriidae. Cornell Univ. Agric. Exp.
Sta. Mem. 274. Ithaca.

 

. 1954. LepidOptera of New York and neighboring
states. Part III. Noctuidae. Cornell Univ. Agric.
Exp. Sta. Mem. 329. Ithaca.

 

. 1960. LepidOptera of New York and neighboring
states. Part IV. Agaristidae through Nymphalidae.
Cornell Univ.

 

Gaertner, J. 1788. De fructibus et seminibus plantarum,
Vol. 1. Stuttgardt.

Gardner, C. A. and H. W. Bennetts. 1956. The toxic plants
of Western Australia. West Australian NeWSpapers
Ltd. Periodicals Division. Perth.

Gates, F. C. 1938. Woody plants, native and naturalized
in Kansas. Trans. Kansas Acad. Sci. 41: 99 - 118.

. 1939. New forms and nomenclatorial combinations
in the Kansas flora. Trans. Kansas Acad. Sci. 42:
135 " 1380

 

Gillis, W. T. 1960. Taxonomic problems in poison ivy.
Pap. Mich. Acad. Sci. 45: 27 - 45.

. 1962. Poison-ivy in northern Michigan. Michigan
801.. l: 17 - 22o

 

480

Goeppert, H. R. 1855. Die Tertiare Flora von Schnossitz
in Schlesien. Gorlitz.

Goin, C. J. 1958. Comments upon the origin Of the herpeto-
fauna of Florida. J. Florida Acad. Sci. 21: 49 - 60.

Gould, F. W. 1962. Texas plants - a checklist and ecologi-
cal summary. Publ. Texas Agric. Exp. Sta. MP-585.

Gray, A. 1846. Analogy between the flora of Japan and
that of the United States. Amer. J. Sci. 2: 135 -
136.

1850. Plantae Lindheimerianae. Part II. (An
account of a collection of plants made by F. Lindheiner
in the western part of Texas in the years 1845 - 6
and 1847 - 8, with critical remarks, descriptions
of new species, etc.) Boston J. Nat. Hist. 6: 141 -
240.

 

. 1856. Manual of the botany Of the northern
United States, ed 2.

 

1858. Diagnostic characters of new species of
phanerogamous plants, collected in Japan by Charles
Wright botanist of the U.S. North Pacific EXploring
Expedition. Mem. Amer. Acad. Arts. 6: 377 - 452.

 

. 1889. The similarity between the flora of Japan
and that of the United States, especially the Atlantic
side. Darwiniana. Appleton Co., New York.

 

Greene, E. L. 1905. Segregates of the genus Rhus. Leafl.
Bot. Observ. and Crit. 1: 114 - 144.

1910. Leafl. Bot. Observ. Crit. 2: 46 - 47.

 

Grimm, J. 1912. Entwicklungsgeschichtliche Untersuchungen
an Rhus und Coriaria. Flora 104: 309 - 334 + 2 p1.

Gronovius, J. F. 1762. Flora virginica. Lugd. Batavia.

Haack, N. H. 1940. Blaartrekkende Destanddeelen van Het
Melksap Uit Enkele Anacardiaceeen (Vesicant constitu-
ents of the milk-sap of some Anacardiaceae), thesis,
University Of Groningen, publ. by J. B. Wolters'
University-Company, Ltd. Batavia (in Dutch).

Hara, H. 1935. Preliminary report on the flora of southern
Hidaka, Hokkaido (Yezo) ix. Bot. Mag. (Tokyo) 49:
861 - 867, (in Japanese).

481

Hara, H. 1939. Critical notes on some East Asiatic
plants, ii. J. Jap. Bot. 15: 452 - 459 (in Japan-
ese .

Harper, R. M. 1906. A phytogeographical Sketch of the
Altamaha Grit region of the coastal plain of Georgia.
Ann. New York Acad. Sci. 17: 1 - 415.

Havard, V. 1885. Report on the flora of western and south-
ern Texas, part 1. Proc. U.S. Natl. Mus. 8: 449 -
533.

Harvey, W. H. and O. Sonder. 1859 - 1865. Flora capensis.
Dublin.

Hayata, B. 1908. Flora Montana Formosae. J. Coll. Sci.
Imp. Univ. Tokyo 25, Art. 19, p. 73 (in Japanese).

1917. General index to the flora Of Formosa, vi.
Supplement to Icones Plantarum Formosarum. Tokyo.

 

Heimsch, C. 1940. Wood anatomy and pollen morphology of
Rhus and allied genera. J. Arnold Arbor. 21:
279 ' 29]..

. 1942. Comparative anatomy of the secondary xylem
in the "Gruinales" and "Terebinthales" of Wettstein
with reference to taxonomic grouping. Lilloa 8:

83 - 198 (+ 942).

 

Hicks, S. D. 1952. Another insect on Rhus. ColeOpt. Bull.
6: 29.

. 1955. A dermestid in poison ivy. ColeOpt. Bull.
9: 9.

 

Hill, G. A., V. Mattacotti, and W. D. Graham. 1934. The
toxic principle of poison ivy. J. Amer. Chem. Soc.
56: 2736.

Horsfield, T. 1798. An experimental dissertation on the
Rhus vernix, Rhus radicans, and Rhus labrum: com-
monly known in Pennsylvania by the names of poison-
ash, poison-vine, and common sumac. Unpublished
dissertation. Univ. of Pennsylvania.

Hosokawa, T. 1954. Outline of the vegetation of Formosa
together with the floristic characteristics.
Angewandte Pflanzensoziologie Festschrift Aichinger,
Vol. 1.

482

Howden, H. F., A. T. Howden, and P. O. Ritcher. 1951.
Insects feeding on poison-oak (Rhus Toxicodendron
L.) ColeOpt. Bull. 5: 17 - 19.

Hultén, E. 1958. The amphi-Atlantic plants and their
phytogeographical connections. Kungl. Svenska
Vetenskapsakad. Handl., Fjarde Series, Vol. 7,
No. 1. Almqvist and Wiksell, Stockholm.

1964. The circumpolar plants 1. Vascular crypto-
gams, conifers, monocotyledons. Kungl. Svenska
Vetenskapsakad. Handl., Fjarde Series, Vol. 8, No. 5.
Almqvist and Wiksell, Stockholm.

 

Ikuse, M. 1956. Pollen grains Of Japan. Hirokawa Publish-
ing Co. Tokyo (in Japanese).

Iltis, H. H. 1965. The genus GentianOpsis (Gentianaceae):
transfers and phytogeographic comments. Sida 2:
129 - 154.

 

1966. The western element in the Eastern North
American flora and its phytogeographical implications.
Abstract of paper given at the meetings of the Botan-
ical Society of America, University Of Maryland,
College Park. 14 - 15 August 1966. Amer. J. Bot.

53: 634.

 

Jepson, W. L. 1936. Flora of California. 2: 445 - 446.

Jones, N. R. 1955. Economic losses due to poison oak derma-
tosis. Proc. Calif. Weed Conf. 7: 3 - 6.

Kalm, P. 1751. A short account of the natural position,
use, and care Of some plants, of which the seeds
were recently brought home from North America for
the service of those who take pleasure in experi-
menting with the cultivation of the same in our
climate. Royal Acad. Sci. translated in Ag. History
13: 33 - 64. 1939 by Esther Louise Larsen.

. 1770. The America Of 1750, also published as
Travels in North America 1: 1772. Translated by
John R. Forster. Wilson-Erickson Co., 1937.

 

Kearney, T. H. Jr. 1894. Some new Florida plants. Bull.
Torrey Bot. Club 21: 486 - 487.

Khittel, J. 1858. Analysis of the leaves of poison-oak
(Rhus Toxicodendron). Amer. J. Pharm. 30: 542 - 544.

 

483

Kimball, C. P. 1965. ArthrOpods Of Florida and neighbor-
ing land areas, Vol. 1 - LepidOptera Of Florida.
Fla. Dept. Ag. Gainesville.

Kligman, A. M. 1958. Poison ivy (Rhus) Dermatitis. A.M.A.
Arch. Derm. 77: 149 - 180.

Knowlton, F. H. 1900. A report on the fossil plants associ-
ated with the lavas of the Cascade Range. 20th
Ann. Rept. U.S.G.S. Part III: 37 - 52.

Koch, K. 1869. Dendrologie: Baume, Straucher, und
Halbstraucher 1: 573 - 582.

Koidzumi, H. 1924. On Rhus rishiriensis. Bot. Mag. (Tokyo)
38: 149 - 150 (in Japanese).

 

Krause, R. W.,and G. N. Kent. 1944. Analysis and correla-
tion of four New Hampshire bogs. Ohio J. Sci. 44:
11 - l7.

Krausel, R. 1919. Nachtrage zur Terifirflora Schlesiens III.
Uber einige Originale Goepperts und neuere Funde.
Jahrb. Preuss. Geol. Landesanst. 40: 363 - 433.

Kremp, G. O. W. 1965. Morphologic encyclOpedia of palyn-
ology. University of Arizona Press. Tucson.

Kryshtofovich, A. 1914. Les derniéres decouvertes des
rests deS flores sarmatique et méotique dans la
Russie méridionale. Bull. Acad. Imp. Sci. St.
Petersbourg 8: 591 - 602 (in Russian).

Kuntze, O. 1891. Revisio Generum Plantarum. Part I.
Leipzig.

Kuprianova, L. A. 1965. The palynology Of the Amentiferae.
Komarov Botanical Institute, Moscow (in Russian).

Laessle, A. M. 1958. The origin and successional relation-
ship of sand-hill vegetation and sand-pine scrub.
Ecol. Monogr. 28: 361 - 387.

Lanjouw, J.,and F. A. Stafleu. 1964. Index Herbariorum.
Part I. The herbaria of the world. ed 5. Regnum
Vegetabile 31. Utrecht.

Lanjouw, J. §t_a1. 1966. International code of botanical
nomenclature. Regnum Veg. 46. Utrecht.

Lavallée, A. 1877. Arboretum segrezianum. Paris.

3 /l /
Pflv mrwrn

LeCont

Li, Huj

Linnaeu

LivingS'

MacGinit

Maisch,

Majima,

I’I—Lf‘l—AH

MarShall ,

484

LeConte, J.,and G. H. Horn. 1883. Classification of the

ColeOptera of North America. Smithsonian Misc.
Collect. 26, NO. 507. Washington, D.C.

Li, Hui-Lin. 1963. Woody plants of Taiwan. Livingston

Publ. CO. Narberth, Pa.

Linnaeus, C. 1738. Hortus Cliffortianus. Amsterdam.

1753. Species plantarum, Vol. I. Facsimile Of

 

the first edition with an introduction by William
T. Stearn. 1957. Ray Society. London.

Livingston, B. E. 1903. The distribution of the upland

plant societies of Kent County, Michigan. Bot.
Gaz. 35: 36 - 55.

1905. The relation of soils to natural vegeta-

 

tion in Roscommon and Crawford Countie,s Michigan.
Bot. Gaz. (Crawfordsville) 39: 22 - 41.

MacGinitie, H. D. 1937. Eocene flora of western America.

Maisch,

Majima,

iii. The flora of the Weaverville beds Of Trinity
County, California. Publ. Carnegie Inst. Wash.

J. M. 1866. On the active principle of Rhus toxi-
codendron. Amer. J. Pharm. 38 (3rd Ser. Vol. 14):

 

4 - 12.

R. 1922a. Uber der Hauptbestandteil des Japan-
Lacks. viii. Mitteilung: Stellung der DOppelbind-
ungen in der Seitenkette des Urushiols und Beweuz-
fUrung, daSS das Urushiol eine Mischung ist. Ber.
Deutsch. Chem. Ges. 55: 172 - 191.

. 1922b. Uber der Hauptbestandteil deS Japan-

 

Lacks. xi. Mitteilung: Chemische Untersuchung der
verschiedenen naturlichen Lackarten, die dem Japan-
Lack nahe verwandt sind. Ber. Deutsch. Chem. Ges.
55: 191 - 214.

, und 8. Cho. 1907. Uber einen Hauptbestandteil

 

des japanischen Lackes. Ber. Deutsch. Chem. Ges.

Marchand, L. 1869. Revision du groupe des Anacardiacées.

J. B. Bailliere et fils. Paris.

Marshall, H. 1785. Arbustrum americanum. Philadelphia.

 

Martin,
Martin,

Masamune

Mason, }

McKelvej

McNair,

MQarnS

485

Martin, A. C., H. S. Zim, and A. L. Nelson. 1951. American
wildlife and plants. McGraw-Hill Book Co., New York.

Martin, W. E. 1959. The vegetation of Island Beach State
Park, New Jersey. Ecol. Monogr. 29: 1 - 46.

Masamune, G. 1934. Floristic and geobotanical studies on
the island of Hakusima, Province Osumi. Mem. Fac.
Sci. Taihoku Imp. Univ. 11: 1 - 638.

Mason, H. L. 1934. Studies Of the Pleistocene paleobotany
of California. Contributions to paleontology iv.
Pleistocene flora of the Tomales formation. Publ.
Carnegie Inst. Wash. 415: 81 - 179.

McKelvey, S. D. 1955. Botanical exploration Of the trans-
Mississippi West, 1790 - 1850. Arnold Arboretum
of Harvard University, Jamaica Plain.

McNair, J. B. 1916. The poisonous principle of poison
oak. J. Amer. Chem. Soc. 38: 1417 - 1421.

. 1921. Lobinol--a dermatitant from Rhus diversiloba
(poison oak). J. Amer. Chem. Soc. 43: 159 - 164.

 

. 1923. Rhus dermatitis, its pathology and chemo-
therapy. University of Chicago Press. Chicago.

 

. 1925. The taxonomy of poison ivy with a note
on the origin of the generic name. Publ. Field Mus.
Nat. Hist. 225, Bot. Ser. Vol. 4, No. 3.

 

. 1936. Five-leaflet poison oak. Bull. Torrey
Bot. Club 63: 473 - 476.

 

McVaugh, R. 1943. The vegetation of the granitic flatrocks
Of southeastern United States. Ecol. Monogr. 13:
1.20 " 1660

. 1945. The genus Triodanis Rafinesque, and its
relationships to Specularia and Campanula. Wrightia
l: 1.3 " 52.

 

. 1957. Personal recollections and military record
Of Colonel Valery Havard. Asa Gray Bull. n.s. 3:
177 - 1920

 

Mearns, E. A. 1902. Two new species of poisonous sumachs
from the states of Rhode Island and Florida. Proc.
Biol. Soc. Wash. 15: 147 - 149.

486

Michaux, A. 1803. Flora boreali-Americana Vol. 1. Paris

et Argent.

Milbrandt, A. C., A. R. Leggett, F. V. Brent, Jr., and

Miller,

W. C. Clark. 1963. Soil survey of Lincoln County,
Mississippi. U.S.D.A. Soil Conservation Service,
Series 1960. NO. 1. 1. U.S.G.P.O. Washington,
D.C.

P. 1754. The Gardeners Dictionary. London.

. 1768. The Gardeners Dictionary, ed. 8. London.

 

, and Thomas Martyn. 1807. The Gardeners and

 

Miquel,

Miyabe,

Moench,

Monk, C.

Morgan,

botanists dictionary. Law and Gilbert. ed. 9.
London.

Guillaume. 1867. Sur 1es affinites de la flore du
Japon avec celles de l'ASie et de 1'Amérique du
Nord. Arch. Néerl. Sci. Exact. Nat. 2: 136 - 156.

K. 1890. The flora of the Kurile Islands. Mem.
Boston Soc. Nat. Hist. 4: 203 - 275.

C. 1794. Methodus Plantae. Marburg.

D. 1960. A preliminary study on the relationships
between the vegetation Of a mesic hammock community
and a sandhill community. Quart. J. Florida Acad.
Sci. 23: 1 - 12.

G. D. 1933. Check list of the butterflies found
in and around Tampa, Florida. Privately printed.
Tampa.

Munting, A. 1713. Phytographia curiosa. Lugd.-Batavia.

Nakai, T. 1909. Flora koreana. J. Coll. Sci. Imp. Univ.

Tokyo 26: 1 - 304.

. 1916a. Notulae ad plantas Japoniae et Koreae.

 

xi. Bot. Mag. (Tokyo) 30: 140 - 148 (in Japanese).
. 1916b. On the nomenclature of Rhus ambigua and

 

B. orientalis. Bot. Mag. (Tokyo) 30: 205 - 206 (in
Japanese).

 

. 1922. Notulae ad plantas Japoniae et Koreae.

 

xxvii. Bot. Mag. (Tokyo) 36: 61 - 73 (in Japanese).

487

Pelton, J. F. 1951. Outline for ecological life history
studies in trees, Shrubs, and stem succulents.
Ecology 32: 334 - 343.

Pfaff, F. 1897. On the active principle of Rhus Toxico-
dendron and Rhus venenata. J. Exp. Med. 2: 181 -
195.

 

Pilar, G. 1883. Flora fossilis Susedana. Zagreb.

Poiret, J. L. M. 1817. Encyclopédie Méthodique, Supple-
ment, Vol. 5. Paris. .

Potbury, S. S. 1932. Studies of the Pleistocene paleo-
botany Of California. ii. A Pleistocene flora
from San Bruno, San Mateo County, California.
Publ. Carnegie Inst. Wash. 415: 25 - 44.

Potzger, J. 1948. A pollen study in the tension zone of
Lower Michigan. Butler Univ. Bot. Stud. 8: 161 -
176.

Powers, S. 1877. Tribes of California. Contr. N. Amer.
Ethnology 3: 1 - 635.

. 1816. Flora Americae Septentrionalis, ed. 2,
Vol. 2. London.

 

Rehder, A. 1939. New Species, varieties, and combinations
from the collections of the Arnold Arboretum.
J. Arnold Arbor. 20: 409 - 431.

, and E. H. Wilson. 1914. Anacardiaceae in Plantae

 

Wilsonianae. An enumeration of the woody plants

collected in Western China for the Arnold Arboretum

of Harvard University during the years 1907, 1908,
and 1910 by E. H. Wilson. Publ. Arnold Arbor. NO.
4, Part iv, Vol. 2: 172 - 184.

Rhoads, A. S. 1918. Some new or little known hosts for

wood-destroying fungi. PhytOpathology 8: 164 - 167.

Richard, A. 1847 - 1851. Tentamen florae Abyssinicae.
Paris.

Ridley, H. N. 1922. Flora of the Malay Peninsula. Vol. 1,

NO. 1. L. Reeve and Co., London.

. 1933. Additions to the flora Of Borneo and other

 

Malay Islands: iv. Kew Bull. 1933: 190 - 202.

 

. a
, , 1
1
a
. 1 I
\
\ ‘
‘
\
‘ r
7 x
0 ‘ '
o n
A V.
v " a
o
. v - e
C
a r) A n s . 1
, ~ 0
N .) I
‘v S
A
a a
T . a
A . 0 -
. s
C
‘ \
. s I c
x
. y o '
o 1 4‘ ‘
u a

 

488

Robinson, 8. L., and M. L. Fernald. 1908. Gray's Manual
of Botany, ed. 7. American Book Co., New York.

Rollins, R. C. 1959. Taxonomy and the international
association. Taxon 8: 277 - 279.

Ross, C. M. 1959. Poison ivy dermatitis; the first South
African cases. S. Afr. Med. J. 33: 657 - 660.

Rostenburg, A., Jr., 1955. An anecdotal biographical
history of poison ivy. A.M.A. Arch. Derm. 72:
438 - 445.

Salisbury, R. A. 1796. Prodromus Stirpium in Horto ad
Chapel Allerton Vigentium. London.

Saporta, G. de. 1888. Origine paléontologique des arbres
cultivés ou utilisés par l'homme. Paris.

Sargent, C. S. 1895. New or little-known plants. Rhus
michauxii. Gard. and Forest 8: 404 - 405.

Sauer, J. 1964. Revision of Canavalia. Brittonia 16:
106 - 181.

Scheele, A. 1848. Beitrage zur Flor von Texas. Linnaea
21: 576 - 602.

Schmidt, F. 1868. Reisen im Amur-lande und auf der Insel
Sachalin. Mem. Acad. Imp. Sci. St. Petersbourg,
Ser. 7, Vol. 12, No. 2: 1 - 227.

Schneider, C. K. 1907. Illustriertes Handbuch der Laub-
holzkunde 2: 148 - 156.

Selling, 0. 1952. Hawaiian pollen statistics 11. Special
Publ. Bernice F. Bishop Mus. 38. Honolulu.

Sessé, M. and J. M. Mocino. 1887. Plantae Nouae Hispaniae.
Bajos de San Augustin NO. 1.

Sharp, A. J. 1966. Some aSpects of Mexican phytogeography.
Ciencia (Mexico) 24: 229 - 232. .

Shirasawa, H. 1895. Die Japanischen LaubhOlzer im Winter-
zustande Bestimmungstabellen. Bull. Coll. Agric.
Imp. Univ. Tokyo 2, No;.5 (in Japanese).

Siebold, P. F. von and J. G. Zuccarini. 1835 - 1844. Flora
japonica. Lugd. Batavia.

 

 

Smi‘

Smi'

Sta

Ste

Ste

Ste

Ste

SWe

489

Smith, H. H. 1928. Ethnology of the Meskwaki Indians.
Bull. Publ. Mus. City of Milwaukee 4: 175 - 326.

. 1932. Ethnology of the Ojibwe Indians. Bull.
Publ. Mus. City of Milwaukee 4: 327 - 525.

 

Smith, J. 1609. Unpublished manuscript later published
(1882) by the Hakluyt Society of London and edited
by Sr. J. Henry Lefroy under the title "The history
Of the Bermudas or Summer Isles."

. 1624. The general history of Virginia, New
England, and the Summer Isles. Book 5.

 

State Of California Dept. of Public Health. 1957. OCCUpa-
tional disease report. Industr. Med. and Surg.

Stearn, W. T. 1957. Preface to the facsimile edition of
Species Plantarum by C. Linnaeus. Ray Society.
London.

Steenis, C. G. G. J. Van. 1962. The land-bridge theory
in botany. Blumea 11: 235 - 542.

Steudel, E. T. 1821. Nomenclator Botanicus, ed. 1.
Stuttgardt and Tubingen.

1841. Nomenclator Botanicus, ed. 2. Stuttgardt
and Tubingen.

 

Steyskal, G. 1951. Insects feeding on plants Of the Toxi-
codendron section of the genus Rhus (poison oak,
ivy, or sumac). ColeOpt. Bull. 5: 75 - 77.

 

Sweet, H. R. and F. A. Barkley. 1936. A most useful plant
family, the Anacardiaceae. Misouri Bot. Garden.
Bull. 24: 216 - 229.

Symes, W. F., and C. R. Dawson. 1954. Poison ivy "Urushiol."
J. Amer. Chem. Soc. 76: 2959.

Tanaka, C. 1916. On the nomenclature of Rhus orientalis.
Bot. Mag. (Tokyo) 30: 153 (in JapaneseL

Tatewaki, M. 1958. Forest ecology of the islands of the
North Pacific Ocean. J. Fac. Agric. Hokkaido Uhiv.

Taylor, W. H. 1917. Poison-ivy and Virginian creeper.
J. Agric. (New Zealand) 15: 162.

490

Thorp, J. §t_§1, 1959. Soil survey Of Creek County, Okla-
homa. U.S.D.A. Soil Conservation Service. Series
1950. NO. 5. Washington, D.C.

Thunberg, C. P. 1784. Prodromus.

Tournefort, J. P. 1700. Institutiones Rei Herbariae. Vol.
1. Paris.

Toyama, I. 1918. Rhus dermatitis. J. Cutan. Dis. Syph.
36: 157 - 165.

Turczaninoff, N. S. 1863. Animadversiones ad Catalogum
primum et secondum Herbarii Universitatis Charko-
viensis. Bull. Soc. Imp. Naturalistes Moscou
36: 612.

Turner, H. 1922. The ecology of Rhus Toxicodendron. Trans.
Illinois State Acad. Sci. 15: 208-211.

Turner, C. E. 1947. Rhus dermatitis as a public health
and health education problem. Amer. J. Publ. Health
NationS Health 37: 7 - 12.

Unger, F. 1849. Einige interessante Pflanzenabdrucke aus
der koni 1. Petrefacten-Sammlung in Munchen. Bot.
Zeitung Berlin) 7: 345 - 353.

Van der Ploeg, D. T. E. 1966. Rhus radicans in Freisland.
Gorteria 3: 30 - 31.(in Dutch).

Vaughan, T. W. 1910. A contribution to the geologic history
of the Floridian plateau. Publ. Carnegie Inst. Wash.
133: 99 - 185. -

Vestal, P. A. 1952. Ethnobotany of the Ramah Navaho. Pap.
Peabody Mus. Amer. Arch. Ethn. Harvard Univ. 40 NO.
4.

Vines, R. 1960. Trees, shrubs, and woody vines of the
Southwest. Univ. Texas Press. Austin.

Warren, L. E. 1910. Contribution to the bibliography of
the poisonous Species of Rhus. Midl. Druggist &
Pharm. Rev. 44: 149 - 150; 218 - 219; 288 - 290.

. 1910. Rhus michauxii: a non-poisonous plant.
Amer. J. Pharm. 82: 499 - 506.

 

 

Wells, B. W., and I. V. Shunk. 1931. The vegetation and
habitat factors Of the coarser sands of the North
Carolina Coastal Plain: an ecological study.
Ecol. Monogr. l: 465 - 520.

White

Willie.

Wilson,

Wodehous
E

Woodson,
p

Wyman, L.

491

White, J. C. 1873. On the action of Rhus venenata and
Rhus toxicodendron upon the human skin. New York
Med. J. 17: 225 - 249.

 

Williams, Howell. 1957. A reconnaissance geologic map of
the central part of the High Cascade Mountains.
State of Oregon Dept. Of Geology and Mineral Indus-
tries in COOperation with the U.S. Geological Survey.

Wilson, E. H. 1920. A phytogeographical Sketch of the
ligneous flora of Formosa. J. Arnold Arbor 2: 25 -
41.

Wodehouse, R. P. 1932. Tertiary pollen. 1. Pollen of the
living representatives of the Green River flora.
Bull. Torray Bot. Club 59: 313 - 340.

Woodson, R. E. 1947. Notes on the "historical factor" in
plant geography. Contr. Gray Herb. 165: 12 - 25.

Wyman, L. C., and S. K. Harris. 1941. Navajo Indian medi-
cal ethnobotany. Bull. Univ. New Mexico AnthrOp.
Ser. Vol. 3 NO. 5.