HARVARD UNIVERSITY

Library of the

Museum of
Comparative Zoology

SREAT BASIN NATURALIST MEMOIR:

nber 2

Brigham Young University

JUN 16 1978

d*\

HARVARD

INTERMOUNTAIN

w££m

BIOGEOGRAPHY:

^v W

A SYMPOSIUM

197

GREAT BASIN NATURALIST MEMOIRS

Editor. Stephen L. Wood, Department of Zoology, Brigham Young University, Provo, Utah
84602.

Editorial Board. Kimball T. Harper, Botany; Wilmer W. Tanner, Life Science Museum;
Stanley L. Welsh, Botany; Clayton M. White, Zoology.

Ex Officio Editorial Board Members. A. Lester Allen, dean, College of Biological and Agri-
cultural Sciences; Ernest L. Olson, director, Brigham Young University Press, Univer-
sity Editor.

The Great Basin Naturalist was founded in 1939 by Vasco M. Tanner. It has been
published from one to four times a year since then by Brigham Young University, Provo,
Utah. In general, only previously unpublished manuscripts of less than 100 printed pages in
length and pertaining to the biological natural history of western North America are ac-
cepted. The Great Basin Naturalist Memoirs was established in 1976 for scholarly works in
biological natural history longer than can be accommodated in the parent publication. The
Memoirs appears irregularly and bears no geographical restriction in subject matter. Manu-
scripts are subject to the approval of the editor.

Subscriptions. The annual subscription to the Great Basin Naturalist is $12 (outside
the United States $13). The price for single numbers is $4 each. All back numbers are in
print and are available for sale. All matters pertaining to the purchase of subscriptions and
back numbers should be directed to Brigham Young University, Life Science Museum, Pro-
vo, Utah 84602. The Great Basin Naturalist Memoirs may be purchased from the same of-
fice at the rate indicated on the inside of the back cover of either journal.

Scholarly Exchanges. Libraries or other organizations interested in obtaining either
journal through a continuing exchange of scholarly publications should contact the Brigham
Young University Exchange Librarian, Harold B. Lee Library, Provo, Utah 84602.

Manuscripts. All manuscripts and other copy for either the Great Basin Naturalist or
the Great Basin Naturalist Memoirs should be addressed to the editor as instructed on the
back cover.

RE AT BASIN NATURALIST MEMOIR!

INTERMOUNTAIN

BIOGEOGRAPHY:

A SYMPOSIUM

Printed in the United States of America

by Brigham Young University Printing Service

3-78 1.5M 29246

REAT BASIN NATURALIST MEMOIRS

Number 2

Brigham Young University

1978

INTERMOUNTAIN

BIOGEOGRAPHY:

A SYMPOSIUM

K. T. Harper

and

James L. Reveal

Symposium Organizers

CONTENTS

Page

Preface 1

The biota of the Intermountain Region in geohistorical context. Arthur Cronquist .... 3

Biogeography of intermountain fishes. Gerald R. Smith 17

Zoogeography of reptiles and amphibians in the Intermountain Region. Wilmer W.

Tanner 43

Avian biogeography of the Great Basin and Intermountain Region. William H.

Behle 55

The flora of Great Basin mountain ranges: Diversity, sources, and dispersal ecology.

K. T. Harper, D. C. Freeman, W. K. Ostler and L. G. Klikoff 81

Alpine phytogeography across the Great Basin. W. D. Billings 105

Phytogeographical variation within juniper-pinyon woodlands of the Great Basin.

Neil E. West, Robin J. Tausch, Kenneth H. Rea, and Paul T. Tueller 119

Patterns of avian geography and speciation in the Intermountain Region. Ned K.

Johnson 137

Explosive evolution of perennial Atriplex in western America. Howard C. Stutz 161

Distribution and phylogeny of Eriogonoideae (Polygonaceae). James L. Reveal 169

Problems in plant endemism on the Colorado Plateau. Stanley L. Welsh 191

Some factors governing plant distributions in the Mojave-Intermountain Transition

Zone. Susan E. Meyer 197

The theory of insular biogeography and the distribution of boreal birds and mam-
mals. James H. Brown 209

Biogeography and management of native western shrubs: A case study, Section Tri-

dentatae of Artemisia. E. Durant Mc Arthur and A. Perry Plummer 229

Applying biogeographic principles to resource management: A case study evaluating

Holdridge's Life Zone model. James A. MacMahon and Thomas F. Wieboldt ... 245
Index 259

No. 2

Great Basin Naturalist Memoirs

Intermountain Biogeography:
A Symposium

Brigham Young University, Provo, Utah

1978

K. T. Harper' and James L. Reveal2

PREFACE

Most of the Intermountain Region is re-
mote from the nation's major transportation
arterials. As a consequence, the region's
beauty and its biological resources are
largely . unappreciated. The area is often
considered to be a wasteland, a desert with
little or no life, and a land of minimal val-
ue. Because few initially understood or ap-
preciated the region, its resources have of-
ten been abused by a variety of human
activities ranging from military weapon
testing to off-road vehicle travel. Addition-
ally, ranchers and governmental land man-
agers have had no precedents to guide their
activities in the unique and often fragile
ecosystems of the region. As a consequence,
the biological landscape has been markedly
altered by grazing, control of wildfires, and
agronomic practices. With time and knowl-
edge, concern for the natural landscape of
the West has grown. National and state leg-
islation now imposes strict guidelines on
most new development activities and re-
quires a balanced analysis of the full impact
of major activities on the land. Even rare
native species have advocates in high places
and now enjoy some legal protection.

Although knowledge has accumulated and
management of the biological resources of
the region has steadily improved, there is
still much to do. New facts must be ac-
cumulated and those now known must be
digested, disseminated, and put to work in
management by individuals and govern-
ments. This volume brings together current
information on a broad spectrum of the na-
tive biota of the region. Most of the authors
Consider the management implications of
their findings. We hope the volume will in-
form and assist resource managers charged
with preserving the ecological health of the
Intermountain West.

For the biologist, this book presents the
first major overview of current biogeogra-
phical research being conducted by a num-
ber of individuals and institutions in the in-
termountain region. Of more importance,
however, this symposium represents the first
attempt to bring both plant and animal sci-
entists and management-oriented researchers
together to discuss and evaluate the bio-
geographic principles at work in the area.
Topics discussed include distribution pat-
terns for fishes, reptiles, amphibians, birds,

'Department of Botany and Range Science, Brigham Young University, Provo, Utah 84602.

'Department of Botany, University of Maryland, College Park, Maryland 20742, and National Museum of Natural Hi!
Washington, DC. 20560.

Smithsonian Institution,

GREAT BASIN NATURALIST MEMOIRS

No. 2

small mammals, and plants within the inter-
mountain region. Special reviews are pres-
ented on Artemisia, Atriplex, and the genus
Eriogonum and its relatives. Several broad
biological problem areas within the region
are reviewed, including the nature of the
floristic transition zone between the Mojave
and Great Basin deserts, the endemic flora
of the Colorado Plateau, the distribution of
the juniper-pinyon community in the Great
Basin, and the evolutionary development of
the alpine biota of the Intermountain Re-
gion. Special considerations are given to the
problems of managing native plant and ani-
mal populations in the area.

The general public will be especially in-
terested in the role biogeographic consid-
erations are now playing in the under-
standing of the present-day distribution of
organisms within the Intermountain Region
and the management options available for
the use and protection of these vital natural
resources. Public land managers will find
that many of the biogeographic principles
discussed will help them to better under-
stand the nature of arid land resources, en-
dangered species, and the impact of man's
technology in the American West.

Each chapter stands as a unit with an in-
troduction, a discussion, and a summary of
pertinent literature. An index completes the
volume.

The Intermountain Biogeography Sym-
posium was held at the University of Mon-
tana, Missoula, 14-15 June 1976. The sym-
posium was sponsored by Brigham Young
University and the Intermountain Forest
and Range Experiment Station of the U.S.
Forest Service, with the cooperation of the
Botanical Society of America and the Pacif-
ic Section of the American Association for
the Advancement of Sciences. The sym-
posium was arranged by Kimball T. Harper
of Brigham Young University and Ralph C.
Holmgren of the U.S. Forest Service, with
the assistance of Dr. George Edmunds of
the University of Utah, Dr. Joseph Murphy
of Brigham Young University, Dr. Neil
West of Utah State University, and Edward
Smith of the Bureau of Land Management.
Major financial support for the symposium
and the publication of these proceedings has
been provided by Brigham Young Univer-
sity and the U.S. Forest Service. We thank
Dr. Stephen L. Wood, editor of the Mem-
oirs series of the Great Basin Naturalist and
the staff of Brigham Young University Press
for their help in bringing the book to pub-
lication. The cover was designed by Kaye
H. Thorne.

K. T. Harper

and

J. L. Reveal

Symposium Organizers.

THE BIOTA OF THE INTERMOUNTAIN REGION IN GEOHISTORICAL CONTEXT

Arthur Cronquist'

Abstract.— The present Great Basin Floristic Province had achieved roughly its present topographic con-
formation by some time in the Miocene epoch and had a climate not too different from the present one, though
probably a little warmer, moister, and less continental. Both the flora and the fauna took on a fairly modern as-
pect during the Miocene, as a result of worldwide evolutionary changes and more specific adaptation to the con-
ditions of the region. Changes in the biota since that time mainly reflect evolution and migration at the level of
species and, to a lesser extent, genera, in response to regional conditions and the repeated fluctuations in climate.
The climatic reversals of the Pleistocene caused repeated inverse migrations of more northern, mesophytic ele-
ments in the flora, on the one hand, and more southern, xerophytic elements on the other. These expansions and
contractions of range favored hybridization and genetic mixing among related plant species. The fauna of the re-
gion, dependent eventually on the flora, must have been subjected to basically the same set of repeated changes
in range and local distribution during the Pleistocene. About 10,000 years ago many of the large mammals in the
Intermountain Region, as elsewhere in North America, rapidly became extinct, perhaps largely through overkill
by primitive man.

A proper understanding of the present is
always facilitated by some knowledge of the
past. Therefore I want to say something
about the geological and biological history
of the Intermountain Region, to help pro-
vide a proper setting for the other papers
of this symposium. Nearly everything that I
have to say is already in the scientific liter-
ature somewhere, but the particular syn-
thesis may be in part new.

As a first approximation of the truth, one
may say that the aspect of the vegetation of
any region is controlled by the climate, and
the taxonomic composition of the flora is
determined by the climate and the history.
The general nature of the fauna is in turn
determined by the vegetation, and the tax-
onomic composition of the fauna is deter-
mined by the vegetation and the history.

It is, of course, also true that the fauna
influences the flora. One of the reasons that
grasses predominate in certain climates is
that they are better adapted to withstand
grazing than are most other herbaceous
plants. Furthermore, evolution of floral
structure is to some extent correlated with

the evolution of pollinating insects (Leppik
1957), and particular species of plants may
become dependent on particular pollinators.
A notable example that may be familiar to
many readers is provided by Yucca and the
Tegeticula moth. Some of the importance of
the influence of the fauna on the flora is
also shown by the devastating effect of the
introduction of goats to some of the islands
off the Pacific Coast of southern North
America. More complex interactions be-
tween plants and animals also occur. Yet
the preponderant control is that exerted by
the food makers (plants) on the food eaters
(animals). Therefore it is reasonably possible
to consider the vegetation and flora of a re-
gion with only secondary attention to the
fauna, whereas any proper consideration of
the fauna must be grounded in a knowledge
of the vegetation. These facts, or what I
take to be facts, are fortunate for me, be-
cause I know a lot more about plants than I
do about animals.

The Intermountain Region may be vari-
ously delimited. For purposes of this dis-
cussion, I take its limits to be those of the

The New York Botaima! Carilen. Bronx, New York 10458.

GREAT BASIN NATURALIST MEMOIRS

No. 2

Great Basin Floristic Province, as defined
by Gleason and Cronquist (1964). In large
part these limits are the same as those of
the Intermountain Flora (Cronquist et al.
1972), but the Great Basin Floristic Pro-
vince extends somewhat further south in-
to Arizona and also includes a part of north-
western New Mexico as well as a sliver of
western Colorado. In addition to the hydro-
graphic Great Basin, the area under consid-
eration also includes the Snake River Plain
and the more westerly segment of the Colo-
rado Plateau. The region has a continental
climate, with fairly hot, dry summers, and
cold, snowy winters. The lowlands and foot-
hills are largely desert and semidesert; a
more mesophytic flora often occupies the
upper elevations. South of the Inter-
mountain Region lie hotter deserts, marked
especially by milder winters. These southern
deserts have a rather different flora, and the
plant communities are often dominated by
Larrea. The southern deserts are not a part
of the Intermountain Region as here de-
fined.

Geologic History

We shall start our consideration of the in-
termountain biota with a summary of the
geologic history of the region from the Cre-
taceous period to the present. Much of the
information in this section comes from pa-
pers by Bateman (1968), Eardley (1968), and
Roberts (1968).

The present Intermountain Region has
been at middle latitudes since before the
beginning of the Cretaceous. North America
has drifted westward, with respect to Eu-
rope and Africa, throughout that time, but
the latitude of our area has changed rela-
tively little (Dietz and Holden 1970, Smith
et al. 1973).

Our region has been subjected to repeat-
ed and almost continuous tectonic distur-
bance, leading to uplift and erosion, from
the beginning of the Cretaceous to the pres-
ent. The terrain throughout that time has
been highly varied, doubtless producing a

diversity of habitats. The Upper Cretaceous,
in particular, was a time of great and pro-
longed uplift in western Utah and eastern
Nevada. There was a large interior drainage
basin in north-central Nevada even in the
late Upper Cretaceous, and in mid-Eocene
time there appear to have been high moun-
tains and large lake basins throughout most
of the present hydrographic Great Basin. In
Oligocene time these lake basins were con-
siderably elevated and themselves subjected
to erosion.

The present Rocky Mountains and Colo-
ado Plateau began to rise early in the Ter-
tiary, and they have continued to rise at
varying rates until the present. The Sierra
Nevada, bounding the Great Basin on the
west, also has a long history. After a rela-
tive quiescence during the Oligocene, the
tilt-uplift of the Sierra Nevada was consid-
erably accentuated during the Miocene. The
concurrent uplift of the Rocky Mountains
and Colorado Plateau shaped the Great Ba-
sin. By some time in the Miocene, it ap-
pears that "basin and range topography ex-
tended from the Wasatch Mountains to the
Sierra Nevada, and most of the area drained
into interior basins" (Roberts 1968). There is
some difference of opinion on the timing,
however, and Axelrod (1950) believes that
the present interior drainage of the Great
Basin dates from near the close of the
Pliocene.

The Snake River Plain, forming a broad
crescent across southern Idaho, belongs to
the Great Basin floristic province but is
geologically distinctive. The Upper Cre-
taceous uplift in western Nevada and east-
ern Utah extended across the present Snake
River Plain as well. During Eocene time
the present Snake River Plain was buried
by lava in a major and prolonged tectonic-
disturbance that formed a volcanic plateau
extending from western Wyoming across
southern Idaho and probably into eastern
Oregon. In Oligocene time the Snake River
basin took shape, possibly "as a tension rift
in the lee of the Idaho batholith" (Axelrod
1968), which began to drift north. Sub-

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

siclence of the basin and outpouring of new
lava flows have continued until the present
time. The youngest deposits in the Craters
of the Moon region at the north edge of the
Snake River Plain are probably only a few
hundred years old.

It is thought that for most of the Cre-
taceous period the climate of the world was
relatively warm and equable, and that trop-
ical and subtropical climates entirely suit-
able for the growth of forests extended from
about 60 degrees north to about 60 degrees
south (Barnard 1973). As late as the Eocene,
the London Clay flora, at 35 degrees north,
is definitely tropical (Hughes 1973). One
may reasonably have some doubt about how
humid the climate may have been in Ne-
vada during the Upper Cretaceous, because
of the presence of an interior drainage ba-
sin, but such Cretaceous fossil floras as we
have from the western United States suggest
the presence of adequate moisture.

The frequent presence of interior drain-
age basins in the Intermountain Region for
many millions of years past tells us some-
thing about the climate. The precipi-
tation/evaporation ratio for much of the re-
gion much of the time must have been
something less than 1. At a p/e ratio of
more than 1, lake basins fill and spill over,
finding external drainage. It is generally
considered that a p/e ratio of not less than
about 1 is required to support a forest.
Therefore, for much or most of its span of
existence the Great Basin is not likely to
have been widely forested. Other parts of
the Intermountain Region appear to have
had a similar climatic regimen.

Island Topography

The topographic diversity of the Inter-
mountain Region, with its associated differ-
ences in temperature and moisture, effec-
tively converts the habitats for many species
of plants and animals into a series of is-
lands. Not only the mountains, but also the
valleys, form such islands for species with-
out good means of dispersal. Birds can trav-
el from one island to another, but small

mammals frequently cannot. Different kinds
of plants likewise differ in the ability to
pass the inhospitable stretches between is-
lands.

On the other hand, these islands do not
have the relative permanence of oceanic is-
lands. The various island habitats in the In-
termountain Region have expanded and
merged, contracted and broken up, dis-
appeared and reappeared, during Pleisto-
cene and post-Pleistocene time because of
changes in the climate. The principles of is-
land biogeography, as expounded for ex-
ample by Mac-Arthur and Wilson (1967), are
pertinent to the Intermountain Region, but
their effect is limited by the climatically
controlled changes in island area.

Early Angiosperm Evolution

The angiosperms appear to have origi-
nated early in the Cretaceous. Since we do
not have fossils to connect the angiosperms
to their necessarily gymnospermous ances-
tors, we cannot say with certainty that they
did not originate somewhat earlier. The fos-
sil record does make it clear that the evolu-
tionary diversification of the group did not
get well started until the Cretaceous. Ang-
iosperms enter the fairly early Lower Cre-
taceous fossil record as an uncommon and
not highly diversified group. Many of these
early angiosperm fossils were at first opti-
mistically identified with modern genera,
leading to the widespread belief that the
angiosperms entered the fossil record full-
blown. We can now say with some assur-
ance that the reverse is true. The pollen
record speaks eloquently to the relative ho-
mogeneity of the early angiosperms, and a
reexamination of the megafossils shows that
their identification with modern genera was
disastrously incorrect. The purportedly Ju-
rassic palm from Utah (Tidwell et al. 1970)
is clearly a palm, but it is not Jurassic. The
stratigraphy of the site where it was collect-
ed is complex, and subsequent careful study
shows that it is of Tertiary age (Scott et al.
1972).

6

GREAT BASIN NATURALIST MEMOIRS

No. 2

The comments made in this paper on
angiosperm evolution in general are heavily
influenced by studies in the past decade by
Dilcher (1969, 1973), Doyle (1969), Hickey
(1973), Walker and Doyle (1975), Doyle and
Hickey (1976), Hickey and Wolfe (1975),
and Wolfe et al. (1975), who are in the
forefront of the ongoing reevaluation of the
early angiosperm fossil record. Their pub-
lished work and my conversations with
them have helped to shape my views, and I
am particularly indebted to Dr. Leo Hickey
for advice and counsel during the prepara-
tion of this paper. Within the Inter-
mountain Region, the work of Axelrod
(1948, 1950, 1952, 1956, 1958, 1964, 1966,
1968, 1975) is of course preeminent. With-
out it, our knowledge of the fossil flora
would be scanty indeed. The interpretation
presented is, as always, my own; those who
helped me are not to be held responsible
for what I might say.

The place of origin of the angiosperms is
still uncertain. It is clear that they are basi-
cally a tropical group, but beyond that the
situation is debatable. We can say that as
early as the Aptian stage of the Lower Cre-
taceous, 125 million or more years ago,
they were well scattered in both Gon-
dwanaland and Laurasia, including North
America, but that they did not begin to
dominate the landscape until the Upper
Cretaceous. There is no reason to suppose
that the Intermountain Region had anything
to do with the origin of the angiosperms,
but at the same time it is clear enough that
it has supported some angiosperms at least
from the Albian stage of the lower Cre-
taceous to the present.

Unfortunately we can not yet see a his-
torical connection between the Cretaceous
and Tertiary angiosperm floras of the Inter-
mountain Region, or indeed of most other
parts of the world. Most of the Cretaceous
genera did not persist long if at all into the
Tertiary, and the limited fossil record does
not show whether our early Teritary genera
originated in situ from the Cretaceous ones
or migrated in from elsewhere. One of the-

few Upper Cretaceous fossil floras definitely
known from within the Intermountain Re-
gion is in the Blackhawk formation in cen-
tral Utah, a member of the Mesa Verde
Group (Parker 1968, as reported by Tidwell
et al. 1972). This flora included some palms
and a number of woody dicotyledons and is
thought to indicate humid lowland condi-
tions under a warm-temperate to sub-
tropical climate. This is in general harmony
with views of the Cretaceous climate of the
region based on other data (e.g. Axelrod
1950).

Beginning with the Paleocene, we have a
more nearly continuous history of the Inter-
mountain flora, but even so there are some
considerable gaps. Well over half of the Pa-
leocene genera of angiosperms in the world
flora are now extinct, and the fossil record
as studied to date rarely shows the origin of
modern genera from the more archaic ones.
It appears that in the Paleocene
the climate of the Intermountain Region
was still reasonably warm and moist, sub-
tropical or warm temperate, as it had been
in the Upper Cretaceous.

Evolution of Floristic Groups
in the Intermountain Region

By the middle of the Eocene, some 50
million years ago, the climate in the Inter-
mountain Region had begun to dry out. The
first indication of this in the fossil record
comes in the Eocene Green River flora of
northwestern Colorado and northeastern
Utah (Axelrod 1950, MacGinitie 1969). This
resembles the early Oligocene Florissant
flora from Colorado (MacGinitie 1953) and
like it may have been a subtropical sa-
vanna-woodland. No closely similar flora ex-
ists today.

Drying of the Intermountain climate con-
tinued, with some fluctuations, throughout
the Tertiary. By early Oligocene the mean
temperature of the world, at least in pres-
ently temperate regions, had begun to drop
(Bowen 1966), and in late Oligocene it
dropped markedly (Wolfe and Hopkins

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

1967); it never again regained the Cre-
taceous levels. Concomitant with increasing
aridity and decreasing mean temperature in
the Intermountain Region was a gradual
trend toward a more continental climate,
with hot, dry summers and cold, somewhat
moister winters, continuing until about the
middle of the Pliocene.

Floristic changes in the Intermountain
Region were, of course, related to the evo-
lutionary diversification of the angiosperms
throughout the world. The monocotyledons
evidently diverged from primitive dicotyle-
dons shortly after the appearance of angio-
sperms in the fossil record, during the
Lower Cretaceous. Palms became important
elements of the world flora during the Cre-
taceous, and grasses in the Oligocene or
earlier. Dicotyledonous herbs were rare
throughout the Cretaceous and on into the
Paleocene and Eocene. They began to be-
come more abundant in the Oligocene, and
they increased dramatically at the beginning
of the Miocene, some 25 million years ago.
During Miocene time the flora of the world
began to take on a fairly modern aspect,
with a great many genera that still exist
today. The increase in dicotyledonous herbs
in the mid-Tertiary is thought to reflect at
least in part the increasing aridity of the
climate throughout much of the world, en-
larging the area not suitable for forests.

It is evident that during the drying of the
climate in western North America the Ter-
tiary flora sorted itself out into a more
northern, mesic flora dominated by trees,
and a more southern, xeric flora with few if
any trees. Fossil floras from near the Oligo-
cene-Miocene boundary in southwestern
Montana suggest a mainly forest vegetation,
with some elements from the drylands to
the south (Becker 1969). Within the dryland
flora there was a further differentiation into
a more northern segment adapted to cold
winters, and a more southern segment
adapted to a warmer climate. The present
Great Basin Floristic Province, representing
the more northern of these two dryland
floras, evidently took shape in the Miocene.

Indeed the Miocene boundary between the
Great Basin flora and the Mohave Desert (a
part of the more southern flora) may have
been about where it is now (Axelrod 1950).

It is not clear how much of the differen-
tiation of the intermountain flora during the
Tertiary represents evolution in situ, and
how much of it reflects immigration from
other regions. Certainly both processes oc-
curred. A similar sorting out occurred in
other parts of the world, and in Eurasia this
involved many of the same families and
even genera. It is not likely that the same
taxonomic groups originated independently
in North America and Asia. There must
have been some interchange.

The genus Artemisia might be considered
in this regard. Although Artemisia tridentata
Nutt. and its immediate allies dominate the
scene in much of the Intermountain Region,
Artemisia is not of American origin. The
tribe Anthemideae of the family Asteraceae
(Compositae), to which Artemisia belongs, is
basically an Old-World tribe, and most of
the species of Artemisia itself occur in the
Old World rather than in the new. Arte-
misia and Juniperus characterize the land-
scape in parts of Armenia, for example, as
well as in the Intermountain Region. Arte-
misia in the western United States is an im-
migrant, although the particular species we
now have may well have originated here
from immigrant ancestors.

Some other members of the Asteraceae
are definitely American. The whole tribe
Heliantheae is clearly so. Its present center
of diversity is in the arid highlands of cen-
tral Mexico, and it seems reasonable to sup-
pose that the tribe is of Mexican or western
American dryland origin. Many members of
the group here will probably be acquainted
with species of Balsamorhiza, Chaenactis,
Enceliopsis, Eriophyllum, Viguiera, and
Wyethia, all members of the Heliantheae,
that grow in the Intermountain Region. The
large genus Haplopappus, in the tribe As-
tereae, is strictly American (North and
South), with one center of diversity in west-
ern North America and another in

8

GREAT BASIN NATURALIST MEMOIRS

No. 2

Chile. Erigeron is another large genus of the
Astereae that has its principal center of di-
versity in western North America and ap-
pears to have originated there. I am not
suggesting that these several genera of He-
liantheae and Astereae originated in the In-
termovmtain Region, but they probably did
not have far to come to get here.

Atriplex, another important genus in the
Intermountain Region, has more species in
the Old World than in the New. The family
Chenopodiaceae, to which Atriplex belongs,
has considerable concentrations of species in
the Mediterranean region, in western and
central Asia, in South Africa, and in Austra-
lia, as well as in the drier parts of both
North and South America.

The Roraginaceae appear to be tropical
and woody in origin, but they are well rep-
resented by numerous herbaceous genera
and species not only in our arid West but
also in the Mediterranean region and in
central Asia. To what extent did our boragi-
naceous intermountain herbs originate in
North America from tropical woody ances-
tors, and to what extent do they reflect im-
migration of herbs from the Old World?
Certain genera, such as Cryptantha and
Plagiobothrys, are clearly American now,
whatever their eventual origin, but others,
such as Lithospermum, are well developed
in Eurasia and may well be immigrants in
North America.

The Rrassicaceae are well represented in
the Intermountain Flora, but they are even
more numerous and diversified in the arid
region from central Asia to the Mediterra-
nean, and the family as a whole is probably
of Old World origin. Such familiar genera
as Lesquerella, Physaria, Stanleya, Strep-
tanthus, and Thelypodium are strictly
American, whatever the origin of the family
as a whole. Cardamine, Lepidium, and Ror-
ripa, on the other hand, are well represent-
ed in the Old World also.

A few families, such as the Hydro-
phyllaceae and Polemoniaceae, evidently
have their principal center of diversification
in western North America, even if the re-

gion of their ultimate origin is not yet
clearly established. Such genera as Phacelia,
in the Hydrophyllaceae, and Cilia, in the
Polemoniaceae (Grant 1959), are clearly at
home in the Intermountain Region. There is
no reason to suppose that they came in
from some other continent.

Axelrod and Chaney have in various pa-
pers (e.g., Axelrod 1958) promoted the
thought that the Tertiary flora of the west-
ern United States can be divided into an
Arcto-Tertiary and a Madro-Tertiary seg-
ment. The Arcto-Tertiary geoflora, domi-
nated by deciduous trees, is considered to
have been very wide-spread, extending
across most of northern North America and
northern Eurasia. The deciduous forest of
the eastern United States is considered to be
the nearest modern American counterpart
and a lineal descendant of the Arcto-Ter-
tiary geoflora. The Madro-Tertiary geoflora,
on the other hand, was adapted to drier,
warmer conditions, with many xeromorphic
shrubs, the trees being restricted to favor-
able habitats, or completely wanting. The
Madro-Tertiary geoflora as so conceived
was geographically more restricted than the
Arcto-Tertiary, being confined to northern
Mexico and the southwestern United States.
The "Madro" part of the name comes from
the Sierra Madre Occidental in north-
western Mexico. The Madro-Tertiary flora
is considered to have originated in situ from
subtropical western American plants that
gradually became adapted (through evolu-
tion) to xeric conditions.

The concept of Arcto-Tertiary and
Madro-Tertiary floras has recently been
challenged by a number of authors, notably
Wolfe (e.g., 1969), and is now in some dis-
repute. The problem, to my mind, is that
some useful generalizations have been taken
too literally and interpreted too rigidly. I
am reminded of Gleason's challenge (1926)
to Clementsian concepts of plant associ-
ations. Most modern ecologists agree with
Gleason that the association is a mental
construct that can be defined only arbi-
trarily. The idea that the community is an

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

9

organism is a good aphorism, but it can
lead to serious misunderstanding if it is
taken literally. Likewise the concept of an
Arcto-Tertiary and a Madro-Tertiary geof-
lora is useful if one conceives of these floras
broadly and loosely and recognizes that
each of them encompasses a considerable
amount of diversity, that some elements
were common to both, and that there was
continuous interchange between them. It is
helpful to think in terms of floristic groups,
but we should keep it constantly in mind
that each species has its own limits of eco-
logical tolerance, its own means of migra-
tion, and its own evolutionary potentialities,
the last being influenced also by hybridiza-
tion with related species. The species that
make up any floristic group have entered
that group, through immigration or through
evolution in situ, at various times in the
past, and species that are now associated
may not' remain associated under some fu-
ture climatic regimen.

I can easily agree with Axelrod that the
modern desert flora of the western United
States and northern Mexico probably "de-
veloped during the Tertiary period by grad-
ual adaptation of more mesic plants to
slowly expanding dry climate" (Axelrod
1950). It seems perfectly logical to suppose
that the present flora of the warm deserts
south of the Intermountain Region is a line-
al descendant of a Madro-Tertiary geoflora
that differentiated originally from American
plants adapted to similarly warm but more
mesic climates. There is no other likely
source. Some of them doubtless originated
instead by adaptation of Arcto-Tertiary taxa
to warmer, drier climates, and some of
these that entered the warm deserts from
the north doubtless take their origin eventu-
ally in Asia, but it strains credulity to de-
rive the bulk of the Madro-Tertiary flora in
such a way.

The origin of the Arcto-Tertiary flora is a
more difficult question. Obviously it repre-
sents an adaptation of tropical or sub-
tropical plants to a cooler but still moist
climate. Since it extended across both North

America and Eurasia, one cannot a priori
assume that it came principally from either
an Old- World or a New World source. It
seems logical to suppose that the Arcto-Ter-
tiary flora originated from the Cretaceous
tropical and subtropical Laurasian flora, but
at the present time that is pure speculation.
We have noted that the angiosperm fossil
record as presently understood does not
provide a good connection between the
Cretaceous and the Tertiary. The problem
is complicated by the fact that during the
Mesozoic era and most of the Tertiary peri-
od North and South America appear to
have been separated, not contiguous. South
America was part of the southern continent,
Gondwanaland, whereas North America was
part of the northern continent, Laurasia.
North America drifted away from Europe
during and after the Cretaceous, but, until
recently, it has mostly been well separated
from South America. I say "mostly," be-
cause the geologic history of the Caribbean
is complex and insufficiently understood,
and the possibility of a direct connection
between North and South America at some
time during the Cretaceous or early Ter-
tiary cannot be completely discounted.

At the present time the tropical part of
the flora of North America (as represented
by southern Florida, the West Indies, south-
ern Mexico, and Central America) is clearly
allied to the flora of South America. If
there is any surviving Laurasian element in
the present tropical North American flora,
it is so thoroughly amalgamated into the
Gondwanaland, South American flora that
no one has yet been able to recognize it. In
making this statement I exclude from con-
sideration some primarily temperate-zone
species and genera that extend into the
tropics at the southern limit of their range.

Although the vegetation of most of the
Intermountain Region is rather similar in as-
pect to that of the deserts farther south, it
is very different in floristic composition. As
Axelrod (1950) has pointed out, some of the
dominant genera in the Intermountain Re-
gion, such as Artemisia, Atriplex, and Cera-

10

GREAT BASIN NATURALIST MEMOIRS

No. 2

toides (Eurotia), apparently relate to the Ar-
cto-Tertiary rather than the Madro-Tertiary
flora. Likewise Astragalus, one of our larg-
est genera in terms of number of species,
has an even larger number of species in
dryland Eurasia. Even if one prefers to
avoid the terms Arcto-Tertiary and Madro-
Tertiary, these genera still relate to Asian
desert plants, presumably by way of a Ber-
ingian connection, rather than to plants
from farther south in western North Ameri-
ca. On the other hand, such large genera as
Penstemon and Eriogonwn are strictly
American, best developed in arid western
North America, without any obvious in-
dication of a more southern (Madro-Ter-
tiary) origin. Haplopappus may well be
from the Madro-Tertiary, as Axelrod sug-
gests, but its derivative Chrysothamnus cen-
ters in the Great Basin. We have already
noted that some of the common genera of
Heliantheae in the Intermountain Begion
may well be of Madro-Tertiary affinity.
Thus it is not possible to assign the charac-
teristic flora of the Great Basin province to
either a chiefly Madro-Tertiary or a chiefly
Arcto-Tertiary origin. Both of these Tertiary
floras clearly contributed to the present
flora of the region.

Thus, by some combination of differen-
tiation from native elements, immigration
from near and far, and proliferation of the
immigrants, the Intermountain Flora ac-
quired its special character during the
Miocene epoch. Xerophytes predominated
especially at lower elevations, but meso-
phytes survived in the moister habitats, of-
ten at higher elevations. These two types
have been in continuous competiton in the
Intermountain Begion since that time.

Tension between Mesophytic and
Xerophytic Communities

Although the Great Basin floristic
province took shape in the Miocene, it was
not immediately so dry as it is now. Axelrod
(1948) considers that open environments ex-
tended through the region in the Middle

Pliocene, but that the plant community was
predominantly grassland, with semidesert
shrubs on the drier slopes. In Miocene and
Pliocene time the presently desert regions
supported species comparable to those in
the pinyon-juniper woodland and oak wood-
land that now occur at slightly higher ele-
vations or around the borders of the desert.

Axelrod (1950) considers that the trend
toward a drier, more continental climate in
the Intermountain Begion, begun early in
the Tertiary, culminated in Middle Pliocene
time, perhaps 4 or 5 million years ago. Lat-
er in the Pliocene the climate probably be-
came a bit cooler and moister. The Pleisto-
cene, as we all know, was marked by
alternating glacial and interglacial stages.
From a long-term geohistorical viewpoint,
the present time may be merely another
Pleistocene interglacial. Actual glaciers in
the Intermountain Begion were largely re-
stricted to upper elevations in the moun-
tains; the continental ice sheet did not
reach that far south in western North
America.

The glacial periods were times of rela-
tively lower temperatures and higher p/e
ratio in the Intermountain Begion, cooler
and more mesic than the interglacials. Dur-
ing the glacial periods, the mesophytes,
many of them of northern floristic affinities,
expanded their distribution at the expense
of the xerophytes; in the interglacials the
process was reversed. The great differences
in elevation, together with the strong local
differences in moisture relations according
to slope and edaphic factors, combined with
the repeated shifts in climate to keep the
species populations in constant turmoil
throughout the Pleistocene. T. M. Barkley
(personal communication) has suggested that
the blurred boundary between Senecio strep-
tanthifolius (a highland species) and Senecio
multilobatus (a lowland, more xerophytic
species) in Utah reflects such hybridization.
Local polyploidy helps such hybrids and
hybrid segregates to persist in appropriate
habitats.

Another example of the advance and re-

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

11

treat of species in the Intermountain Region
due to climatic changes is provided by the
oaks. In north-central Utah there exist today
clones of oak that have been conclusively
demonstrated to be hybrids between
Quercus gambelii and O. turbinella. Quercus
gamhelii is common in the area today, but
Q. turbinella reaches its present northern
limits more than 250 miles to the south of
these hybrids. It is reasonably believed that,
during the postglacial hypsithermal period,
some five or six thousand years ago, the
range of Q. turbinella extended north into
north-central Utah, permitting the forma-
tion of the hybrids (Cottam et al. 1959).

Alpine fir, Abies lasiocarpa, provides an
example in the reverse direction. According
to Cottam et al. (1959), fossils discovered in
1957 by D. J. Jones demonstrate that alpine
fir grew along the shores of Lake Bonne-
ville at a time when the lake level stood
well below the Provo stage. Recent fluctua-
tions in the level of Great Salt Lake remind
us that p/e ratios in the Intermountain Re-
gion continue to fluctuate, but up until now
the climatic changes during the relatively
short time for which we have formal, writ-
ten records do not approach the magnitude
of the changes that occurred during geolog-
ic time.

Present-Day Correlation of Elevation
with Floristic Groups

Elevation is closely correlated with mois-
ture relations as well as with temperature in
the Intermountain Region. As one goes
higher into the mountains, the temperature
drops and the p/e ratio increases, and one
finds a progressively more northern element
in the flora. Many years ago I read some-
where that in the western United States one
can roughly equate one mile of latitude
with four feet of altitude. In my own expe-
rience, this conversion factor works fairly
well, although there are, of course, always
modifying factors to be taken into account.
At moderately high elevations in the moun-
tains, one finds many species similar or

identical to those of the northern coniferous
forest, and above timberline one finds many
species similar or identical to those of the
modern circumboreal arctic flora. The
spruce-fir forests of midupper elevations in
the Intermountain Region represent a south-
ern extension of the northern coniferous for-
est. Even though the dominant species are
different, they compare closely with species
from the northern forest. Abies lasiocarpa
compares with Abies balsmnea, Picea engel-
mannii and P. pungens compare with P.
glauca, and Pinus contorta compares with P.
banksiana. Pseudotsuga menziesii, on the
other hand, does not have a boreal equiva-
lent.

North of the Intermountain Region, in
the northern Rocky Mountains of Canada
and the northwestern United States, a very
large proportion of the high-mountain spe-
cies can be related directly to something
from the holarctic or the northern con-
iferous forest. As one goes progressively
southward, a larger and larger proportion of
the alpine species are evidently highland
derivatives from common lowland elements.
In the Intermountain Region both of these
types are well represented at upper alti-
tudes. Alpine and subalpine species of Are-
naria, Gentiana, Myosotis, Pedicularis,
Ranunculus, and Saxifraga are likely to
have boreal affinities. On the other hand,
montane species of Allium, Eriogonum, Hul-
sea, Hymenoxys, Lomatium, and Penstemon,
even at the highest elevations, generally re-
late to species of lower elevations, often of
dry habitats. Some common montane gen-
era, such as Erigeron, do not fit into either
of these patterns. Erigeron is best developed
in the western American cordillera, but the
species of the dry lowlands are evidently
advanced, and the more primitive species
are distinctly mesophytic.

Evolutionary History of Mammals

The evolutionary history of the mammals
parallels in many ways that of flowering
plants. Although the group takes its origin

12

GREAT BASIN NATURALIST MEMOIRS

No. 2

from therapsid reptiles in the Triassic peri-
od, the placental mammals do not enter the
fossil record until late in the Cretaceous.
Placental mammals diversified explosively
during the Paleocene, and they have been
the dominant animals in terrestrial ecosys-
tems since that time. Evolution of mammals
in North America is closely correlated with
that in Eurasia, but not well correlated with
that in South America, because of the essen-
tial separation of North and South America
until relatively recent times.

The animals that may have had the most
important influence on the plants during
the Tertiary period were the grazing ani-
mals—ungulates, in the broad sense. Grazing
mammals began to evolve in the Paleocene
or Eocene, and they reached full flower in
the Oligocene and Miocene (Jones and Arm-
strong 1973). One may reasonably suppose
that there is a relationship between the evo-
lution of grazing mammals and the rise of
grasses during the same general time.
Grasses originated no later than the Oligo-
cene, and by Miocene time they were com-
mon. The intercalary meristem of the grass
leaf can reasonably be interpreted as an
adaptation to grazing pressure. Thus, al-
though it may be true that at any given
time the nature of the fauna is more depen-
dent on the flora than vice versa, in the
long rim the evolution of plants is strongly
influenced by animals.

The most startling feature of the evolu-
tionary history of mammals in North Ameri-
ca was the rapid extinction of a great many
of the large mammals about ten thousand
years ago. There is no real parallel in the
evolutionary history of plants. Similar ex-
tinction occurred to varying degrees in oth-
er parts of the world, least of all in Africa.
Both climatic changes and the influence of
early man have been invoked to explain the
massive extinctions. In North America, the
case for the predominant influence of man
is very good (Martin 1967), although the
subject still evokes considerable debate and
difference of opinion (Axelrod 1967). The
large mammals had survived much more ex-

tensive climatic changes during the Pleisto-
cene, and their disappearance from the
scene appears to be closely correlated with
the spread of man. Human hunters killed
the large herbivores, and many of the large
predators disappeared along with their prey.
Bison, camels, elephants, and horses were
abundant in the Intermountain Region dur-
ing the Pliocene and Pleistocene (Axelrod
1950), but of these only the bison survived
the human onslaught. It is clear enough that
horses, at least, are well adapted to modern
conditions in the Intermountain Region, and
burros do very well a little farther south.
On the other hand, the camels introduced
into our southwest more than a century ago
did not make the grade, although they
might well have done so in the absence of

Evolutionary History of Birds

The birds apparently originated in the
upper Jurassic and began to radiate in the
Cretaceous, but nearly all the Cretaceous
families are now extinct. After the extinc-
tion of the dinosaurs and before the evolu-
tion of large carnivorous mammals, there
were some large flightless birds, which
played the ecological role later taken over
by large mammalian carnivores. In the
northern hemisphere these birds were com-
mon from the Upper Paleocene to the
middle of the Eocene. An ecologically sim-
ilar but taxonomically distinct group of
large, flightless, predatory birds was com-
mon from early Eocene to middle Pliocene
time in South America, an area into which
the large carnivores made a relatively late
entry. Again the geographic separation of
North America from South America during
most of the time from the Cretaceous until
late in the Tertiary had a profound effect
on evolutionary patterns.

By the end of the Eocene the birds were
highly diversified, and all living families and
orders can be traced back at least that far.
In Miocene time the avifauna began to take
on a more modern aspect, and most of the

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

13

modern genera had come into existence by
Pliocene time (Storer 1974). There was no
great wave of recent extinction comparable
to that of the large mammals.

The avifauna of the Intermountain Re-
gion has no endemic species and is distin-
guished mainly by what isn't there. The
species are all more or less widespread. Dis-
tinctively Californian, southern Rocky
Mountain, and Mohavean species mostly do
not extend into the Intermountain Region
(W. H. Behle, this symposium).

Evolutionary History of Insects

The evolutionary radiation of insects, like
that of flowering plants, mammals, and
birds, goes back many millions of years. The
Coleoptera (beetles) are well known as fos-
sils as far back as the Permian period. The
Diptera and Hymenoptera date from the
Jurassic period, but only in forms such as
midges and crane flies (Diptera) and saw
flies (Hymenoptera), which are not and pre-
sumably never were important pollinators.
The bees and the higher Diptera, which are
now important pollinators, first appear in
the fossil record in early Tertiary time, al-
though they may well have originated
somewhat earlier (Carpenter 1953, Baker
and Hurd 1968). Most or all of the early
Tertiary bees belong to extinct genera, and
one may legitimately speculate that the ev-
olution of modern bees was intimately re-
lated to the evolution of structurally com-
plex, bee-pollinated flowers during the
Tertiary. The Lepidoptera originated no lat-
er than the late Cretaceous (MacKay 1970)
and had already diversified to some extent
in early Tertiary time, but here again the
important pollinators are apparently not an-
cient types. Drawing upon the recent dis-
coveries of Cretaceous fossil insects report-
ed by Rodendorf and Zherikhin in 1974,
Doyle (1976) visualizes "major extinctions of
'Jurassic' groups within a relatively brief in-
terval of the Late Cretaceous, and a some-
what slower rise of groups now associated
with angiosperms." The coevolution of

structurally complex flowers and insects ca-
pable of recognizing complex patterns rep-
resents another example of major evolution-
ary interaction between plants and animals
(Leppik 1957, Baker and Hurd 1968).

The faunistic differentiation between
Laurasia and Gondwanaland shows up at
least in the aquatic insects of the Inter-
mountain Region. Species of Gondwanaland
ancestry occur mostly in the warmer wa-
ters, or their eggs hatch relatively late in
the summer. Some species of Laurasian af-
finity connect to Eurasia through Beringia,
and others through Europe (G. F. Edmunds,
personal communication).

Literature Cited

Axelrod, D. I. 1968. Climate and evolution in
western North America during Middle Pliocene
time. Evolution 2: 127-144.

1950. Evolution of desert vegetation in west-
em North America. Carnegie Inst. Washington
Publ. 590: 215-306.

1952. A theory of angiosperm evolution. Ev-
olution 6: 29-60.

1956. Mid-Pliocene flora from west-central

Nevada. Univ. Calif. Publ. Geol. Sci. 33: 1-321.

1958. Evolution of the Madro-Tertiary geof-

lora. Bot. Rev. 24: 433-509.

1964. The Miocene Trapper Creek flora of

southern Idaho. Univ. Calif. Publ. Geol. Sci. 51:
1-148.

1966. The Eocene Copper Basin flora of

northeastern Nevada. Univ. Calif. Publ. Geol.
Sci. 59: 1-125.

1967. Quaternary extinctions of large mam-
mals. Univ. Calif. Publ. Geol. Sci. 74: 1-42.

1968. Tertiary floras and topographic history

of the Snake River Basin, Idaho. Geol. Soc.
Amer. Bull. 79: 713-734.

1975. Evolution and biogeography of Madr-

ean-Tethyan sclerophyll vegetations. Ann. Mis-
souri Bot. Gard. 62: 280-334.

Baker, H. G., and P. D. Hurd, Jr. 1968. Intrafloral
ecology. Ann. Rev. Entom. 13: 385-414.

Barnard, P. D. W. 1973. Mesozoic floras. In: N. F.
Hughes (ed.), Organisms and continents through
time. Publ. Syst. Assoc. 9: 175-187.

Bateman, P. C. 1968 Geologic structure and history
of the Sierra Nevada. Univ. Missouri Rolla J. 1:
121-131.

Becker, H. F. 1969. Fossil plants of the Tertiary
Beaverhead basins in southwestern Montana.
Palaeontographica 127B: 1-142.

14

GREAT BASIN NATURALIST MEMOIRS

No. 2

Billings, W. D. 1974. Adaptations and origins of
alpine plants. Arctic Alp. Res. 6: 129-142.

Bo wen, R. 1966. Paleotemperature analysis. Else-
vier. Amsterdam.

Carpenter, F. M. 1953, The evolution of insects.
Amer. Sci. 41: 256-270.

Cottam, W. P., J. M. Tucker, and R.
Drobnick. 1959. Some clues to Great Basin
postpluvial climates provided by oak distribu-
tion. Ecology 40: 361-377.

Cronquist, A., A. H. Holmgren, N. H.
Holmcren, and J. L. Reveal. 1972. Inter-
mountain Flora. Vol. 1. Hafner Publ. Co., New
York.

Dietz, R. S., and J. C. Holden. 1970. The breakup
of Pangaea. Sci. Amer. 223(4): 30-41.

Dilcher, D. L. 1974. Approaches to the identi-
fication of angiosperm leaf remains. Bot. Rev.
40: 1-157.

Dorf, E. 1969. Paleobotanical evidence of Mesozoic
and Cenozoic climatic changes. Proc. N. Amer.
Paleont. Conv. Sept. 1969 D: 323-346.

Doyle, J. A. 1969. Cretaceous angiosperm pollen
of the Atlantic Coastal Plain and its evolution-
ary significance. J. Arnold Arb. 50:1-35.

1973. Fossil evidence on early evolution of

the monocotyledons. Quart. Rev. Biol. 48:
399-413.

1976. Man bites botanical dogma. Paleobiol-
ogy 2: 265-271.

Doyle, J. A. and L. J. Hickey. 1976.
Pollen and leaves from the mid-Cretaceous
Potomac group and their bearing on early ang-
iosperm evolution, pp. 139-206. In: C. B. Beck
(ed.), Origin and early evolution of angiosperms.
Columbia University Press, New York.

Eardley, A. J. 1968. Major structures of the Rocky
Mountains of Colorado and Utah. Univ. Mis-
souri Rolla J. 1: 79-99.

Gleason, H. A. 1926. The individualistic concept of
the plant association. Bull. Torrey Bot. Club.
53: 7-26

Gleason, H. A., and A. Cronquist. 1964. The
natural geography of plants. Columbia Univer-
sity Press, New York.

Grant, V. 1959. Natural history of the Phlox fam-
ily. Martinus Nijhoff. The Hague, Netherlands.

Hickey, L. J. 1973. Classification of the architecture
of dicotyledonous leaves. Amer. J. Bot. 60:
17-33.

Hickey, L., J, and J. A. Wolfe. 1975. The bases
of angiosperm phylogeny: vegetative orphology.
Ann. Missouri Bot. Gard. 62: 538-589.
62: 538-589.

Hughes, N. F. 1973. Mesozoic and Tertiary distri-
butions, and problems of land plant evolution.
In: N. F. Hughes (ed.), Organisms and conti-

nents through time. Publ. Syst. Assoc. 9: 188-
198.

Jones, J. K., Jr., and D. M. Armstrong.
1974. Mammalia; pp. 401-416 in Encycl. Brit-
annica, ed. 15, vol. 11:401-416.

Leppik, E. E. 1957. Evolutionary relationship be-
tween entomophilous plants and anthophilous
insects. Evolution 11: 466-481.

MacArthur, R. H., and E. O. Wilson. 1967. The
theory of island biogeography. Princeton Univ.
Press.

MacGinitie, H. D. 1953. Fossil plants of the Floris-
sant beds, Colorado. Carnegie Inst. Wash. Publ.
465: 83-151.

1969. The Eocene Green River flora of

northwestern Colorado and northeastern Utah.
Univ. Calif. Publ. Geol. Sci. 83: 1-203.

MacKay, M. K. 1970. Lepidoptera in Cretaceous
amber. Science 167: 379-380.

Martin, P. C. 1967. Prehistoric overkill. In: P. S.
Martin and H. E. Wright, Jr. (eds.), Pleistocene
extinctions: the search for a cause. Proc. VII
Congr. Internat. Assoc. Quaternary Res. 6: 75-
120.

Raven, P. H., and D. I. Axelrod. 1974. Angio-
sperm biogeography and past continental move-
ments. Ann. Missouri Bot. Gard. 61: 539-673.

Roberts, R. J. 1968. Tectonic framework of the
Great Basin. Univ. Missouri Rolla J. 1: 101-119.

RODENDORF, B. B., AND V. V. ZhERIKHIN.

1974. Paleontoogiya i okhrana prirody. Priroda
1974(5): 82-91.

Scott, R. A. P. L., Williams, L. C. Craig, E. S.
Barghoorn, L. J. Hickey, and H. D.
MacGinitie. 1972. "Pre-Cretaceous" an-
giosperms from Utah: evidence for Tertiary age
of the palm woods and roots. Amer. J. Bot. 59:
886-896.

Smith, A. G., J. C. Briden, and G. E.
Drewry. 1973. Phanerozoic world maps. In:
N. F. Hughes (ed.), Organisms and continents
through time. Publ. Syst. Assoc. 9: 1-42.

Storer, R. R. 1974. Bird. Encycl. Britannica, ed.
15, vol. 2: 1053-1062.

TlDWELL, W. D., S. R. RUSHFORTH, J. L. REVEAL, AND

H. Behunin. 1970. Palmoxylon simperi
and Palmoxylon pristina: Two pre-Cretaceous
angiosperms from Utah. Science 168: 835-840.

TlDWELL, W. D., S. R. RUSHFORTH, AND D.

Simper. 1972. Evolution of floras in the In-
termountain Region, pp. 19-39. In: A.
Cronquist, A. H. Holmgren, N. H. Holmgren
and J. L. Reveal, Intermountain Flora, vol. 1,
Hafner Publ. Co., New York.
Walker, J. W., and J. A. Doyle.

1975. The bases of angiosperm phylogeny: Pa-

1978 INTERMOUNTAIN BIOCEOGRAPHY: A SYMPOSIUM 15

lynology. Ann. Missouri Bot. Gard. 62: 664-723. geny: Paleobotany. Ann. Missouri Bot. Gard. 62:

Wolfe, J. A. 1969. Neogene floristic and vegeta- 801-824.

tional history of the Pacific Northwest. Wolfe, J. A., and D. M. Hopkins.

Madrono 2: 83-110. 1967. Climatic changes recorded by Tertiary

Wolfe, J. A., J. A. Doyle, and V. M. land floras in northwestern North America.

Page. 1975. The bases of angiosperm phylo- Symp. Pacific Sci. Contr. 25: 67-76.

BIOGEOGRAPHY OF INTERMOUNTAIN FISHES

Gerald R. Smith1

Abstract.— Eighty-three species of fishes belonging to 26 genera live in the area bounded by the Sierra Ne-
vada, Grand Canyon, Rocky Mountains, and Snake River Plain. The waters inhabited by these fishes are part of
the Great Rasin, Colorado River, Snake River, upper Pit River and upper Klamath River drainages. The adapta-
tions and distribution patterns of these fishes have been shaped by extensional faulting and volcanic activity in
the Great Basin, uplift of the surrounding ranges and plateaus, and cyclic fluctuations of the Pleistocene pluvials
and interpluvials. Fossil evidence indicates that in the Pliocene many of the lineages had established distributions
broadly inclusive of the present-day patterns, and the subsequent trends have been extinction and some species
differentiation.

Analysis of the fauna is based on designation of 48 barrier-bounded, faunally homogeneous drainage units and
quantitative evaluation of patterns among species distributions and faunal similarities of drainages. Cluster analy-
sis of species based on correlations among their patterns revealed the existence of a basic northern intermountain
fluvial fauna consisting of Cottus bairdi, Prosopium williamsoni, Catostomus platyrhynchus, Rhinichthys cata-
ractae, Richardsonhis balteatus, and their vicariants. These fishes have similar ecology and dispersal patterns.
They are ecologically associated with Salmo clarki (upstream) and Rhinichthys osculus (downstream), but these
two species have broader distributions, probably because of more frequent colonization via stream capture by the
former and extinction resistance by the latter. Rhinichthys osculus is the most widespread intermountain fish,
being found in 32 of the 48 drainage units; Gila bicolor is next most widespread, being found in 21 units. Fifty-
one of the 83 species are found in only one drainage unit; 18 of these are endemic to that unit (33 are more
widespread outside the study area).

Species distributions are broader in the north, and northern and peripheral units have more species. The spe-
cies:area curve for the Great Basin shows a steep slope (z = .59), with especially low species density in small
drainages, indicating high extinction and low colonization. Postpluvial aridity, especially in the south, is the ma-
jor cause of extinction and a major cause of isolation. Principal components and cluster analysis of drainage units,
based on shared species, show a high correspondence between faunal similarity and geographic proximity (and
weakness of barriers) and also reveal the effects of extinction in erasure of patterns. The Wasatch Range, Sierra
Nevada, and southern divide boundary of the Snake River Plain have been strong barriers, leading to intensive
faunal differentiation. Strong barriers and concomitant differentiation also exist in eastern Nevada, near the origi-
nal center of Basin and Range tectonism. Two dozen examples of vicariant species are found to be associated
with stronger-than-average barriers.

The colonization rate and extinction rate have both been accelerated in postsettlement time by introductions
and habitat destruction. Most drainages and populations have been affected. Five species and many local popu-
lations have become extinct. Eighteen species and many more populations are vulnerable or threatened. Reversal
of the trend will require sound ecosystem management of watersheds and restriction of exotic introductions.

The distribution patterns of fishes in the survival has been dependent upon continu-

intennountain region of the western United ity of lakes, marshes, springs, and streams in

States offer an unusual opportunity to study isolated basins for many thousands of years,

the evolutionary results of long-term ecolog- Dispersal has been dependent on occasional

ical changes because of the insular nature of continuity of suitable habitat among basins,

the drainage basins, the relatively well- Compared to most kinds of organisms, the

known geological history, and the close de- restrictions on freshwater fishes are more

pendency of these fishes on the continuity rigid; compared to most habitats, Great Ba-

of their habitat in time and space. Their sin aquatic habitats have been less stable

'Museum of Zoology and Museum of Paleontology, University of Michigan, Ann Arbor, Michigan 48106.

17

L8

GREAT BASIN NATURALIST MEMOIRS

No. 2

and more confined. These restrictions pro-
vide relative constancy of some variables of
interest and some replication in a quasi ex-
periment of evolutionary responses to
changing ecological conditions.

The predominant environmental factors
are the degree of isolation of the drainage
basins and the late Cenozoic history of fluc-
tuating pluvial and interpluvial episodes.
The isolation has greatly restricted dispersal
and colonization; the fluctuating climate has
subjected the fish populations to extreme
conditions, forcing adaptation or extinction.
Fortuitously, the dominant aquatic habitats
are also agents dominant in geomorpho-
logical processes, and the lakes and streams
that supported the fish populations have left
stark and beautiful records in their wake
(Gilbert 1890, Russell 1885). This paper is a
summary of the relationship of fish distribu-
tions to geological history and the contribu-
tion of fish distributional data to the under-
standing of that history, as reconstructed by
Cope (1883), Snyder (1908, 1917), Hubb's
and Miller (1948a,b), Miller and Hubbs
(1960), Miller (1945, 1948, 1958, 1965),
Hubbs, Miller, and Hubbs (1974) Evermann
(1897), and La Rivers (1962).

Geological Setting

The Great Basin includes more than 150
drainage basins among approximately 160
regularly spaced, roughly parallel mountain
ranges. The ranges and valleys trend north
or northeast and are bounded by steeply
dipping normal faults. Faulting began as
early as Eocene or Oligocene, but most of
the ranges were formed during the past 20
million years by crustal extension. The
Great Basin crust is relatively thin and is
bounded to the east, north, and west by
zones of much thicker crust. Extension has
been generally northwest-southeast and has
been estimated between 50 and 300 km.
Geophysical evidence indicates that the
driving force may be a rising, spreading,
semimolten body (diapir), whose origin is
related to the early and Middle Cenozoic

subduction of the Farallon plate at the
West Coast trench (see Scholz et al. 1971,
Atwater 1970). The structure and processes
of the system are essentially those of an in-
terarc basin. Volcanic evidence for this in-
terpretation (from Armstrong et al. 1969)
involves an episode of andesitic volcanism
with much silicic ash in east-central Nevada
40-30 million years ago, abruptly changing
to basaltic volcanism which radiated toward
the margins of the Great Basin during the
past 20 million years. Scholz et al. suggest
that the outward radiation of basaltic vol-
canism tracked the spreading margins of the
underlying diapir and its concomitant crus-
tal extension, the whole process being in-
itiated at the release of compressional stress
when the subduction of the Farallon plate
was complete.

The late Cenozoic tectonic relationship
between the Great Basin and surrounding
areas has been deduced from seismic data
by Smith and Sbar (1974). An arcuate pat-
tern of zones of shallow earthquakes defines
the Great Basin boundary along the Snake
River plain on the north, the Wasatch front
on the east, and from about Cedar City
west 200 km across southern Nevada on the
south. Fault plane solutions indicate a slight
counterclockwise rotation of the Great Ba-
sin subplate accompanied by rifting along
the Snake River Plain. Volcanic activity has
proceeded eastward along the Snake River
Plain at a rate of about 4 cm per year, the
current focus being in the Yellowstone re-
gion (Armstrong et al. 1975). This activity is
interpreted by Eaton et al. (1975) and
Smith and Sbar (1974) as marking the prog-
ress of a mantle plume of molten magma
being overridden by the westward-moving
North American plate.

This brief tectonic outline may be
roughly correct, at least as far as the emp-
irical surface history is concerned, and pro-
vides us with an indication of several of the
geological variables that probably played a
part in the development of patterns of dis-
tribution. It is not our intent to fit the dis-
tributions to this history, but to recognize

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

19

the antiquity of the general topographic-
pattern, the large potential for instability of
drainage on a local scale, and the pervasive
part played by volcanism.

The evolution of structural features of
Great Basin and Wasatch Range topography
set the stage for increasing isolation of
drainage units. The degree of isolation of
basins has also increased with generally in-
creasing aridity (Axelrod 1950). During
Pleistocene pluvial stages, however, consid-
erable connectedness of drainages can be as-

sumed. Figure 1 shows maximum stands and
fluvial connections of many of the larger
Pleistocene lakes of the region, in addition
to possible Pliocene coverage of Lake
Idaho. Depicted connections and high
stands were not all contemporaneous.

Little is known of early and middle
Pleistocene pluvials. Older pre-Bonneville
and pre-Lahontan lacustrine sediments are
correlated with Cedar Ridge (Kansan) gla-
ciation in the Rocky Mountains and are
overlain by the Pearlette (restricted, type-0)

Fig. 1. Noncontemporaneous maximum extent of Late Pleistocene lakes and known fluvial connections in intermountain western North America and
possible maximum extent of Pliocene Lake Idaho. Compiled and modified from Miller (1946a), Hubbs and Miller (1948), Trimball and Carr (1961). Feth
(1961), Bright (193), Snyder et al. (1964), Morrison (1965), and Hubbs et al. (1974).

20

GREAT BASIN NATURALIST MEMOIRS

No. 2

ash, which is 600,000 years old (Morrison
1965, Richmond 1970). Well-developed soils
separate the older pre-Bonneville and pre-
Lahontan sediments from younger pre-Bon-
neville and pre-Lahontan lacustrine sedi-
ments, believed to be contemporaneous
with the Sacagawea Ridge (Illinoian) glacia-
tion. These units are also overlain by sub-
aerial units and soils believed correlated
with the last great interglacial (Sangamon),
for which dates are around 130,000 years
BP (Richmond 1970).

The Alpine formation of the Bonneville
Basin and Aetza formation of the Lahontan
Basin are thick units of lacustrine deposits
frequently interrupted by soils and subaerial
zones. They appear to span a time begin-
ning perhaps as early as 70,000 years BP,
correlative with the Bull Lake glaciation.
The latter portion of this period was appar-
ently a time of intermediate lacustrine oc-
cupation of other pluvial basins as well,
e.g., Searles (Smith 1968) and Yellowstone
(Birkeland et al. 1971). The period from
about 35,000 to 25,000 was apparently
dominated by widespread subaerial deposi-
tion and soil formation (Morrison 1965, Bir-
keland et al. 1971).

The most recent pluvial episode, corre-
lated with the Pinedale glaciation, has left
the most abundant and clear record (e.g.,
probably most of the Great Basin lakes
shown in Figure 1). Radiocarbon dates for
high and low stages of five lakes appear to
be only partly synchronous. Data for Bonne-
ville (Morrison 1965, Bright 1966), Lahon-
tan (Broecker and Kaufman 1965, Morrison
1965), Searles (Smith 1968), Mojave (Ore
and Warren 1971), and Yellowstone (Rich-
mond 1970) agree in showing development
of generally, but not consistently, high lake
levels during the period between 25,000
and 13,000 years BP. In each lake, a low
stand is marked at about 11,000 years, fol-
lowing a high stand in the previous one or
two thousand years. High levels are record-
ed again in the interval between 11,000 and
10,000 years, followed by low stand at
about 9,000 years BP and unstable-low

stands throughout the warm period of
8,000- 4,000 years BP and to the present.
Lake Malheur, in the Harney Basin, shows a
broadly similar pattern (Hansen 1947). Lo-
cal details will be mentioned below, in con-
nection with special distributional problems.
Although there is a tendency to look to
the 13,000 and 11,000 BP lacustrine highs
and their inferred associated climates for ex-
planation of distributional patterns, it is im-
portant to remember that they represent
only an unstable, late episode among a
number of intermittent pluvial periods. The
role of geologic and climatic factors in the
manipulation of barriers and habitats will
be examined by analyzing fish distributions
to discover the major patterns and possible
determinants of the patterns.

Methods

Of the 70 or more major drainage basins,
many are fishless and will not be considered
in this study. The basins with fishes are di-
vided into barrier-bounded, faunally homo-
geneous units. Criteria for barriers are pres-
ence of mountains or deserts uncrossed by
continuous aquatic habitat or at least one-
way restriction of fish dispersal, such as bar-
rier falls. Criteria for homogeneity include
the restriction that a drainage unit should
not have two or more barrier-separated
areas of endemism within it. For practical
purposes, contiguous but separated small
areas were often joined as one if they con-
tained the same fauna. These criteria al-
lowed the relatively objective designation of
48 drainage units, which are outlined and
labeled in Fig. 2. To aid in ease of refer-
ence, drainage units are given initials in-
dicating general region and faunal affinity:
(B) Bonneville, (S) Snake, (O) Oregon lakes
(some of which are in Nevada and Califor-
nia), (K) Klamath, (G) Goose (L) Lahontan,
(R) Ruby group (east-central Nevada), (D)
Death Valley (with Owens River and Mo-
jave Basin), and (C) Colorado. Subsequent
initials in each label refer to the specific
area or some valley or ancient lake in it.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

21

In the analyses, the drainage units are
clustered and ordinated according to the
similarity of their fish faunas. The presence
or absence of fish species and higher taxa
are the characteristics by which drainage
units are evaluated. General dispersal pat-
terns are also sought by clustering species
according to the similarity of their geogra-
phic ranges. The methods used are single
linkage cluster analysis (the similarity
coefficients used are Jaccard's coefficient

and correlation coefficients) and principal
components (of the among-species cov-
ariance matrix), from the general field of
numerical taxonomy (Sneath and Sokal
1973).

Variation in species density and its causes
are examined by regression of species num-
bers on area and by graphs examining spe-
cies density in drainage units and breadth of
distribution of species among units. Histori-
cal perspectives are sought through an ex-

Fig. 2. Outline and identification of 48 drainage units as defined in this analysis. BONNEVILLE group, Utah, Idaho, and Nevada: BB— Bear R.-We-
ber R. dr., BT-Thousand Springs-northwest Great Salt Lake dr., BP-Provo R.-Utah L.-Jordan R. dr., BD-Deep Cr. dr., BSn-Snake V. and related dr.,
BSv-Sevier R. dr., BSh-Shoal Cr. (Pine Canyon Cr.) dr.; SNAKE R. group, Idaho, Wyoming, Utah, Nevada, Oregon: SU-Upper Snake above Shoshone
Falls, SL-Lost R.-Camas Cr. dr., SW-Wood R. dr., SM-Middle Snake R. (below Shoshone Falls and Wood R. falls); OREGON LAKES group, Oregon,
Nevada, California: OH-Harney Basin, OAk-Alkali L. dr., OAb-Abert L. dr., OSm-Summer L. dr., OFR-Fort Rock-Christmas L. Basin, OSi-Silver
L. dr., OW-Wamer L. dr., OC-Catlow V. dr., OAl-Alvord dr., OL-Long V. (Massacre L.) dr., OSp-Surprise V (Alkali lakes) dr.; G-GOOSE LAKE,
California, Oregon; K-KLAMATH system (upper), Oregon, California; LAHONTAN group, Nevada, California, Oregon: L-Lahontan (Humboldt R.,
Pyramid L„ L. Tahoe, Walker L., Crescent V., Grass V.), LDm-Diamond V., LE-Eagle L. dr., LMd-Madeline Plains, LDx-Dixie V., LT-Toiyabe
(Big Smokey V.), LF-Fish L. dr., LN-Newark V., LC-Clover-Independence V. (Snowater L.). RUBY group, Nevada: RF-Ruby-Franklin dr., RB-
Butte V., RW-L Waring (Goshute V.), RSt-Steptoe V.; COLORADO group, Utah, Nevada, Arizona: CRR-Railroad V., CW-White R. (Muddy R.)
dr., CM-Meadow Valley Wash dr., CV-Virgin R. dr., C-upper Colorado dr. (above Virgin R.), CLV-Las Vegas dr.; DEATH VALLEY group, Califor-
nia, Nevada: DO— Owens dr., DMo— Mojave dr., DMn— Amargosa-Manly dr., DT-Amargosa-Tecopa dr., DP-Pahrump V.

22

GREAT BASIN NATURALIST MEMOIRS

No. 2

animation of the fish fossil record for data
on past distributions, species density, and
amounts of taxonomic change.

Fish Fauna

The fishes inhabiting the Intermountain
Region are generally classified (Bailey et al.
1970) in 83 species, 26 genera, and 7 fami-
lies; 46 of the species (55 percent) and 6 of
the genera (13 percent) are endemic (Table
1). In the section that follows, the species
are listed with a statement of the range
within and beyond the study area (abbrevia-
tions refer to drainage units in Figure 2),
fossil record, and other special information.
Fossil species from the region that have liv-
ing representatives in the region are includ-
ed in the list and marked with a + .

Petromyzontidae

Lampetra minima Bond and Kan 1973.
Miller Lake Lamprey. K (Miller L., Ore.),
Apparently extinct.

Lampetra tridentata (Gairdner in Richard-
son) 1836. Pacific lamprey. SM, K, G; coast-
al drainages, Alaska to S Calif., and E Asia.

Lampetra lethophaga Hubbs 1971. Pit-
Klamath brook lamprey. K (upper Klamath
dr., Ore.); Pit R. dr., Calif.

Table 1. Numbers of families, genera, and species of
native fishes in the Intermountain Region, as considered
in this report.

Families

Petromyzontidae
lampreys

Genera

1

Species

3

(1 endemic)

AC1PENSEBIDAE

1

1

sturgeons

Salmonidae

trouts, whitefish

4

10

(3 endemic)

Cyprimdae
minnows

13

(4 endemic)

28

(14 endemic)

Catostomidae
suckers

3

21

(11 endemic)

Cyprinodontidae
killifishes

3
(2 endemic)

8
(8 endemic)

Cottidae
sculpins

1

12

(6 endemic)

Acipenseridae

Acipenser transmontanus Richardson
1836. White Sturgeon. SM; coastal streams,
Alaska to Calif.

Salmonidae

Salvelinus malma (Walbaum) 1792. Dolly
Varden (the interior "bull trout" popu-
lations may be a distinct species [Cavender
1969]). SL (Hubbs and Miller 1948b), SM
(Miller and Morton 1952), K (Bond 1973),
possibly B (Rostlund 1951), but not counted
in this study; freshwater and anadromous,
NW N America and E Asia. Rare in Inter-
mountain Region. Pliocene relative, west
central Nevada (Cavender 1969).

Salmo clarki Richardson 1836. Cutthroat
trout. BB, BP, BSv, BD, BSn, SU, SL, SW,
SM, K, G, OA1, L, LE, CV, C; Western
North America from headwaters of South
Saskatchewan, Missouri, South Platte, Ar-
kansas, Pecos, and Rio Grande drainages to
Eel R. in N Calif, and to SE Alaska. Much
local differentiation. Threatened or extinct
over most of former range in Great Basin;
genetically modified by introgression from
cultured and introduced upper Snake River
forms and Salmo gairdneri (Miller 1961,
1977). Pleistocene, L. Bonneville (Smith et
al. 1968). Pliocene S. esmeralda LaRivers
(1966), Esmeralda Co., Nev., may be a rela-
tive.

Salmo gairdneri Richardson 1836. Rain-
bow trout. SM, K; Pacific drainage, N
America, from N Mexico to W Alaska, and
probably headwaters of Peace and Ath-
abaska drainage. Widely introduced.

Salmo sp. Redband trout. SM, G, OH,
OAb, OSi, OW, OC; McCloud and Pit dr.,
Calif. (Bond 1973, LeGendre et al. 1972,
LeGendre 1976).

+ Salmo sp. Pliocene, Lake Idaho (SM),
Smith 1975, may be a relative of one or
more of the above trouts.

Salmo apache Miller 1972. Arizona trout.
C (upper L. Colorado dr.); Salt R. dr., Ari-
zona. Status: threatened (Miller 1977).

Oncorhyncus tshawytscha (Walbaum)

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

23

1792. Chinook Salmon. SM; Pacific dr. from
S Calif, to Hokkaido, W Arctic dr. Ana-
dromous. Pliocene congener, O. salax Smith
1975, L. Idaho (SM).

Prosopium williamsoni (Girard) 1857a.
Mountain whitefish. BB, BP, BSv, SU, SL,
SW, SM, OH, L, C; northward in head-
waters of upper Missouri, Milk, Saskatche-
wan, Peace, and Stikine drainages (Scott
and Crossman 1973).

Prosopium abyssicola (Snyder) 1919, Bear
Lake whitefish. BB (Bear L., Utah and
Idaho).

Prosopium spilonotus (Snyder) 1919. Bon-
neville whitefish. BB (Bear L.) Pleistocene,
L. Bonneville, Smith et al. 1968.

Prosopium gemmiferum (Snyder) 1919.
Bonneville cisco. BB (Bear L.) Pleistocene,
L. Bonneville, Smith et al. 1968.

+ Prosopium prolixus Smith 1975, Plioc-
ene, L. Idaho (SM).

Cyprinidae

Acrocheilus alutaceus Agassiz and Picker-
ing 1855. Chiselmouth. SM, OH; northward
in the Columbia and Fraser drainages.
Pliocene relative, A. latos Cope, L. Idaho
(SM), (Smith, 1975).

Eremichthys acros Hubbs and Miller
1948a. Desert dace. L (Soldier Meadows).
Status: rare (Miller 1977).

Gila atraria (Girard) 1857b. Utah chub.
BB, BT, BP, BSv, BSn, BSh, SU. Extensive
local geographic variation in body size,
number of gill rakers, color, etc. Pleisto-
cene, L. Bonneville, Smith et al. 1967.

Gila coerulea (Girard) 1857b. Blue chub.
K; Klamath B. Pliocene G. milleri Smith
1975, is a related form from L. Idaho (SM).

Gila robusta Baird and Girard 1854a.
Boundtail chub. CV, CW, C. Some geo-
graphic variation.

Gila elegans Baird and Girard 1853. Bo-
nytail. C. Status: endangered (Miller 1977).
A related form is known from the Pliocene
of Arizona (Uyeno and Miller 1965).

Gila cypha Miller 1946b. Humpback
chub. C. Status: endangered (Miller 1977).

Gila (Siphateles) bicolor (Girard) 1857b.

Tui chub. K, G, OH, OAb, OSm, OSi, OW,
OC, OFB, OAk, L, LCI, LDm, LN, LE, LF,
LDx, LT, DO, DMo, CRB. Considerable ge-
ographic variation in gill rakers, meristics,
osteology, size, etc.; many well-differen-
tiated subspecies (Snyder 1917, Miller 1973,
Hubbs and Miller 1972, Hubbs et al. 1974).
Gila bicolor pectinifer is an ecologically and
morphologically divergent form that is ap-
parently partially reproductively isolated
from sympatric G. b. bicolor under favor-
able ecological circumstances in the Lahon-
tan Basin (Hubbs et al. 1974, Hopkirk and
Behnke 1966). Pliocene Gila turneri (Lucas
1900), ( = G. esmeralda LaRivers 1966), Es-
meralda Co., Nev., may be a relative.

Gila (Siphateles) alvordensis Hubbs and
Miller 1972. Alvord chub. OA1.

Gila (Snyderichthys) copei (Jordan and
Gilbert) 1880. Leatherside chub. BB, BP,
BSv, SU, SW.

Hesperoleucus symmetricus (Baird and Gi-
rard) 1855. California roach. G; Sacramento
dr., coastal streams, Calif.

Iotichthys phlegethontis (Cope) 1874.
Least chub. BB, BP, BSv, BSn. Status: vul-
nerable (Miller 1977).

Richardsonius egregius (Girard) 1858. La-
hontan redside L,LE.

Richardsonius balteatus (Richardson)
1836. Redside shiner. BB, BT, BP, BSv, BSh,
SU, SW, SM, OH; northward through the
Columbia dr. to the Nass R., B. C, and the
Peace R., B. C. and Alberta. The form of
the Bonneville, Upper Snake, Palouse, Har-
ney Basin (except Silvies R.) and some iso-
lated headwaters of the M Snake, is differ-
entiated, with fewer rays in the anal fin (R.
b. hydrophlox). Richardsonius durranti of
Pliocene L. Idaho (SM) is a relative (Smith
1975).

Ptychocheilus oregonensis (Richardson)
1836. Northern squawfish. SM, OH; north-
ward through Umpqua, Siuslaw, and Colum-
bia drainages to the Nass R., B. C, and the
Peace R., B. C. and Alberta. Ptychocheilus
arciferus of Pliocene L. Idaho (SM) is a rel-
ative (Smith 1975).

Ptychocheilus lucius Girard 1857b. Colo-

24

GREAT BASIN NATURALIST MEMOIRS

No. 2

rado squawfish. C. Status: endangered (Mill-
er 1977). P. prelucius, Pliocene of Arizona
(C) is a relative (Uyeno and Miller 1965).

Mylocheilus ca minus Richardson 1836.
Peamouth. SM; northward through the Co-
lumbia drainage to the Nass R. and the
Peace R., B. C. M. robustus, Pliocene L.
Idaho, is a relative (Smith 1975).

Rhinichthys cataractae (Valenciennes)
1842. Longnose dace. BB, BP, SU, SW, SM,
OH; widespread throughout northern U.S.
and southern Canada, north to the Mac-
kenzie and south to N Mexico along the
Rockies and from Ungava drainage south to
Tennessee and North Carolina in the East.

Rhinichthys falcatus (Eigenmann and Ei-
genmann) 1893. Leopard dace. SM; Colum-
bia, Fraser drainages.

Rhinichthys osculus (Girard) 1857b.
Speckled dace. BB, BT, BP, BSv, BD, BSn,
BSh, SU, SW, SM, K, G, OH, OAb, OSi,
OW, OFR, OSp, OL, L, LCI, LE, LMd,
LDm, LT, DO, DT, CV, CW, CM, CLV,
C; elsewhere in Pacific drainage N. A. from
the Columbia dr. of extreme southern B. C.
(Scott and Crossman 1973) to northwestern
Mexico (Sonora), New Mexico, Arizona and
Calif. Geographically variable in meristics,
color, size, and proportions (Hubbs et al.
1974).

Rhinichthys sp. BSn (Bonneville desert).

Relictus solitarius Hubbs and Miller 1972.
Relict dace. RF, RSt, RW, RB.

Moapa coriacea Hubbs and Miller 1948a.
Moapa dace. CW Status; endangered (Mill-
er 1977).

Lepidomeda albivallis Miller and Hubbs
1960. White River spinedace. CW (Preston
and Lund Spr., Nevada). Status: vulnerable
(Miller 1977).

Lepidomeda altivelis Miller and Hubbs
1960. Pahranagat spinedace. CW (Ash Spr.
and upper Pahranagat L., Nev.). Status: ex-
tinct.

Lepidomeda mollispinis Miller and Hubbs
1960. Virgin spinedace. CV, CM (distinct
subspecies in Meadow Valley Wash, ex-
tinct). Status: vulnerable (Miller 1977).

Lepidomeda vittata Cope 1874. Little

Colorado spinedace. C. (Little Colorado
dr.). Status: vulnerable (Miller 1977).

Plagopterus argentissirnus Cope 1874.
Woundfin. CV; (and formerly from the Gila
R. dr.). Status: endangered (Miller 1977).

Catostomidae

Catostomus ardens Jordan and Gilbert
1880. Utah sucker. BB, BP, BSv, BD, BSn,
SU. Pleistocene Lake Bonneville (Smith et
al. 1968).

Catostomus macrocheilus Girard 1857b.
Largescale sucker. SM, OH; northward
through Columbia drainage to Nass R.,
B.C., and Peace R., B. C. and Alberta. (?)
Pliocene and Pleistocene relatives, Lake
Idaho (Smith 1975).

Catostomus occidentalis Ay res 1854. Sac-
ramento sucker. G; Sacramento dr., and
coastal drainages, Calif.

Catostomus sp. OSp (trib. Surprise Valley,
Nev.).

Catostomus warnerensis Snyder 1908.
Warner sucker. OW. Status: endangered
(Miller 1977).

Catostomus tahoensis Gill and Jordan in
Jordan 1878. Tahoe sucker. L, LE.

Catostomus fumeiventris Miller 1973.
Owens sucker. DO.

Catostomus insignis Baird and Girard
1854b. Sonora sucker. CV (formerly); lower
Colorado drainage.

Catostomus latipinnis Baird and Girard
1854b. Flannelmouth sucker. CV, C; lower
Colorado.

Catostomus snyderi Gilbert 1898. Kla-
math largescale sucker. K; Klamath dr.

Catostomus catostomus (Forster) 1773.
Longnose sucker. SU; northern U.S., Can-
ada, Alaska, NE Asia.

Catostomus (Pantosteus) columhianus (Ei-
genmann and Eigenmann) 1893. Bridgelip
sucker. SW, SM, OH; Columbia and Fraser
drainages. C. arcnatus of Pliocene Lake
Idaho is a relative (Smith 1975).

Catostomus (Pantosteus) discobolus Cope
1872. Bluehead sucker. BB, SU, C.

Catostomus (Pantosteus) clarki Baird and
Girard 1854b. Desert sucker. CV, CW, CM;

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

25

lower Colorado drainage.

Catostomus (Pantosteus) platyrhynchus
(Cope) 1874. Mountain sucker. BB, BP, BSv,
BD, BSh, SU, SM, L, C; upper Missouri,
Saskatchewan, Columbia, and Fraser drain-
ages.

Catostomus (Deltistes) luxatus Cope 1879.
Lost R. sucker. K; Klamath dr., Ore., Calif.
Catostomus owyhee of Pliocene L. Idaho is
a relative (Smith 1975).

Clxasmistes brevirostris Cope 1879. Short-
nose sucker. K; Klamath dr. Ore., Calif.
Status: vulnerable (Miller 1977).

Chasmistes cujus Cope 1883. Cui-ui. L.
Status: endangered (Miller 1977).

Chasmistes liorus Jordan 1878. June suck-
er. BP. Status: threatened or extinct (Miller
1977).

Chasmistes sp. SU. Status: extinct. C. spa-
tulifer, Pliocene L. Idaho, is a relative
(Smith 1975).

+ Chasmistes spp. Undescribed Plio-
Pleistocene fossil Chasmistes are known
from four other localities: Owens Valley dr.
(DO), Madeline Plains (OMd), Fossil Lake
(OFR), and the Thatcher Basin, Idaho (BB)
(Miller 1965; Bright 1967).

Xyrauchen texanus (Abbott) 1861. Hump-
back sucker. C. Status: vulnerable (Miller
1977).

Cyprinodontidae

Empetrichthys merriami Gilbert 1893. Ash
Meadows killifish. DT. Status: extinct.

Empetrichthys latos Miller 1948. Pahrump
killifish. DP. Status: endangered (in refuges;
extinct in native habitat) (Miller 1977). £.
erdisi (Jordan) 1924a from the Pliocene of
Ridge Basin, L.A. Co., Calif., is a relative
(Uyeno and Miller 1962).

Crenichthys baileyi (Gilbert) 1893. White
River springfish. CW. Status: special con-
cern (Miller 1977).

Crenichthys nevadae Hubbs 1932. Rail-
road Valley springfish. CRR. Status: special
concern (Miller 1977).

+ Fundulus spp.— Five species of Late
Cenozoic Fundulus are known from S Calif,
and Nevada (Death Valley, Mojave dr., La-

hontan dr., and Ridge Basin); they are rela-
tives of Empetrichthys and Crenichthys
(Uyeno and Miller 1962).

Cyprinodon salinus Miller 1943. Salt
Creek pupfish. DMn.

Cyprinodon nevadensis Eigenmann and
Eigenmann 1889. Amargosa pupfish. DMn,
Dt. Geographically variable (Miller 1948,
LaBounty and Deacon 1972). Status: several
subspecies rare or endangered, one extinct
(Miller 1977).

Cyprinodon diabolis Wales 1930. Devils
Hole pupfish. DT. Status: endangered (Mill-
er 1977).

Cyprinodon radiosus Miller 1948. Owens
pupfish. DO. Status: endangered (Miller
1977).

+ Cyprinodon breviradius Miller 1945,
from the Tertiary of Death Valley is related
to the above four species.

Cottidae

Cottus bairdi Girard 1850. Mottled scul-
pin. BB, BT, BP, BSv, BD, BSn, SU, SM,
OH, C; Columbia drainage north to British
Columbia, east through Missouri, L. Winni-
peg, Hudson Bay and Ungava drainages,
and across northern U.S. south to Oregon,
Nevada, Utah, New Mexico, Montana, Iowa,
Missouri, Alabama, and Georgia. Late
Pleistocene fossil, L. Bonneville (Smith et al.
1968). Geographically variable (Bisson and
Bond 1971, Bond 1963).

Cottus confusus Bailey and Bond 1963.
Shorthead sculpin. SL, SM; Columbia, Pu-
get Sound, and Flathead R. drainages.

Cottus extensus Bailey and Bond 1963.
Bear L. sculpin. BB (Bear L.). Late Pleisto-
cene fossil, L. Bonneville (Smith et al.
1968).

Cottus echinatus Bailey and Bond 1963.
Utah L. sculpin. BP (Utah L.). Status: ex-
tinct.

Cottus beldingi Eigenmann and Eigen-
mann 1891. Piute sculpin. BB, SU, SL, SM,
C (Grand R. dr., Colorado, only; not includ-
ed in calculations), L; Columbia dr. Plio-
cene fossil, Lahontan basin (Jordan 1924,
Hubbs and Miller 1948b).

26

GREAT BASIN NATURALIST MEMOIRS

No. 2

Cottus leiopomus Gilbert and Evermann
1894. Wood River sculpin. SW.

Cottus greenei (Gilbert and Culver) 1898.
Shoshone sculpin. SM (Thousand Springs
area, mouth of Salmon Falls Cr., Idaho).

Cottus pitensis Bailey and Bond 1963. Pit
sculpin. G; Pit R. dr., California and
Oregon. May be extinct in Oregon (Bond
1973). C. calcatus Kimmel 1975 of the
Miocene-Pliocene Deer Butte fm., SE Ore-
gon, (SM), is a relative.

Cottus princeps Gilbert 1898. Klamath
Lake sculpin. K (Klamath L.).

Cottus klamathensis Gilbert 1898. Mar-
bled sculpin. K (Klamath basin); upper Pit
R., Calif.

Cottus tenuis (Evermann and Meek) 1898.
Slender sculpin. K (Klamath basin).

Cottus rhotheus (Smith) 1883. Torrent
sculpin. SM; Columbia dr. and nearby
coastal streams, Puget Sound dr., Fraser dr.
southern B.C.

Significance of the Fossil Record

Consideration of the fossil evidence, as
noted among the species accounts, leads to
several general conclusions that serve as a
background for analysis of history of the
fauna. Those conclusions are as follows.

(1) Many genera of the intermountain
fish fauna were in the region, and in
some of the areas presently occupied, by
Pliocene time (e.g., Prosopium, Salve-
linus, Salmo, Oncorhynchus, Acrocheilus,
Gila, Mylocheilus, Ptychocheilus, Richard-
sonius, Catostomus [3 subgenera], Chas-
mistes, Empetrichthys, and Cottus).

(2) Pliocene forms are specifically differ-
ent from Recent forms.

(3) Pleistocene forms are not usually spe-
cifically distinguishable from Recent
counterparts.

(4) The general latitudinal gradients that
exist today are also reflected in the fossil
occurrences; i.e., Cottus was restricted to
the northern regions, cyprinodontids
were restricted to the southern regions,
except that in pluvial periods some

northern forms (e.g., Chasmistes) were
much further south, and in the Pliocene
some faunas included forms now dis-
placed to subarctic regions (e.g., Myo-
xocephalus) as well as forms now dis-
placed to the south (e.g., Orthodon,
Mylopharodon, Archoplites; Lake Idaho,
Smith 1975).

(5) During some times in the past, distri-
butions were broader and faunas larger
than at present (corollary of No. 4; e.g.,
Chasmistes; Lake Idaho fauna).

Analysis of Fish Distributions

What are the major patterns of distribu-
tion? From the basic data of records of na-
tive occurrence of intermountain fishes
among the 48 drainage units, a matrix was
computed showing the correlation of each
species distribution with every other within
the region. Correspondence of species pat-
terns can be examined by similarity indices
and by correlation coefficients. In this case,
the product-moment correlation coefficient
was found to be less sensitive to simple
widespread abundance and was chosen as
the basis for discussion. The correlations are
summarized by a single-linkage phenogram,
in which species are clustered together ac-
cording to the correlation among their dis-
tributions (Fig. 3).

The general nature of the phenogram
may be described in five general sections,
(1) a large, rather closely clustered Kla-
math-Snake R. -Bonneville group; (2) a
group inhabiting the Colorado Basin; (3) a
small group characterizing the Lahontan
Basin; (4) Owens Basin and Death Valley
forms; and (5) isolated species with patterns
unlike any others.

The apparent similarity of patterns of
Klamath and middle Snake species is exag-
gerated somewhat by the fact that many of
their species appear only in one or both of
those units within the study area, but have
diverse distributions outside. Nevertheless,
the similarity of patterns among the north-
ern drainages, including the Bonneville and

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

27

the Colorado, is much higher than among
the southern drainages. Factors contributing
to this phenomenon will be discussed below.
A second observation is that species of
the pluvial Bonneville Basin are scattered in
a number of small diverse clusters, sug-
gesting an unusually complex history for
this assemblage. Lahontan species show the
same trend in a less extreme way; here,
however, the scattered forms are those
widespread species that are found in numer-
ous other drainages (e.g., Salmo clarki, Pros-
opium williamsoni, Rhinichthys osculus, and
Gila bicolor).

Relictus solitarius, which inhabits isolated
basins of east central Nevada, RF, RSt, RW,
and RB, has a unique distribution pattern.
Of the 14 species connecting at a level
more remote than .5, all but 2, Rhinichthys
osculus and Gila bicolor, are restricted to
very isolated, depauperate basins.

At least one cluster of species comprises
widespread, ecologically similar forms,
whose dispersal modes and histories might
be inferred to be somewhat similar. Cottus
bairdi, Prosopium williamsoni, Richardsonius
balteatus, Rhinichthys cataractae, and Ca-
tostomus platyrhynchus are the nucleus of a
Bonneville-Snake medium- and small-stream
fauna; some of these, or their vicariants, are
also found in the Lahontan and Colorado
systems and other northern mountain drain-
ages. They are often ecologically associated
with Salmo clarki (upstream) and Rhii-
nichthys osculus (downstream). It can also
be inferred that any habitat that supports
the majority of this group might well be ex-
pected to support the others. (The success
of introduced Richardsonius balteatus in the
Green River drainage is an example.) These
forms, especially the trout and sculpin, oc-
cupy headwaters and as such are subject to
a higher-than-average incidence of dispersal
by stream capture. (A process by which
"dispersal" to a new drainage system occurs
without the individual fishes necessarily
leaving a limited home range.) If dispersal
by stream capture were the primary factor

determining the breadth of distribution of
these fishes, we should see a correlation be-
tween degree of headwater habitat prefer-
ence and number of drainages occupied

SIMILARITY .9 .8 .7 .6 .5 .4 .3 .2

Colorado R. 6 sp.

Catostomus latipinnis

Gila robusta

Catostomus insignis —
Plagop. argentissimus-
Catostomus clarki
Lepidom. mollispinis •
White R. 4 sp.

Goose L. 3 sp.

Bear L. 4 sp.

Chasmistes sp.

Ca tos tomus ca tos tomus

Cottus leiopomus

Klamath 9 sp. !

Lampetra tridentata —

Salmo gairdneri

Salvelinus malma

Cottus confusus

Middle Snake R. 6 sp.|
Catost. columbianus — — i

Catost. macrocheilus — ■

Acrocheilus aJutaceus — M
Ptycho. oregonensis — '
Cottus bairdi
Prosopium williamsoni
Richardson, balteatus
Rhinichth. cataractae
Catost. plati/rhynch

Gila copei

Gila atraria

Catostomus ardens
Idtich. phlegethont

Cottus beldingi

Ca tos t . di scobol us ■

Salmo clarki

Salmo sp. (redband)

Rhinichthys sp.

Cottus echinatus ■
Chasmistes liorus
Chasmistes cujus
Eremichthys acros
Richardson, egregius
Catost. tahoensis
Catost. warnerensis

Gila bicolor

Rhinichthys osculus

Gila alvordensis

Catost. fumeiventris
Cyprinodon radiosus
Crenichth. nevadae
Empetrichth. merriami
Cyprinodon di
Cyprinodon nevadens
Cyprinodon sa

Catostomus sp.

Empetrichthys latos
Relictus solitarius

mi I

:us— Jl
:ae 'V

ius '

'B^

"-T3

us 2) r

k

merriami — |

abolis — I I

vadensis r

linus I

Fig. 3. Cluster analysis of 83 species of inter-
mountain fishes based on correlation of their distribu-
tions among 48 drainage units. Species with similar
patterns cluster together; species with unique or dis-
tinct patterns cluster at remote levels.

28

GREAT BASIN NATURALIST MEMOIRS

No. 2

(compare species list and Figure 4). The
correlation does not exist, suggesting that
breadth of habitat occupiable by a species,
and extinction resistance, are about as im-
portant as frequency of colonization events.
The most widespread species are Rhi-
nichthys osculus, 32/48 drainages, and Gila
bicolor, 21/48 drainages. These are not par-
ticularly vagile or otherwise prone to colo-
nization, but are interpreted to be extinc-
tion-resistant generalists with ability to
persist in small streams and desert spring
habitats as well as other aquatic environ-
ments. It is inferred that the distribution of
these and other intermountain fishes has
been shaped by a few successful coloniza-
tions and many extinctions.

The model of MacArthur and Wilson
(1963, 1967) is applicable to the study of a
system such as that described in the preced-
ing paragraph. Figure 4 shows the relation-
ship between species and the number of
drainages occupied. Fifty-one of the 83 spe-
cies occupy only one drainage unit in the

study area. Eighteen of these are endemic
to single areas in the study; 33 are more
widespread outside the area, generally in
the Columbia, Pit, Klamath, or Colorado
drainages. The extremely concave curve
suggests a nonequilibrium situation with ex-
tinction heavily predominating over coloni-
zation.

It was noted above that longitudinal dis-
tribution of species in the north is distinctly
broader than that of species in the south.
This could result from any of several pro-
cesses: (1) extinction may have been more
severe in the south, leading to reduced
ranges, (2) there may simply be more longi-
tudinally continuous aquatic habitat in the
north (omitting the Snake and Colorado
drainages from the comparison) because of
drainage orientations and the general north-
ward increase in precipitation /evaporation
ratios, or (3) barriers may be more extreme
in the south. The fossil record, though in-
complete, suggests that extinction of species
has been at least as common in the north

jF

-/

9

t~~ju:

2 3 4 5 6 7 8 9 10 II 12 13 14 15 16

NUMBER OF DRAINAGES OCCUPIED BY SPECIES

-v

r¥

^V

Fig 4 Breadth of distribution of 83 species among 48 drainage units. Species are distributed according to the number of units they occupy. Fifty-
one species (61 percent) are found in only one unit; 18 of these are endemic to that unit, 33 have additional range outside the study area. The steep
concavity of the curve indicates much extinction and limited colonization.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

29

and that the phenomenon exists indepen-
dently of extinction, though possibly in-
tensified by it. Factor (2) is observable and
intuitively obvious as a causal agent in dis-
tribution; factor (3) involves a nonobvious
principle that also may be important to the
understanding of distribution patterns. Be-
cause of the well-known interaction among
latitude, altitude, and temperature (e.g., Jan-
zen 1967), the habitat of cool-stream fishes
tends to be at lower altitudes in the north
and higher altitudes in the south. Many
northern salmonids, suckers, minnows, and
sculpins show this pattern (e.g., Fig. 5). The
result is that mountains of a given altitude
are not as "high" in terms of barrier effects
in the south, in that passes with drainage
connections at 8000 ft may be occupied and
accessible at 40 degrees north but not at 44
degrees north. Similarly, to these species,
lowlands and valleys may be barriers in the

south, but not in the north. Conversely, to
lowland, warmwater fishes such as cyprino-
donts, suitable habitats "pinch out" against
the hillsides at lower elevations in the north
in much the same way that suitable habitats
for sculpins, etc., "pinch out" at upper lim-
its of mountain aquatic habitat in the south.
That cyprinodonts and sculpins, for ex-
ample, have been so ecologically and evolu-
tionarily limited by this phenomenon in the
Intermountain Region, as opposed to east-
ern North America, is significant (cyprino-
donts range north to Canada and sculpins
range south to Alabama in the east; there is
almost no latitudinal overlap in the Inter-
mountain Region).

It could be concluded that in the Inter-
mountain Region mountain barrier effects
on stream fishes decrease southward, except
that an offsetting corollary also exists: the
latitude-altitude-temperature effect that

9

-w

i

8

••

•:; •

Cot t us

bl

Hfdi

7

• •••

• •

•

•

• •

• •

•

•

•

t

••

•

•

ST 6

o

• • •

•

•
• •

X

•

•

•

» 5

•

•
•

/. •

•

•

•

•

V

4

3
<-

i

•

• •• •

•

•

•

z
<

C

o

5

•

O

Q.

z 3

o

>

PANG

9
o

at

•

•

1

LU 2

"

a.

•

•

1 .

•

•

6

1

1

CO

L-£

•

•

1

38° 39°

0° 41° 4 2° 4 3° 44° 45° 46° 4 7

°N Latitude

5. Ecological gradient in habitat of 99 samples of Cordis bairdi from low elevations in the north to high elevations in the south

48°

30

GREAT BASIN NATURALIST MEMOIRS

No. 2

brings temperature optima high into the
mountains in the south also brings arid
desert conditions and higher climatic varia-
bility to the tops of many ranges, thus pro-
viding the ultimate barrier because aquatic
habitat is eliminated entirely. These are
places where an observer can stand beneath
dark clouds and see rain falling but not
reaching the ground.

Analysis of Drainage Units

The 48 drainage imits will be compared
from the standpoints of numbers of species
present, similarity (shared species) of the

faunas of pairs of units, and strengths of
barriers among units. The basic zoogeo-
graphic data have been accurately known
since the C. L. Hubbs expeditions of the
1930s and 1940s, though different tax-
onomic or drainage interpretations change
the numbers slightly and, in a few cases, it
is not certainly known whether species are
indigenous or introduced (Hubbs and Miller
1948b, Hubbs et al. 1974). It is clear from
Figures 6 and 7 that northern drainages
have more species, on the average. Also,
large basins support more species than small
basins (Fig. 8). Peripheral areas have more
species on the average than those of the in-

Fig. 6. Map of drainages showing major existing fluvial and lacustrine habitats, basin outlines as defined in this study, numbers of species of native
fishes in each drainage unit large, central numerals), and numbers of species of fishes limited to only one or the other side of each drainage divide bar-
rier (small numerals Deal drainage divides). Breaks in dividing lines indicate probable late-pluvial connections (compare Fig. 1). See Fig. 2 for names of
drainage units. Basins with no native fishes are unlabeled. Basins with single species are marked with a subscript to the numeral one indicating the in-
- Rfunichthys oscultu. s - Hilutw, solitarius. b - Gila bicolor, E - Empetrichthys latos.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

31

terior, probably as a result of their proxim-
ity to colonization source areas (see Mac-
Arthur and Wilson 1967), but also because
areas associated with the Sierra Nevada and
Wasatch mountains have higher precipi-
tation/evaporation ratios and more aquatic
habitat.

The pattern of species density among
drainage areas (Figs. 4 and 7) is indicative
of the degree of isolation and the frequency
of extinction among Great Basin fishes.
Brown (1971, 1978) showed that the distri-
bution of boreal mammals on Great Basin
mountaintop islands was characterized by
fewer species, especially on small islands,
than expected on the basis of the theory of
extinction-colonization equilibrium. He fur-
ther suggested (1971:477) that the fishes,
like the mammals, colonized extensively
during pluvial (and boreal flora) maxima,
with subsequent isolation, reduced coloniza-
tion, and increased extinction rates. A graph
of log number of species of fishes against

log drainage area (Fig. 8) confirms Brown's
suggestion and provides an even more ex-
treme example. The z (slope) value for the
fish example is 0.59 (r = .66). This value is
considerably higher than the value (z =
.43) for Great Basin mammals, indicating an
extremely high extinction rate (see also
Hubbs et al. 1974:79). The paucity of spe-
cies in the majority of basins, especially
compared with potential source areas (Figs.
6 and 7), is evidence for an extremely low
colonization rate.

The strengths of the barriers among ba-
sins is shown by the small numerals next to
barriers in Figure 6. The numerals give the
number of species whose distribution does
not cross the barrier. For example, 21 spe-
cies of fishes are restricted above or below
the falls of the Snake R. in units SU and
and SM; of the 14 species in the Colorado
drainage (C) and the 13 in the Utah Lake
drainage (BP), 17 species fail to cross the
Wasatch Mountain barrier. (The number in

15
14
13
12
I I

2 "o

<r

« 9
u.

° 8

ce

CD 7

2
i 6

5

4

3

2

I

/ ( (

4

{

/

"

(

ilH

8 9 10 II 12 13 14 15
NUMBER OF SPECIES

16 17 18 19 20 21 22

Fig. 7. Distribution of drainage units according to their species density. Of the 48 drainage units with native fishes. 24 contain only one or two spe-
cies. The extreme extinction indicated by the concavity of this curve was probably intensified during the altithermal 8000-4000 years B.P. Local habi-
tats most affected tend to single species, with sympatry relatively rare (Hubbs et al. 1974, p. 76).

32

GREAT BASIN NATURALIST MEMOIRS

No. 2

common between BP and C is ((13+14) —
17)/2 = 5).

The number of species limited by the
barrier can be converted to an index show-
ing the relative strength of the barrier, in-
dependent of species density, by dividing
the number of species stopped, x, by the to-
tal number of species in the two basins. The
total number, T = (a + b + x)/2, where a
and b are the numbers of species in the two
areas, and x is the number stopped at the
barrier. Thus, the barrier index, B = x/T.
The barrier index between the White Biver
(CW) drainage and Bailroad Valley (CBB)
= 9/((7 + 2 + 9)/2) = 1.0, i.e., all species
stopped by the barrier. That between the
northern Bonneville (BB) and the upper
Snake (SU) = 7/((17+14 + 7)/2) = .37.

With the exception of the last example
given, for which there is a striking geologi-
cal explanation in the late Wisconsin Bon-
neville flood (Malde 1968, Bright 1963), and
the Harney Basin (B = .52) for which there
is evidence for a stream capture (see Bisson

and Bond 1971), the index values for bar-
riers separating the Colorado and Snake
drainages C, Sm, from the Great Basin
drainages are uniformly high (B =
.73— .96). This can be taken as strong evi-
dence for the rarity of colonization across
the Wasatch divide and the southern divide
of the Snake Biver. There is no evidence
for a north-south gradient in the strength of
mountain barriers. Barrier strength within
the Lahontan and Bonneville basins, where
differences must be largely due to differen-
tial extinction after break-up of the system,
are low (B > .6, except where extinction
has been severe but differential.) Barrier
strength between the Bonneville and Lahon-
tan Basins is high (B = .92), but artificially
elevated by severe extinction in the north-
west Bonneville Thousand Springs unit (BT).
Barriers in the Mojave-Owens-Death Valley
systems are strong (B = .75-1.0) because of
the high level of taxonomic differentiation
in this system (Miller 1948) and because of
severe extinction.

17
13
10

uj 7 -
o

UJ

Q- c
</) 5

u.

O 4
cr

3 -

mi2 xlO3 .5
km2xl03

I 5

5

DRAINAGE AREA

20

AO

25

Fig. 8. The number of species in isolated Great Basin drainage units shown as a function of drainage basin area (as outlined in Fig. 1). Slope of the
double logarithmic plot equals 0.59, fitted by least squares regression (r = .66). (Compare with similar graph in Hubbs et al. 1974:79. which shows the
size of numerous fishless basins, but for which basins were defined differently and include hierarchical overlap.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

33

In the Oregon lakes section, the values
for barrier strength show the full range of
possible values, probably because of differ-
ent combinations of extinction. Except for
the Alvord Basin, there is little evidence for
isolation sufficient to produce taxonomic
differentiation. The Harney Basin (OH) is a
moderately isolated composite of several
stages of connections to the Snake River,
and at least one colonization (Gila bicolor)
from the south (Bison and Bond 1971). The
barriers between the Oregon lakes and the
Klamath (K) and Goose (G) systems are rel-
atively strong (B = .63-.93). Finally, the
barriers associated with the central Great
Basin, around the Ruby group of drainages
(R-), Railroad valley (CRR), the White Riv-
er system (CW), and Meadow Valley Wash
(CMV) are strong (B = .63-1.0) and sepa-
rate strongly differentiated faunas, in-
dicating an old belt of isolation and sub-
stantial' differentiation (Hubbs et al. 1974).
The Virgin River system is associated with
the southern part of this belt, showing a re-
markable degree of isolation and differen-
tiation from the upper Colorado drainage
(14 species, B = .78). Part of this must be
explained by recent absence of large-river
fishes (Gila elegans, Gila cyplia, Ptycho-
cheilus lucitis, Xyraachen texanus) from the
Virgin River, but part is the result of a cen-
ter of differentiation probably dating from
before the present configuration of the Col-
orado River in the Grand Canyon area
(Hunt 1969, Metzger et al. 1973, Smith
1966).

In the above consideration of barrier ef-
fects on similarity among drainage units, no
allowance has been made for phylogenetic
information above the species level and its
value in discovering relatively older pat-
terns of dispersal and vicariance. The fol-
lowing method is developed here for adding
phylogenetic information to the system. Of
the various groups under consideration, only
the catostomids have been studied strictly
cladistically (Smith and Koehn 1969), so
species groups, subgenera, and genera of
other groups as currently accepted in recent

traditional taxonomic studies have been
used as the source of phylogenetic informa-
tion. The method is simply to add the
higher taxonomic units, e.g., the Cottus
bairdi species group (Bailey and Bond
1963), the genus Richardsonius, the sub-
genus Pantosteus of Catostomus, etc., to the
taxonomic list as single taxa, recording in
the data matrix each drainage unit in which
they are represented. In this way, 22 higher
taxa were added to the 83 species in the
study— taxa which should contain evidence
of older patterns of drainage connections.
For example, the species in the genus Chas-
mistes individually imply no special pattern
except endemism in their local areas of oc-
currence, but the genus Chasmistes as a tax-
on describes an ancient dispersal pattern of
fundamental significance (Miller 1965) and
seminal significance in zoogeographic meth-
odology (Taylor 1960). Such data, when
added to the matrix, contribute their pat-
tern information to derived faunal indices,
correlations, and summarization.

In the present study, the phylogenetically
augmented data matrix was used in two
ways to analyze similarity of faun-
al/drainage units. First, taxon-by-taxon cor-
relation and covariance matrices were cal-
culated (as for Fig. 3) and the principal
components of these matrices were com-
puted.

Principal component scores for drainage
units were obtained by multiplication of the
principal component matrices by the taxon-
by-drainage data matrix. Second, Jaccard
similarity coefficients were calculated to
show the similarity and differences among
drainage units as revealed by the shared
taxa of fishes found in each. The distance
(difference) matrix was summarized by a
single-linkage phenogram of the drainage
units.

Figure 9 combines the information of the
principal component analysis and the drain-
age phenogram to provide a summary of
faunal similarity among the drainage units.
Principal axes I and II represent separate,

34

GREAT BASIN NATURALIST MEMOIRS

No. 2

uncorrelated, trends among the units, and
placement on the graph is according to cal-
culated scores relative to those trends. The
lines indicate levels of connection based on
degree of dissimilarity (indicated by numer-
als) in the distance matrix. The dominant
trend, representing 29 percent of the infor-
mation in the data matrix, separates the
rich fauna of the Snake R., Bonneville Ba-
sin, upper Colorado, Lahontan Basin, Har-
ney Basin, and Klamath and Goose lakes,
from the depauperate faunas of the more

southern basins and the small Oregon lakes
units. The loadings of the taxa on principal
component I indicate that the correlated
distributions on which this trend is based
are those of the following taxa: Salmonidae,
Cottidae (especially the bairdi species
group), Catostomus, Prosopium, Salmo, and
their northern species. The fewer of these
taxa present in a basin, the lower its score
on principal component axis (P.C.) I. The
strong latitudinal correlation with P.C. I re-
sults partly from dispersal, limitation, and

Fig 9 Principal components (axes) and cluster analysis (lines and numerals) of patterns of fish-faunal similarity among intermountain drainage units.
The letter symbols indicate the position of drainage units on the axes representing major trends in faunal similarity. The lines indicate clusters and the
difference levels of branches in the clusters. Free arrows indicate separate connections to the main cluster, at the difference levels indicated by the nu-
merals (corresponding to the outer, basal connections as in Fig. 3).

1978

INTERMOUNTAIN BIOGEOGRAPHV. A SYMPOSIUM

35

southern extinction of the basic northern In-
termountain fauna.

The second trend in the principal com-
ponents analysis accounts for 12 percent of
the total information in the data matrix and
is correlated with longitude. The Klamath,
Goose, Middle Snake, Lahontan, and most
Oregon lakes basins are distinguished from
the Bonneville, Colorado, and associated
systems. Basins with only Rhinichthys os-
culus and (or) one or two endemic species
are at the extreme (lower) right of the
graph. The trend is based primarily on the
distribution of the subgenus Siphateles (bico-
lor and alvordensis), and to a lesser extent
on the genus Gila (s.l.) and its relatives. The
distribution of salmonids, especially the
rainbow and redband trouts, are also corre-
lated with this axis.

The cluster analysis superimposed on Fig-
ure 9 is generally corroborative of the
structure of the data revealed by principal
components analysis. The faunal similarities
of the Bonneville group, especially its Bear
River section, and the upper Snake, the
Harney Basin-Middle Snake, Lahontan-
Eagle Lake, and Virgin River-Meadow Val-
ley Wash, are instructively depicted. The
cluster of depauperate basins (low on P.C.
I) is apparently united on the basis of gen-
eral absence of most of the northern inter-
mountain fauna. The Silver Lake-Abert
Lake fauna (low on P.C. II) consists of Rhi-
nichthys oscahis, Gila bicolor, and the red-
band trout; the Warner drainage has,, in ad-
dition, Catostomas warnerensis; the Catlow
drainage has only G. bicolor and the red-
band trout. A cluster of six drainages (low
on P.C. I; central on P.C. II) is character-
ized by the presence of Gila bicolor, only.
These are the Alkali and Summer basins,
Oregon; the Newark, Fish Lake, and Dixie
basins, Nevada; and the Mojave drainage.
Adjacent to these in the cluster analysis are
four basins with G. bicolor and R. osculus:
Clover Valley, Diamond Valley, and
Toiyabe drainage, Nevada, and Fort Rock
Basin, Oregon. Three basins have R. os-
culus, only: Long Valley, Nevada (of the

Oregon Lakes group); the Madeline Plains
drainage, California; and the Las Vegas
drainage, Nevada. Three unique drainages
complete the large, depauperate cluster:
Surprise Valley (California and Nevada)
with R. osculus and a species of Catos-
tomus; the Owens Basin, California, with G.
bicolor, R. osculus, Catostomus fumeiventris,
and Cyprinodon rudiosus; and Railroad Val-
ley, with G. bicolor and Crenichthys ne-
vadae. This large cluster is clearly based on
similarity resulting from extinction, not dis-
persal. Data at the subspecies and racial
level would be necessary for the method to
discriminate and cluster in a way that
would reflect fine taxonomic affinity and
dispersal.

Basins with few, but distinctive, endemic-
species, like those of the Ruby group (War-
ing, Steptoe, Butte, and Franklin) and three
Death Valley units (Manly, Tecopa, and
Pahrump) are extreme in all aspects of the
analysis, both principal components scores
and level of clustering.

It is interesting that the Lahontan Basin
(with Eagle Lake) joins the cluster of de-
pauperate basins surrounding it at the 0.4
level and is immediately joined by Goose
Lake. That complex joins the Bonneville-
Upper Snake cluster at 0.5. The other ba-
sins then join into the Great Basin cluster at
the levels indicated. The analysis clearly in-
dicates the intimate association of the Snake
R., Goose L., Colorado R., and Klamath sys-
tems to the Great Basin, notwithstanding
the long isolation and evolution of endem-
ism among these systems. The relationships
and distinctive endemism of the lower Colo-
rado group, CV, CM, and CW, are in-
dicated by their positions and clustering
level in the graph.

The trends and clusters illustrated in Fig-
ure 9 effectively summarize the faunal rela-
tionships among the drainages. The posi-
tions of drainages in the analytical space
based on shared taxa is strikingly similar to
their distribution in geographic space. The
departure from this correspondence can be
interpreted as an indication of the role of

36

GREAT BASIN NATURALIST MEMOIRS

No. 2

extinction in the shaping of distribution pat-
terns of intermountain fishes.

VlCARIANCE

To what extent do the barriers separate
vicariant sister species in a manner sugges-
tive of the classical allopatric speciation
model? The following examples, involving
two dozen species, seem to fit the model
satisfactorily. In the north, Richardsonius
balteatus of the Bonneville-upper Snake is a
sister species to R. egregius of the Lahon-
tan; Gila (Siphateles) bicolor of the Lahon-
tan and associated drainages is a sister spe-
cies to G. alvordensis of the Alvord basin;
Catostomus ardens of the Bonneville and
upper Snake is a sister species to C. macr-
ocheilus of the Columbia system below the
falls of the Snake; C. warnerensis (Warner
V.), C. sp. (Surprise V.), and C. fumeiventris
(Owens V.) are sister species to C. tahoensis;
Cottus echinatus (Utah L.) is a sister species
to C. extensus (Bear L.).

In the south, the three species of Lepido-
meda in the Virgin, Meadow Valley Wash,
and White (Muddy) River drainages are sis-
ter species; Crenichthys nevadae and C.
baileyi in the White River Valley and Rail-
road Valley are sister species; Empetrichthys
merriami and E. latos in Tecopa Valley and
Pahrump Valley are sister species; the
five or six species of Cyprinodon in the
Death Valley system are sister species; and
Catostomus (Pantosteus) discobolus and
clarki of the upper and lower Colorado sys-
tems are sister species. Numerous examples
of more complex relationships exist (e.g.
Smith 1966; Smith and Koehn 1969). The
above estimates are based on cladistic infer-
ences and phenetic similarity. Some esti-
mates will be in error because of paral-
lelism, convergence, or missing sister
groups, just as estimates of fossil ancestors
could be in error for the same reasons. The
strength of 13 barriers involved in the
above separations range from B = .53 to
1.0, with a mean of .83. Barriers not in-

volved in separation of sister species are, as
expected, much lower (.23-.82, x = .58) for
nine barriers in the eastern Bonneville and
upper Snake drainages.

Of these examples, the five species of
Cyprinodon and two species of Emp-
etrichthys in the Death Valley region (Mill-
er 1948) and the Cottus from Bear Lake
and Utah Lake (Smith et al. 1968) seem to
represent postpluvial speciation. The time
frame for the other examples is not readily
known. A more detailed study of degree of
differentiation and barrier chronology is re-
quired. Studies of differentiation and vicari-
ance at subspecific levels in relation to bar-
riers will provide detailed information about
isolation effects and rates of evolution
(Hubbs et al. 1974).

Several species show extensive geographic
variation. Gila bicolor in the Lahontan, Mo-
jave, Owens drainage, and Oregon lakes,
and Gila atraria of the Bonneville and up-
per Snake are outstanding examples, show-
ing variation in size, color, shape, osteology,
gill-raker number and other meristic and
morphometric characters (Hubbs and Miller
1948b, Hubbs et al. 1974, Miller 1973). The
widespread Rhinichthys osculus varies in
size, shape, colors, morphometric and meris-
tic characters, and ecology (Hubbs et al.
1974). Other variable species are Salmo
clarki (LeGendre et al. 1972), Prosopium
williamsoni (Holt 1960), mountain suckers
(Smith 1966), and Cottus bairdi (Bond
1963). The nature of the barriers and habi-
tat gradients suggest that all species found
in more than one system will show differen-
tiation.

Sister species that are found sympatrically
provide striking examples of an alternative
speciation model (or possibly reinvasion af-
ter allopatric speciation). The three endem-
ic species of Prosopium in Bear Lake and
the sympatric ecological races of Gila bico-
lor in the Lahontan Basin (Hubbs et al.
1974) are the best examples. Cottus in the
Klamath system and Gila in the Colorado
drainage might be additional examples.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

37

Historical Changes

In the past 100 years, extinction and col-
onization have greatly intensified as a result
of activities of people. Five species have
become extinct and 18 species are rare, vul-
nerable, endangered, or threatened. The
most serious cause of environmental deterio-
ration was the elimination of vegetative
cover by overgrazing in the period 1880-
1900, which greatly reduced the stability of
rivers and other aquatic habitats (Cottam
1961, Miller 1961). There are a number of
ironic examples of fish populations elimi-
nated by flood, and many examples of
streams and lakes desiccated by lack of
summer flow. Diversion of surface waters
and overutilization of groundwaters have
depleted or eliminated many habitats. Over-
protection can also be a danger: at least
one population became extinct when emer-
gent vegetation crowded out the aquatic
habitat after grazers were excluded by a
protective fence (J. Brown, pers. comm.).

Introduction of exotic species, mostly
from the species-rich Mississippi Basin
faima, has created a situation in which most
intermountain waters now have more in-
troduced species than native species. The
native species, having spent the last ten
thousand or more years under extreme cli-
matic selection, in the absence of much
competition or predation, have proved ill-
adapted to compete and avoid predation
(Hubbs et al. 1974).

A sample of field notes and records rep-
resenting collections taken over the years
1930-1960 (biased toward habitats with na-
tive populations) shows an interesting distri-
bution (Table 2). In trout waters there are
usually 0-2 native species and 1-4 in-
troduced species. In other waters, there is a
tendency for a moderate correlation be-
tween the numbers of introduced and native
species. Habitat that supports a diversity of
native forms tends also to support a diver-
sity of introduced forms. Here, the coloniza-
tion-extinction curve is in a new nonequi-
librium: (manmade) colonizations are
increasing and extinctions are accelerating
almost proportionately.

Discussion

The purpose of this paper is the summary
and interpretation of information about dis-
tribution of intermountain fishes. Quan-
titative methods have been used to make
the evaluation of generalized and unique
patterns more objective. Some of the analy-
ses carried out are not described in the
above section, but should be mentioned for
those who wish to examine the methods in
more detail.

The cluster analysis of species was per-
formed on the 83-species data set as well as
the set based on 105 taxa (species and high-
er taxa). The latter analysis is not figured,
but differs from Figure 3 in that phyloge-

Table 2. Summary of numbers of native and introduced species of fishes in 374 collections from the Great Ba-
sin, Upper Snake, and Upper Colorado drainages, during 1930-1960. Collections contained a mean of 2.4 native
species and 0.9 introduced species.

^ c

||

13

1 *>

(Column
Totals)

4

3

2
1
0

40

108

67

55

43

49

10

2

3

1

1

1

1

6

2

2

2

5

1

6

18

8

7

7

7

1

25

37

25

17

14

23

4

52

32

29

20

13

3

1

2 3 4 5 6

Number of Native Species per Collection

7

18

54
145
150

(Row
Totals)

38

GREAT BASIN NATURALIST MEMOIRS

No. 2

netically related forms have a slightly high-
er tendency to cluster near each other,
partly because they tend to link through
their species-group or genus if they do not
have a more similar species distribution
with which to cluster. Otherwise, the
phenograms are identical. Without the phy-
logenetic information in the matrix, the re-
sults are more ecological and less evolution-
ary.

Principal components analyses were used
because the results are readily interpretable
and reveal meaningful structure. Theo-
retically the method is inappropriate with
categorical data, but the disadvantage is in-
significant, compared to the insights provid-
ed by the method. Components were calcu-
lated from the correlation and covariance
matrices. The latter differ from Figure 9 in
that the structure was summarized in small-
er steps: the first six components accounted
for 52 percent of the total information (27
percent in the first two), in ordinations that
separately discriminated different drainages
on the basis of the uniqueness of their
faunas. The components of the correlation
matrix were broader in this summarization,
revealing more general trends as shown in
Figure 9 (the first six components account
for 66 percent of the total information).

Use of these quantitative methods is justi-
fied by the summarizations they afford and
by the otherwise nonobvious insights that
they provide. Examples of the latter are the
discovery of the basic northern fluvial faun-
al unit and its relation to other species pat-
terns (Fig. 3) and to the drainage units (dis-
cussion of Fig. 9). A second example is the
factoring out and display of the extinction
and dispersal effects by distortion of posi-
tions of the units in Figure 9 relative to
their positions in geographic space.

The barrier analysis (Figure 6) quantified
a basic (and intuitively obvious) distinction
between postpluvial desiccation barriers and
mountain barriers, though severe extinction
in some units artificially increased the cal-
culated barrier strength. The high barrier
indices for the Wasatch Range and the

southern edge of the Snake River Plain
mark zones of extreme differentiation that
are also zones of tectonic interest in that
they are the boundaries of the Great Rasin
subplate (Smith and Sbar 1974). A second
zone of great barrier strength and extreme
faunal differentiation is that running north-
south in eastern Nevada from the Ruby
Mountains to the White River drainage and
Meadow Valley Wash. This zone is near the
region that may mark the center from
which tectonism leading to basin and range
structure began and radiated outward (Arm-
strong et al. 1969, Scholz et al. 1971). The
principal components and barrier analyses
also emphasized the distinction between the
faunas of the Virgin, Meadow Valley Wash,
and White River drainages and that of the
rest of the Colorado R., particularly that
above Grand Canyon. Origin of the dis-
tinction very likely dates from the Pliocene
drainage pattern associated with the inter-
ruption of the Colorado fluvial system by
Hualpai Lake near the mouth of the present
Colorado River Canyon in the Grand Wash
Cliffs (Hunt 1956, 1969).

The problem of the Colorado River con-
nections to the Mojave Desert is still un-
solved (Hubbs and Miller 1948b) but the
fact that the fauna is limited to fishes with
brackish water tolerance (except for the
three species from the north) suggests that
the timing of the connections could date
back to the Rouse embayment, some time in
the Pliocene (Metzger et al. 1973:34). This
suggestion is based on the assumption that a
connection after recession of the embay-
ment and the establishment of the Colorado
River would have allowed colonization by
lower Colorado River primary freshwater
fishes as well as brackish water forms.

Finally, this analysis has proved to be a
study of patterns of extinction as well as
patterns of dispersal. The results indicate
that the two processes are not easily sepa-
rated analytically, and that studies of dis-
persal based strictly on cladism, ignoring ex-
tinction, are in peril of error due to missing
faunas and missing sister groups.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

Unfortunately the process of extinction is
not restricted to the history of the present
example, but is accelerating owing to land
misuse and water use conflicts and is avail-
able for study as a dynamic process. Intense
attention is currently focused on protection
of a few individual species, but attention
must soon shift to long-term maintenance of
stability of aquatic habitats through man-
agement of surface waters and ground wa-
ters at the level of watershed ecosystems.
Greater restraint in the introduction of
exotic species would also enhance the survi-
val of native species.

Acknowledgments

Robert R. Miller provided much useful
information and helpfully reviewed the
manuscript. Ruth L. Elder prepared the
maps and Kama Steelquist drafted graphs
and photographed the figures. Beverly
Smith typed the manuscript. Dwight W.
Taylor provided helpful and thought-pro-
voking discussion.

Literature Cited

Abbott, C. C. 1861. Descriptions of four new species
of North American Cyprinidae. Proc. Acad.
Nat. Sci. Philadelphia 12: 473-474.

Agassiz, J. L. R. 1855. Synopsis of the iehthyological
fauna of the Pacific slope of North America,
chiefly from the collections made by the expe-
dition under the command of Capt. C. Wilkes,
with recent additions and comparisons with
eastern types. Amer. J. Sci. 19: 71-99, 215-231.

Armstrong, R. L., E. B. Ekben, E. H. McKee, and D.
C. Noble. 1969. Space-time relations of Ce-
nozoic silicic volcanism in the Great Basin of
the western United States. Amer. J. Sci. 267:
478-490.

Abmstrong, R. L., W. P. Leeman, and H. E
Malde. 1975. K-Ar dating, Quaternary and
Neogene volcanic rocks of the Snake River
Plain, Idaho. Amer. J. Sci. 275: 225-251.

Atwater, T. 1970. Implications of plate tectonics for
the Cenozoic tectonic evolution of western
North America. Bull. Geol. Soc. Amer. 81:
3513-3536.

Axelrod, D. I. 1950. Evolution of desert vegetation
in western North America. Publ. Carnegie Inst.
Wash. 590: 217-306.

Aybes, W. O. 1854. [Description of new fishes from
California.] Proc. Calif. Acad. Sci. 1: 3-22.

Bailey, R. M., J. E. Fitch, E. S. Herald, E. A.
Lachner, C. C. Lindsey, C. R. Robins, and W.
B. Scott. 1970. A list of common and scien-
tific names of fishes from the United States and
Canada. Amer. Fish. Soc. Spec. Publ. 6: 1-150.

Bailey, R. M., and C. E. Bond. 1963. Four new
species of freshwater sculpins, genus Cottus,
from western North America. Occas. Pap. Mus.
Zool. Univ. Michigan 634: 1-27.

Baird, S. F., and C. Girard. 1854a. Descriptions of
some new fishes from the River Zuni. Proc.
Acad. Nat. Sci. Philadelphia 1853: 368-369.

1854b. Descriptions of new species of fishes

collected by Mr. John H. Clark on the U.S. and
Mexican Boundary Survey, under the direction
of Lt.-Col. Jas. D. Graham. Proc. Acad. Nat.
Sci. Philadelphia 1853: 387-390.

1855. [Cyprinidae of Heermann's collection.]

Proc. Acad. Nat. Sci. Philadelphia 7: 135-138.

BlRKELAND, P. W., D. R. CRANDELL, AND G. M.

Richmond. 1971. Status of correlation of
Quaternary stratigraphic units in the western
conterminous United States. Quaternary Res. 1:
208-227.

Bisson, P. A., and C. E. Bond. 1971. Origin and
distribution of the fishes of Harney Basin, Ore-
gon. Copeia 2: 268-281.

Bond, C. E. 1963. Distribution and ecology of fresh-
water sculpins, genus Cottus, in Oregon. Un-
published doctoral dissertation, University of
Michigan Library, Ann Arbor.

1973. Keys to Oregon freshwater fishes. Ore-
gon State Univ. Agric. Exp. Sta., Tech. Bull. 58:
1-42.

Bond, C. E., and T. T. Kan. 1973. Lampetra (En-
tosphenus) minima, n. sp., a dwarfed parasitic
lamprey from Oregon. Copeia 3: 568-574.

Bright, R. C. 1963. Pleistocene lakes Thatcher and
Bonneville, southeastern Idaho. Unpublished
doctoral dissertation, University of Minnesota
Library, Minneapolis.

1966. Pollen and seed stratigraphy of Swan

Lake, southeastern Idaho: Its relation to region-
al vegetational history and to Lake Bonneville
history. Tebiwa 9(2): 1-47.

1967. Late-Pleistocene stratigraphy in That-
cher Basin, southeastern Idaho. Tebiwa 10(1):
1-7.

Broecker, W. S., and A. Kaufman. 1965.
Radiocarbon chronology of Lake Lahontan and
Lake Bonneville II, Great Basin. Bull. Geol.
Soc. Amer. 76: 537-566.

Brown, J. H. 1971. Mammals on mountaintops:
Nonequilibrium insular biogeography. Amer.
Naturalist 105: 467-478.

40

GREAT BASIN NATURALIST MEMOIRS

No. 2

1978. The theory of insular biogeography

and the distribution of boreal birds and mam-
mals. Great Basin Naturalist Mem. 2: 209-227.

Cavender, T. M. 1969. An early Pliocene Salvelinus
from Nevada with comparative notes on Recent
species. Abstracts, American Society of Ich-
thyologists and Herpetologists 49th annual
meeting, pp. 29-30.

Cope, E. D. 1872. Recent reptiles and fishes, pp.
432-443. In: F. V. Hayden, Preliminary report
of the U.S. Geological Survey of Wyoming and
portions of contiguous territories, part 4. U.S.
Government Printing Office, Washington, D.C.

1874. On the Plagopterinae and the ichthyol-
ogy of Utah. Proc. Amer. Philos. Soc. 14: 129-
139.

1879. The fishes of Klamath Lake, Oregon.

Amer. Naturalist 13: 784-785.

1883. On the fishes of the Recent and Plio-
cene lakes of the western part of the Great Ba-
sin, and of the Idaho Pliocene lake. Proc. Acad.
Nat. Sci. Philadelphia 1883: 134-166.

Cottam, W. P. 1961. Our renewable wild lands: A
challenge. University of Utah Press, Salt Lake
City.

Eaton, G. P., R. L. Christiansen, H. M. Iyer, A. M.
Pitt, D. R. Mabey, H. R. Blank, Jr., I. Zietz,
and M. E. Gettings. 1975. Magma beneath
Yellowstone National Park. Science 188: 787-
796.

ElGENMANN, C. H., AND R. S. ElGENMANN.

1889. Description of a new species of Cyprino-
don. Proc. Calif. Acad. Sci., Ser. II. 17: 270.

1891. Cottus beldingi, sp. nov. Amer. Natu-
ralist 25: 1132-1133.

Evermann, B. W. 1897. A report upon salmon in-
vestigations in the headwaters of the Columbia
River, in the state of Idaho, in 1895, together
with notes on fishes observed in that state in
1894 and 1895. Bull. U.S. Fish Commiss. 16:
149-202.

Evermann, B. W., and S. E. Meek. 1898. A report
upon salmon investigations in the Columbia
River basin and elsewhere on the Pacific Coast
in 1896. Bull. U.S. Fish Commiss. 17: 15-84.

Feth, J. H. 1961. A new map of western con-
terminous United States showing the maximum
known or inferred extent of Pleistocene lakes.
U.S. Geol. Surv. Prof. Pap. 424-B: 110-112.

Forster, J. R. 1773. An account of some curious
fishes, sent from Hudson's Bay; in a letter to
Thomas Pennant, Esq. Philos. Trans. 63: 149-
160.

Gilbert, C. H. 1893. Report on the fishes of the
Death Valley Expedition collected in southern
California and Nevada in 1891, with descrip-

tions of new species. N. Amer. Fauna 7:229-
234.

1898. The fishes of the Klamath River Basin.

Bull. U.S. Fish Commiss. 17: 1-13.

Gilbert, C. H., and B. W. Evermann. 1894. A re-
port upon investigations in the Columbia River
Basin, with descriptions of four new species of
fishes. Bull. U.S. Fish Commiss. 14: 169-204.

Gilbert, G. K. 1890. Lake Bonneville. U.S. Geol.
Surv. Monogr. 1:1-438.

Girard, C. 1850. A monograph of the freshwater
Cotti. Proc. Amer. Assoc. Advanc. Sci. 1849:
409.

1857a. Contributions to the ichthyology of

the western coast of the United States, from
specimens in the museum of the Smithsonian
Institution. Proc. Acad. Nat. Sci. Philadelphia 8:
131-137.

1857b. Researches upon the cyprinoid fishes

inhabiting the fresh waters of the United States
west of the Mississippi Valley, from specimens
in the museum of the Smithsonian Institution.
Proc. Acad. Nat. Sci. Philadelphia 8:165-213.

1858. Fishes. Vol. 10, part IV: 1-400. Report

of explorations and surveys to ascertain the
most practicable and economical route for a
railroad from the Mississippi River to the Pacif-
ic Ocean. House of Representatives, 33rd Con-
gress, 2nd Session, Ex. Doc. No. 91 (1859).

Hansen, H. P. 1947. Postglacial forest succession,
climate, and chronology of Pacific Northwest.
Trans. Amer. Philos. Soc. 37: 1-130.

Holt, R. D. 1960. Comparative morphometry of
mountain whitefish. Prosopium williamsoni.
Copeia 1960(3): 192-200.

Hopkirk, J. D., and R. J. Behnke. 1966. Additions
to the known native fish fauna of Nevada.
Copeia 1966(1): 134-136.

Hubbs, C. L. 1932. Studies of the fishes of the order
Cyprinodontes. XII. A new genus related to
Empetrichthys. Occas. Pap. Mus. Zool. Univ.
Michigan 252: 1-5.

1971. Lamptetra (Entosphenus) lethophaga,

new species, the nonparasitic derivative of the
Pacific lamprey. Trans. San Diego Soc. Nat.
Hist. 16: 125-163.

Hubbs, C. L., and R. R. Miller. 1948a. Two new,
relict genera of cyprinid fishes from Nevada.
Occas. Pap. Mus. Zool. Univ. Michigan 507: 1-
30.

1948b. The zoological evidence: Correlation

between fish distribution and hydrographic his-
tory in the desert basins of western United
States. In: The Great Basin, with emphasis on
glacial and postglacial times. Bull. Univ. Utah,
Biol. Ser. 10: 17-166.

1972. Diagnoses of new cyprinid fishes of

isolated waters in the Great Basin of western

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

41

North America. Trans. San Diego Soc. Nat.
Hist. 17: 101-106.

HUBBS, C. L., R. R. MlLLEB, AND L. C.

Hubbs. 1974. Hydrographic history and relict
fishes of the north-central Great Rasin. Mem.
Calif. Acad. Sci. 7: 1-259.

Hunt, C. R. 1956. Cenozoic geology of the Colo-
rado Plateau. U.S. Geol. Surv. Prof. Pap. 279:
1-99.

1969. Geologic history of the Colorado River.

U.S. Geol. Surv. Prof. Pap. 669-C: 59-130.

Janzen, D. H. 1967. Why mountain passes are high-
er in the tropics. Amer. Naturalist 101:
233-249.

Jordan, D. S. 1878. Contribution to North American
ichthyology. III.R. A synopsis of the family Ca-
tostomidae. Rull. U.S. Natl. Mus. 12: 97-237.

1924a. Miocene fishes from southern Califor-
nia. Rull. S. Calif. Acad. Sci. 23: 42-50.

1924b. Description of a recently discovered

sculpin from Nevada regarded as Cottus bel-
dingi. Proc. U.S. Natl. Mus. 65(6): 1-2.

Jordan, D. S., and R. W. Evermann. 1896-1900.
The fishes of north and middle America. Rull.
U.S. Natl. Mus. 47: 1-3313.

Jordan," D. S., and C. H. Gilbert. 1880. Notes on a
collection of fishes from Utah Lake. Proc. U.S.
Natl. Mus. 3: 459-465.

Kimmel, P. G. 1975. Fishes of the Miocene-Pliocene
Deer Rutte formation, southeastern Oregon.
Univ. Michigan Pap. Paleontol. 14: 69-87.

LaRounty, J. F., and J. E. Deacon. 1972.
Cyprinodon milleri, a new species of pupfish
(family Cyprinodontidae) from Death Valley,
California. Copeia 4: 769-780.

LaRivers, I. 1962. Fishes and fisheries of Nevada.
Nevada State Fish and Game Commission, Car-
son City.

1966. Paleontological miscellanei. Occas.

Pap. Riol. Soc. Nevada 11: 1-8.

Legendre, P. 1976. An appropriate space for clus-
tering selected groups of western Northern
American Salmo, Syst. Zool. 25: 13-195.

Legendre, P., C. R. Schreck, and R. J. Rehnke.
1972. Taximetric analysis of selected groups of
western North American Sahno with respect to
phylogenetic divergences. Syst. Zool. 21:
292-307.

Lucas, F. A. 1900. A new fossil cyprinoid, Leu-
ciscus turned, from the Miocene of Nevada.
Proc. U.S. Natl. Mus. 23: 333-334.

MacArthur, R. H., and E. O. Wilson. 1963. An
equilibrium theory of insular zoogeography. Ev-
olution 17: 373-387.

1967. The theory of island biogeography.

Princeton University Press, Princeton, New Jer-
sey.

Malde, H. E. 1968. The catastrophic late Pleisto-
cene Ronneville flood in the Snake River Plain,
Idaho. U.S. Geol. Surv. Prof. Pap. 596: 1-52.

METZGER, D. G., O. J. LOELTZ, AND R.

Irelna. 1973. Geohydrology of the Parker-
Rlythe-Cibola area, Arizona and California. U.S.
Geol. Surv. Prof. Pap. 486-G: 1-130.
Miller, R. R. 1943. Cyprinodon salinus, a new spe-
cies of fish from Death Valley, California.
Copeia 1943: 68-78.

1945. Four new species of fossil cyprinodont

fishes from eastern California. J. Wash. Acad.
Sci. 35: 315-321.

1946. Correlation between fish distribution

and Pleistocene hydrography in eastern Califor-
nia and southwestern Nevada, with a map of
the Pleistocene waters. J. Geol. 54: 43-53.

1948. The cyprinodont fishes of the Death

Valley system of eastern California and south-
western Nevada. Misc. Publ. Mus. Zool. Univ.
Michigan 68: 1-155.

1958. Origin and affinities of the freshwater

fish fauna of western North America, pp. 187-
222. In: C. L. Hubbs (ed.), Zoogeography. Publ.
Amer. Assoc. Advanc. Sci. 51: 1-509.

1961. Man and the changing fish fauna of

the American southwest. Pap. Michigan Acad.
Sci. 46: 365-404.

1965. Quaternary freshwater fishes of North

America, pp. 569-581. In: H. E. Wright, Jr.,
and D. G. Frey (eds.), The Quaternary of the
United States. Princeton University Press,
Princeton, New Jersey.

1972. Classification of the native trouts of

Arizona with the description of a new species,
Salmo apache. Copeia 3: 401-422.

1973. Two new fishes, Gila bicolor snyderi

and Catostomus fumeiventris, from the Owens
River Rasin, California. Occas. Pap. Mus. Zool.
Univ. Michigan 667: 1-19.

1977. Freshwater fishes. Red Data Rook.

Vol. 4, Pisces. International Union for the Con-
servation of Nature and Natural Resources, Sur-
vival Service Commission, Paris.

Miller, R. R., and C. L. Hubbs. 1960. The spiny-
rayed cyprinid fishes (Plagopterini) of the Colo-
rado River system. Misc. Publ. Mus. Zool. Univ.
Michigan 115: 1-39.

Miller, R. R., and W. M. Morton. 1952. First
record of the dolly varden, Salvelinus rnalma
from Nevada. Copeia 1952: 207-208.

Morrison, R. B. 1965. Quaternary geology of the
Great Basin, pp. 265-285. In: H. E. Wright, Jr.,
and D. G. Frey (eds.), The Quaternary of the
United States. Princeton University Press,
Princeton, New Jersey.

Ore, H. T., and C. N. Warren. 1971. Late Pleisto-
cene-Early Holocene geomorphic history of

42

GREAT BASIN NATURALIST MEMOIRS

No. 2

Lake Mojave, California. Bull. Geol. Soc. Amer.
82:2553-2562.

Richardson, J. 1836. Fauna boreali-americana; or,
the zoology of the northern parts of British
America, containing descriptions of the objects
of natural history collected on the late northern
land expeditions, under the command of Sir
John Franklin, R. N., vol. 4.

Richmond, G. M. 1970. Comparison of the Qua-
ternary stratigraphy of the Alps and Rocky
Mountains. Quaternary Res. 1: 3-28.

Rostlund, E. 1951. Dolly varden, Salvelinus malma
(Walbaum)? Copeia 4: 296.

Russell, I. C. 1885. Geological history of Lake La-
hontan, a Quaternary lake of northwestern Ne-
vada. U.S. Geol. Surv. Monogr. 1-288.

Scholz, C. H., M. Barazanci, and M. L.
Sbar. 1971. Late Cenozoic evolution of the
Great Basin, western United States, as an en-
sialic interarc basin. Bull. Geol. Soc. Amer. 82:
2979-2990.

Smith, G. I. 1968. Late Quaternary geologic and
climatic history of Searles Lake, southeastern
California, pp. 93-310. In: R. B. Morrison, and
H. E. Wright, Jr. (eds.), Means of correlating
Quaternary succession. University of Utah Press,
Salt Lake City.

Smith, G. R. 1966. Distribution and evolution of
the North American catostomid fishes of the
subgenus Pantosteus, genus Catostomus. Misc.
Publ. Mus. Zool. Univ. Michigan 129: 1-132.

1975. Fishes of the Pliocene Glenns Ferry

formation, southwest Idaho. Univ. Michigan
Pap. Paleontol. 14: 1-68.

Smith, G. R., and R. K. Koehn. 1969. Phenetic and
cladistic studies of biochemical and morpho-
logical characteristics of Catostomus. Syst. Zool.
20: 282-297.

Smith, G. R., W. L. Stokes, and K. F.
Horn. 1968. Some late Pleistocene fishes of
Lake Bonneville. Copeia 4: 807-816.

Smith, R. B., and M. L. Sbar. 1974. Contemporary
tectonics and seismicity of the western United
States with emphasis on the intermountain seis-

mic belt. Bull. Geol. Soc. Amer. 85: 1205-1218.

Smith, R. 1883. Description of a new species of
Vranidea (U. rhothea) from Spokane R., Wash-
ington Territory. Proc. U.S. Natl. Mus. 6: 347-
348.

Sneath, P. H. A., and R. R. Sokal. 1973.
Numerical taxonomy; the principles and prac-
tices of numerical classification. W. F. Free-
man, San Francisco.

Snyder, C. T., G. Hardman, and F. F.
Zdenek. 1964. Pleistocene lakes in the Great
Basin. U.S. Geol. Surv. Misc. Geol. Inv. Map I-
416.

Snyder, J. O. 1908. Relationships of the fish fauna
of the lakes of southeastern Oregon, Bull. U.S.
Bur. Fisheries 27: 69-102.

1917. The fishes of the Lahontan system of

Nevada and northeastern California. Bull. U.S.
Bur. Fisheries 35: 31-86.

1919. Three new whitefishes from Bear

Lake, Idaho and Utah. Bull. U.S. Bur. Fisheries
36: 1-9.

Taylor, D. W. 1960. Distribution of the freshwater
clam Pisidium ultrarrwntanwn; a zoogeographic
inquiry. Amer. J. Sci. 258A: 325-334.

Trimble, D. E., and W. J. Carr. 1961. Late Qua-
ternary history of the Snake River in the Ameri-
can Falls region, Idaho. Bull. Geol. Soc. Amer.
72: 1739-1748.

Uyeno, T., and R. R. Miller. 1965. Middle Plio-
cene cyprinid fishes from the Bidahochi forma-
tion, Arizona. Copeia 1965: 28-41.

Valenciennes, A. 1842. In G. Cuvier, and A. Va-
lenciennes (eds.), Histoire naturelle des poissons.
16: 1-472.

Walbaum, J. J. 1792. Petri Artedi renovati, Pars, i-
v. Ichthyologica: Genera Piscium, part iii. Bib-
liotheca et philosophia ichthyologica cura Jo-
hannis Julii Walbaumii edidit.

Wales, J. H. 1930. Biometrical studies of some
races of cyprinodont fishes from the Death Val-
ley region, with description of Cyprinodon dia-
bolis, n. sp. Copeia 1930: 61-70.

ZOOGEOGRAPHY OF REPTILES AND AMPHIBIANS
IN THE INTERMOUNTAIN REGION

Wilmer W. Tanner1

Abstract.— Few, if any, amphibians and reptiles are endemic to Utah. This is also true for much of the Great
Basin, upper Colorado Plateau, southern Idaho, and Wyoming. Many species that would seemingly survive in this
inland, mountainous area are not here. Only one widespread salamander and a few frogs and toads have occu-
pied suitable habitats in the area. Lizards and snakes, like the amphibians, provide few distributions that extend
throughout the area. A migration which presumably followed the Pleistocene Ice Age brought most of the species
into the area as climatic conditions warmed.

Distribution maps of our modern species and subspecies indicate rather clearly that these vertebrates have in-
vaded the Intermountain Region in relatively recent geological time. Only the periphery of Utah and adjoining
states to the east and west have been penetrated by many of the species in the regional fauna.

Few, if any, of the amphibians and rep-
tiles now present in Utah are endemic. This
is perhaps also true for all intermountain
states except those in the south.

There is evidence that for a period of
time at the close of the Pleistocene the
southern Great Basin deserts were more hu-
mid than at present. Studies of fossil pack
rat middens (Neotoma lepida) by Wells and
Jorgensen (1964), indicate that the low
desert "ranges in the vicinity of Frenchman
Flat (Nevada Test Site) were significantly
less arid than at present. Middens now
found in areas where the dominant desert
shrubs are Larrea and Coleogyne have the
leaves, seed, and twigs of Juniperus os-
teosperma imbedded in the crystalline
urine." Wells and Jorgensen (1964) suggest
that the present zonal position of the pi-
nyon-juniper forest in southern Nevada is
about 600 meters above its position of
about 10,000 years ago.

The desert valleys of southern New Mexi-
co, Arizona, and California must have been
considerably more moist and humid during
at least part of each year at about the same
geological time as the valley floors and low
foothills of southern Nevada were covered

by forests of pinyon and juniper. We as-
sume that climatic conditions then existed
which permitted considerable movement of
both reptiles and amphibians.

Ballinger and Tinkle (1972) and Larsen
and Tanner (1975) have assumed that there
were Pleistocene refugia in the southern
deserts of the United States and /or Mexico
which maintained the ancestral stock from
which many of the present species and sub-
species of intermountain amphibian and
reptilian fauna have arisen. The disjunct
populations scattered throughout the "is-
land" mountains of New Mexico and Ari-
zona are highly suggestive of widespread
populations being forced from the low val-
leys into cooler, moister mountains as post-
Pleistocene drying slowly but continuously
changed the valleys into uninhabitable
deserts for many species. The xeric condi-
tions were, however, an invitation for other
species to move in, so that the lower Sono-
ran valleys and their associated mountain
refugia now support a rich and varied series
of amphibians and reptiles.

If we accept the hypothesis that there
was a period of time between the cold, wet
Pleistocene and the dry hot conditions of

'Life Science Museum, Brigham Young University, Provo, Utah 84602.

43

44

GREAT BASIN NATURALIST MEMOIRS

No. 2

today when much of the southwestern
United States was warm but still more hu-
mid and moist than at present, we can envi-
sion a time in which a great migration of
amphibians and reptiles moved toward the
plateaus and the mountainous areas of the
west central United States. Nevertheless,
many species of the south did not penetrate
to environments in the Intermountain Re-
gion which we might expect to be compatible
with their needs.

Many North American species that would
seemingly survive today in intermountain
environments are not here. In June 1942 I
had the privilege of escorting Dr. A. H.

z j \

~~~\r

r~~S_ \-Js0at\t^

-~L\i ttZ

1 I <^L

\r^\ ft jSBISl

a

Fig. 1. (a) Known western fossil records of Oph-
isaurus attenuates; (b) isolated populations of two
plethodontid salamanders: upper, Plethodon neo-
mexicanus, lower Aneides hardyi.

Wright into various habitats in central
Utah. He was indeed disappointed not to
find salamanders in the debris and leaf mold
in and adjacent to mountain springs. Some
species of Plethodontidae have reached
northern Idaho and the mountains of central
New Mexico. Those in Idaho are related to
species along the coast from Washington
south into California. We can only surmise
that the ancestral stock of these species in-
vaded our area from north central North
America after the Ice Age while the north-
ern tier of states was moist enough to per-
mit their movement. It is puzzling why
some did not persist in the Snake River Val-
ley and the Uinta and Wasatch mountains
of Utah.

At the close of the Pleistocene, some spe-
cies expanded their ranges westward across
the Great Plains. One example is cited by
Etheridge (1961) in which fossil remains of
Ophisaurus attenuatus were found in glacial
deposits some distance west of their present
range (Fig. la). Unfortunately, few fossils
are available to substantiate the movement
of other species. The disjunct ranges of
many species are highly suggestive however,
of the westward movement which was ap-
parently interrupted on the high plains by
rapid drying as the ice receded. A few
Great Plains species did reach the moun-
tains and are now found as isolated popu-
lations (Fig. lb). Two genera of salamanders
reached New Mexico and are found in the
Jemez and Sacramento mountains of the
southern part of that state: the genera are
represented by Aneides hardyi and Pletho-
don neomexicanus. One may speculate as to
why these or other genera from this large
family did not reach Colorado. The best an-
swer available is that the warming at the
close of the Ice Age provided moist, warm
climatic conditions in what is now the
southern Great Plains but less favorable
conditions on the central plains. With the
ice receding from the mountains, climatic
conditions in the southern Great Plains of
New Mexico were apparently more moist
than at present and similar to the situation

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

45

in southern Nevada about 10,000 to 15,000
years ago.

We are aware of only two plethodontid
salamanders that have survived. The wide-
spread tiger salamander may have been
here during the Pleistocene. If it did exist
here during the Pleistocene, it may repre-
sent one of the few species able to survive
that period by remaining in the region.
Once the valleys became dry, the salaman-
ders were isolated in the mountaintops with
no opportunity to expand their range. Pre-

sumably, their isolated mountain distribution
and the rapidly drying conditions prevented
them from reaching Colorado. To the west,
north, and south of Utah the deserts devel-
oped rapidly. Glacial lakes which were
present in many Great Basin valleys dis-
appeared or were reduced to salty rem-
nants; associated vegetation changes isolated
amphibians in small areas around waterways
and desert springs (Figs. 2c, 2d, and 9).

The drying out and warming of the
southern portions of this vast inland area

Fig. 2. Distribution of some amphibians in the southwestern United States: (a) Canyon treefrog, Hyhi
arenicolor; (b) Woodhouse's toad, Bufo woodhoasei; (c) red-spotted toad, Bufo punctatus; and (d) south-
western toad, Bufo microscaphus.

46

GREAT BASIN NATURALIST MEMOIRS

No. 2

provided an opportunity for species in the
southern deserts (northern Mexico and per-
haps some areas in Sonora, Chihuahua, and
Coahuila) to expand their ranges northward.
It is now possible to detect some such range
expansions along valleys running north and
west from the international border.

Time does not permit an examination of
all species, but we can examine one. The
leopard lizard, Crotaphytus wislizeni, ap-
pears to have emerged from a refugium in
the Chihuahua-Coahuila area and followed
routes approximately as indicated by the ar-
rows in Figure 3a. The migration resulted

Fig. 3. (a) The theorized flow distribution for the
leopard lizard Crotaphytus wislizeni; (b) geographical
distribution as known today.

Fig. 4. Present-day distribution of southwestern rep-
tiles: (a) desert iguana, Dipsosaurus dorsalis; (b) desert
tortoise, Gopherus agassizi, (c) zebratailed lizard, Calli-
saurus draconoides; and (d) banded gecko, Coleonyx
variegatus.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

47

in the distributional pattern for the species
shown in Figure 3b. In this and other spe-
cies occupation of some areas produced
semi-isolation, and geographical subspecies
have evolved in such areas as the Colorado
Plateau, Virgin River Valley, and the Great
Basin of Utah and Nevada. This is true not
only for the leopard lizards but for most of
the species that have extended their ranges
into the northern and western valleys.

Not all species moved as far or perhaps

as fast. An examination of present-day
ranges are the best indicators of the general
movements that occurred. The following list
of 23 species (Figs. 2 and 4-8) all show
range extensions into the Great Basin and
the valleys of the Colorado drainage. Some
species have either expanded their range
more rapidly or have been able to cross ele-
vation barriers of 5,000 feet or higher (Figs.
5 and 6) while others have not (Figs. 4 and
8).

Fig. 5. Distribution of southwest reptiles: (a) Desert spiny lizard, Sreloporus magister; (b) tree lizard,
Urosaurus ornata; (c) side-blotched lizard, Uta stansburiana; and (d) less earless lizard, Holbrookia macu-

48

GREAT BASIN NATURALIST MEMOIRS

No. 2

1. Southwestern toad (Bufo microscaphus)

2. Woodhouse toad {Bufo woodhousei)

3. Red-spotted toad (Bufo punctatus)

4. Canyon treefrog (Hyla arenicolor)

5. Desert tortoise (Gopherus agassizi)

6. Banded gecko (Coleonyx variegatus)

7. Desert iguana (Dipsosaurus dorsalis)

8. Zebra-tailed lizard (Callisaurus draconoides)

9. Lesser Earless lizard (Holbrookia maculata)

10. Desert spiny lizard (Sceloporus magister)

11. Side-blotched lizard (Uta stansburiana)

12. Tree lizard (Urosaurus ornata)

13. Western Whiptail (Cnemidophorus tigris)

14. Western Blind Snake (Leptotyphhps humilis)

15. Western Patch-nosed Snake (Salvadora hexdepis)

16. Coachwhip Snake (Masticophis flagellum)

17. Glossy Snake (Arizona elegans)

18. Common Kingsnake (Lampropeltis getulus)

19. Black-necked Garter Snake (Thamnophis cyrtopsis)

20. Western Ground Snake (Sonora semianulata)

21. Mojave Rattlesnake (Crotalus scutulatus)

22. Speckled Rattlesnake (Crotalus mitchelli)

23. Sidewinder (Crotalus cerastes)

Some species must have survived the
Pleistocene in refugia that lay between the

Coachwhip
WESTERN STATES

Fig. 6. Distribution of southwestern reptiles: (a) Western patchnosed snake, Salvadora hexalcpis; (b)
western blind snake, Leptotyphhps humilis; (c) coachwhip, Masticophis flagellum; and (d) western whip-
tail, Cnemidophorus tigris.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

49

cold climates of the mountains and the
drier, mild climates of southern plains. Such
species include:

1. The Western Toad (Bufo boreas)

2. Spotted Frog (Rana pretiosa)

3. Rubber Boa (Charina bottae)

4. Western Garter Snake
(Thamnophis

Such species seemingly have moved north
as climatic conditions permitted. They oc-
cupied only mountain valleys in the south-
ern parts of their present range (Fig. 9).

Other species have apparently moved
into the Intermountain Region from the

northern or central Great Plains. Such spe-
cies include:

1. The Chorus Frog (Pseudacris tnseriata)

2. Smooth Green Snake (Opheodrys vernalis)

3. Racer (Coluber constrictor)

These are eastern species which seem to
have entered through the northern Great
Plains (Fig. 10). Presumably, the smooth
green snake had a much wider distribution
in earlier times than at present. This is in-
dicated by its disjunct distribution.

If there are reptile species that survived
the Pleistocene in the lower valleys of the
Great Basin of Utah and Nevada, the short-

Fig. 7. Distribution of southwestern reptiles: (a) Glossy snake Arizona elegans; (b) black-necked garter
snake, Tluimnophis cyrtopsis; (c) western ground snake, Sonora semianulata; and (d) common kingsnake,
Lampropeltis getulus.

50

GREAT BASIN NATURALIST MEMOIRS

No. 2

horned lizard (Phrynosoma dougkissi), the
western skink (Eumeces skiltonianus), Fig.
10a), and the sagebrush lizard (Sceloporus
graciosus) are the species most likely to
have done so. These species now range
from the valleys up to at least 9,000 feet in
the mountains and plateaus.

In summary, we can conclude that the
ancestral stocks of the great majority of
present day intermountain amphibians and

Mojave Rattlesnake
STERN STATES

s \i*

f^.

'V

\

_ L I

Sidewinder
WESTERN STATE!

■Aj

Speckled \ N

Rattlesnake
WESTERI

Fig. 8. Distribution of southwestern reptiles: (a) Mo-
jave rattlesnake, Crotalus scutulatus; (b) sidewinder,
Crotalus cerastes; and (e) speckled rattlesnake, Crotalus
mitchelli.

reptiles originated either to the south or
east of the area in question. By far the
greater numbers of both amphibians and
reptiles came from the south or the south-
east.

The Intermountain West is a good, if not
a classical, example of the David Starr Jor-
dan theory. He stated that animals have
three alternatives if radical changes occur in
the environment:

1. They can follow the environment and thus remain
constant.

2. They can remain and adapt to the new environ-
ment.

3. If they can do neither, they will become extinct.

It may not be possible to cite an example
of a reptile or amphibian that has remained
constant. Yet we do have some that have
wide distributions and little morphological
divergence. Two examples are the Charina
bottae and the Bufo boreas. Both have wide
distribution with little external variation.

Without a fossil record, we do not know
how many amphibians and reptiles existed
in our area since late Pleistocene time and
were unsuccessful in the struggle for survi-
val. Presumably, we have had during the
last ten thousand to fifteen thousand years
substantial environmental changes that of-
fered challenges beyond the ability of some
species to adapt. Other species have extend-
ed their range through adaptive radiation,
which increased the number of geographical
subspecies or morphological clines in the
species as isolated habitats were occupied.

There is reason to believe that the move-
ment north is still occurring. The estab-
lishment of populations of Crotalus mitchelli
and C. scutulatus in Utah appear to be re-
cent (Fig. 8). The first specimen of C.
scutulatus was taken in 1954 and the first
C. mitchelli in 1960. Both were taken on
the southwest slope of the Beaver Dam
Mountains, only a few miles inside Utah.
Since then, these species have apparently
expanded their ranges and are seen more of-
ten by field workers.

The northern plateau Lizard (Sceloporus

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

51

u. elongatus) has in recent times crossed the occurs along the foothills extending north to
central plateaus of Utah through the low Ephraim and south to Monroe. We do not
areas between Emery and Salina and now have records of this species west of the Se-

Western Toad

,'l STATES

Fig. 9. Amphibians and reptiles with a more northern distribution but with populations extending south
into parts of the Intermountain West and the Great Basin: (a) Spotted frog, Rana pretiosa; (b) rubber boa,
Charina bottae; (c) western toad, Bufo boreas; and (d) common garter snake, Ttiamnophis sirtalis.

52

GREAT BASIN NATURALIST MEMOIRS

No. 2

■ELL

1978

INTEKMOL .STAIN BIOGEOGRAPHY! A SYMI'OSH M

53

vier River. A specimen of the long-nosed
snake ( Rhniocheilus lecontet) was recentl)
taken south of Dragerton- an indication that
this species is still expanding its range.

In conclusion, it should be noted that
many areas in Utah souk- local, others ex-
tensive) have had their reptile populations
reduced by human activity. The most com-
mon disruptive influence has been over-
grazing on some private, Bureau of Land
Management, and state lands.

Figures lb, 3b, and 4-10 are taken largely
from Stebbins 1966). Figure fa is from
Etheridge fr961j. Even though the distribu-
tion maps have not been brought up to date
for 10 years, the ranges of species used in
this study have changed little.

LlTERAl i HI. Cll ED

Ballinceb, R. K., and D. VV. Tinkle
1072. Systematica and evolution of tin- genus
Ufa Sauna [guanidae] Misc Publ Mu* Zool
Univ. Michigan 1 i~ I 83

Etheridce R L961 Late Cenozou glass lizards

OphisaUTUi from the fOUtheni Great Plains.

Herpetologica IT J T'j 186.
I K R and W. W. Tanneb L975 Evolution

of the- scleroporine lizards [guanidae ' Jreal

Basin Nat 35 1 20

Stebbins R C L966 A held guide to western rep-
tiles and amphibians Houghton Mifflin Co

Boston

Wells B. v., and C. D. Jobcensek 1964
Pleistocene wood rat middens and < limatic

changes in the Moja ) i re ord of juni-

per woodlands. Science 143 1171-1173.

Fig. 10. Amphibians and reptiles which bav< irthem distribution hut which seemingly have

entered the interrnouritain and Great Basin areas from the central Great Bla.

mec&s skiltonumu.% this distribution is more comparable to those in Fig. 9 ; b smooth green snake, ftyh-
eodrys i.errui': Cohtber "jnArutor: and d chorus frog. Pseudocrii mgfita.

AVIAN BIOGEOGRAPHY OF THE GREAT BASIN AND INTERMOUNTAIN REGION

William H. Behle'

Abstract.— There are no endemic species of birds in the Great Basin. Nevertheless, a distinctive Great Basin
avifauna exists which contains components of the Mojave Desert, Rocky Mountain, and Great Plains avifaunas as
well as species obligate to sagebrush and the pinyon-juniper forest. Seemingly there has been little spread of Cal-
ifornia and Sierra Nevada species eastward, but a westward extension from the Rocky Mountains of several spe-
cies is indicated. While several Rocky Mountain species reach their western limits on the eastern edge of the
Great Basin, others have extended into the eastern portion. Two Great Plains representatives are late arrivals,
namely the Baltimore Oriole and Indigo Bunting, with evidence of introgression now occurring with related
western species. A similar but longstanding situation exists for the flickers. A zone of hybridization occurs in
northern Utah between two species of junco. A rather abrupt junction zone between the Great Basin and Mojave
Desert avifaunas exists in southern Nevada and extreme southwestern Utah. Several species representing the Mo-
jave Desert avifauna have extended their ranges in recent years into southern Utah. Geographically variable birds
show diverse patterns of distribution along with much clinal variation and intergradation. A center of differen-
tiation for four species occurs in western Utah in the eastern portion of the Great Basin while two more occur in
the western portion of the basin. The Wasatch Front is a dividing area between western and eastern races in
several species. Extreme southwestern Utah constitutes a transition area where several species are represented by
different races or intergradational populations. A study of the avifaunas of 14 boreal "islands" in isolated moun-
tain ranges in western and southeastern Utah in comparison with the Rocky Mountain "continent" in central and
northern Utah shows a close correlation between number of species present and habitat diversity. In addition, a
low correlation exists between the number of species that are permanent residents on isolated mountains and the
distance of those mountains from the "continent."

Biogeography is concerned with the dis-
tribution of organisms in time and space.
Applying this to birds and the Great Basin
region, it is the consensus among students of
avian paleontology that most species of
modern birds arose during the Pleistocene
(Selander 1965), but there is virtually no
fossil record of birds for the Great Basin
during that interval of time. Trimble and
Carr (1961) mention that bones of birds as
well as of several kinds of mammals and
molluscs have been found in gravel over-
lying the Raft Formation of American Falls
Lake bed in southern Idaho which represent
the late Quaternary, but no identities of the
birds are given. A number of bird bones as-
sociated with prehistoric human habitations
in caves in the Great Basin have been
found, two of the best-known sites being
Danger Cave near Wendover (Jennings

'Department of Biology. University of Utah, Salt Lake City, Utah 84112.

1957) and Hogup Cave near the north-
western corner of Great Salt Lake (Aikens
1970), but all the bird bones and feathers
represent living species. Hall (1940) de-
scribes an ancient nesting site of White
Pelicans at Rattlesnake Hill on the north-
eastern edge of the town of Fallon, Nevada,
containing bones of White Pelicans, Double-
crested Cormorants, and a Canada Goose.
The bones were situated beneath a water-
formed calcareous layer, which indicated
that the bones had been under water at
least once; but whether this was before, at,
or after the time when Lake Lahonton at-
tained its maximum level was not ascer-
tained. The implication from the find is that
the avian associates in this prehistoric time
were about the same as one finds today at
the colony on Anaho Island in nearby Pyra-
mid Lake. Despite the virtual absence of a

55

56

GREAT BASIN NATURALIST MEMOIRS

No. 2

fossil record, it is probably safe to assume
that the species of birds present in the
Great Basin in the Quaternary were essen-
tially the same as those present in the re-
gion today. With different climatic condi-
tions, however, from time to time there
have doubtless been different assemblages of
birds and different distributional patterns
than are seen at present. Thus, in the ab-
sence of a fossil record for the region under
consideration, reliance must be placed on
an analysis of the distribution of today's
species in the search for clues to dispersal
routes and subspecific differentiation.

Before considering the spatial dimension
of the biogeography of birds of the region
under consideration, it may be well to note
two special items in connection with birds.
One is that some birds are migratory. Thus
a distinction must be made between sum-
mer residents and permanent residents. The
migratory summer residents are able to eas-
ily traverse distances between mountain
ranges and so are less subject to the effects
of isolation than are the sedentary per-
manent residents. The second point is that
there is a wealth of data pertaining to the
distribution of birds in the collections of
many museums, with much of the data
readily available in published reports. For
the region under consideration the following
constitute the principal sources of informa-
tion on the distribution of birds: for Califor-
nia, Grinnell and Miller (1944); for Nevada,
Linsdale (1936 and 1951) and Johnson
(1965, 1973, 1974); for Idaho, Burleigh
(1972); for Utah, Behle (1943, 1955, 1958,
1960), Behle, Bushman and Greenhalgh
(1958), Behle and Ghiselin (1958), Behle and
Perry (1975), and Hay ward, Cottam, Wood-
bury, and Frost (1976); for Colorado, Bailey
and Niedrach (1965). Phillips (1958) has dis-
cussed many special problems having to do
with the collecting of birds and the short-
comings of museum collections. Even
though the material available falls short of
the need, birds are still one of the best
known groups of animals in terms of
biogeography.

Great Basin Avifauna

Turning now to the spatial dimension, an
important initial consideration is whether
there is a distinctive avifauna in the Great
Basin. Are the kinds of birds that occur in
western Utah and Nevada different en
masse from those found in the California-
Sierra Nevada region on the west or the
Colorado-Rocky Mountain region to the
east? This query pertains only to land birds,
since water birds are widespread in their
occurrence throughout North America and
generally show few regional distinctions ex-
cept for relative abundance of particular
species. An analysis of the distribution of
the land birds indicates that the great ma-
jority that occur in the Great Basin range
widely throughout western North America.
There are about 154 kinds of resident birds
in this wide-ranging category. Any dis-
tinctions, then, pertain to a relatively few
kinds, but mostly it is a matter of a differ-
ent combination of species in the Great Ba-
sin as compared with the assemblages in
neighboring regions. Udvardy (1963: 1157)
includes a Great Basin avifauna in his treat-
ment of the bird faunas of North America,
stating that the species fall geographically
and ecologically into two groups, namely
(1) the sagebrush-arid woodland faunal
group and (2) the northwestern arid wood-
land faunal group. Miller (1951) in his anal-
ysis of the distribution of the birds of Cali-
fornia includes a Great Basin avifauna as
one of four faunal groups represented in the
state, one that is intrusive into northeastern
California east of the Sierran crest. He
states that the Great Basin avifauna consists
of two categories: (1) species of interior
continental derivation that occur south of or
below the boreal areas, and (2) geographic-
races that have differentiated in the Great
Basin at austral levels. He designates 35
kinds as belonging to the Great Basin avi-
fauna. Johnson (1975) followed Miller in his
treatment of a Great Basin avifauna.

Probably the most distinctive feature
about the Great Basin avifauna is the pres-

1978

INTERMOUNTAIN BIOGEOGRAPHYI A SYMPOSIUM

57

ence of certain birds that are associated
with two plant formations that occur widely
throughout the region, namely big sage (Ar-
temisia tridentata) and the pinyon-juniper
woodland. Birds that occur almost exclu-
sively in stands of sagebrush are the Sage
Grouse, Sage Thrasher, and Sage Sparrow.
Birds that occur chiefly, if not exclusively,
in the pygmy woodland, which itself has
much sage interspersed with the junipers
and pinyon pines, are the Cassin's Kingbird,
Gray Flycatcher, Scrub Jay, Pinyon Jay,
Plain Titmouse, Bush-tit, Blue-Gray Gnat-
catcher, Cedar Waxwing, Gray Vireo,
Black-throated Gray Warbler, and Brewer's
Sparrow. However, the pygmy woodland
occurs throughout the Southwest so these
associated species of birds occur in areas
beyond the Great Basin. To properly char-
acterize the Great Basin avifauna, com-
parisons with surrounding regions are neces-
sary.

There are about 30 kinds of distinctive
birds that occur in the California-Pacific
Coast-Sierra Nevada region that are not
known to occur in either the Great Basin or
the Rocky Mountains. Many are endemic to
the West Coast area and constitute the
most conspicuous elements of the California
avifauna. Some of these, such as the Moun-
tain Quail and White-headed Woodpecker,
occur in the Sierra Nevada on the western
rim of the Great Basin, but I find little evi-
dence of these distinctive California forms
spilling over eastward into the mountain
ranges in the Great Basin. Several northern
birds reach the southern limits of their
ranges, at least in part in the Great Basin.
These are the Marsh Hawk, Roughed
Grouse, Sharp-tailed Grouse, Sage Grouse,
Lewis Woodpecker, Tree Swallow, Swain-
son's Thrush, Water Pipit, American Red-
start, and Fox Sparrow. Many southern
birds reach their northern limits, in at least
part of their range, in the Great Basin.
These are the Whip-poor-will, Black
Phoebe, Gray Flycatcher, Plain Titmouse,
Bewick's Wren, Bendire's Thrasher, Blue-
gray Gnatcatcher, Gray Vireo, Virginia's

Warbler, Black-throated Gray Warbler,
Painted Redstart, Scott's Oriole, Lesser
Goldfinch, Black-throated Sparrow, Gray-
headed Junco, and Black-chinned Sparrow.
There are about 25 kinds representing the
Mojave Desert avifauna that occur in south-
eastern Nevada and southwestern Utah but
which do not penetrate any farther north
into the Great Basin except on an acciden-
tal basis. These are discussed in the follow-
ing section of this paper.

Nineteen kinds of birds occur in Colorado
that do not occur as breeders in either the
Great Basin or California areas. Mostly
these are species of the Great Plains avi-
fauna that reach the western limits of their
ranges along the east base of the Rocky
Mountains. One species is endemic to the
mountains of Colorado, namely the Gray-
crowned Rosy Finch. There are several spe-
cies that occur as breeders in both the
Rocky Mountains and the Great Basin
which do not occur in the California-Sierra
Nevada area. Thus they reach their western
limits within the Great Basin. These are the
Northern Three-toed Woodpecker, Catbird,
Brown Thrasher, Veery, Water Pipit, Black
Rosy Finch and Indigo Bunting. None of
these are common in the Great Basin and at
least two, the Brown Thrasher and Indigo
Bunting, appear to be late arrivals in the
region west of the Rocky Mountains. Three
species are found at the eastern edge of the
Great Basin in Utah but are not known to
occur in the basin per se. These are the
Purple Martin, Gray Jay, and Pine Gros-
beak. Several kinds are essentially restricted
in Utah in their breeding range to the Colo-
rado River drainage system, but occasionally
individuals occur in the Great Basin as acci-
dentals. These are the Gambel Quail, Costa
Hummingbird, Roadrunner, Bendire's
Thrasher, and Blue Grosbeak. Finally, I
know of no species of bird that is endemic
to the Great Basin.

From all this we can conclude that there
is a distinctive Great Basin avifauna but it
is one that is not characterized by endemic-
species. Rather it is recognizable on the

58

GREAT BASIN NATURALIST MEMOIRS

No. 2

basis of a different assemblage of birds,
many of which are intrusive from surround-
ing regions. There is more evidence of a
western spread of eastern species into the
Great Basin than there is of an eastward
spread from the Sierra Nevada-California
area. Because of the lack of endemics, the
Great Basin avifauna is not as distinctive as
surrounding avifaunas, but it is more sharp-
ly confined, being delimited on the west by
the Cascade-Sierra cordillera and on the
east by such outlying ranges of the Rockies
as the Wasatch Mountains of northern Utah
and the high plateaus of central Utah. On
the south the Great Basin avifauna meets
the Mojave Desert avifauna in a rather dis-
tinct and narrow junction zone. There is no
comparable junction zone or mountain bar-
rier at the northern limits of the Great Ba-
sin. Here the Great Basin species gradually
merge with those of either the western
woodland edge or those of the open Palouse
country east of the Cascades.

Relations of Mojave Desert and

Great Basin Avifaunas in Southwestern

Utah

and Southeastern Nevada

The northern limits of the Mojave Desert
Biome in Nevada have been mapped by
Gullion et al. (1959: 279). Areas included
are Meadow Valley Wash, Muddy River,
and Pahranagat Valley. In southwestern
Utah, the warm southern desert occurs
along the floor of the Virgin River Valley
to the mouth of Zion Canyon near Spring-
dale (including Coal Pits Wash) as well as
along the lower stretches of tributary
streams such as La Verkin, Ash, and Santa
Clara creeks and Beaver Dam Wash on the
west side of the Beaver Dam Mountains. In
Arizona it occurs along the Virgin Riv-
er Valley. There are 28 kinds of summer
resident birds in this region that are repre-
sentatives of the Mojave Desert avifauna.
Fifteen of these are known to occur in Utah
only in this area. The other 13 occur there
regularly but a few extralimital records exist

elsewhere in the state. The Mojave Desert
avian indicators are the Black Hawk, Gam-
bel's Quail, White-winged Dove, Ground
Dove, Inca Dove, Roadrunner, Lesser
Nighthawk, Costa's Hummingbird, Rivoli's
Hummingbird, Ladderbacked Woodpecker,
Wied's Crested Flycatcher, Black Phoebe,
Vermilion Flycatcher, Verdin, Cactus Wren,
Le Conte's Thrasher, Crissal Thrasher,
Black-tailed Gnatcatcher, Phainopepla,
Bell's Vireo, Lucy's Warbler, Painted Red-
start, Hooded Oriole, Scott's Oriole, Sum-
mer Tanager, Blue Grosbeak, Abert's Tow-
hee, and Rufous-crowned Sparrow. Several
of these species seem to have extended their
ranges into southwestern Utah in recent
years, namely, the Black Hawk, White-
winged Dove, Inca Dove, Rivoli's Hum-
mingbird, Wied's Crested Flycatcher, Black-
tailed Gnatcatcher, Summer Tanager, and
possibly the Rufous-crowned Sparrow, al-
though the latter may represent an over-
looked species associated with a relict grass-
land habitat. A summary of records and
details of distribution for this complement
of birds has recently been presented by
Behle (1976b) for the three-state region. Im-
mediately to the north of this Mojave
Desert or Lower Sonoran area and at high-
er elevations in the region in the pinyon-
juniper belt, birds are found that represent
the Great Basin avifauna.

There are some aspects of subspecies dis-
tribution and intergradation in extreme
southwestern Utah that are significant in
terms of southern derivations of the popu-
lation. These are discussed elsewhere in this
paper. The hybridization that produces in-
tergradation in these several species, as well
as increased variability in the populations,
suggests the presence of a suture zone, us-
ing the terminology of Remington (1968).
He defined a suture zone as "a band,
whether narrow or broad, of geographic-
overlap between major biotic assemblages,
including some pairs of species or semi-
species which hybridize in the zone." As
Uzzell and Ashmole (1970) further note, su-
ture zones stand to biotas as zones of sec-

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

ondary intergradation stand to pairs of pop-
ulations. Unless one prefers to regard the
Gilded Flicker as a separate species from
the Red-shafted Flicker, to my knowledge
no hybridization occurs in extreme south-
western Utah at the species level. Rather,
the crossing is between representatives of
different subspecies producing intermediate
and highly variable populations where, in
addition to the intergrades, typical repre-
sentatives of the two parental stocks occur.
In the region to the north of the Virgin
River Valley, some cases are known where
introgression has taken place. These are dis-
cussed in another section of this paper.

Boreal Islands and Effects of Isolation

One of the most significant aspects of
zoogeography in the Great Basin and Inter-
mountain Region pertains to the dis-
continuous occurrence of boreal species on
the many isolated mountaintops of the re-
gion. The distribution of birds on 31 such
islands has been discussed by Johnson (1975)
in a study patterned after similar studies by
Brown (1971) on mammals of the Great Ba-
sin ranges and Vuilleumier (1970) on birds
in the paramo islands in the northern
Andes. Although Johnson had data available
from several of my reports for certain is-
lands in western Utah which constituted the
eastern fringe of his study area, additional
data for Utah have been mobilized for this
paper to extend Johnson's study. Although I
have followed his procedures, our data are
not precisely comparable because of region-
al differences in the avifauna, my elimi-
nation of water birds from the boreal cate-
gory, and the circumstance that I have
followed Brown's approach of considering
as boreal species those that occur above
7500 feet elevation rather than attempting
to determine the lower edge of the forest
woodland. The 80 species that I have desig-
nated as boreal are listed in the Appendix
along with an indication of their presence
or absence on the 14 boreal islands studied
in western and southeastern Utah. The basic-

data for the several islands are presented in
Table 1. These data were first subjected to
a normality check which showed that they
fit a normal distribution in an untrans-
formed condition. The data were then ana-
lyzed by means of a partial correlation
analysis which showed that three variables,
namely elevation of highest peak, total area,
and habitat diversity score (HDS) were
highly correlated. Then a stepwise multiple
regression study showed that HDS had the
highest correlation with the total number of
bird species occurring. The R-value (correla-
tion coefficient) was .86, which was signifi-
cant at the .001 level. Because of the inter-
correlation among the three independent
variables, the multiple regression analysis
was run first with HDS included in the
equation while excluding elevation and total
area. Then it was run excluding HDS but
including elevation and area. Finally calcu-
lations were made including all three varia-
bles. As is indicated by the data summa-
rized in Table 3 and the adjusted R2 values,
the effect of multicolinearity is present
when all three variables are included in the
equation. Although not as strong, these re-
sults follow those of Johnson closely. Be-
cause of the multicolinearity when all three
independent variables were included in the
equation, the R2 value given in Figure
2 (which is .75) is the value derived from
the equation excluding elevation and total
area. The results shown in Figure 2 are sim-
ilar to those of Johnson (1975: 553).

Although there is a high correlation be-
tween the number of kinds of birds occur-
ring and general habitat diversity as repre-
sented by the habitat diversity scores, there
is the complication that the HDS involves
many environmental variables. In attempt-
ing to identify particular aspects of the hab-
itat community structure that control the
kinds and number of species present, John-
son (1975: 555) analyzed the species compo-
sition of the boreal birds and their ecologic
rolls in the community. He divided them
into two groups: "Restricted," which oc-
curred in 5 or fewer of his 31 sample areas,

60

GREAT BASIN NATURALIST MEMOIRS

No. 2

Fig. 1. Map of Utah showing locations of Boreal Islands and Rocky Mountain Continent area above 7500 feet
elevation. 1. Raft River Mts., 2. Deep Creek Mts., 3. Stansbury Mts., 4. Oquirrh Mts., 5. House Range,
6. Needle Range, 7. Wah Wah Mts., 8. Frisco Mts., 9. Mineral Mts., 10. Pine Valley Mts., 11. La Sal Mts.,
12. Abajo Mts.-Elk Ridge, 13. Henry Mts., 14. Navajo Mtn., 15. Wasatch-Uinta-Tushar-High Plateau Continent.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

61

and "Standard," which occurred in 28 or
more. Birds in the "Standard" category are
presumed to have generalized boreal re-
quirements in contrast to specialized re-
quirements for the "Restricted" group.
Johnson noted Willson's (1974) work that
deals with aspects of habitat structure in re-
lation to species and numbers of birds. A
more recent paper along similar lines is
Flack's (1976) study of bird populations in
the aspen forests in western North America.
The approach of Willson and Flack focuses
attention on the significance of particular
environmental variables presently covered
by Johnson's habitat diversity score.

The next highest correlation shown by
my data is with width of barrier, but this is
significant only in connection with the cate-
gory of permanent residents (R = .43). In
other words, for the summer residents there
is no correlation between number of kinds
occurring on an island and distance from
the nearest island or continent, while for
the permanent residents the number of
kinds decreases with remoteness from the
continental area. The distance correlation is
minor, however, compared with that for
habitat diversity. Again my results are es-

sentially the same as those of Johnson
(1975). He expressed the opinion that the
distance factor in the case of birds operates
through impoverishment of habitat rather
than through ease of access.

A low correlation shows up for my data
between number of species and total area of
the island (see Table 2). This is contrary to
the results of both Brown and Johnson as
well as the postulate of MacArthur and
Wilson (1963, 1967) that area and environ-
mental diversity are closely related and that
total area serves as a good general predictor
of habitat variety. The lack of correlation
between number of species and size of area
for the islands that I studied was probably
influenced by the disparate results for the
two smallest islands, namely the Frisco
Mountains with a size of 11 square miles
and only 19 kinds of birds as compared to
Navajo Mountain with 13 square miles and
49 kinds of birds. I gave a habitat diversity
score of 3 to the Frisco Mountain area and
a 5 to Navajo Mountain. The Frisco range
is very dry and has a sparse coniferous for-
est. Navajo Mountain is also lacking in sur-
face accumulation of water, yet supports
much more forest covering. Environmental

Table 1. Data for Boreal Islands and the Rockv Mountain Mainland in Utah."

Area

Mountain

No.

Range

N,

N2

N3

AR

WB

DM

EHP

LHP

HDS

1

Raft River Mts.

61

22

39

64

48

79

9892

41.92

10

2

Deep Creek Mts.

52

20

32

223

9

104

12101

39.83

11

3

Stansbury Mts.

44

18.

26

54

16

39

11031

40.27

8

4

Oquirrh Mts.

50

20

30

82

16

19

10676

40.22

9

5

House Range

25

9

16

25

35

63

9725

39.09

4

6

Needle Range

29

10

19

92

11

65

9783

38.16

4

7

Wah Wah Mts.

42

15

27

54

11

53

9065

38.33

6

8

Frisco Mts.

19

8

11

11

11

.38

9669

38.31

3

9

Mineral Mts.

34

12

22

24

25

11

9619

38.20

5

10

Pine Valley Mts.

46

18

28

79

39

10

10325

37.32

10

11

La Sal Mts.

64

25

39

314

28

42

13089

38.26

12

12

Abajo Mts.—
Elk Ridge

42

22

20

368

38

70

11445

37.50

9

13

Henry Mts.

41

17

24

108

38

21

11615

38.07

7

14

Navajo Mtn.

49

22

27

13

62

58

10416

37.02

5

15

Wasatch-Uinta
Mainland

80

33

47

-

-

-

13498

40.77

18

°N, = total number boreal species found; N2 = number of these permanently resident; N3 = no. summer residents; AR = total area above 7500
feet in square miles; WB = width of interisland lowland desert barrier, e.g., distance from closest boreal island; DM = distance from mainland; EHP
= elevation of highest peak; LHP = latitude of highest peak; HDS = Habitat Diversity Score.

GREAT BASIN NATURALIST MEMOIRS

No. 2

patchiness or some other aspect of the more
extensive woodland on Navajo Mountain
presumably accounts for the greatly in-
creased number of species present. In these
two instances, at least, total area is not as
good a predictor of number of kinds of
birds as is total forest woodland area with

all the attendant attributes, whatever they
may be.

From his study of boreal mammals on the
mountaintops of the Great Basin ranges,
Brown (1971) concluded that their diversity
and distribution could not be explained in
terms of an equilibrium between coloniza-

Table 2. Results of partial correlation analysis of island data. Upper number indicates the correlation
coefficient; lower number is the level of significance. Meaning of symbols is the same as in Table 1.

N,

N2

N3

AR

WB

DM

EHP

LPH

HDS

1.000

N,

.001

.933

1.000

N2

.001
.970

.001
.816

1.000

N3

.001
.442

.001
.581

.001
.313

1.000

AR

.114

.029

.275

.001

.319

.428

.220

-.023

1.000

WB

.267

.127

.449

.938

.001

.167

.169

.153

.329

-.018

1.000

DM

.569

.564

.602

.251

.952

.001

.581

.673

.474

.771

.055

.128

1.000

EHP

.029

.008

.087

.001

.851

.662

.001

.329

.145

.430

-.132

-.204

.305

-.016

1.000

LHP

.250

.622

.125

.653

.485

.289

.956

.001

.867

.832

.824

.655

.091

.138

.714

.323

1.000

HDS

.001

.001

.001

.011

.757

.638

.004

.260

.001

Table 3. Summary of results from stepwise multiple regression analysis showing relationship of total number
of bird species (the dependent variable) to independent variables. HDS = habitat diversity score; WB = width
of interisland barrier; EHP = elevation of highest peak; AR = total area; DM = distance from mainland; LHP
= latitude of highest peak.

Variable

Multiple R

R2

R2 Change

Simple r

Treatment A.

EHP and AR excluded as independent

variables

HDS

.86693

.75156

.75156

.86693

WB

.89979

.80963

.05807

.31877

LHP

.90712

.82286

.01323

.32918

(constant)

Treatment B.

HDS exc

luded

as an independent variable

AR

.44173

.19512

.19512

.44173

LHP

.58987

.34795

.15283

.32918

WB

.72392

.52405

.17611

.31877

DM

.74207

.55067

.02661

.16689

(constant)

Treatment C.

All variables included in the analysis

HDS

.86693

.75156

.75156

.86693

WB

.89979

.80963

.05807

.31877

AR

.91085

.82964

.02002

.44173

DM

.91693

.84076

.01111

.16689

(constant)

.91870

.84402

.00326

.58097

EHP

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

63

tion and extinction. His interpretation was
that boreal mammals reached all the islands
during the Pleistocene and since then there
have been extinctions but no colonizations.
In his study of boreal birds, Johnson (1975)
concluded that a similar nonequilibrium sit-
uation prevails for the permanent resident
species, but for the summer residents the
equilibrium theory of island species number
does apply since species are excluded by
habitat deficiencies rather than barriers.

Subspecies of Geographically
Variable Species in Utah

An aspect of biogeography that is of pri-
mary interest to the systematist is the geo-
graphic distribution of different subspecies

or races of geographically variable species.
Twenty-three species present systematic-
problems in Utah. Of these, 7 are montane
or boreal forms, 14 are valley or austral
species, and 2 are wide-ranging types that
extend from the valleys up to the mountain-
tops. Of the total, 14 are represented by 2
breeding races and 4 by 3 races, with possi-
bly another in the last category. Another 4
species are represented by only one race in
the state, but each has an intergrading pop-
ulation in some part of Utah that is transi-
tional with another race in surrounding re-
gions. The distribution of the races and
populations in nine geographic regions in
the state is indicated in Table 4 except for
the Red Crossbill, about which a decision as
to the number of races represented in Utah

Habitat Diversity Score

Fig. 2. Relationship between habitat diversity score (HDS) and total number of birds (N\) occurring on boreal
islands in Utah. Numbers of sample areas correspond to those used in Fig. 1 and Table 1. R2 value shown is the
adjusted value because of the small number of sample areas.

64

GREAT BASIN NATURALIST MEMOIRS

No. 2

awaits the results of a pending systematic
review by Allen Phillips. Areas of inter-
gradation of varying extent occur between
the races. Two instances of a minor barrier
effect have been revealed. No uniform pat-
tern of distribution prevails. Rather, there
are several situations indicated whereby 2
or more species show racial changes in
about the same general area. The picture of
variation is more indicative of broad
changes on a regional basis than of differen-
tiation in isolated mountain ranges, as is of-

ten the case with more sedentary groups
such as mammals.

One distributional pattern is where differ-
ences occur between populations in the
west desert portion of northern Utah and
those of the Wasatch and Uinta mountains.
This is seen in the Dusky Grouse, Cliff
Swallow, Mountain Chickadee, Brown Cree-
per, Scrub Jay, and Steller's Jay. Cliff Swal-
lows represent an extreme case of gradual
clinal variation, with only specimens from
the ends of the cline in extreme western

Tahi.k 4: Subspecies of geographically variable birds or intermediate populations represented in various geography

House,

Pine Valley

Raft River,

Stansbury,

Needle,

Mountains

Deep Creek

Oquirrh

Wah Wah

Vircin River Valley

Mountains

Mountains

Mountains

Beaver Dam Wash

Ruteo jamaicensis

calurus

calurus

calurus

calurus

Red-tailed Hawk

Dendragapus ohscurus

oreinus

oreinus >

-

obscurus

Blue Grouse

obscurus

Ottts aslo

inyoensis

inyoensis

inyoensis

inyoensis >

Screech Owl

yumanensis

Bubo virginianus

occidentalis

occidentalis

occidentalis

patlescens

Great Homed Owl

Chordeiles minor

hesperis

hesperis

hesperis

henryi

Common Nighthawk

Picoides vilbsus

leucothorectis >

monticola

leucothorectis

leucothorectis

Hairy Woodpecker

monticola

Empidonax traillii

adastus

adastus

adastus >

extimus

Willow (Traill's)

extimus

Flycatcher

Eremophila alpestris

utahensis

utahensis

utahensis

-

Homed Lark

Petrochelidon pyrrhonota

hypopolia

hypopolia >

hypopolia >

tachina

Cliff Swallow

pyrrhonota

pyrrhonota

Cyanocitta stelleri

-

macrolopha

macrolopha

macrolopha

Steller's Jay

Aphelocoma cocrulescens

nevadae

nevadae

nevadae

nevadae

Scrub Jay

Parus atricapillus

nevadensis

nevadensis

nevadensis

nevadensis

Black-capped Chickadee

Parus gambeli

inyoensis

inyoensis >

inyoensis >

inyoensis

Mountain Chickadee

wasatchensis

wasatchensis

Certhia familiaris

leucosticta

leucosticta

leucosticta

leucosticta

Brown Creeper

Catherpes mexicanus

-

-

-

conspersus

Canyon Wren

Suriiu mexicana

_

-

-

-

Western Bluebird

/ anius ludovicianus

gambeli

gambeli

gambeli

gambeli

Loggerhead Shrike

(leothlypis trichas

occidentalis

occidentalis

occidentalis

occidentalis >

Common Yellowthroat

scirpicola

Agelaius phoeniceus

fortis >

fortis

fortis

fortis >

Red-winged Blackbird

nevadensis

nevadensis

Molothnis aler

artemisiae

artemisiae

artemisiae >

obscurus

Brown-headed Cowbird

obscurus

Ctirpixlacus mexicanus

solitudinus

solitudinus

solitudinus

solitudinus

House Finch

Melospiza meUxiiu

montanus

montanus

montanus

fallax

Song Sparrow

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

65

and eastern Utah sufficiently different to be
assigned to separate races (see Behle 1976a).
In two species more of a step cline is repre-
sented. In one of these, the Dusky Grouse,
specimens from the Deep Creek Mountains
near the Utah-Nevada border, are typical of
the race Dendragapus obscurus oreinus.
Those from the Oquirrh Mountains are clos-
est to oreinus but show an approach to ob-
scurus. In the Wasatch Mountains the
grouse represent the race obscurus. A sim-
ilar situation exists in the Mountain Chick-

adee. Those from the Deep Creek Moun-
tains represent the race Varus gambeli
inyoensis. Those from the Stansbury and
Oquirrh mountains are closest to inyoensis
but show an approach to wasatchensis
which occurs in the Wasatch Mountains and
thence east to the Uinta Mountains. In the
Brown Creeper, representatives from all the
west desert ranges represent the race Cer-
thia familiaris leucosticta. Those from the
Wasatch Mountains are a highly variable lot
of intergrades but as a whole stand closest

Wasatch Mountains
Wasatch Plateau

Pavant,
Tushab
Mountains

Aquarius,

PaUNSAUGUNT,

Mabkacunt

Plateaus

I'inta Mountains—

Tavaputs

Plateau

La Sal— Ahajo

Henhv

Mountains

calurus

calurus

calurus

fuertesi

"

obscurus

obscurus

obscurus

obscurus

obscurus

inyoensis

inyoensis

inyoensis

inyoensis

inyoensis

occidentalis

occidentalis

occidentalis

occidentalis

pallescens

hesperis

hesperis

henryi

howelli

henryi

monticola

monticola

monticola >
leucothorectis

monticola

luecothorectis

adastus

adastus >
extimus

adastus >
extimus

adastus

adastus >
extimus

utahensis

leucolaema

leucolaema

leucolaema

leucolaema
occidentalis

hypopolia >
pyrrhonota
macrolopha >
annectens
woodhouseii >
nevadae

hypopolia >

pyrrhonota

macrolopha

woodhouseii >
nevadae

hypopolia >

pyrrhonota

macrolopha

woodhouseii >
nevadae

pyrrhonota >

hypopolia

macrolopha

woodhouseii

hypopolia >

pyrrhonota

macrolopha

woodhouseii

nevadensis

nevadensis

nevadensis

garrinus

garrinus

wasatchensis

wasatchensis

wasatchensis

wasatchensis

gambeli

montana >
leucosticta

montana

montana

montana

montana

conspersus >

griscus

conspersus

conspersus
bairdii

conspersus
occidentalis

conspersus
bairdii

gambeli

gambeli

gambeli

gambeli >
excubitorides

gambeli >
excubitorides

occidentalis

occidentalis

occidentalis

occidentalis

occidentalis

fortis

fortis

fortis

fortis

fortis

artemisiae

artemisiae

artemisiae

artemisiae

artemisiae

solitudinus >
frontalis

solitudinus >
frontalis

solitudinus >
frontalis

solitudinus >
frontalis

frontalis

montanus

montanus

montanus

montanus

montanus

GREAT BASIN NATURALIST MEMOIRS

No. 2

to the race montana. The Steller's Jays of
the west desert ranges including the
Oquirrh Mountains are typical of the race
Cyanocitta stelleri macrolopha, while those
from the Wasatch Mountains constitute an
intergrading population between mac-
roloplia and annectens, a northern race. The
Scrub Jays of the Oquirrh Mountains and
other west desert ranges are typical of the
race Aphelecora caerulescens nevadae, but
those from the Wasatch are intergrades be-
tween nevadae and woodhouseii, closest to
the latter. Thus in these several species a
break occurs along the west escarpment of
the Wasatch Mountains dividing west desert
races from intergrading populations in the
Wasatch Mountains and eastward. The Jor-
dan Valley between the Oquirrh and
Wasatch Mountains, only about 25 miles
across, thus seems to act as a weak barrier
for the montane forms.

The second pattern is for the break along
a west-east cline to occur farther east be-
tween the Wasatch and Uinta mountains.
Here there is not even a valley to serve as
the line of demarcation. This situation is
seen in the Red-tailed Hawk and Black-
capped Chickadee. For the hawk, the popu-
lation in the Wasatch and all of western
Utah represents the race Buteo jamaicensis
calurus, while those from the Uinta Basin
and Tavaputs Plateau region are closest to
fuertesi, a race which extends southeast into
Texas. The Black-capped Chickadee of the
Oquirrh and Wasatch ranges represents the
race Parus atricapillus nevadensis. By the
time the Uinta Basin is reached the popu-
lation represents garrinus.

The third pattern is for a race or popu-
lation to be represented in northern Utah
and a different one in the southern part of
the state. Exemplifying this are the Steller's
Jay, Hairy Woodpecker, and Great Horned
Owl. As previously noted, the Steller's Jays
from the Wasatch Mountains represent an
intergrading population between the races
annectens and macrolopha, closest to the
latter. In southern Idaho the jays are closest
to annectens. South of Mount Nebo at the

southern end of the Wasatch Mountains, the
jays are typical of macrolopha. In the case
of the Hairy Woodpecker the break occurs
south of the Aquarius, Paunsaugunt, and
Markagunt plateaus. This is farther south
than the transition area for the Steller's
Jays. An unexpected distributional feature of
the Hairy Woodpecker is that the southern
race Picoides villosus leucothorectis ex-
tends farther north in the isolated mountain
ranges of the Great Basin in western Utah
than it does in the plateaus and mountains
of central Utah. This suggests that the prop-
agules for the west desert ranges came from
the southeast rather than directly west from
the Wasatch, a situation similar to that of
the Three-toed Woodpecker (Johnson 1975:
548). Whatever the direction of spread, leu-
cothorectis and monticola seem to have met
in the Snake and Deep Creek ranges where
an intergrading population occurs. In con-
trast, a sharp break in the distribution of
the two subspecies occurs between the
Tushar and Mineral mountains in south-
western Utah. The population of the Tushar
Mountains is monticola while that of the
Mineral Range about 25 miles to the west
with Beaver Valley between is typical leu-
cothorectis. In the case of the Horned Owls,
the race extending across southern Utah is
Bubo virginianus pallescens but its range
swings north in eastern Utah to include the
La Sal Mountain-Moab region.

The fourth situation is found in extreme
southwestern Utah along the Virgin River
Valley, where, in addition to the numerous
indicator species of the Mojave Desert avi-
fauna previously discussed, there are differ-
ences at the subspecies level for several
kinds of birds. In three geographically vari-
able species there are races that do not oc-
cur elsewhere in the state. These are a sub-
species of Cliff Swallow (Petrochelidon
pyrrhonota tachina), a race of Brown Cow-
bird (Molothrus ater obscurus), and a race of
Song sparrow Zonotrichia melodia fallax).
These three races are of southern origin. In
four other species, the populations are inter-
gradational toward southern races. This is

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

67

the case for the Screech Owl, where the
population is Otus asio inyoensis toward
yumanensis. The Gilded Flicker (formerly
Colaptes chrysoides but now considered to
be C. aaratus chrysoides) has been observed
in Beaver Dam Wash in Utah, and one
specimen has been obtained that is inter-
mediate between that race and the Red-
shafted Flicker (C. a. cafer). In the Rough-
winged Swallow, the population is Stelgi-
dopteryx ruficoUis serripennis toward psam-
mochroa. In the Yellow- throat, the popu-
lation is Geothlypis trichas occidentalis
toward scirpicola. Yet another intergrading
population occurs in the region in the case
of the Red-winged Blackbird, but the race
with which intergradation occurs is a west-
ern race. The population is Agelaius phoe-
niceus fortis toward nevadensis.

A fifth distributional pattern shows a dif-
ferent population in the Red Rock country
of southeastern Utah as compared to the
rest of the state. This is seen in the Night-
hawks and Horned Larks. For the former,
the race in southeastern Utah is Chordeiles
minor henryi as opposed to howelli to the
north in the Uinta Basin and hesperis in
western Utah. For the Horned Larks there
is an intergrading population between the
race Eremophila alpestris leucolaema and
occidentalis in southeastern Utah that is
closest to leucolaema.

Finally, a situation occurs in the Horned
Larks that is unlike the racial distribution of
any other geographically variable species in
the state. One race, leucolaema, occurs in
subalpine meadows in the plateaus of cen-
tral Utah and in alpine tundra of the Tush-
ar Mountains, while a different race, uta-
hensis, occupies the desert floor of the
valleys below. The high elevation race is
the same as the lowland race of the Uinta
Basin in northeastern Utah. This same phe-
nomenon of two races at different altitudes
in the same general region is found in the
Sierra Nevada (see Behle 1942), where the
race sierrae occurs in montane meadows as
opposed to different races in the lowland
vallevs both east and west of the mountains.

In contrast, in the Raft River Mountains of
northwestern Utah, Horned Larks taken
from the top of the mountain at 9500 feet
represent the same race as in the lowlands,
namely utahensis; and, in the Colorado
Rockies, the race leucolaema ranges from
the valleys up to the Arctic tundra at over
11,000 feet.

Clinal Variation

Clines are essentially a phenomenon of
geographic variation but they are also part
of the picture of biogeography inasmuch as
they would not be evident if samples were
not present from many geographic areas.
Clinal variation is manifest in the characters
of many kinds of birds in the Great Basin
and Utah. Occasionally similar clines appear
in unrelated species, which suggests that
some common environmental influence is
exerting a selective influence. Clines in
some instances extend in a north-south di-
rection, while in others they extend from
east to west or northeast to southwest. Phil-
lips (1958) mentions the Song Sparrow (Zono-
trichia melodia) in the Great Basin as an
example where two clines cross per-
pendicularly. One cline toward longer wings
and darker color extends northward while
another toward short wings, large bill, and
heavy breast-spotting proceeds westward. In
connection with his work on the birds of
Nevada, Linsdale (1938: 175) itemized the
changes observed for several variable spe-
cies, then generalized that for many birds
there is a decrease in size toward the south.
The largest individuals occur in the north-
eastern corner of the state. The bill be-
comes shorter and stubbier toward the east
and smaller toward the south. The wings
and tail are generally longer toward the
east. General coloration becomes paler and
grayer toward the east and sometimes
brighter and darker in the vicinity of the
Colorado River.

In Utah, clines are most evident in size
characters. The usual pattern is for birds in
the northern part of the state to be of

68

GREAT BASIN NATURALIST MEMOIRS

No. 2

larger size than those in the southern part,
with a smooth gradient occurring the length
of the state. The gradient in Utah is usually
a portion of a more extensive cline extend-
ing throughout western North America. A
recent study that I made of the White-
throated Swift (Behle 1973) revealed clinal
variation nicely. Measurements of popu-
lations from Montana south to Arizona were
analyzed. Clinal variation was most appar-
ent in wing length, which is regarded in or-
nithological systematics as a good indicator
of overall size. Clinal variation was less evi-
dent in tail length and virtually nonexistent
in bill and tarsal lengths. For wing length,
the means for the several populations mea-
sured showed a gradual transition from
143.2 mm in the Montana sample to 136.5
in the Arizona-New Mexico sample, a dif-
ference of 6.7 mm. While there was a gen-
eral decrease in wing length from north to
south in Utah samples, a mosaic pattern of
variation was shown in the several semi-iso-
lated populations represented. For example
specimens from the Raft River Mountains in
northwestern Utah have the longest wings
in the state (average wing length 146.0
mm). They are larger than those from cen-
tral northern Utah, northeastern Utah, or
Colorado and are closest to the Montana
population in size. Swifts from the Beaver
Dam Wash in extreme southwestern Utah
have the shortest wing length (wing 134.2).
They are smaller than samples from central
southern and southeastern Utah and are
even smaller than the Arizona-New Mexico
sample. These extreme Utah populations
differ in average wing length by 11.8 mm,
which is greater than that between Mon-
tana and Arizona-New Mexico birds (6.7
mm). The circumstance that northern swifts
have longer wings than do southern swifts
may be correlated with the behavioral fea-
ture that northern individuals migrate dur-
ing the winter from their breeding areas
while those in the southern part of their
range are sedentary. Another case of north-
south clinal variation in size is seen in the
Cliff Swallows in western North America

(Behle 1976a). Clinal variation in size in
Utah has become apparent from our studies
of the Great Horned Owls and Hairy
Woodpeckers (unpublished data).

Clines are also evident in Utah birds in
color characters. A west-east gradient oc-
curs in several species in northern Utah
whereby paler-colored birds occur in the
desert Great Basin portion of the state, with
a transition eastward to darker birds in the
Wasatch and Uinta mountains. Such clines
are most evident in dorsal coloration. Spe-
cies showing this phenomenon are the Dus-
ky Grouse, Screech Owl, Common Night-
hawk, Cliff Swallow, Horned Lark, Scrub
Jay, Mountain Chickadee, and Creeper. Of
the lot, the phenomenon seems to be most
pronounced in the Dusky Grouse. Represen-
tatives are pale and gray in the ranges of
eastern Nevada and in the Deep Creek
Mountains of western Utah. In the Oquirrh
Mountains they start to be slightly darker,
showing more brown. The darkening is ac-
centuated in the Wasatch Mountains and
continues to a still greater degree in the
Uinta Mountains and eastward into Colo-
rado. In general, east-west clinal variation
in Utah is the reverse of that for Nevada,
since the birds become paler and grayer in
the western part of the state, where the
Great Basin occurs. Clinal variation in color
from darker birds in the north to lighter
birds in the south shows up in a few birds
such as the Steller's Jay. In Utah, as in Ne-
vada, brighter coloration occurs in some
species in the valley of the Virgin River.
Examples showing this are Yellow-throats
and Song Sparrows.

Secondary Contact of Species

IN THE INTERMOUNTAIN REGION

In recent years several studies have been
made of secondary contact of pairs of close-
ly related species or subspecies of birds in
North America (see Selander 1965: 536, for
a listing of kinds and sources of informa-
tion). For most cases, the area of contact is
the Great Plains, but in four instances the

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

phenomenon shows up in Utah. Three of
these involve eastern and western kinds
hybridizing. The fourth involves a northern
and a southern species meeting. For two of
the four the contact has probably been
brought about during the past few decades;
the other two are of longer duration. The
first case pertains to the Indigo Bunting
(Passerina cyanea, a species that originally
occurred only in eastern North America),
and the Lazuli Bunting (P. amoena), a west-
ern species. Apparently the planting of trees
and shrubs in cities and parks across the
plains states bridged the former grasslands
hiatus separating the two species, and a
highway was thus provided for dispersal of
the Indigo Bunting westward into the range
of the Lazuli Bunting. Historical records
suggest that the Indigo Bunting arrived in
Utah about 40 years ago. That hybridization
of the two species has occurred is indicated
by two intermediate specimens. One, in the
Cornell University collection, was taken
near Ogden, Weber County, Utah, on 12
August 1945 (Sibley and Short 1959: 447).
The other is in the University of Utah col-
lection and was taken along Minnie Maud
Creek, 2 miles east of Nutter's Banch
Duchesne County, Utah, on 30 June 1966.
The Indigo Bunting is now fairly common
in southern Utah, where it exists sympa-
trically with the Lazuli Bunting. Whitmore
(1975) has recently discussed the inter-
specific behavioral competition now in evi-
dence in this region between the two spe-
cies.

The second instance of recently estab-
lished contact in Utah pertains to two kinds
of oriole, the Baltimore Oriole, formerly
called Icterus galbula, which is an eastern
type, and the Bullock's Oriole, formerly des-
ignated as I. bullockii, a common western
kind. Worthen (1973) reported an example
of the Baltimore Oriole taken 2 miles south
of Milford, Beaver County, Utah, on 27
June 1964. It was one of a series of several
orioles obtained at this location. Although
in worn plumage, the specimen represents a
"pure" first-year male. While this particular

specimen shows no tendency toward the
Bullock's Oriole, some others in the series
do show evidences of hybridization. Sibley
and Short (1964) have shown that hybridiza-
tion in the two orioles is now common
throughout the Great Plains. As a result, the
two orioles are presently considered as races
of one species, e.g., I. g. galbula and 7. g.
bullockii.

The third case of hybridization in Utah
pertains to flickers. There are three types of
flickers in North America: the Yellow-
shafted Flicker, essentially an eastern bird
formerly designated as Colaptes auratus; the
Bed-shafted Flicker of the west, formerly
called C. cafer; and the Gilded Flicker of
the southwest and lower California, for-
merly called C. chrysoides. The Yellow-
shafted and Bed-shafted forms for over 100
years have been known to hybridize in a
broad montane belt in western North Amer-
ica extending from British Columbia south-
ward throughout the Bocky Mountain re-
gion. Short (1965) interprets the picture of
speciation as follows. He postulates a geog-
raphic separation of the ancestral auratus-
cafer population during the Illinois glacial
age or earlier. The separation continued
during subsequent periods of glaciation (ex-
cept for possible hybridization between the
two differentiated stocks during interglacial
periods). With the waning of the last major
advance of the Wisconsin period of glacia-
tion, the eastern Yellow-shafted Flicker,
auratus, was able to expand its range west-
ward and northwestward into British Co-
lumbia. In contrast, the Bed-shafted Flicker,
cafer, remained restricted to the area south
of the glaciers in the western United States.
Eventually the two populations made con-
tact and hybridized along the length of the
Bockies. Utah is west and south of the main
zone of introgression and Short scarcely
mentions the area in his discussion, but
there is evidence of much crossing taking
place in Utah. A recent study by Bich
(1967) of 137 specimens in the University of
Utah collection revealed that 85 are "pure"
cafer, 4 are "pure" auratus, and 48 are in-

70

GREAT BASIN NATURALIST MEMOIRS

No. 2

termediates. This is a surprisingly large
number of intergrades with so few auratus
seemingly present. It suggests that there is-
little or no selective pressure against the
characters produced by auratus genes. The
greatest flow of auratus genes into the cafer
population pool in Utah appears to be oc-
curring in northwestern Utah, as indicated
by the highest incidence of intermediates.
In contrast, there are fewer intermediates
from eastern Utah, suggesting that the east-
west gradient from the Great Plains area is
not of great significance in Utah. In other
words, the Yellow-shafted Flickers in Utah
have seemingly come mostly from the
northwest rather than the east. Inter-
mediates occur throughout the Great Basin
mountain ranges.

Short (1965) conceives of all the North
American flickers belonging to a single spe-
cies, Colaptes auratus, and, following his
lead, the Yellow-shafted Flicker is now
known as C. a. auratus, the Red-shafted
Flicker is C. a. cafer, and the Gilded Flick-
er is C. a. chrysoides. The latter apparently
hybridizes with the Red-shafted Flicker in
extreme northwestern Arizona and south-
western Utah; Wauer and Russell (1967) re-
port a specimen taken at the Terry Ranch
in Beaver Dam Wash, Washington County,
on 28 April 1965 as being a hybrid of chry-
soides X cafer derivation.

The last case of secondary contact in
Utah involves two species of junco that
come together in extreme northern Utah,
with a relatively restricted zone of hybridi-
zation extending in an east-west direction
across the state. One population is a north-
ern form, a race of the Dark-eyed (Oregon)
Junco (Junco hyemalis mearnsi). The other
is a southern form, the Gray-headed Junco
(/. caniceps caniceps). The contact of these
two kinds was originally detected by Miller
(1941: 200). At a locality 10 miles east of
Kamas, all examples that he collected were
pure caniceps. There was a shift in frequen-
cy of characters of /. h. mearnsi northward
indicated by specimens from 20 miles north
of Kamas, in the Uinta Mountains, then in

succession Woodruff, Randolph, and Garden
City, until nearly pure populations of
mearnsi were found near the Utah- Idaho
border. Since then breeding hybrid juncos
have been taken in the Wasatch Mountains
east of Salt Lake City and in the Uinta
Mountains. Hybrids extend westward in
northern Utah to the Raft River Mountains
and beyond into northeastern Nevada.

Centers of Differentiation in the
Great Basin

On a previous occasion (Behle 1963), I
studied the distribution of the races of 50
geographically variable species of birds
whose ranges include or impinge upon the
Great Basin and its flanking regions. The re-
sults showed that the Great Basin is not in
itself one large center of differentiation. In-
stead, several distribution areas were re-
vealed that either occur in portions of the
Great Basin or are situated in nearby sur-
rounding regions. The areas were designated
as the Warner Region, Sierra Nevada,
Western Great Basin, Eastern Great Basin,
Rocky Mountains, Northern Idaho, Inyo Re-
gion, Mojave Desert, Colorado Desert, and
Navajo Country. The races of the geograph-
ically variable species occurring in each of
the 10 distribution zones were listed. Since
many species occupied each area, it was the
different combinations of races along with
common transition zones or areas of inter-
gradation between races that served to
characterize and delimit the several areas.
No one species showed conformance in the
distribution of its races with the various dis-
tribution areas outlined, although the
horned lark approached this in slight de-
gree. In only a few instances were races en-
demic in any one area. The Great Basin is
too diversified in terms of environmental
factors to have influenced in some common
way all the geographically variable birds
that are found in the region. The differen-
tiation and distribution of the races is pre-
sumably correlated largely with localized
different environments, but, in addi-

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

71

tion, barrier effects and individual histories
of the various kinds of birds in terms of
their point of origin, dispersal, and depen-
dency on particular plant associations have
played a role.

There are indications of three centers of
differentiation in the Great Basin region,
one in the eastern part, and two in the
western portion. The latter two have been
evaluated by Miller (1941). One is the
White Mountain area of eastern California
in the southwestern portion of the Great
Basin. The other is the Warner Mountain-
Warner Valley region of southern Oregon
and northeastern California in the north-
western portion of the Great Basin. More
recently Johnson (1970) presented additional
data for the avifauna of the Warner Moun-
tains and has reevaluated the affinities of
the boreal elements. He gives different re-
sults than I presented (Behle 1963). The
center of differentiation in the eastern part
of the Great Basin rests on four races that
have fairly common, though not identical,
ranges. Three of these were described in the
course of our fieldwork at the University of
Utah, namely a race of Dusky Grouse (Den-
dragapus obscurus oreinus), a race of Horn-
ed Lark (Eremophila alpestris utahensis),
and a race of Fox Sparrow (Passerella iliaca
swarthi). The fourth is a race of Black-
capped Chickadee (Parus atricapillus neva-
densis) described by Linsdale of the Univer-
sity of California.

Theoretical Historical Aspects of

Distribution

of Birds in the Great Basin

To my knowledge, no direct evidence has
been detected of the influence of Pleisto-
cene or Holocene climates on the distribu-
tion of birds in the Great Basin or Inter-
mountain Region. Still, some inferences can
be drawn. Resident birds present in the re-
gion today are closely associated in their
occurrence with particular biotic commu-
nities, especially the plant components.
Since climatic change has resulted in altera-

tion of community types, the avian associ-
ates have almost certainly been affected
too, either being forced out of areas where
the plant habitat has disappeared or in-
vading new areas as their requisite habitat
has become established. Dispersal resulting
from population pressure has also resulted
in extensions of ranges. In some instances
former allopatric species have become sym-
patric, as in the cases of the flickers and
buntings. With contractions of ranges, for-
mer sympatric species conceivably have be-
come allopatric. Such movements would be
in the nature of long-term adjustments.

Of particular interest in this connection is
the present-day discontinuous distribution of
the coniferous forest on the mountaintops of
the Great Basin and the attendant boreal
species of birds. Johnson (1975: 556) has
noted the two theories that have been of-
fered to account for this. One proposes that
during the Pleistocene cold climates pre-
vailed with relatively more moisture and
less evaporation than in the region today.
These conditions induced the formation of
glaciers in the mountains and the pluvial
lakes Bonneville, Lahonton, and a host of
lesser lakes on the floor of the Great Basin.
At the same time, the coniferous forest pre-
sumably extended altitudinally down into
the valleys, bordering on the lakes, and be-
came distributed more or less continuously
throughout the Great Basin. Boreal species
of birds presumably accompanied the con-
iferous forest and also occurred more or less
continuously at lowland elevations. Sub-
sequent climatic change to the warmer and
drier conditions of today resulted in melting
of the glaciers, disappearance or diminution
of the lakes, and retreat of the coniferous
forest up into the mountains. The boreal
birds presumably were also forced to move
up into the mountains, where they occur as
breeders today. Martin and Mehringer
(1965) have mobilized the evidence from
pollen studies in support of such climatic
changes. In accord with this line of reason-
ing such avian species as the northern
Three-toed Woodpecker, Water Pipit, and

72

GREAT BASIN NATURALIST MEMOIRS

No. 2

Black Rosy Finch were presumably formerly
much more widespread and abundant but
have subsequently been confined to the
mountaintops where Hudsonian zone or al-
pine-arctic conditions prevail. Concurrently
the lowland valleys were invaded by low-
land species from surrounding regions, spe-
cies adapted to the warmer, xeric conditions
that have come to prevail there.

The second point of view is that the
montane pockets of boreal forest and at-
tendant faunas have come about through
dispersal over long distances from parental
stocks such as those in the Sierra Nevada
and Rocky mountains. Some evidence per-
taining to the vegetation to support this in-
terpretation has been presented by Wells
and Berger (1967) and Critchfield and Al-
lenbaugh (1967), and Johnson (1975) has
noted the probable role of certain species of
boreal birds such as the Band-tailed Pigeon,
Pinyon Jay, and Clark's Nutcracker in long-
distance colonization through transporting
and/or burying seeds of conifers. In accord-
ance with this theory, the Three-toed
Woodpecker, Water Pipit, and Black Rosy
Finch have extended their ranges westward
from the Rocky Mountain continental area
only to certain boreal islands in the eastern
part of the Great Basin. Present indications
are that the Rosy Finch has progressed the
farthest, being known from the Jarbidge
Mountains, the Ruby Mountains, and the
Wheeler Peak area of the Snake Range in
eastern Nevada. The Water Pipit stops at
the Deep Creek Range in extreme western
Utah. The Northern Three-toed Wood-
pecker is known from the Snake Range.
However, I suspect that if more collecting
were done, all three species would be found
at Wheeler Peak and the Ruby Mountains
in Nevada. I favor the relict mountaintop
theory as Brown (1971) does for mammals.
As Johnson notes, these two hypotheses are
not mutually exclusive. Both processes could
have occurred in the past. Extensions of
range are occurring at present, as indicated
by the historic record for certain species.
The diversity of distribution patterns that

have been noted for Utah leads to the infer-
ence that each species has had its own par-
ticular distributional history determined by
its habitat requirements and the habitat
changes experienced.

M

Imi

ANAGEMENT IMPLICATIONS

Birds come into the picture of natural re-
source management in the Great Basin in
several ways. In connection with environ-
mental impact studies, special consideration
is being given to threatened and endangered
species such as the Bald Eagle, Peregrine
Falcon, and Osprey. All of these occur in
the Great Basin. Indeed, studies of raptors
by Brigham Young University biologists
have revealed high populations for many
species in remote areas of the Great Basin.
Even subspecies are important in terms of
endangered species, because it is the south-
ern race of Bald Eagle and the southern
race of Peregrine Falcon that are endan-
gered. Another area where subspecies are
important in management pertains to the
introduction of exotic game species. If more
should be introduced (and there are serious
objections to the practice) stock should be
selected representing races whose native
habitat is most nearly like that where the
introduction is to be made. Another point is
that populations at type localities, such as
the Dusky Grouse (which is a game species)
from the Deep Creek Mountains, should be
protected. Regional avifaunistic reports,
such technical papers as descriptions of new
forms and systematic revisions, and sym-
posia such as the present one, especially the
published proceedings, constitute valuable
resource material for those charged with
making evaluations. They provide baseline
data to compare against in future years to
establish long-term changes.

In the 130 years since settlement of the
Great Basin and other parts of the Inter-
mountain West, many well known changes
have occurred in the vegetation as pointed
out by Cottam (1947) and others. Con-
comitant changes have occurred in the bird

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

73

life. For example, as the grassland was es-
sentially extirpated from Utah through over-
utilization, the habitat for the Grasshopper
Sparrow and Sharp-tailed Grouse was re-
moved, the result being the near extermina-
tion of these kinds of birds in the region
today. As pertains to the grouse, certain
protected areas containing the little remain-
ing requisite habitat are needed for its sur-
vival. One such tract is east of Wellsville,
Cache County, Utah. As more and more
sagebrush is removed for land cultivation,
Sage Grouse and other sage-inhabiting spe-
cies are being affected. The Conservation
Committee of the Wilson Ornithological So-
ciety (see Braun et al. 1976) has recently re-
ported on the extent of alteration of this
community and attendant deleterious effects
on the associated bird life. Chaining out of
junipers and pinyon pine in the Great Basin
is of less consequence in terms of the bird
life because the extent of the habitat is so
vast. Nevertheless, I feel that some typical
areas should remain undisturbed to serve as
study areas. They should be large enough to
preserve habitat diversity and maintain spa-
tial relations intact. Yet extensive areas of
continuous forest may not be as effective in
preserving communities and species as
would numerous, smaller, diversified, irregu-
lar areas. A large, essentially undisturbed
and diversified area is the Wheeler Peak re-
gion in the Snake Range in eastern Nevada.
At one time the area was proposed for a
Great Basin natural park. I would like to
see this area so designated. This would af-
ford some measure of protection. The Leh-
man Cave National Monument has already
been established there. The Bureau of Land
Management is, I believe, considering the
designation of the Deep Creek Mountain in
western Utah as a quasi-primitive area. A
complication is that part of that range is In-
dian reservation. The Beaver Dam Wash
area of extreme southwestern Utah is
unique in its flora and fauna and needs pro-
tection—especially from collectors.

Regarding such theoretical considerations
as the application of island biogeography

theory to conservation practice as has been
advocated in the design of wildlife refuges,
Simberloff and Abele (1976) express the
opinion that the application of the general
principle is premature at the present time.
They feel that, in this particular instance,
broad generalizations have been based on
limited and insufficiently validated theory
and on field studies of taxa which may be
idiosyncratic. The implication is that much
more research is needed. I suggest that the
boreal islands of the Great Basin constitute
propitious areas for further avian research
as a sequel to Johnson's and my work.

Summary and Conclusions

Although there are no endemic species of
birds in the Great Basin region of western
North America, nevertheless a distinctive
avifauna exists there by virtue of a different
combination of birds as well as the presence
of many species associated with sagebrush
and the pinyon-juniper forest. Physiographic
boundaries determine the limits of the
Great Basin avifauna on the west and east,
while on the south there is a sharp junction
zone with the Mojave Desert avifauna that
occurs in southern Nevada and in extreme
southwestern Utah. To the north there is a
gradual blending with the avifauna of the i
Palouse prairie and the northern montane
woodland. About 30 kinds of distinctive
birds that occur in the California-Pacific
Coast-Sierra Nevada region are not known
to occur in the Great Basin, suggesting that
relatively little eastward spread has oc-
curred. In contrast, seven Rocky Mountain
species reach their western limits within the
Great Basin. Some of the latter group,
namely the Yellow-shafted Flicker, Balti-
more Oriole, and Indigo Bunting, are recent
arrivals and introgression has occurred with
western congeners. Instability of present-day
ranges for many species of birds is further
indicated by the finding in recent years of
several other kinds, mostly in southwestern
Utah, that are new to the state list.

Ten northern species reach their southern

74

GREAT BASIN NATURALIST MEMOIRS

No. 2

limits in at least part of their ranges in the
Great Basin. A zone of hybridization be-
tween a race of the northern Dark-eyed
Junco (/. h. mearnsi) and the southern Gray-
headed Junco (/. c. caniceps) occurs across
northern Utah and northeastern Nevada.
Sixteen southern species reach their north-
ern limits in the Great Basin, while an addi-
tional 25 species stop at a distinct Great
Basin— Mojave Desert junction zone in
southern Nevada and extreme southwestern
Utah. Three avifaunas are represented in
the Great Basin region today, namely the
Bocky Mountain, Great Basin, and Mojave
Desert.

Montane species, which are mostly associ-
ated with the coniferous forest, are dis-
continuously distributed in boreal islands on
the tops of isolated mountain ranges in the
general region. An analysis of the avifaunas
of 14 such islands in western and south-
eastern Utah, as compared with that of the
Bocky Mountain continent in central and
northern Utah, shows a close correlation be-
tween the number of species present and
habitat diversity. A slight, negative correla-
tion shows up for permanent residents with
distance from the continent. The results are
similar to those of Johnson (1975) for a dif-
ferent set of islands mostly located in Ne-
vada.

An analysis of the distribution of races of
22 species in Utah represented by more
than one race in the state reveals a variety
of patterns. For several a break occurs
along the Wasatch Front on the east side of
the Great Basin between a west desert race
and either an eastern montane race or an
intergrading population toward a different
race in eastern Utah. In a few others, the
break is farther east between the Wasatch
and Uinta mountains. Another situation is
for there to be one race in northern Utah
and a different race in the southern part of
the state. In three species, there are differ-
ent races or populations in southeastern
Utah; but southwestern Utah is the most
distinctive transitional area where, in three
species, different races are represented and

in five others intergradational populations
occur. For the Horned Lark, one race oc-
curs in subalpine meadows in central Utah,
and a different race is a summer resident in
the desert region at the base of the moun-
tains. In some species intergrading popu-
lations occur over broad areas; in others the
phenomenon is confined to a narrow zone.

A center of differentiation occurs in west-
ern Utah in the eastern portion of the Great
Basin where four races of geographically
variable birds have ranges that somewhat
coincide. This is similar to the White
Mountain and Warner Mountain centers in
the western portion of the Great Basin.
Clinal variation occurs in many species, in-
volving both size and color characters.
Some clines run north and south and others
run east and west. Some are gradual; others
are step clines. Past climatic change has
doubtless influenced the distribution of spe-
cies and avifaunas in the region. It is infer-
red that during cold intervals of the Qua-
ternary boreal birds occurred in lowland
valleys, but with a warming trend they
have retreated to the mountaintops, where
they are found today. This would account
for the current distribution of the Water Pi-
pit and Black Bosy Finch, although the pos-
sibility exists of a westward spread of these
species from the Bocky Mountains.

Acknowledgments

I am indebted to many people for help in
various ways in the preparation of this pa-
per. Tom Boner and Marie Magleby com-
piled the lists of species for the various bo-
real islands. John Wyckoff mobilized the
data for physical features in Table 1, han-
dled the statistical treatment of the data,
and constructed the graph for Figure 2.
Dave Prouse and Marie Magleby worked on
the map. Magleby and William Pingree
measured hundreds of birds, and Pingree
helped in making subspecies determinations.
Norma Fernley did the typing. Many indi-
viduals have helped in the field work

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

75

throughout the years,
team effort.

It has indeed been a

Literature Cited

Aikens, C. M. 1970. Hogup Cave. Univ. Utah Anthro-
pol. Pap. No. 93: 1-286.

Bailey, A. M., and R. J. Niedrach. 1965. Birds of
Colorado, Vol. 1, 2. Denver Museum of Natural
History, Denver.

Behle, W. H. 1942. Distribution and variation of
the horned larks (Otocoris alpestris) of western
North America. Univ. Calif. Publ. Zool. 46:
205-316.

1943. Birds of Pine Valley Mountain region,

southwestern Utah. Univ. Utah Biol. Ser. 7 (5):
1-85.

1955. The birds of the Deep Creek Moun-
tains of central western Utah. Univ. Utah Biol.
Ser. 11 (4): 1-34.

1958. The birds of the Raft River Mountains,

northwestern Utah. Univ. Utah Biol. Ser. 11 (6):
1-40.

1960. The birds of southeastern Utah. Univ.

Utah Biol. Ser. 12 (1): 1-56.

1963. Avifaunistic analysis of the Great Basin

region of North America. Proc. 13th Intern. Or-
nithol. Congr.: 1168-1181.

1973. Clinal variation in White-throated

Swifts from Utah and the Rocky Mountain re-
gion. Auk 90: 299-306.

1976a. Systematic review, intergradation,

and clinal variation in Cliff Swallows. Auk 93:
66-77.

1976b. Mohave Desert avifauna in the Virgin

River Valley of Utah, Nevada, and Arizona.
Condor 78: 40-48.

Behle, W. H., J. B. Bushman, and C. M.
Greenhalgh. 1958. Birds of the Kanab area
and adjacent high plateaus of southern Utah.
Univ. Utah Biol. Ser. 11 (7): 1-92.

Behle, W. H., and J. Ghiselin. 1958. Additional
data on the birds of the Uinta Mountains and
Basin of northeastern Utah. Great Basin Nat.
18:1-22.

Behle, W. H., and M. L. Perry. 1975. Utah birds:
Checklist, seasonal and ecological occurrence
charts, and guides to bird finding. Utah Mu-
seum of Natural History, Salt Lake City.

Benson, S. B. 1935. A biological reconnaissance of
Navajo Mountain, Utah. Univ. Calif. Publ. Zool.
40: 439-456.

Blanchard, J. F. Jr. 1973. An ecologic and system-
atic study of the birds of the Oquirrh Moun-
tains, Utah. Unpublished master's thesis, Uni-
versity of Utah, Salt Lake City.
Braun, C, M. F. Baker, R. L. Eng, J. S. Gashwiler,

and M. H. Schroeder. 1976. Conservation
Committee report on effects of alteration of
sagebrush communities on the associated avi-
fauna. Wilson Bull. 88: 165-171.

Brown, J. H. 1971. Mammals on mountaintops:
nonequilibrium insular biogeography. Amer.
Naturalist 105: 467-478.

Burleigh, T. D. 1972. Birds of Idaho. Caxton Print-
ers, Ltd., Caldwell, Idaho.

Critchfield, W. B., and G. L. Allenbaugh.
1969. The distribution of Pinaceae in and near
northern Nevada. Madrono 19: 12-26.

Cottam, W. P. 1947. Is Utah Sahara bound? Bull.
Univ. Utah, Biol. Ser. 37 (11): 1-40.

Flack, J. A. D. 1976. Bird populations of aspen for-
ests in western North America. Amer. Ornithol.
Union, Ornithological Monogr. 19: 1-97.

Grinnell, J., and A. H. Miller. 1944. The distri-
bution of the birds of California. Pacific Coast
Avif. 27: 1-608.

GULLION, G. W., W. M. PULICH, AND F. G.

Evenden. 1959. Notes on the occurrence of

birds in southern Nevada. Condor 61: 278-297.
Hall, E. R. 1940. An ancient nesting site of the

White Pelican in Nevada. Condor 42: 87-88.
Hayward, C. L., C. Cottam, A. M. Woodbury, and

H. H. Frost. 1976. Birds of Utah. Great Ba-
sin Nat. Mem. 1: 1-229.
Jennings, J. D. 1957. Danger Cave. Univ. Utah An-

thropol. Pap. 27: 1-328.
Johnson, N. K. 1965. The breeding avifaunas of the

Sheep and Spring ranges in southern Nevada.

Condor 67: 93-124.
1970. The affinities of the boreal avifauna of

the Warner Mountains, California. Occas. Pap.,

Biol. Soc. Nevada 22: 1-11.
1973. The distribution of boreal avifaunas in

southeastern Nevada. Occas. Pap., Biol. Soc.

Nevada 36: 1-14.
1974. Montane avifaunas of southern Ne-
vada: historical change in species composition.

Condor 76: .334-337.
1975. Controls of number of bird species on

montane islands in the Great Basin. Evolution

29: 545-563.
Linsdale, J. M. 1936. The birds of Nevada. Pacific

Coast Avif. 23: 1-145.
1938. Bird life in Nevada with reference to

modifications in structure and behavior. Condor

40: 173-180.
1951. A list of the birds of Nevada. Condor

53: 228-249.
MacArthur, R. H., and E. O. Wilson. 1963. An

equilibrium theory of insular zoogeography.

Evolution 17: 373-387.
1967. The theory of island biogeography.

Princeton University Press, Princeton, N.J.

76

GREAT BASIN NATURALIST MEMOIRS

No. 2

Martin, P. S., and P. J. Mehringer, Jr.
1965. Pleistocene pollen analysis and biogeo-
graphy of the southwest, pp. 433-451. In: H. E.
Wright, Jr. and D. G. Frey (eds.), The Qua-
ternary of the United States. Princeton Univ.
Press, N.J.

Miller, A. H. 1941. Speciation in the avian genus
Junco. Univ. Calif. Publ. Zool. 44: 173-434.

1951. An analysis of the distribution of the

birds of California. Univ. Calif. Publ. Zool. 50:
531-644.

Phillips, A. R. 1958. The nature of avian species. J.
Arizona Acad. Sci. 1: 22-30.

Remington, C. L. 1968. Suture-zones of hybrid in-
teraction between recently joined biotas. Evolu-
tionary Biol. 2: 321-428.

Rich, W. J. 1967. The woodpeckers of Utah. Un-
published master's thesis, University of Utah,
Salt Lake City.

Selander, R. K. 1965. Avian speciation in the Qua-
ternary, pp. 527-542. In: H. E. Wright Jr. and
D. G. Frey (eds.), The Quaternary of the United
States. Princeton Univ. Press, N.J.

Short, L. L., Jr. 1965. Hybridization in the flickers
(Colaptes) of North America. Bull. Amer. Mus.
Nat. Hist. 129: 307-428.

Sibley, C. C, and L. L. Short, Jr. 1959.
Hybridization in the buntings (Passerina) of the
Great Plains. Auk 76: 443-463.

1964. Hybridization in the orioles of the

Great Plains. Condor 66: 130-150.
Simberloff, D. S., and L. G. Abele. 1976. Island

biogeography theory and conservation practise.

Science 191: 285-286.

Trimble, D. E., and W. J. Carr. 1961. Late Qua-
ternary history of the Snake River in the Ameri-
can Falls region, Idaho. Bull. Geol. Soc. Amer.
72: 1739-1748.

Udvardy, M. D. F. 1963. Bird faunas of North
America. Proc. 13th Intern. Ornithol. Congf. 13:
1147-1167.

Uzzell, T., and N. P. Ashmole. 1970. Suture-
zones: an alternative view. Systematic Zool. 19:
197-199.

Vuilleumier, F. 1970. Insular biogeography in con-
tinental regions. I. The northern Andes of South
America. Amer. Naturalist 104: 373-388.

Wauer, R. H., and R. C. Russell. 1967. New and
additional records of birds in the Virgin River
Valley. Condor 69: 420-423.

Wells, P. V., and R. Berger. 1967. Late Pleisto-
cene history of coniferous woodland in the Mo-
jave Desert. Science 155: 1640-1647.

Whitmore, R. C. 1975. Indigo Buntings in Utah
with special reference to interspecific com-
petition with Lazuli Buntings. Condor 77:
509-510.

Willson, M. F. 1974. Avian community organiza-
tion and habitat structure. Ecology 55:
1017-1029.

Woodbury, A. M., and H. N. Russell, Jr. 1945.
Birds of the Navajo country. Univ. Utah Biol.
Ser. 9 (1): 1-160.

Worthen, G. L. 1968. The taxonomy and distribu-
tion of the birds of the southeastern Great Ba-
sin, Utah. Unpublished master's thesis, Univer-
sity of Utah, Salt Lake City.

1973. First Utah record of the Baltimore

Oriole. Auk 90: 677-678.

1978 INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM 77

Occurrence of boreal birds on montane islands and portion of Rocky Mountain Continent in Utah.

1 *
i 1

I I I I I I J J 1 I i | 1 i S

i I " S ^ - >° U I a "S ^ i

a.

cc Q

Si 3 c b « ^ "C j: .S «-£ « m^

g

"Goshawk

(Accipiter gentilis) XX X XX XXXX

"Sharp-shinned Hawk

(Accipiter striatus) XXXXX X XXXXXX

"Cooper's Hawk

(Accipiter cooperii) XXXX X X XX XX

"Red-tailed Hawk

(Buteo jamaicensis) XXXXX XXXXXX XX

"Golden Eagle

(Aquila chrysaetos) XXXXX X XX XX

"American Kestrel

(Falco sparverius) XXXX XXXXX

"Dusky Grouse

(Dendragapus obscurus) XXXX XXXX X

"Ruffed Grouse

(Bonasa utnbellus) X X

"Band-tailed Pigeon

(Columba fasciata) , XXXXXX

"Flammulated Owl

(Otus flammeolus) X XXX

"Great Horned Owl

(Bubo virginianus) XXX XXX XX

"Pygmy Owl

(Glaucidium gnoma) X

"Spotted Owl

(Strix occidentalis) X X

"Long-eared Owl

(Asia otus) XX X

"Saw-whet Owl

(Aegolius acadicus) XXX X

White-throated Swift

(Aeronautes saxatalis) XX X XXXXXXXXX

Species ° = Permanent residents. Others are summer residents.

78 GREAT BASIN NATURALIST MEMOIRS No. 2

Black-chinned Hummingbird

(Archilochus alexandri)

Broad-tailed Hummingbird

(Selasphorus platycercus)

Calliope Hummingbird

(Stellula calliope)
"Common Flicker

(Colaptes auratus)
"Pileated Woodpecker

(Dryocopus pileatus)
"Yellow-bellied Sapsucker

(Sphyrapicus varius)
"Williamson's Sapsucker

(Sphyrapicus thyroideus)
"Hairy Woodpecker

(Picoides villosus)
"Downy Woodpecker

(Picoides pubescens)
"Northern Three-toed Woodpecker

(Picoides tridactylus)

Hammond's Flycatcher

(Empidonax hammondii)

Dusky Flycatcher

(Empidonax oberholseri)

Western Flycatcher

(Empidonax difficilis)

Western Wood Peewee

(Contopus sordidulus)

Olive-sided Flycatcher

(Contopus borealis)

Horned Lark

(Eremophila alpestris)

Violet-green Swallow

(Tachycineta thalassina)

Tree Swallow

(Tachycineta bicolor)

Purple Martin

(Progne subis)
"Gray Jay

(Perisoreus canadensis)
"Steller's Jay

(Cyanocitta stelleri)
"Clark's Nutcracker

(Nucifraga columbiana)
"Black-capped Chickadee

(Parus atricapillus)
"Mountain Chickadee

(Parus gambeli)
"White-breasted Nuthatch

(Sitta carolinensis)
"Red-breasted Nuthatch

(Sirta canadensis)
"Pygmy Nuthatch

(Strta pygmaea)
"Brown Creeper

(Certhia familiaris)

House Wren

(Troglodytes aedon)

XX X XX

XXXXX X XXX XXX

X

xxxxxxxxxxxxxxx

X

X XX xxxxxxx

X X

xxxx xxxxxxxxxx

X X X X X X

XXXX X

XXX X XXXX X

XXXXXXX X XXX X

XXXX X X X XX

XXXX xxxxxxx

XX XX XXXX

X XX

xxxxxxxxxxxxxxx

X X XX XX

X
XX X

XXXX X xxxxxxx
xxxxxxxx xxxxxx

XXXX XX X

xxxxxxxxxxxxxxx

XX xxxxxx

xxxxxxxxxxxxxxx

X xxxxxx

XXXX X XXXX XX

xxxxxxx xxxxxxx

1978 INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM 79

Rock Wren
(Salpinctes obsoletus)
American Robin
(Turdus migratorius)
Hermit Thrush
(Catharus guttatus)
Swainson's Thrush
(Catharus ustulatus)
Veery

(Catharus fuscescens)
Western Bluebird
(Sialia mexicana)
Mountain Bluebird
(Sialia currucoides)
Townsend's Solitaire
(Myadestes townsendi)
Golden-crowned Kinglet
(Regulus satrapa)
Ruby-crowned Kinglet
(Regulus calendula)
Water Pipit
(Anthus spinoletta)
Solitary Vireo
(Vireo solitarius)
Warling Vireo
(Vireo gilvus)
Orange-crowned Warbler
(Vermivora celata)
Virginia's Warbler
(Vermivora virginiae)
Yellow-rumped Warbler
(Dendroica coronata)
Grace's Warbler

(Dendroica graciae) X X

MacGillivray's Warbler
(Geothlypis tolmiei)
Wilson's Warbler
(Wilsonia pusilla)
Western Tanager
(Piranga ludoviciana)
Black-headed Grosbeak
(Pheucticus melanocephalus)
Cassin's Finch
(Carpodacus cassinii)
"Pine Grosbeak
(Pinicola enucleator)
Black Rosy Finch
(Leucosticte atrata)
"Pine Siskin
(Carduelis pinus)
"Red Crossbill
(Loxia curvirostra)
Green-tailed Towhee
(Pipilo chlorura)
Rufous-sided Towhee
(Pipilo erythrophthalmus)
Vesper Sparrow
(Pooecetes gramineus)

xxxxxxx xxxxxx
xxxxxxxxxxxxxxx
xxxxxxxxxxxxxxx

XX XX

X X

X XX
XXXXXXXXXXXXX X
XX X XX XXXX X
XXX X XX

XXXXXXXXXXXX XX
X XX

XXXX X XXX XX
XXXXXXX XXX XXX
XXXX X X X

XXXX XX XXX XXX

xxxxxxxxxxxxxxx

XXXX XX xxxxxx

X
XXXX XX xxxxxxx
XXXX X xxxxxxx

XXXXXXXX XXX XX

X X

X XX

XXXXXXXXXXX XXX

XXXX xxxxx

XXXXXXX XXX X X

XXX X XXX XXX

XX XXXX

80 GREAT BASIN NATURALIST MEMOIRS No. 2

Dark-eyed Junco

(Junco hyemalis) X

Gray-headed Junco

(Junco caniceps) XXXX XXXXXXXXXX

Chipping Sparrow

(Spizella passerina) XXXXXXXXXXXXXXX

White-crowned Sparrow

(Zonotrichia leucophrys) XXX X XX X

Fox Sparrow

(Zonotrichia iliaca) XX XX

Lincoln's Sparrow

(Zonotrichia lincolnii) X XX

Song Sparrow

(Zonotrichia melodia) XX XX X

Totals 61 52 44 50 25 29 42 19 .34 46 64 42 41 49 80

THE FLORA OF GREAT BASIN MOUNTAIN RANGES:
DIVERSITY, SOURCES, AND DISPERSAL ECOLOGY

K. T. Harper', D. Carl Freeman1, W. Kent Ostler1, and Lionel G. Klikoff2

Abstract.— The high elevation floras of 9 mountainous "mainlands" (3 in the Sierra-Cascade system and 6 in
the High Plateau-Wasatch-Teton system) and 15 isolated mountain "islands" in the Intermountain Region have
been analyzed. Mainland floras support more species per unit area and show a smaller increase in diversity as
area is increased than islands. In this respect, the isolated mountains behave as true islands. The number of en-
demics is low on the islands (never exceeding 5 percent of any flora), however; and the island floras are over-
whelmingly dominated by species with no apparent modifications for long-range dispersal. Furthermore, the east-
ern mainland has exerted a far greater influence on the flora and the vegetation of the islands than has the
western mainland, despite the fact that the former is downwind of the islands. Thus, evidence from endemics,
dispersal ecology, and sources of the floras suggests that the isolated mountains have not acquired their full floras
by long-range dispersal. We conclude that although the floras of the islands have many insular characteristics,
they were less isolated in the relatively recent past than at the present. The island floras do not appear to be in
equilibrium in the sense that immigrations equal extinctions.

The biogeography of disjunct segments of
similar habitat has intrigued biologists since
the days of Charles Darwin (1859) and A.
R. Wallace (1880). Their pioneering obser-
vations were based primarily on oceanic is-
lands, but others have analyzed the biology
of such habitats as caves (Culver, Holsinger,
and Baroody 1973), woodlots (Curtis 1956),
fresh water lakes (Barbour and Brown
1974), and isolated patches of herb land in
high-elevation forests (Vuilleumier 1970).

The appeal of islandlike environments to
biologists is partially explained by the fact
that complete inventories of selected taxa
can be prepared for several disjunct points
in a reasonably short time. Furthermore, is-
land systems are ideally suited for the anal-
ysis of such dynamic processes as dispersal,
competition, and evolution. Basic principles
of community structure and trophic dynam-
ics also appear to have been better demon-
strated and more easily studied in island
systems than in larger, more heterogeneous
environments (Lindeman 1942, Simberloff

and Wilson 1970, Brown 1971a, Heatwole
and Levins 1972, and MacArthur, Diamond,
and Karr 1972).

In this paper we consider the vascular
plant floras of islandlike enclaves of mesic
environment on high mountains in the
deserts of the Great Basin. In the strictest
sense, these high mountains are less isolated
than oceanic islands, since dispersing prop-
agules or their carriers may rest in the
desert and survive to move on again. Also,
species of the mountain islands could evolve
(and apparently often have) from the floras
of the unfavorable environments that sepa-
rate the islands (Billings 1977). Furthermore,
evidence suggests that at varying times in
the Pleistocene many of the islands were
connected by vegetation similar to that now
confined to the slopes of the mountains
(Wells and Jorgensen 1964 and Wells and
Berger 1967). Nevertheless, the tops of the
high mountains of the arid West provide
disjunct patches of habitat that may have
much in common with real islands.

'Department of Botany and Range Science, Brigham Young University, Provo, Utah H4WI2
'Department of Biology, Allegheny College, Meadville, Pennsylvania 16335.

81

82

GREAT BASIN NATURALIST MEMOIRS

No. 2

Methods

This paper is based entirely on published
floras or checklists of workers who have
collected extensively on specific mountain
ranges. We utilize 9 floras from the more-
or-less continuous mountain systems that
flank the Great Basin on the west and east
and floras or checklists for 15 mountain
ranges in or near the Great Basin (Fig. 1,
Table 1). We have assumed that the floras
of the relatively continuous flanking moun-
tain systems (the Cascade-Sierra system in

California and the Teton-Wasatch-High
Plateau system in Wyoming, Idaho, and
Utah) have long had relatively free access
to large floras adapted for life at high ele-
vations and thus qualify as mainland floras
in the parlance of island biogeographers.

The mountain islands have been assigned
discrete boundaries which are defined by
the 7500 foot contour line. The size and
elevation of these islands and their distance
from the mainlands were taken from topo-
graphic maps. Island-to-mainland distances
were computed by summing the distances

Fig. 1. Location of the floras considered.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

83

is*

Is

II

> ^

o a>

,0. <u

cs fct,

O Si

t- t- i<

t- t- rr hini-hininsot-

C75 CX CX 05J!OJO^J3C505J)

l P 1 | l

r~ as

23

m « a;

t~- <N Q —i

1 3 1 1 1 1 1

1 1 § 1 s 1 1-

- * J 8 Is « .S

-h oi co -<t in cd r^

J J U 3 S J J

^ (N -h <N (N fH

CI

3

C

8

CO

oo

co

OC

5

s.

CI

X

oo

-

O)

3

OS

^H

a>

~H

—

05

oo m oo cd m

2 5

in h co tj< i— i cd
<-h t~- r~ oo in co

CO ■"* ^ <N CO <N

a -a a

S2S
111

ooffiOHontm

t- r- t- t-

G5 G5 CX G5

tj 0)

R

o

&_>

o

c

£ «

DC

<

J? PC

E .£

O Q

r- c o m

o> in oo co

-*r t- a> (N

o* -T as oo"

|| Is o i

II U £ 1°-* 1

c ■£ m -"= 2 t3 «»
S Z PC oo on D

S3

84

GREAT BASIN NATURALIST MEMOIRS

No. 2

across inter-island barriers of desert (areas
below 7500 feet) along the shortest route
possible from a particular island to the
nearest edge of each mainland.

It should be noted that our island areas
and distances to mainlands do not always
agree with those reported by Brown
(1971a), Johnson (1975), and Behle (1977),
who have used some of the same islands
that we have. Those discrepancies arise
from the manner in which the mainland
and island borders are defined by the sever-
al authors. Brown (1971a), for example,
combined the White and Inyo ranges, but
the flora used in our work (Lloyd and Mit-
chell 1973) covers only the White Moun-
tains. Johnson (1975) let the lower edge of
forest or woodland serve as the edge of his
islands, while we have followed Brown
(1971a) and used the 7500 foot contour as
the island edge. In Johnson's (1975) work,
the Pine Valley Mountains were considered
to be part of the mainland, but our criteria
dictate that those mountains be considered
an island.

As noted elsewhere in this symposium
(West et al. 1977), distance to the nearest
mainland is a weak ecological variable in
the Great Basin, since each mountain range
has probably received migrant species from
both mainlands. We measured the width of
valley barrier between each mountain sys-
tem and both mainlands in an effort to ob-
tain a better understanding of the bio-
geographic consequences of distance.

Johnson (1975) and Harner and Harper
(1976) demonstrated that habitat diversity
exerts a strong influence on diversity of
birds and vascular plants, respectively. John-
son (1975) used plant criteria to quantify
habitat diversity for birds on Great Basin
mountains, but his criteria for habitat diver-
sity would lead to circular logic if they
were used to help explain plant diversity.
Conceivably, one could devise a habitat di-
versity measure based on physical character-
istics of the sample areas, but a useful mea-
sure would probably require more
information about individual mountain

ranges than is now available. Accordingly,
we have used only area, elevation, and loca-
tion in our analysis of factors controlling
plant diversity.

The component species of each checklist
have been individually considered for in-
clusion in our study. We have eliminated
species from the checklists which are not
known to occur above 7500 feet. Species
that are potentially able to survive and re-
produce in desert environments have also
been excluded. This latter criterion was
used to improve the likelihood that the is-
lands considered are at least currently func-
tioning as islands. We experienced difficulty
in rigidly applying this last criterion, since
some species which occur above 7500 feet
along the eastern edge of the Great Basin
do not extend above that elevation in the
Sierras. We have included all species which
occur above 7500 feet on the eastern edge
of the Great Basin (that do not tolerate
deserts) but normally occur below that ele-
vation on the western mainland.

For each species included in the study,
we have noted lifeform, likely means of dis-
persal, and geographic range. The lifeform
categories recognized are: 1) annual, 2) per-
ennial forb, 3) perennial graminoid, 4)
shrub, or 5) tree. Categories of dispersal in-
clude: 1) mega wind, 2) miniwind, 3) stick-
tight, 4) fleshy fruit, or 5) no apparent mod-
ification. Species were placed in the
following groups with respect to geographic
range: 1) occurring on both mainlands, 2)
confined to the western mainlands and a
few isolated mountains, 3) confined to the
eastern mainlands and a few isolated moun-
tains, or 4) known only from one or a few
mountains in this study. Because the authors
of the several checklists were uneven in
their treatment of taxa of subspecific rank,
we have ignored such taxa.

It will be recognized that many arbitrary
decisions are required to classify all of the
species in respect to the foregoing charac-
teristics. We have followed the lifeform
classification given in the index of Hol-
mgren and Reveal (1966), except that we

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

85

have separated annual plants from perennial
herbs. We consider megawind propagules to
be dust-like seeds (as in orchids) and seeds
with large plumose appendages (as in milk-
weeds) that can be expected to be regularly
transported over a mile by wind. Miniwind
propagules are considered to include such
fruits as winged utricles of some chenopods,
samaras of maples, grass caryopses that have
large surface-to-volume ratios, and winged
seeds of conifers. Normal dispersal distance
of miniwind propagules is probably no more
than a few yards. The sticktight category
includes "hitchhiker" fruits such as those of
Xanthium, Arctium, Circaea, and Bidens
which are presumably adapted for dispersal
on the fur or feathers of vertebrates. Under
the heading of fleshy fruits, we include
drupes, pomes, berries, and fleshy cones
such as those borne by Juniperus. We as-
sume that such propagules appeal to and
are often dispersed by birds. Propagules des-
ignated as having no modifications for dis-
persal are produced by a great variety of
dry-fruited species in which seeds are rela-
tively large, have a small surface-to-volume
ratio, and are without wings or plumose ap-
pendages.

The categorization of individual species
according to geographic range also present-
ed difficult problems. Once the floras were
recorded on computer cards, the species
were separated into the four floristic groups
previously mentioned. Examination of the
lists thus compiled demonstrated that some
of the species that supposedly occurred only
on western mainlands did in fact also occur
infrequently on the eastern mainlands, even
though they were not encountered on any
of the checklists. In like manner, some spe-
cies on the list of taxa found only on check-
lists from the eastern mainlands are known
to occur (usually sparingly) on the western
mainland. Finally, species that occur on is-
land checklists but not on mainland lists are
rarely local endemics, but are instead north-
ern or southern species or uncommon main-
land species that have reached some of the
isolated mountains. Despite these defi-

ciencies of the geographic range lists, we
have used them for certain analyses that
would have been otherwise impossible to
make.

We have used Holmgren and Reveal
(1966) as our nomenclatural authority for all
species occurring in the Great Basin. No-
menclature of species that occur in Califor-
nia but do not occur in the Great Basin fol-
lows Munz and Keck (1959). Species
mentioned that occur to the south of the
study area but not in California are named
according to Kearney and Peebles (1951) or
Clokey (1951). Problems of synonymy were
largely resolved with the Holmgren and Re-
veal (1966) checklist.

Results

The Study Areas

Our floristic samples are drawn from 6
states and from areas ranging in size from 1
to 3,630 square miles. The mainland floras
are distributed across a north-south gradient
of about 450 miles in the west (3 floras) and
600 miles in the east (6 floras). The 15 is-
lands are geographically centered on the
Great Basin and are spread across more
than 400 miles of distance in both north-
south and east-west directions (Fig. 1). Max-
imum elevation varies from 14,495 to 9105
ft above sea level among mainland areas
and from 14,246 to 8235 ft among islands
(Table 1).

Unfortunately, few climatological stations
are maintained at high elevations in the re-
gion. The few data that are available sug-
gest that the climates of eastern and west-
ern mainlands are somewhat similar in
respect to annual precipitation and poten-
tial evaporation at comparable elevations,
while the island areas tend to receive less
precipitation and to experience greater po-
tential evaporation than either mainland.
Conditions conducive to aridity appear to
be maximal on the more southerly of the
mountain islands considered (United States
Department of Interior 1970).

GREAT BASIN NATURALIST MEMOIRS

No. 2

The Flora

A total of 2,225 different species occur
above 7500 ft elevation in the 24 floras
considered in this paper. Approximately 27
percent of those species occur on both
mainlands and on occasional mountain
ranges between the mainlands. Some 29
percent of the species appear on the west-
ern but not the eastern mainland, and
roughly 30 percent of the species are repre-
sented on the eastern but not the western
mainland. The remaining species (about 14
percent) were recorded only on island
checklists (Table 2).

Species representative of those occurring
on both mainlands include the following:

Aconitum columbianum Nutt.

Balsamorhiza sagittate. (Pursh) Nutt.

Carex aurea Nutt.

Carex lanuginosa Michx.

Elymus glaucus Buckl.

Epilobium angustifolium L.

Equisetum arvense L.

Fritillaria atropurpurea Nutt.

Geum macrophyllum Willd.

Glyceria eUita(Na.sh) A. S. Hitchc.

Hackelia floribunda (Lehm.) I. M. Johnst.

Lonicera involucrata (Rich.) Bank

Qsmorhiza chilensis Hook. & Am.

Populus tremuloides Michx.

Pinus ponderosa Laws.

Purshia tridentata (Pursh) DC.

Ribes cereum Dougl.

Sitanion hystrix (Nutt.) J. G. Smith

Thalictrum fendleri Engelm.

Viola adunca J. G. Smith

Table 2. General distributional character-
istics of the flora considered.

Total species

Species occurring on checklists

from both mainlands
Species confined to western mainland

or occurring on western mainland

and some islands but not on

eastern mainland
Species confined to eastern mainland

or occurring on eastern mainland

and some islands but not on

western mainland
Species recorded only on islands

2.225

613

678
288

Species confined to the western mainland or
that occur on the mainland and a few iso-
lated mountains include the following:

Agropywn pringlei (Scribn. & Sm.) Hitchc.

Allium obtusurn Lemmon

Artemisia douglasiana Bess.

Bromus breviaristatus Buckl.

Carex amplifolia Boott

Carex tahoensis Smiley

Cheilanthes gracillima D.C. Eaton

Cryptantha mohavensis (Greene) Greene

Hulsea brevifolia Gray

Libocedrus decurrens Torr.

Mimulus torreyi A. Gray

Oryzopsis kingii (Bol.) Beal

Pinus jefferyi Grev. & Balf.

Populus trichocarpa Torr. & Gray

Prunus emarginata (Dougl.) Walp.

Sequoiadendron giganteum (Lindl.)

Stipa californica Merr & Davy

Taxus brevifolia Nutt.

Trifolium andersonii A. Gray

Tsuga mertensiana (Bong.) Carr.

Species confined to the eastern mainland or
to that mainland and a few islands are rep-
resented by the species listed below.

Abies lasiocarpa (Hook.) Nutt.
Acer grandidentatum Nutt.
Balsamorhiza macrophyllum Nutt.
Besseya wyomingensis (A. Nels.) Rydb.
Calamagrostis scopulorum M. E. Jones
Ceanothus martini M. E. Jones
Chlorocrambe hastata (S. Wats.) Rydb.
Clematis columbiana (Nutt.) Torr. & Gray
Erigeron ursinus D.C. Eaton
Geum rossii (R. Br.) Ser.
Hierochloe odorata (L.) Beauv.
Mertensia arizonica Greene
Moldavica parviflora (Nutt.) Britton
Orthocarpus tolmiei Hook. & Arn.
Pinus edulis Engelm.
Picea pungens Engelm.
Primula parryi A. Gray
Quercus gambelii Nutt.
Ribes wolfii Rothrock
Thermcypsis montana Nutt.

Species occurring on the checklists of some
of the isolated mountains but on neither
mainland include local endemics as well as
more widespread species that penetrate our
area from primarily northern or southern
floras. Representatives of each of these
groups are listed below.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

87

Endemics Listed by Location
Ruby Mountain Area

Castilleja linoides Gray

Eriogonum kingii Torr. & Gray

Primula capillaris Holmgren & Holmgren
Spring Range

Angelica scabrida Clokey & Mathias

Antennaria soliceps Blake

Castilleja clokeyi Pennell

Cirsium clokeyi Blake

Opuntia charlestonensis Clokey

Penstemon keckii Clokey

Potentilla beanii Clokey

Silene clokeyi C. L. Hitchc. & Maguire

Synthyris ranunculina Pennell

Tanacetum compactum Hall
Toiyabe Mountains

Draba arida C. L. Hitchc.

Mertensia toyabensis Macbr.
Wheeler Peak

Eriogenum Holmgrenii Reveal
Species Entering from North

Castilleja viscidula A. Gray

Cymopterus nivalis S. Wats

Erigeron watsoni (A. Gray) Cronq.

Selaginella selaginoides (L.) Link
Species Entering from South

Agastache pallidiflora (Heller) Rydb.

Antennaria marginata Greene

Aqailegia triternata Payson

Arenaria confusa Rydb.

Eleocharis montana (H.B.K.) Roem. & Schult.

Festuca arizonica Vasey

Muhlenbergia wrightii Vasey

In respect to lifeform characteristics, the
floras of the mainlands and islands (both

close to and well removed from the main-
lands) do not differ significantly (Table 3).
We had anticipated that since perennial
forbs show a preference for more mesic
sites (Harner and Harper 1973) and the is-
land habitats appear to be more xeric than
the mainlands, such species might be under-
represented on the isolated mountains. The
data lend no support to that idea. Woody
species and graminoides are also uniformly
distributed among the floristic groups re-
ported in Table 3.

The number of annual species is consid-
erably higher on the western as opposed to
the eastern mainland (Table 3). In fact, if
only herbaceous species are considered, Chi-
square analysis demonstrates that the num-
ber of annual species on the two mainlands
departs significantly from random expecta-
tions. Also, significantly fewer annual spe-
cies occur in the combined flora of the is-
lands than on the western mainland, but the
island flora does not differ from that of the
eastern mainland in this respect. Chabot
and Billings (1972) have noted that annual
species are more common in the alpine
flora of the Sierras than in other alpine
floras of North America.

Floristic Diversity Considerations

Species-area relationships for the total
flora and various lifeform subsamples there-

Table 3. Lifeform relationships of the floras considered. The criterion for separation of near and far is-
lands was a barrier width of less than or greater than 100 miles. The following four islands constitute the
"far islands" category: Deep Creek, Jarbidge, Santa Rosa, and Spring. Expected numbers of species in
each category (assuming random distribution of lifeform classes among floras) is enclosed by parentheses.

Floristic
Group

Shrubs

Lifeform Class
Perennial Herbs
Forbs Graminoides

Annuals

Total

W. mainlands

27

111

696

226

181

1,241

(28.1)

(114.4)

(734.9)

(214.7)

(148.9)

E. mainlands

27

119

754

213

139

1,252

(28.4)

(115.4)

(741.4)

(216.6)

(150.2)

Near islands

30

108

776

204

147

1,265

(28.7)

(116.6)

(749.1)

(218.9)

(151.7)

Far islands

18

77

440

136

73

744

(16.9)

(68.6)

Summation

(440.6)

Chi-Square = 18.285

(128.7)

(89.2)

(Not a significant departure from random expectations at 12 degrees of freedom and the 0.95 probability level.)

88

GREAT BASIN NATURALIST MEMOIRS

No. 2

of are shown for both mainlands and islands
in Figure 2. Three generalizations can be
drawn from that figure: 1) there are con-
sistently more species per unit area on the
mainlands than on the islands, 2) floristic di-
versity increases faster on islands than main-
lands as area increases, and 3) area usually
accounts for more of the variation in spe-
cies diversity on islands than on mainlands
(i.e., correlation coefficients for species-area
relationships are usually larger for islands
than for mainlands). Observations 1 and 2
have been duplicated in numerous island
biogeographic studies (MacArthur and Wil-
son 1967) and are commented on here only
to emphasize that the isolated mountains

under study do exhibit strong similarities
with true islands.

The third observation may be partially
attributable to the classification of a single
flora. We have treated the Bryce Canyon
flora as mainland, but Figure 2 demon-
strates that its flora and lifeform subsamples
consistently fall on the species-area trend
line for islands and well below the trend
line for mainlands. Correlation coefficients
for both mainlands and islands would have
been improved had we classified Bryce
Canyon as an island. The area lies at the
southern extremity of the more-or-less con-
tinuous system of highlands extending south
from northern Utah and along the western

AREA ISO MILES)

Fig. 2. Species-area relationships for mainland and island floras. Relationships for the total flora and
various lifeform subsets thereof are shown. Mainland data are represented by dots; insular floras are shown
with triangles. The individual floras are identified in the diagram for total species combined: numbers cor-
respond to specific floras identified in Table 1. Subscript c indicates mainland correlation coefficients or
regression equations; subscript i indicates island coefficients and equations. S represents number of species
and A represents area.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

edge of the Colorado Plateau. We initially
considered the habitat breaks along the
highland corridor to be short and in-
consequential as migration barriers and thus
settled on the mainland classification for the
area. In retrospect, it seems likely that the
narrowness of the corridor has combined
with general climatic differences and unusu-
al soils to effectively filter out numerous
northern taxa that would otherwise be ex-
pected in the area.

Slopes (Z-values) for the species-area
regression lines of Figure 2 are shown as
the exponents of area (A) in the equations
associated with the figure. The Z-value of
0.11 for total flora on the mainlands is
slightly smaller than values commonly re-
ported (e.g., 0.12-0.17 by MacArthur and
Wilson 1967). The average Z-value of .19
reported for nested quadrats in pinyon-
juniper ecosystems of Utah and New Mexi-
co (Harner and Harper 1976) should prob-
ably not be compared to the Z-values ob-
tained for mainlands in this study, since it
seems likely that Z-values for nested quad-
rats where the largest sample area is only a
few acres will always be larger than values
for regional floras from areas ranging in size
from a few to several hundred square miles.

The Z-value of 0.31 for the total flora of
islands (Fig. 2) is well within the range of
values (0.20-0.35) reported for a variety of
kinds of biota on true islands and close to
the theoretically expected value of
0.26-0.27 (MacArthur and Wilson 1967).
We call attention in passing to the fact that
woody plants have flatter species-area re-
gression lines than perennial herbs on both
mainlands and islands.

The flatness of species-area regression
lines for mainlands has been attributed to
the fact that small sample areas there carry
individuals of many species that are poorly
adapted to the sample area but nevertheless
occur there because vigorous populations of
each such taxon exist in nearby, suitable
habitats (MacArthur and Wilson 1967). The
steepness of species-area trend lines for is-
lands is related to at least two factors: 1)

decreasing likelihood of an island being col-
onized by dispersing taxa as size decreases
and 2) increased likelihood of local extinc-
tion of small populations on little islands.

Brown (1977) reports Z-values of 0.165
for boreal birds and 0.326 for boreal mam-
mals on sites of isolated Great Basin moun-
tains. He has previously reported a Z-value
of 0.428 for boreal mammals, using a more
restricted group of species and a different
set of mountains (Brown 1971a). Our Z-
value for vascular plants on isolated moun-
tains thus lies between those for boreal
birds, which seem definitely to be in equi-
librium on the moimtains (i.e., neither in-
creasing or decreasing in respect to number
of species per unit area over long time peri-
ods), and small boreal mammals, which are
believed to be losing species by local ex-
tinction faster than new taxa can colonize.
Plants in general appear to behave more
like mammals than like birds on the moun-
tains considered, and perennial herbs yield
Z-values that are especially steep and ap-
proach the values reported for mammals.

Both area and maximum elevation of the
mountain ranges were strongly positively
correlated with total vascular species on
those ranges in this study (Table 4). There
was a weak negative correlation between
number of species and distance to the near-
est mainland. In multiple correlation analy-
sis, only area makes a large contribution to
the coefficient of multiple determination
(R2). Elevation appears to be so closely cor-
related with area (r = 0.66) that it brings
little new information into the multiple cor-
relation analysis. Distance also enters the
multivariate equation; but it, like elevation,
contributes only slightly over 0.01 to the
Revalue (Table 4).

The overwhelming dominance of area in
the multiple correlation analysis is, in all
probability, an illusion. Wyckoff (1973) and
Harner and Harper (1976) have demon-
strated that both environmental favorability
(annual precipitation and/ or soil texture)
and environmental heterogeneity (variation
in soil characteristics, elevation, and/or ex-

90

GREAT BASIN NATURALIST MEMOIRS

No. 2

posure) exert a strong influence on the
number of vascular plant species per unit
area. However, since area subsumes all of
these variables, it alone consistently ac-
counts for a highly significant amount of
the variation in floral diversity in almost
any suite of samples. Unfortunately, data on
environmental favorability and hetero-
geneity are not available for the sample of
mountains considered here. We have thus
resorted to the use of the less definitive but
nevertheless useful variables of area, eleva-
tion, and distance.

We commented earlier on the com-
plicating effect of two close mainlands in is-
land biogeography studies. In order to bet-
ter evaluate the influence of distance
between island and mainland on floristic di-
versity of the islands, we have measured the
width of unfavorable habitat separating
every island from each mainland. Then, by
using only species that appear to be con-
fined to one mainland or to one mainland
and a few islands (i.e., species common to
both mainlands or unique to islands were
excluded), we used simple and multiple cor-
relation to analyze the relative influence of
area, elevation, and distance from mainland
on the number of species from either east-
ern or western mainlands on the 15 islands.
The results (Table 5) show that distance
now becomes the major factor influencing
the number of western mainland species on
the islands. For eastern mainland species,
distance is not significantly correlated with

number of species in simple correlation
analyses, but it makes a sizeable contribu-
tion in the multiple correlation analysis.
The dissimilar results for species number-
distance relationships for species of western
or eastern mainland origin may be related
to the fact that the islands considered
are on the average more distant from the
western mainland (149 miles) than from the
eastern (117 miles). In any event, the results
in Table 5 seem to suggest that use of a
single distance (distance to nearest main-
land) as in Table 4 may obscure the impor-
tance of distance in studies of island floras.

We recognize also that a decrease in spe-
cies of any given checklist is to be expected
as one moves away from the center of the
geographical area sampled for the checklist.
Such a decrease with distance would be ex-
pected even in large continental areas of
relatively uniform climate, topography, geo-
logical substratum, and geological history
and may have nothing to do with dispersal
habits of the species. The decrease may re-
flect nothing more than the difficulty expe-
rienced by locally evolved taxa as they at-
tempt to expand their range through
established vegetations.

Sources of the Flora

In this section we consider the question
of source of the floras of the isolated moun-
tains. How important a contribution do lo-
cal endemics make to the floras of the iso-
lated ranges? Are the island floras derived

Table 4. Factors influencing the number of vascular plant species on the 15 mountain islands. Distance
is measured to the nearest mainland area having an elevation over 7500 ft.

Factor

Area of island

Elevation of highest peak

Distance to mainland

Contribution to

Simple Correlation

Coefficient of

Coefficient (r) with

Multiple

Number of Species

Determination (R2)

.879

.777

.668

.014

-.091

.013

Total.799

R =

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

91

equally from eastern and western mainlands,
or is one source more important than the
other?

In respect to endemics, the data suggest
that their contribution to the floras of the
isolated mountains is comparatively minor.
The number of endemics of moderate-to-
high elevations appears to be considerably
larger on the Spring Range (the Charleston
Mountains which Clokey [1951] studied are
part of this range) than on any other range
considered here. Yet even on the Spring
Range, which Clokey (1951) considered to
be about five million years old, endemics
account for only about 5 percent of the
flora above 7500 ft. Endemics account for
less than 2 percent of the White Mountain
flora (Lloyd and Mitchell 1973). In contrast,
plant endemics on many remote oceanic is-
lands account for over 50 percent of the
flora (Carlquist 1974). Such data force one
to conclude that the mountain ranges con-
sidered are far less isolated than remote
oceanic islands such as St. Helena, the Ha-
waiian Islands, or New Caledonia, where
the majority of the flora is endemic.

In order to evaluate the relative contribu-
tion of western and eastern mainland floras
to individual islands, we have separated out
species unique to western as opposed to
eastern mainlands (see Table 2). The rela-
tive contribution of uniquely western or
eastern species on individual islands is
plotted against distance to the respective
mainlands in Figure 3. The data demon-
strate that the eastern source area con-
sistently contributes many more species to
the islands than does the western source
area. On only one island (the White Moun-
tains) does the western mainland contribute
a larger percentage of the total flora than
the eastern. As will be shown later (Fig. 4),
the preeminence of the eastern source area
in island floras can be demonstrated for all
dispersal types.

To further illustrate the relative contribu-
tion of the respective mainlands to the is-
land floras, we have compiled a similarity
matrix for all possible combinations among
the 24 floras (Table 6). Various inter-
relationships among floras are summarized
in Table 7. At first glance, the low sim-

Table 5. Factors influencing the number of vascular plant species on the islands when species occur-
ring on both mainlands and on islands only are excluded. Width of barrier (distance) separating an island
from each mainland has been determined for all islands.

Factor

Western Mainland Species

Simple Correlation
Coefficient (r) with
Number of Species

Contribution to

Coefficient of

Multiple

Determination (R2)

Area of island

Elevation of highest peak

Distance to W. mainland

Factor

.502
.644
.646

Eastern Mainland Species

Simple Correlation
Coefficient (r) with
Number of Species

Total
R =

.092
.417
.509
.714

Contribution to

Coefficient of

Multiple

Determination (R2)

Area of island

Elevation of highest peak

Distance to E. mainland

.590
.306
.137

Total
R =

.348

.156
.504
.710

92

GREAT BASIN NATURALIST MEMOIRS

No. 2

ilarity values seem to indicate little com-
monality among floras, but those values
must be evaluated in light of the way in
which they are computed: for example, the
37 percent similarity value between the
Ruby and Deep Creek Mountains represents
223 species common to those ranges. Read-
ers are referred to the similarity equation
given in the legend for Table 7 for details
of computation.

Several relationships reported in Table 7
merit attention: 1) internal similarity of the
floras from the western mainland is almost
identical to the comparable figure for east-
ern mainland floras, 2) the island floras are
less similar to each other than are floras
from either mainland, 3) island floras are,
on the average, more similar to eastern
mainland floras than to western mainland
floras, and 4) even islands closest to the
western mainland have slightly closer floris-
tic affinities with the eastern, rather than

the western, mainland. The second of the
foregoing items indicates, as one might ex-
pect, that the flora of individual mountain
islands tends to be a more random assem-
blage of species than is found in individual
floras on either mainland. Items 3 and 4 in-
dicate that the island floras have been more
influenced by the eastern than the western
mainland, despite the fact that they lie
"downwind" (in this case, the prevailing
westerly winds) from the western mainland.
This last fact is visually conspicuous in the
field since many of the dominant plants of
most of the isolated mountain ranges have
eastern affinities. Examples of such domi-
nant, or at least abundant, plants include
the following:

Agropijron spicatum (Pursh) Scribn. & Smith
Amelanchier alnifolia (Nutt.) Nutt.
Amelanchier atahensis Koehne
Artemisia arbuscula Nutt.
Artemisia tridentata Nutt.

40,—

W 30

m

Q

LU

20

rs= .84

^

.56

A /

A

A
A

A

A

A A,

A

44

_EASJEBfcL

A

•

^

\^A

•
1 1

•

-___•

•

•

1

10

o

300

50
250

100 150 200 250 3 OO WESTERN

200 150 100 so o EASTERN

DISTANCE FROM MAINLAND (MILES)

Fig. 3. Percent of the flora contributed by species that appear to have immigrated from the western as
opposed to the eastern mainland. The western contribution is shown by dots, the eastern by triangles. The
revalue represents the correlation coefficient for the curvilinear correlation between percent of species
contributed by the western mainland and distance from that mainland. The rw-value is the correlation
coefficient for the curvilinear relationship between contribution of eastern species and distance (length of
low elevation barrier between the islands and the eastern mainland). Individual islands can be identified
in the figure by referring to island-mainland distances in Table 1.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

93

Bromus anomalus Rupr.

Caltha leptosepala DC.

Ceanothus martini M. E. Jones

Delphinium occidentale (S. Wats.) S. Wats.

Geranium fremontii Torr.

Holodiscus dumosus (Hook.) Heller

Juniperus osteosperma (Torr.) Little

Lathyrus pauciflorus Fern.

Lewisia rediviva Pursh

Oenothera eaespitosa Nutt.

Pachistima myrsinites(Puish) Raf.

Phlox longifolia Nutt.

Primula pamji A. Gray

Ranunculus jovis A. Nels.

Valeriana occidentalis Heller

Since others (McMillan 1948 and Major
and Bamberg 1967) have speculated about
the relative effectiveness of northern and
southern migration lanes from the western
outliers of the Rocky Mountains in provid-
ing species for interior Great Basin moun-
tains, we have investigated that problem us-
ing the similarity matrix of Table 6. Below
we have summarized the relations of four
interior ranges in the Basin (Deep Creek,
Ruby, Toiyabe, and Wheeler Peak) with
three northern sources (northern Wasatch,
Mount Timpanogos, and Red Butte Canyon)
and two southern sources (Bryce Canyon
National Park and Pine Valley Mountains).

Average Percent Similarity
with

Three Two

Northern Southern

Mountain Range Sources Sources

Deep Creek

31.7

21.0

Rubv

29.7

15.0

Toiyabe

25.0

, 18.5

Wheeler Peak

24.3

21.5

The data demonstrate that although both
northern and southern routes have fed spe-
cies onto the isolated mountains, the north-
ern route seems consistently to have been
more effective than the southern. The low
similarity of the four interior mountain
ranges with the East Tintic Mountains and
their higher similarity with mountain ranges
such as the Jarbidge to the north suggests
that migration from the western outliers of
the Rockies has been primarily along the
northern rim of the Great Basin and south-

ward along the north-south-oriented moun-
tain ranges rather than westward across the
dry basins that separate the ranges of cen-
tral Utah and Nevada. That hypothesis is
strengthened by the low similarity shown by
the East Tintic Mountains with the three
northern sources (average similarity of 18
percent).

Certain species seem clearly to have
reached the interior mountain islands of the
Great Basin via the northern route, while
others have apparently reached those islands
via the southern route. Species representa-
tive of each route are noted below.

Northern Route

Ceanothus velutinus Dougl.

WESTERN SPECIES

EASTERN SPECIES

150 200
DISTANCE FROM MAINLAND (MILES)

Fig. 4. Regression lines relating percent saturation of
eastern or western floristie elements on the 15 islands
to distance to mainland. Distance is defined as in Fig.
3. Regression coefficients (r-values) larger than .514 are
significant at the 0.05 probability level. The b-values
are slopes for the regression lines.

94

GREAT BASIN NATURALIST MEMOIRS

No. 2

Kalmia polifolia Wang.
Ledum glandulosum Nutt.
Finns albicaulis Engelm.
Rubus parviflorus Nutt.
Southern Route

ArctosUiphtjlos patula Creene
Nieotiinui attenuata Torr.
Pivaphullum ramosissimum (Nutt.) Rydb.
Pinus aristata Engelm.
Pinus ponderosa Laws.

Dispersal Ecology

Plant dispersal habits in both mainland
and island floras are dominated by types
which have no apparent modifications for
dispersal and types with weak modifications
for dispersal by wind (Table 8). For conven-
ience, we refer to the latter category as the
"miniwind" modification. Species whose
propagnles have no apparent modifications
for dispersal account for from 50.4 to 53.7
percent of the species in the floras consid-

ered in Table 8. Species having miniwind
propagules contribute between 28.8 and
33.5 percent of the species. Together, these
two dispersal types account for almost 85
percent of the species considered. On the
average, species having propagules modified
for long-range dispersal by wind (megawind
dispersal type) contribute almost 7.5 per-
cent of all species in our floras. Fleshy
fruited species contribute slightly fewer spe-
cies (average 6.1 percent of all species), and
species dispersed by sticktights contribute
the few remaining species (about 2.5 per-
cent).

Our data indicate that dispersal types
modified for long-range movement (i.e.,
fleshy fruit and megawind categories) show
no tendency to be overrepresented on the
remote islands (Table 8). In contrast,
Carlquist (1967) has shown that as many as
58 percent of the plant species that reach

Table 6. Similarity among the 24 floras as determined with the Jaccard (1912) similarity index. Values report-
ed are percent similarity for all possible pairs of floras. Checklist numbers correspond to those assigned to each
area in Table 1.

Checklist No

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22 23 24

1

Albion

2

21

Cassia

3

27

15

Deep Cr.

4

17

11

18

Tintic

5

28

16

32

15

Jarbidge

6

17

11

23

13

19

Kaibab

7

16

12

22

17

18

19

Pine V.

8

34

22

41

18

33

20

21

Raft R.

9

25

14

37

13

40

19

15

29

Ruby

10

26

19

21

17

.30

12

14

27

23

Santa Rosa

11

14

8

23

15

16

25

22

19

17

13

Spring R.

12

20

12

29

17

28

20

20

29

27

25

25

Toiyabe

13

19

16

18

11

23

14

14

21

19

22

11

19

Wamer

14

27

15

39

19

31

21

20

33

36

25

22

29

18

Wheeler

15

16

9

25

12

21

19

18

22

24

17

27

31

18

25

White

16

19

12

20

18

17

27

19

21

15

16

25

17

12

23

15

Bryce

17

14

9

17

7

23

11

13

16

21

17

12

17

23

14

17

9

Lassen

18

25

14

32

18

33

21

21

31

32

18

18

23

18

25

21

19

18

Timpanogos

19

24

14

35

15

34

24

21

32

32

18

19

28

19

25

22

18

22

47

N. Wasatch

20

26

17

28

21

32

19

22

32

25

20

17

24

21

23

16

17

17

44

40

Red Butte

21

15

12

18

8

23

11

12

16

19

15

10

19

25

15

18

9

38

18

21

17

Sagehen Cr.

22

11

6

16

6

20

12

11

14

21

10

13

18

15

14

28

9

29

17

24

15

24

Sequoia

23

21

11

31

11

31

24

17

28

30

15

18

22

17

26

20

19

17

37

39

28

16

20 Uinta

24

17

9

24

9

31

17

14

23

29

14

12

20

16

,20

19

12

21

32

38

26

19

23 35 Yel

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

95

remote oceanic islands are dispersed inter-
nally by birds (mostly fleshy fruits). Prop-
agules borne externally on birds by virtue of
being held in place by barbs or prickles
(sticktights) also account for many (fre-
quently over 20 percent) of the in-
troductions. He found windborne seeds to
be poorly represented (usually less than 10
percent of the flora) on all save the closest

of the remote islands. He considered ecolog-
ical conditions on the island to be strong
determinants of the dispersal types that suc-
ceeded.

In contrast to Carlquist's findings, Hed-
berg (1970) found wind-dispersed plants to
represent almost 30 percent of the flora
above about 7900 ft on the mountains of
east Africa. In Hedberg's study, plants dis-

Table 7. Floristic similarity relations among the floras considered. The index of similarity used is that
of Jaccard (1912). Jaccard's index is computed as follows:

C

SI

X 100.

In the equation, C represents the number of species common to the two floras, A is the number of species
in flora A, and B is the number in flora B.

Areas Considered

No. of

Floras

Involved

No. of

Comparisons

Averaged

Average
Percent

Similarity

Western mainland (internal similarity)
Eastern mainland (internal similarity)
Mountain islands (internal similarity)
W. mainland compared with islands
E. mainland compared with islands
W. mainland compared with E. mainland
Four closest islands to W. mainland

compared to W. mainland
Four closest islands to W. mainland

compared to E. mainland
Four closest islands to E. mainland

compared to W. mainland
Four closest islands to E. mainland

compared to E. mainland

3

3

30.3

6

15

30.1

15

105

21.0

18

45

15.0

21

90

21.4

9

18

17.3

7

12

18.1

10

24

18.8

7

12

11.4

10

24

21.0

Table 8. Plant dispersal habits of the floras considered. Expected number of species in each category
(assuming random distribution of lifeform classes among floras) is enclosed by parentheses. Island groups
are defined as in Table 3.

Floristic

Mega-

Mini-

Fleshy

Stick-

No

Group

wind

wind

Fruits

tights

Modification

Total

W. mainland

93

.358

88

.36

666

1,241

(92.3)

(395.8)

(75.3)

(30.9)

(646.7)

E. mainland

104

414

77

26

631

1,252

(93.2)

(399.3)

(75.9)

(31.1)

(652.4)

Near islands

83

415

65

.30

672

1,265

(94.1)

(403.5)

(76.7)

(31.5)

(659.2)

Far islands

55

249

43

20

377

744

(55.4)

(237.3)

(45.1)

(18.5)

(387.7)

Summation Chi-Square =

15.501

(Not a significant departure from random expectations at 12 degrees of freedom and the 0.95 probability level.)

96

GREAT BASIN NATURALIST MEMOIRS

No. 2

persed internally by birds accounted for
only 1 to 2 percent of the alpine flora of
east Africa.

The relationship between various dis-
persal types and island-to-mainland distance
is presented in Figure 4. There, we regress
percent saturation of species of various dis-
persal habits (i.e., the number of species of
a given dispersal habit on each island is ex-
pressed as a percentage of the number of
species of that dispersal habit that would be
expected in an area of comparable size on
the appropriate mainland) against distance.
As expected, the regression lines all have
negative slopes, and there is a slight (but
statistically nonsignificant) tendency for dis-
persal types that are easily dispersed over
long distances (megawind and fleshy fruit
types) to have regression lines with gentler
slopes than are obtained for species that are
less likely to be dispersed far from the par-
ent plant. Average slope values for western
and eastern mainlands and each dispersal
type are shown below.

Dispersal
Type

Average
Slope Value

Megawind

Fleshy Fruits

Miniwind

Sticktight

No Modification

.03

.05
.08

.(MS

.07

The data in Figure 4 also support our
earlier conclusion that the eastern mainland
has exerted a greater influence on the
mountain islands than has the western main-
land. Every dispersal type shows greater
saturation for eastern species than for spe-
cies from the western mainland. Since the
number of species originating from each of
the two mainlands is roughly equal (see
Table 2), the results in Figure 4 suggest that
species from the eastern mainlands have
been about four times as effective in reach-
ing and surviving on the islands as those
from the west. On the average island, west-
ern species have a saturation value of 8 per-
cent, but the comparable value for species
from the eastern mainland is 36 percent.

The great disparity between correlation
coefficients for saturation-distance analyses
for eastern and western species in Figure 4
is noteworthy. In five of the six analyses the
r-values are much larger for western spe-
cies. It seems possible that those values re-
flect a differential in age of the two floristic
elements on most of the islands. If the
Rocky Mountains are much older than the
Sierras, as Billings (1977) reports, it is pos-
sible that the eastern floristic element has
dispersed essentially to its limit and is now
poorly related to distance, while the west-
ern element is still actively dispersing.

Finally, we call attention to a con-
spicuous relationship between range limits
of species and plant lifeform. Our data
demonstrate that woody plants and per-
ennial graminoid species are over-
represented in the broad-range category
(i.e., occurring on both mainlands) and un-
derrepresented in the category of species
unique to islands (Table 9). Perennial forbs,
on the other hand, display a significant ten-
dency toward underrepresentation in the
broad-range category and over-
representation in the island-only class. An-
nual species show no significant trends in
this respect. It seems possible that the pat-
terns observed reflect evolutionary rather
than dispersal processes. In general, woody
plants and graminoides appear to be ecolog-
ically broad niched and to have the ability
to become community dominants. In con-
trast, many perennial forb genera seem to
be narrow niched and to rarely achieve a
dominant place in their community.

Discussion

Mountains as Islands

One might expect an island flora to be
distinguished from that of the nearest main-
land in a variety of ways. As we began this
study, it seemed to us that insular floras
should display 1) an overrepresentation of
species modified in one way or another for

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

97

long-distance dispersal, 2) fewer species per
unit area than observed on the mainland, 3)
steeper species-area curves than for main-
land floras, 4) uneven stocking of species
ecologically preadapted for existence on
available islands, and 5) higher rates of en-
demism than the mainland.

Our results demonstrate that the isolated
mountains of the Intermountain West satisfy
some of our preconceived notions and thus
qualify as islands, but they fail to qualify on
other counts. The islands do indeed have
fewer species per unit area than adjacent
mainlands, and species-area trend lines for
islands are steeper than those for main-
lands (Fig. 2). Although the amount of en-
demism is low on the islands (always less
than 5 percent), the amount still appears to
be higher than on areas of comparable ele-
vation and size on the mainlands. Too,
there is uneven stocking of species on the
islands. The Pine Forest Mountains of ex-
treme northwestern Nevada, for example,
are stocked by Pinus albicaulis Engelm., the
Santa Rosas by Pinus flexilis James, while
the Jarbidge and Ruby Mountains to the
east and the Sierras to the west have both.
The observed distribution pattern for these
and many other species [e.g., Abies concolor
(Gord. & Glend.) Lindl. and Picea engel-
mannii Parry ex Engelm.] seems explainable
only in terms of randomness of colonization

and/ or extinction (See Critchfield and Al-
lenbaugh 1969 for range details for these
and other conifers in the Great Basin.)

Our expectations relative to an over-
representation of species modified for long-
range dispersal on the islands in large part
failed. The isolated mountains are over-
whelmingly dominated by species with no
obvious means for being dispersed great dis-
tances. Furthermore, there is no tendency
for species with modifications for long-dis-
tance dispersal to be overrepresented on
even the most distant islands (Table 8). Our
data do, however, show a weak tendency
for percent saturation of poorly dispersed
species (i.e., no-modification, miniwind, and
sticktight categories) to decline faster and
more reliably (larger r-values) with distance
than for megawind and fleshy-fruited spe-
cies, which are probably more easily dis-
persed (Figure 4).

Recent literature references demonstrate
that at least some of the species that we
have classified as unmodified for dispersal
are, in fact, highly adapted for dispersal by
vertebrate animals. Although we placed all
conifers with unwinged seeds in the unmo-
dified-for-dispersal category, a recent paper
by Vander Wall and Balda (1977) shows
that the Clark's Nutcracker regularly dis-
perses the seeds of several pines (P. edulis,
P. albicaulis, and P. flexilis) in a sublingual

Table 9. Plant lifeform relative to the range limits of the species considered. Expected number of spe-
cies appears in parentheses in each category.

Range
Category

Woody
Plants

Lifeform Class

Perennial Herbs
Forbs Graminoides

Annuals

Total
Species

Occurring on
both mainlands

94
(70.0)

333
(368.6)

114

(98.9)

72
(75.5)

613

Occurring on one
mainland only

140

(151.1)

795
(796.2)

215

(213.6)

174
(163.0)

1,324

Occurring on
islands only

20

(32.9)

210

(173.2)

Summation Chi-Square

30

(46.5)
= 36.020°°

28
(35.5)

288

lificant departure from random expectations at 6 degrees of freedom and the 0.99 probability level.)

98

GREAT BASIN NATURALIST MEMOIRS

No. 2

pouch and caches them in soil suitable for
their germination and growth. In addition,
the Nutcracker is known to occasionally
feed on the winged seeds of Pinus aristata
and Pinus ponderosa in northern Arizona.
Vander Wall and Balda (1977) have evi-
dence for the dispersal of seeds over 13
miles in a single flight by the Nutcracker.
In California, the Nutcracker regularly
feeds on and caches the seeds of Pinus mon-
ophylla Torr. & Frem. and Pinus jefferyi as
well as Pinus albicaulis and Pinus flexilis
(D. Tomback, personal communication).
Johnson (1975) suggests that the Pinon Jay
and the Band-tailed Pigeon may also be in-
volved in long-distance transport of con-
iferous tree seed. J. Pederson (personal com-
munication) reports that the Band-tailed
Pigeon has been taken several miles from
the nearest Quercus gambelii in south-
eastern Utah with a crop full of unbroken
acorns. Staniforth and Cavers (1977) demon-
strate that some seeds of two Polygonum
species (P. lapathifolium L. and P. pensyl-
vanicum L.) retain viability after passing
through the digestive tract of the cottontail
rabbit in eastern Canada. The foregoing
data lead us to suspect that large seeds from
the dry fruits of many species will eventu-
ally be shown to be dispersed by vertebrate
animals.

The foregoing discussion is an acknowl-
edgement that we have underestimated the
number of plant species that are modified
for long-range transport on our islands.
Nevertheless, the number of species in the
no-modification and miniwind categories is
so great on the islands that we are still
forced to conclude that the vast majority of
the species there did not reach those sites
by long-range dispersal. Although the high
elevation community types may never have
been able to survive on the valley floors at
any time during the Pleistocene, as Wells
and Berger (1967) argue, many of the com-
munity components may have been able to
migrate directly across valley floors during
that period. Also, as Billings (1977) empha-
sizes, climatic cooling would have signifi-

cantly narrowed the barriers between is-
lands.

Our discussion of mountains as islands
would not be complete without some com-
ment on the question of equilibrium of spe-
cies number on the islands. Brown (1977)
contends that birds are and small mammals
are not in equilibrium on isolated mountains
in our study area. Are the plants in equilib-
rium? It will be recognized that the equilib-
rium argument is based on two assumptions:
1) local extinctions do occur, and 2) new in-
troductions occur as often as extinctions on
each island. Both assumptions are difficult,
if not tactically impossible, to test con-
clusively. A definitive test would require
that we know of every population of every
species on every island, and that we mon-
itor each island regularly enough (preferably
annually) in order to know when a species
became extinct or immigrated and became
established there. Obviously, such data are
not available for any island in our study. As
a consequence, any statement about the
status of our islands relative to the equilib-
rium question must be based on inferences,
not facts.

With respect to extinctions, there is con-
clusive evidence that Pinus aristata and
Pinus flexilis coexisted with Abies concolor
and Juniperus osteospenna on Clark Moun-
tain in southeastern California about 25,000
years ago (Mehringer and Ferguson 1969).
Today neither of these pines occurs there.
Similarly, Pinus monophylla and Juniperus
osteosperma existed on the Turtle Range
14,000 years ago (Wells and Berger 1967),
but do not occur there now. The relatively
steep species-area curves for herbs (Fig. 2)
may indicate extinctions, but we can offer
no evidence in support of that possibility.

Concerning new immigrations onto the
isolated mountains, there are abundant re-
cords of exotic species invading at lower
elevations (Young, Evans, and Major 1972).
Nevertheless, we know of no documented
cases of unaided immigrations onto the
mountains of species that cannot survive in
at least some microsites on the valley floors.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

There is strong evidence that species
modified for long-range dispersal are not
overrepresented on the islands relative to
the mainlands (Table 8). If extinctions and
immigrations had been in equilibrium, even
since the close of the Pleistocene, one might
have expected long-range dispersal types to
be at least somewhat overrepresented on is-
lands; but even that tendency is not ob-
served (Table 8). As noted above, there is a
weak tendency for percent saturation of
long-distance dispersal types to decline less
rapidly against distance from the mainland
than for supposedly less well-dispersed taxa.
These two bits of evidence lead us to ten-
tatively conclude that the flora of the iso-
lated mountains is not in equilibrium, even
though some species do appear to be mov-
ing about in the area.

If the islands are not in equilibrium, the
extinction rate must be low for all plant
groups and especially so for the woody taxa.
We draw this inference from the relative
flatness of the species-area curve for most
plant groups (Fig. 2) in contrast to mam-
mals (Brown 1977). Intuitively, this infer-
ence seems valid since herbaceous plants as
primary producers should be able to main-
tain larger populations than their vertebrate
consumers. Woody plants (especially trees)
would be expected to maintain smaller pop-
ulations than their vertebrate consumers,
but would have far greater longevity. Tro-
phic position and longevity likely have
much to do with the relative extinction rate
of vertebrates and plants. Plant groups of
differing trophic habit (e.g., vascular sap-
rophytes and nongreen parasites such as Co-
rallorhiza and Orobanche, respectively, ver-
sus photosynthetic forms) and longevity
should show different extinction rates.

We had not expected to find the eastern
mainland (Rocky Mountains) floristic ele-
ment to be so much more successful than
the western mainland (Sierra) element on
the Great Basin mountains. As others have
noted in this symposium, the Rocky Moun-
tain element also dominates the avian fauna
(Behle 1977 and Johnson 1977) and the al-

pine flora (Billings 1977) of the isolated
mountains. The evidence seems to imply
that three basic factors have combined to
give the Rocky Mountain element an ad-
vantage over that from the Sierra. Those
factors are: 1) time, 2) geological parent
material, and 3) climate.

As Billings (1977) has noted, most of the
Great Basin mountains are younger than the
Rockies and older than the present Sierra
Nevada and Cascade ranges. Thus, species
from the east have had longer to colonize
the isolated mountains than high-elevation
taxa from the Sierra, since that flora must
have arisen much later than the first. In ad-
dition, propagules of species unique to the
western mainlands would have had great
difficulty establishing themselves on the
mountain islands even after reaching them,
since most habitats would have already
been occupied by eastern taxa.

Plants originating at higher elevations on
the western mainland could generally be ex-
pected to be adapted to acidic soils, since
the Sierra Nevada is primarily composed of
acidic, igneous rock (Major and Bamberg
1967). Soils on the isolated Mountains, how-
ever, have prevailingly basic to circum-
neutral soils. Again, taxa from the eastern
mainland would have an advantage in colo-
nizing the islands, since the western outliers
of the Rockies are prevailing formed from
calcareous rocks. In this connection, it
is significant that Billings (1950) found as-
semblages of Sierra plants in the western
Great Basin to be confined to acidic habi-
tats on hydrothermally altered rocks.

Finally, western plants have evolved in
an environment that is less continental (i.e.,
more moist and thermally less variable) than
that associated with the isolated mountains
of concern or the western outliers of the
Rockies. Johnson (1977) considers the cli-
matic variable to be highly influential in
confining western bird species to the
Sierras. We believe that continentality may
similarly increase the difficulty of estab-
lishment of western plant species that are
dispersed to the mountain islands. As in the

100

GREAT BASIN NATURALIST MEMOIRS

No. 2

preceding cases, species from the east
would be better preadapted for life on the
islands.

Niche Expansion

Brown (1971b) has shown that the altitu-
dinal range of a normally low-elevation
chipmunk (Eutamias dorsalis) expands up-
ward on Great Basin mountains which lack
a high elevation congener (£. umbrinus). In
the course of our work on isolated moun-
tains in the Region, we have observed sev-
eral cases in which plant species also dis-
play a niche expansion in the absence of
normal competitors. Although quantitative
data are lacking, we take this opportunity
to put such anecdotal evidence as is avail-
able on record.

An apparent case of niche expansion is
presented by Abies lasiocarpa in the Jar-
bidge Mountains. There, in the absence of
its common coniferous competitors (e.g.,
Abies concolor, Picea engelmannii, Picea
pungens, and Pseudotsuga menziesii (Mirb.)
Franco), Abies lasiocarpa plays a major role
in forest vegetation from the sagebrush-grass
and streambank communities at low eleva-
tions to timberline. We know of no other
place where this species succeeds in such a
variety of habitats.

A double zone of Artemisia tridentata oc-
curs on mountainsides of Nevada and west-
ern Utah. There the species commonly
dominates a wide belt both below and
above the juniper-pinyon zone. It appears
likely that Artemisia has simply moved into
a zone that is elsewhere dominated by
larger mesophytes such as Quercus gambelii,
Pinus ponderosa, or a rich mixture of moun-
tain brush species.

In the northern Wasatch Mountains, the
range of Acer grandidentatum extends manv
miles farther north than that of its common
associate in the south, Quercus gambelii. In
mixed stands of Acer and Quercus, Acer is
normally conspicuous only on slope bases
and ravine edges. North of the limits of
Quercus, however, Acer dominates both
slopes and depressions. The phenomenon

can be seen with particular clarity in the
southwest corner of Cache Valley, Utah.

Although Chamaebatiaria millefolium
(Torr.) Maxim, occurs on both of the main-
lands recognized in this study, it is rarely a
conspicuous component of the vegetation
on either. On the remote islands, however,
Chamaebatiaria is often common and a con-
spicuous part of the vegetation.

Finally, West et al. (1977) review evi-
dence suggesting that the anomalously high
upper elevation of the juniper-pinyon zone
on many of the isolated mountains of the
Great Basin may be attributable to the low
diversity of the high-elevation flora and the
paucity of well-adapted competitors. They
note also that the niche of both juniper and
pinyon appears to be severely compressed
on the west flank of the southern and
middle Wasatch Range where Quercus gam-
belii and Acer grandidentata combine to
form a dense woodland. Both juniper and
pinyon occur in the flora there, but neither
is an important part of the vegetation.

Adequacy of Checklists

In the inception of this study, we were
concerned that the checklists on which our
work would be based would be too in-
complete to give meaningful results. In ret-
rospect, we acknowledge that all of the lists
are probably incomplete. Undoubtedly, ad-
ditional effort will add a few species to
some lists and many to others. Nevertheless,
the lists have yielded results that seem rea-
sonable and defensible. Furthermore, the
sample on hand is already sufficiently large
to minimize the possibility that new collec-
tions will seriously alter species-area rela-
tionships or lifeform and dispersal-type
spectra for the floras.

Management Implications

Species-area curves reveal much that
should be useful to natural resource man-
agers. The curve for trees on islands in Fig-
ure 2, for example, suggests that the Santa
Rosa Mountains are drastically understocked
with trees. Could trees be successfully in-

1978

INTEHMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

101

troduced there to provide shelter for ani-
mals or construction materials for man?
Since other islands in the study support so
many more tree species than that range, we
suspect that introduction of one or a few
carefully selected tree species into favorable
sites would have a high probability of suc-
cess there.

Species-area curves also have many useful
implications for conservation programs for
unusual and rare plant and animal species.
Managers will find the basic theory relative
to rare species and size of reserves nicely
capsulized in the following short, non-
technical papers: Terborgh (1974 and 1976),
Diamond 1976, Whitcomb et al. (1976), and
Simberloff and Abele (1976).

Johnson (1975) developed a habitat diver-
sity index that accounted for a major por-
tion of the observed variation in number of
bird species on isolated mountains in the
Great Basin. Behle (1977) has verified that
the index is a useful indicator of bird diver-
sity throughout the Basin. Since that index
is based on various plant parameters and
the presence or absence of free flowing wa-
ter, it has relevance to our discussion here.
Many of our small, arid mountain ranges in
the Intermountain West have only a few
acres of complex forest habitat (a prime
variable in Johnson's index) in a single loca-
tion and but a few score feet of flowing
water. Since the index shows that bird di-
versity is highly dependent upon such habi-
tat, it would seem prudent for developers
interested in preserving the natural biotic
diversity of the environment to insure that
roads, campgrounds, or buildings not in-
fringe upon such habitats. Yet, unfortu-
nately, our developments often are centered
directly on such microenvironmental rari-
ties. By so locating developments, we al-
most insure that we will lose some and per-
haps many plant and animal species from
the entire range. The campground at Blue
Lake on the Pine Forest Bange in north-
western Nevada is a prime example of such
faulty planning. With foresight, the devel-
opment could have been placed well away

from the lake but still in the open pine
groves. Water could have been piped to the
campground with minimal disturbance to
the natural system around the lake. Instead,
the current plan places every visitor in a
position to disturb the several unusual plant
and animal species that perhaps occur at
only that spot on the entire range.

ACKOWLEDGMENTS

We are indebted to Dr. Noel Holmgren
(1972) for his exhaustive pioneer work on
the biogeography of the Intermountain
West. It has been an invaluable aid as we
have planned and pondered this paper. Dr.
William E. Evenson prepared most of the
computer programs used in our analyses.
Our research has been supported in part by
a grant from the National Science Founda-
tion (GB-39272) and another from the Be-
search Division, Brigham Young University.
A grant from the U.S. Bureau of Beclama-
tion through the Utah State Division of
Water Besources has helped defray pub-
lication costs for this paper.

Literature Cited

Allred, K. W. 1975. Timpanogos flora. Unpub-
lished master's thesis, Brigham Young Univer-
sity, Provo, Utah.

Arnow, L. A. 1971. Vascular flora of Red Butte
Canyon, Salt Lake County, Utah. Privately
printed by the author, Salt Lake City, Utah.

Atwood, N. D., and L. C. Higgins. 1976. Checklist
of plants of Pine Valley Mountains, Utah. Un-
published report.

Behle, W. H. 1977. Avian biogeography of the
Great Basin and Intermountain Region. Great
Basin Naturalist Mem. 2: 55 -80.

Billings, W. D. 1950. Vegetation and plant growth
as affected by chemically altered rocks in the
western Great Basin. Ecology 31: 62-74.

1977. Alpine phytogeography across the

Great Basin. Great Basin Naturalist Mem. 2:
105-117.

Brown, J. H. 1971a. Mammals on mountaintops:
nonequilibrium insular biogeography. Amer.
Naturalist 105: 467 -478.

1971b. Mechanisms of competitive exclusion

between two species of chipmunks (Eutamias).
Ecology 52: 305-311.

102

GREAT BASIN NATURALIST MEMOIRS

No. 2

1977. The theory of insular biogeography

and the distribution of boreal birds and mam-
mals. Great Basin Naturalist Mem. 2: 209-227.

Buchanan, H. 1973. Checklists of vascular plants of
the Albion, Black Pine, Cassia, and Sublette
Mountain ranges of southern Idaho. Unpub-
lished report.

Buchanan, H., and G. T. Nebeker. 1971. Checklist
of higher plants of Bryce Canyon National Park,
Utah. Weber State College Printing Depart-
ment, Ogden, Utah.

Carlquist, S. 1967. The biota of long-distance dis-
persal. V. Plant dispersal to Pacific Islands.
Bull, of the Torrey Bot. Club 94: 129-162.

1974. Island biology. Columbia University

Press, New York.

Chabot, B. F., and W. D. Billings. 1972. Origins
and ecology of the Sierran alpine flora and veg-
etation. Ecol. Monogr. 42: 163-199.

Clokey, I. W. 1951. Flora of the Charleston Moun-
tains, Clark County, Nevada. University of Cali-
fornia Press, Berkeley.

Critchfield, W. B., and G. L. Allen-
baugh. 1969. The distribution of Pinaceae in
and near northern Nevada. Madrono 19: 12-26.

Culver, D. C, J. R. Holsinger, and R. Baroody.
1973. Toward a predictive cave biogeography:
the Greenbrier Valley as a case study. Evolution
2: 689-695.

Curtis, J. T. 1956. The modification of mid-latitude
grasslands and forests by man, pp. 721-736. In:
W. L. Thomas (ed.), Man's role in changing the
face of the earth. University of Chicago Press,
Chicago, Illinois.

Darwin, C. 1859. On the origin of species by
means of natural selection (reprint of the first
edition). Watts and Co., London.

Despain, D. G. 1975. Flora of Yellowstone National
Park. Yellowstone Library and Museum Associ-
ation, West Yellowstone, Idaho.

Diamond, J. M. 1976. Island biogeography and con-
servation: strategy and limitations. Science 193:
1027-1029.

Gillett, G. W., J. T. Howell, and H. Leschke.
1961. The flora of Lassen Volcanic National
Park, California. Wasmann J. Biology 19(1):
1-185.

Graham, E. H. 1937. Botanical studies in the Uinta
Basin of Utah and Colorado. Ann. Carnegie
Museum, Pittsburgh, Pennsylvania, 26: 1-432.

Harner, R. F., and K. T. Harper. 1973. Mineral
composition of grassland species of the eastern
Great Basin in relation to stand productivitv.
Can. J. Bot. 51: 2037-2046.

1976. The role of area, heterogeneity, and fa-

vorability in plant species diversity of pinyon-
juniper ecosystems. Ecology 57: 1254-1263.

Heatwole, H., and R. Levins. 1972. Trophic struc-
ture stability and faunal change during recoloni-
zation. Ecology 53:531-534.

Hedberg, O. 1970. Evolution of the Afroalpine
flora. Biotropica 2: 16-23.

Holmgren, A. H. 1972. Vascular plants of the
Northern Wasatch, 4th ed. Privately printed by
the author, Logan, Utah.

Holmgren, A. H., and J. L. Reveal. 1966.
Checklist of the vascular plants of the Inter-
mountain Region. U.S. Forest Service Research
Paper INT-32.

Holmgren, N. H. 1972. Plant geography of the In-
termountain Region, pp. 77-161. In: A.
Cronquist, A. H. Holmgren, N. H. Holmgren,
and J. L. Reveal. Intermountain flora, vol. I.
Hafner Publ. Co., New York.

Jaccard, P. 1912. The distribution of the flora in
the alpine zone. New Phytol. 11: 37-50.

Johnson, N. K. 1975. Controls of numbers of bird
species on montane islands in the Great Basin.
Evolution 29: 545-567.

1977. Patterns of avian geography and speci-

ation in the Intermountain Region. Great Basin
Naturalist Mem. 2: 137-159.

Kearney, T. H., and R. H. Peebles. 1951. Arizona
flora. University of California Press, Berkeley.

Lewis, M. E. 1971. Flora and major plant commu-
nities of the Ruby-East Humboldt mountains
with special emphasis on Lamoille Canyon. Re-
port to Humboldt National Forest, Elko, Ne-
vada.

1973. Wheeler Peak Area species list. Report

to Intermountain Region, U.S. Forest Service,
Ogden, Utah.

1975. Plant communities of the Jarbidge

Mountain complex, Humboldt National Forest.
Report to Intermountain Region, U.S. Forest
Service, Ogden, Utah.

1976. Flora of the Santa Rosa Mountains,

Humboldt National Forest. Report to Inter-
mountain Region, U.S. Forest Service, Ogden,
Utah.

Lindeman, R. L. 1942. The trophic-dynamic aspect
of ecology. Ecology 23: 399-418.

Linsdale, M. A., J. T. Howell, and J. M.
Linsdale. 1952. Plants of the Toiyabe Moun-
tains area, Nevada. Wasmann J. Biol. 10:
129-200.

Lloyd, R. M., and R. S. Mitchell. 1973. A flora of
the White Mountains, California and Nevada.
University of California Press, Berkeley.

MacArthur, R. H., J. M. Diamond, and J. Karr.
1972. Density compensation in island faunas.
Ecology 53: 330-342.

MacArthur, R. H., and E. O. Wilson. 1967. The
theory of island biogeography. Princeton Uni-
versity Press, Princeton, New Jersey.

1978

INTERMOUNTA1N BIOGEOGRAPHY: A SYMPOSIUM

103

MacDougall, W. B. 1947. Plants of Grand Can-
yon National Park. Grand Canyon Natural His-
tory Association, Grand Canyon, Arizona.

Major, J., and S. A. Bamberg. 1967. Some cordille-
ran plants disjunct in the Sierra Nevada of Cali-
fornia and their bearing on Pleistocene ecologi-
cal conditions, pp. 171-188. In: H. E. Wright
and W. H. Osburn (eds.), Arctic and alpine en-
vironments. Indiana University Press, In-
dianapolis, Indiana.

McMillan, C. 1948. A taxonomic and ecological
study of the flora of the Deep Creek Mountains
of central western Utah. Unpublished master's
thesis, University of Utah, Salt Lake City.

Mehringer, P. J., Jr., and C. W. Ferguson.
1969. Pluvial occurrence of bristlecone pine
(Pinus aristata) in a Mojave Desert mountain
range. Interim Besearch Beport No. 14, Dept.
of Geochronology, University of Arizona, Tuc-
son.

Millican, M. T. 1969. Transect flora of the Eagle
Peak Begion, Warner Mountains, Modoc Coun-
ty California. Unpublished master's thesis,
Humboldt State College, Areata, California.

Munz, P. A., and D. D. Keck. 1959. A California
flora. University of California Press, Berkeley.

Nebeker, G. T. 1975. Manual of the flora of the
East Tintic Mountains, Utah. Unpublished mas-
ter's thesis, Brigham Young University, Provo,
Utah.

Preece, S. J., Jr. 1950. Floristic and ecological fea-
tures of the Baft Biver Mountains of north-
western Utah. Unpublished master's thesis, Uni-
versity of Utah, Salt Lake City.

Bockwell, J. A., and S. K. Stocking. 1969.
Checklist of the flora of Sequoia-Kings Canyon
National Park. Sequoia National History Associ-
ation, Three Bivers, California.

Savage, W. 1973. Annotated list of vascular plants
of Sagehen Creek drainage basin, Nevada Coun-
ty, California. Madrono 22: 115-139.

Simberloff, D. S., and L. G. Abele. 1976. Island
biogeography and conservation: strategy and
limitations. Science 193: 1032.

Simberloff, D. S., and E. O. Wilson. 1970.
Experimental zoogeography of islands. A two-

year record of colonization. Ecology 51:
934-937.

Staniforth, B. J., and P. B. Cavers. 1977. The im-
portance of cottontail rabbits in the dispersal of
Polygonum spp. J. Applied Ecology 14: 261-268.

Terborgh, J. 1974. Preservation of natural diversity:
the problem of extinction-prone species. Bio-
Science 24: 715-722.

1976. Island biogeography and conservation:

strategy and limitations. Science 193:
1029-1030.

United States Department of Interior, Geological
Survey. 1970. The National atlas. U.S. Geo-
logical Survey, Washington, D.C.

Vander Wall, S. B., and B. P. Balda.
1977. Coadaptations of the Clark's Nutcracker
and the Pinon pine for efficient seed harvest
and dispersal. Ecol. Monogr. 47: 89-111.

Vuillemier, F. 1970. Insular biogeography in con-
tintental regions. I. The northern Andes of
South America. Amer. Naturalist 104: 373-388.

Wallace, A. B. 1880. Island life. MacMillan and
Co., London.

Wells, P. V., and B. Berger. 1967. Late Pleisto-
cene history of coniferous woodland in the Mo-
jave Desert. Science 155: 1640-1647.

Wells, P. V., and C. D. Jorgensen. 1964.
Pleistocene woodrat middens and climatic
change in the Mojave Desert: a record of juni-
per woodlands. Science 143: 1171-1173.

West, N. E., B. J. Tausch, K. H. Bea, and P. T.
Tueller. 1977. Phytogeographical variation
within juniper-pinyon woodlands of the Great
Basin. Great Basin Naturalist Mem. 2: 119-136.

Whitcomb, B. F., J. L. Lynch, P. A. Opler, and C. S.
Bobbins. 1976. Island biogeography and con-
servation: strategy and limitations. Science 193:
1030-1032.

Wyckoff, J. W. 1973. The effects of soil texture on
species diversity in an arid grassland of the
eastern Great Basin. Great Basin Nat. 33:
163-168.

Young, J. A., B. A. Evans, and J. Major. 1972.
Alien plants in the Great Basin. J. Bange Man-
agem. 25: 194-201.

ALPINE PHYTOGEOGRAPHY ACROSS THE GREAT RASIN

W. D. Billings'

Abstract.— Alpine vegetations and floras are compared in two transects across the Intermountain Region. The
first extends from the Beartooth Mountains in the central Rocky Mountains to the central Sierra Nevada some
1200 km to the southwest. It includes six mountain ranges. The second transect crosses the Mojave Desert from
Olancha Peak in the southern Sierra Nevada to Charleston Peak and thence to San Francisco Peaks in northern
Arizona. The largest numbers of arctic-alpine species are in the Beartooth and Ruby mountains, indicating migra-
tions of these species along the Rocky Mountain cordillera. The lowest numbers of arctic-alpine species are in
the central and western Great Basin and in the Sierra Nevada. Sdrensen's Index of Floristic Similarity was calcu-
lated for all possible pairs of the nine alpine areas. There is little correlation of floristic similarity with alpine
proximity across the Intermountain Region. Rather, any such correlation seems to be in a north-south direction;
this is stronger in the eastern part of the region. Insularity and uniqueness of alpine floras seem to increase to-
ward the western part of the basin. This is probably due to evolution of alpine endemics from preadapted low-
land taxa.

The middle-latitude mountains of North
America north of Mexico, for simplicity and
convenience, may be grouped into four
large systems. With few exceptions, the
mountains in these systems trend north to
south, a fact of considerable importance in
the phytogeography of arctic and alpine
plants. The four systems are the Appala-
chians in the eastern part of the continent,
the Rocky Mountains, the Cascades-Sierra
Nevada, and last, but not least, the Great
Rasin ranges. The latter three systems domi-
nate the western third of the continent.

In general, the mountain ranges, in their
present forms, are younger the closer they
are to the Pacific Coast. The Appalachians
constitute a very old mountain complex dat-
ing from Permian and Triassic times. Much
of the large Rocky Mountain system origi-
nated in the Laramide Revolution in late
Cretaceous and early Paleocene. The oldest
basin ranges also rose during the Laramide
Orogeny, but most of these ranges, particu-
larly in the west, have been upthrust during
a period of time from the Oligocene to the
Pleistocene. Additionally, the whole basin

floor has been uplifted in the Pleistocene.
Orogenic activity continues at present. Even
though the Sierran batholith is rather old,
the present Sierra Nevada is primarily a
product of uplift during the Pliocene and
Pleistocene (Axelrod 1962 and pers. comm.
1973; Rateman and Wahrhaftig 1966). Fossil
lobed oak leaves at 2850 m in the lower al-
pine zone on Elephant's Rack, south of Car-
son Pass, lend additional evidence of recent
Sierran uplift. The high volcanoes of the
Cascades are also of similarly recent age.

Alpine Islands in the Great Rasin

There are some 200 individual mountain
ranges within the Great Rasin. Most of
these trend in a general north to south di-
rection and are separated by broad desert
valleys. In the days when mountain ranges
were shown on maps by hachures, Dutton
described the pattern as similar to an "army
of caterpillars marching to Mexico" (Morri-
son 1965). These basin ranges, with eleva-
tions which vary from about 1800 m to
over 4300 m, are by no means alike either

Department of Botany, Duke University, Durham, North Carolina 27706.

105

106

GREAT BASIN NATURALIST MEMOIRS

No. 2

geologically or botanically. Holmgren (1972)
divided the region, on the bases of floristics
and geology, into four main divisions made
up of 16 sections. The mountain ranges
within any one section have certain charac-
teristics in common but also there are some
rather remarkable ecological differences be-
tween mountains within the same section.
Holmgren notes that there is greater varia-
tion in floristic composition of the alpine
zone across the Great Basin from one peak
to another than occurs in any other zone. I
agree.

For many years, biogeographers working
in mountain areas have compared isolated
mountain ranges and summits to islands
(Wallace 1880, Willis 1922). With increased
interest in island biogeography (MacArthur
and Wilson 1967), several important papers
have appeared which provide quantitative
information on the geographical relation-
ships among the biota on isolated montane
"islands." Notable among these are those of
Hedberg (1970) on the Afroalpine floras, F.
Vuilleumier (1970) on paramo avifaunas of
the northern Andes, B. S. Vuilleumier (1971)
and Simpson (1974) on paramo floras,
Brown (1971) on mountaintop mammals in
the Great Basin, and, recently, Johnson
(1975) on bird species on montane islands in
the Great Basin. MacArthur (1972) also car-
ried his theory over to montane islands in
the Appalachians in his study of the distri-
butions of thrush species. It is notable that
two of these papers (Brown's and Johnson's)
are concerned with the islandlike distribu-
tion of vertebrates in the Great Basin. Nor
have plant geographers ignored the island
nature of the Great Basin montane islands
(P.V. Wells, pers. comm. July 2, 1968, and,
of course, Harper et al. in this symposium).

Both Brown and Johnson have viewed the
Great Basin desert "sea" and its montane is-
lands as being bordered on the east and
west by large cordilleras, the Rocky Moun-
tains and Sierra Nevada, which they desig-
nate as "mainlands" or "continents." The
desert sea is open to the south as far as the
real sea off Mexico. However, to the north,

the desert sea eventually diminishes until it
is blocked by the jumbled mountain masses
of British Columbia which connect the
coastal mountains and the Rockies.

Johnson (1975) used the lower edge of
forest-woodland as the perimeter of his is-
lands. This is a real biological boundary. On
the other hand, Brown (1971), Harper et al.
(this symposium), and I have defined the
lower boundaries of the montane islands
rather arbitrarily. Harper et al. and Brown
have used an elevation of 7500 ft (2286 m)
while I have used 9000 ft (2743 m) as an
approximation of the extreme lower eleva-
tion of alpine plants (Fig. 1). This latter fig-
ure is a somewhat liberal estimate of the
lower limits of alpine islands in the Inter-
mountain Region, but some alpine sites do
exist this low, particularly in cirques. Tim-
berline is usually higher than this and is of-
ten very ragged at its upper limits. Upper
timberline is frequently used as the bound-
ary between subalpine and alpine vegeta-
tion in North America. However, on most
mountains of the earth it is not a particu-
larly good boundary, and it is not a good
boundary on most American mountains ei-
ther, including those of the Great Basin.
Timberlines almost always exist at much
higher elevations than the lowest patches of
alpine vegetation. This is often true around
glacial valleys, both those with and those
without glaciers at the present time. For ex-
ample, timberline around the Athabaska
Glacier in the Canadian Rockies is fully 675
m above the terminus of the glacier, where
in the morainal gravels there are a number
of arctic-alpine plant species but no trees.
The reasons for this ubiquitous phenomenon
are rather simple: those factors which de-
fine the lower limits of alpine species are
not necessarily those which limit the up-
ward distribution of trees.

As Figure 1 indicates, even conservative
alpine islands in the Great Basin are much
smaller and more isolated from each other
than the mountain ranges themselves. But
these alpine islands have been both larger
and smaller in the past than they are at

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

107

120°W

46"N-

■j:r

ftMi BJ V,,^ \i

° c,?c

V
18W

Fig. I. Map of the Intermountain Region showing, in black, areas above 2750 m in elevation. These represent
areas which are wholly or partially "alpine" at the present time. The dashed line at the 1850 m contour is an
approximate estimate of what could have been the lower limit of alpine vegetation at full-glacial. Alpine regions
used in the northern transect are indicated from left to right by letters: P, Piute Pass (Sierra Nevada); W, Pelli-
sier Flats (White Mts.); T, Toiyabe Mts.; R, Ruby Mts.; DC, Deep Creek Mts.; B, Beartooth Mts. Those in the
southern transect are: O, Olancha Pk and two nearby peaks in the southern Sierra Nevada; C, Spring Mts. (Char-
leston Peak); and SF, San Francisco Peaks.

108

GREAT BASIN NATURALIST MEMOIRS

No. 2

present. The areas of such islands during
glacial and hypsithermal times have had a
great deal to do with their present floristics.
For example, Simpson (1974) showed that
the larger sizes of Andean paramos as they
existed at full glacial are more highly and
significantly correlated with numbers of
plant species per paramo than are the
present sizes of each paramo. Also, the dis-
tance between paramos at full glacial is al-
most statistically significant in regard to
present-day floristic richness— but the effect
of present distance between paramos upon
floristic composition is not statistically sig-
nificant.

Present-day alpine climates on the Great
Basin mountains are quite cold during the
winter. In this respect, basin alpine climates
are probably comparable in winter temper-
ature to alpine areas of the Rocky Moun-
tains and the Sierra, although very few al-
pine weather data exist to substantiate this.
In summer, the basin mountains are far
cooler than the desert below. Daytime sum-
mer maximum air temperatures are lowest
on crests and ridges and nighttime minima
are lowest in cirques and canyons; the same
relative conditions exist in the Rocky Moun-
tains and Sierra Nevada.

These Great Basin alpine areas are con-
siderably moister than the desert valleys in
both winter and summer. However, except
for the Ruby Mountains and nearby ranges,
they are drier on an annual basis than al-
pine regions of the Rocky Mountains or the
Sierra Nevada because they receive less
snow. But they do receive enough snow to
form persistent drifts in the lee of cliffs and
ridges. It is such drifts that shorten the al-
pine growing season, keep plants well wa-
tered, and allow the growth of alpine plants
of certain species. The ranges toward the
southwestern part of the basin are relatively
drier than the rest because they lie in the
most extreme part of the Sierran precipi-

tation shadow. There is a trend toward
more winter snow and summer rain in a
traverse across the basin in an easterly di-
rection toward the Rocky Mountains. The
summer rains in the east are usually in the
form of freshening thundershowers, which, I
believe, have much to do with the survival
so far south of populations of certain arctic
species in the Rocky Mountains and eastern
basin ranges.

These mountains were much colder and
wetter during glacial times up to at least
12,000 years BP. Permanent snow was
abundant, and many of the ranges had val-
ley glaciers. These glaciated ranges include
the White Mountains, Toiyabe, Santa Rosa,
Independence, Jarbidge, East Humboldt,
and Ruby Mountains. In the latter range,
valley glaciers emerged from the mountains
onto the now sagebrush-covered plains.
Even far to the south, there were glaciers
on the Spring Mountains2 and San Francisco
Peaks. Even though these glaciers were
small as compared to glaciers and icefields
on the Sierra and in the Rocky Mountains,
they did create cirques which are now the
refugia for a number of alpine species. The
colder, wetter climate also depressed and
telescoped vegetational zones and lowered
timberlines an estimated 600 to 1200 m
(Baker 1970, Loope 1969, Wells and Berger
1969). Even though timberline in itself is
neither the only nor the best indicator of al-
pine conditions, such a depression would
have greatly increased the size of basin
range alpine islands and decreased the dis-
tances between them. This situation would
have increased the chances of establishment
of arctic-alpine species by long-distance dis-
persal and, in certain instances, by direct
migration over tundralike terrain. Since tim-
berline depression has been variously esti-
mated and since it did undoubtedly vary
from one part of the Basin to another, I
have compromised by showing in Figure 1

'In botanical literature, the Spring Mountains are frequently, but incorrectly, called the "Charleston Mountains" due mainly to the title of Clokey's
(1951) flora of the central portion of the range dominated by Charleston Peak.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

109

what a 900 m depression would do to the
sizes and proximity of alpine islands and
continents at full glacial.

Not only have the alpine islands been
larger in the past, they have also been
smaller. During the hypsithermal, about
7500 to 4500 years BP, timberlines of Pinus
longaeva in the White Mountains and the
Snake Range advanced upward to a level
about 100 to 150 m above where they now
stand (LaMarche and Mooney 1967). Such a
long period of warmth and aridity could
have eradicated some alpine species by for-
est shading or by eliminating their snowy
refugia on the higher peaks.

The hard rock geological history of the
Basin Ranges is a complex and varied one. I
shall not go into it here. Suffice to say that
sedimentary rocks, and thus calcareous sub-
strata, are more abundant in the eastern
ranges while the western rocks tend to be
igneous or metamorphic, and yet with some
dolomites and other calcareous types. To
many alpine plants, the chemical natures of
these different kinds of substrates are all-
important in marginal habitats.

Alpine Vegetations and Floras

In any study of vegetation and flora, it is
important to distinguish clearly between
"origin" and "maintenance" of each. Origin
is very difficult to pinpoint, and particularly
so with alpine floras because they leave al-
most no macrofossils. However, the pres-
ence of diploids in widespread polyploid
arctic-alpine species may play the role of
"cytological fossils." The origin of a par-
ticular flora on any mountaintop is the re-
sult of the interaction of many factors: past
geologic events, past climates, migrations of
each species, polyploidy, evolution, and
even man's activities. Hypotheses are nu-
merous; answers are few.

While maintenance of an alpine flora and
its vegetation is somewhat easier to under-
stand, there is still much work involved.
One must measure the characteristics of
x>th the physical and biological aspects of

environment, preferably along meso- and
microtopographic gradients throughout the
year (Billings 1973). Also, it is necessary to
know the tolerance ranges of plant popu-
lations in any particular place. This requires
many field measurements and much labora-
tory experimentation under controlled con-
ditions. We have made only a beginning in
understanding maintenance of some alpine
plant species. Such studies require the inter-
action of many people; they are not one-
person jobs.

Alpine Vegetations

In trying to reach at least a partial un-
derstanding of alpine phytogeography and
ecology in the Great Basin, it is helpful to
start by describing alpine vegetations as
they now exist from the Rocky Mountains
across the basin ranges to the Sierra Ne-
vada. A good approach is to look at these
vegetations along a northeast to southwest
transect from the Beartooth Mountains on
the Montana-Wyoming line to the central
Sierra Nevada. Such a transect crosses many
of the basin ranges, but of particular inter-
est are the Ruby Mountains, the Toiyabe
Range, and the White Mountains. These
represent the eastern, central, and western
basin ranges, respectively.

Quantitative analyses of alpine vegetation
have been made in the Beartooth Mountains
(Johnson and Billings 1962), the northern
Ruby Mountains (Loope 1969), the White
Mountains (Mooney 1973), and the central
Sierra Nevada (Chabot and Billings 1972). I
do not know of any quantitative alpine veg-
etational data from the Toiyabe or Toquima
Ranges; perhaps there are some. The floris-
tic information and photographs in Linsdale
et al. (1952) are of some help, as are per-
sonal qualitative observations which I made
in 1949.

Space does not allow the presentation of
those long tables of vegetational composi-
tion which do exist; one simply can refer to
the publications listed above. However, the
alpine vegetation of the Beartooth Moun-

110

GREAT BASIN NATURALIST MEMOIRS

No. 2

tains is typically Rocky Mountain alpine
with large expanses of alpine tundra in the
true sense: Geum rossii turf in mesic sites,
Deschatnpsia caespitosa meadow in moist
sites, and Carex scopulorum in wet bogs.
The crests and ridges are occupied by
rather dense stands of cushion and rosette
plants with Silene acaulis, Carex elynoides,
and many other species dominating. Early
and late snowbeds are abundant and have
characteristic species surrounding them.

Alpine vegetation in the Ruby Mountains
lacks the broad expanses of tundra charac-
teristic of the Reartooth. However, south of
Lake Peak along the divide there is fairly
extensive tundra at elevations from 3140 m
to 3300 m. This alpine vegetation is charac-
terized by Silene acaulis and Carex pulvi-
nata. Similar alpine vegetation exists at the
same elevation on the north slope of Wines
Peak, with Geum rossii and Silene acaulis
being the dominants. Carex scopulorum
grows in dense stands around alpine ponds.
Some of the best-developed alpine vegeta-
tion in the Rubies is in the cirque floors. In
Island Lake cirque, the vegetation is domi-
nated mainly by Erigeron peregrinus, Salix
arctica, Caltha leptosepala, Geum rossii, Sib-
baldia procumbens, and Polygonum bistor-
toicles. Oxyria digyna is common on rocky,
moist sites, particularly around snowbanks.
One is struck with the remarkable similarity
in species composition and site character-
istics to the alpine vegetation of the Rear-
tooth some 700 km to the northeast. Essen-
tially, the alpine vegetation of the Ruby
Mountains is a small, only slightly attenu-
ated, isolated example of Rocky Mountain
alpine vegetation.

In strong contrast, the alpine vegetation
of the Toiyabe Range, only 250 km south-
west of the Rubies, apparently bears little
resemblance to that of the northern Rubies
or to that of the Rocky Mountains. It seems
to be an open, rocky vegetation with a few
scattered alpine grasses such as Trisctum
spicatum and various species of Draba and
Eriogonum, some of which are endemic.
There is not the great variety of alpine veg-

etation types which one sees in the Rocky
Mountains. More vegetational work is
needed in the small and little-known alpine
areas of the central Great Rasin; further
study may change our ideas of these regions
of what might be termed alpine desert.

Another 150 km farther southwest is the
long and high massif of the White Moun-
tains. In contrast to the Toiyabe and To-
quima Ranges, this has been rather in-
tensively studied environmentally, vege-
tationally, and floristically. The alpine
vegetation has been well-described by
Mooney et al. (1962), Mitchell et al. (1966),
and Mooney (1973). Over most of the exten-
sive alpine area in the White Mountains,
the vegetation is extremely sparse; Mooney
(1973) reports a plant cover of only 1.5 per-
cent at 4175 m on the side of White Moun-
tain Peak. Mitchell et al. (1966) say that
vegetational cover on windy, gravelly flats
on Pellisier Flats, an area of 21 km2 near
the north end of the range, rarely exceeds
15 percent. However, on seepage banks and
meltwater runs, vegetational cover ranges
from 10 to 95 percent. Pellisier Flats at
4100 to 4430 m has active frost polygons
and miniature solifluction steps reminiscent
of the Reartooth Plateau. The most striking
vegetational feature of the White Moun-
tains, however, is the sharp distinction in
vegetation and flora between the dolomite
barrens and granite or quartzite fell-fields.
The dolomite barrens are very desertlike
and, although in places they may have veg-
etational cover up to 12 percent, in other
places there are almost no plants. Phlox cov-
illei and Eriogonum gracilipes are character-
istic species on the dolomite. A granite fell-
field at 3870 m had a vegetational cover of
50 percent and was dominated by Trifolium
monoense and Koeleria cristata; the first is
essentially endemic to the White Mountains
and the second is a cosmopolitan species.
The alpine vegetation of the White Moun-
tains is decidedly unlike that of the Rubies—
but both are Basin ranges.

There have been several vegetational
studies made on alpine vegetation in the

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

111

Sierra Nevada. These vary considerably be-
tween the northern end of the range and
the southern, where the peaks are higher
and the climate drier. Only 48 km south-
west of the White Mountain crest and with-
in sight of it, is Piute Pass at 3540 m on the
Sierran white granite. Here, Chabot and
Billings (1972) found the low alpine vegeta-
tion covering less than 30 percent of the
rocky ground and characterized by Lupinus
breweri, Antennaria rosea, and Carex helleri.
Other common Sierran alpine species in
similar locations on nearby peaks and ridges
are Phlox caespitosa, Penstemon davidsonii,
Ivesia pygmaea, and Draba lemmonii. At
higher elevations, Oxyria digyna, Polerno-
nium eximium, and Hulsea algida are con-
stant members of the alpine vegetation usu-
ally on scree slopes. The most luxuriant
vegetation near here is along snowbank
meltwater brooks where Dodecatheon jef-
freyi, Lewisia pygmaea, and Sedum rosea
occur. On very thin granitic soils which dry
out after snowmelt, only Calyptridium um-
bellatum, the dwarf Koenigia-like endemic
Polygonum minimum, and one or two other
species can exist. There is a great difference
in alpine vegetation here, not only from
that of the Beartooth but also from that of
the White Mountains on the near horizon.

Alpine Floras

While there are a few floras for whole
mountain ranges between the Bocky Moun-
tains and the Sierra, e.g., Lloyd and Mit-
chell (1973) for the White Mountains, the
main sources of alpine floras per se appear
to be journal papers or theses. Beferring
again to Figure 1, I have assembled at least
tentative figures on the numbers of alpine
species for the same NE-SW transect along
which the trend in alpine vegetation has
been described. These figures will certainly
be changed with additional collecting and
publication, particularly of more volumes of
Cronquist et al. Intermountain Flora and
Howell's proposed Sierran Flora. The
sources are the Beartooth Mountains (John-
son and Billings 1962), the Deep Creek

Mountains (McMillan 1948), the Buby
Mountains (Loope 1969), the Toiyabe Bange
(Linsdale et al. 1952), the White Mountains
(Mitchell et al. 1966), and the central Sierra
Nevada (Chabot and Billings 1972). Addi-
tionally, I have included floristic data from
a series of three isolated alpine areas across
the southern desert region. From west to
east these are Olancha Peak and its neigh-
bors in the Sierra Nevada (Howell 1951),
the Spring Mountains (Clokey 1951), and
San Francisco Peaks (Little 1941).

Many alpine plant species also occur in
the Arctic. As a basis for whether or not
our species do, I have checked each against
Polunin's (1959) Circumpolar Arctic Flora.
Species which are not arctic-alpine may be
endemic to a continental region of the
middle-latitudes or to the mountain range
itself or its near neighbors. While we need
much more information on both scores, the
absolute figures on arctic and endemic spe-
cies in each small alpine area can tell us
something about migrations into an area
and possibly about the evolution of new
species after such migrations from afar or
from the arid regions below. Nor should we
forget that some endemic species may be
relicts or that some widespread arctic-alpine
species may have become extinct during-
xerothermic times, or that some may never
have reached certain alpine areas because
of lack of adaptation for long-distance dis-
persal.

The upper part of Table 1 (the
Beartooth-Sierra transect) illustrates immedi-
ately that there are many alpine species in
the central Bocky Mountains that also occur
in the Arctic. In the Beartooth Mountains,
91 out of 194, or almost half the alpine
flora, also occurs in the Arctic. This is in
contrast to the 24 arctic-alpine species in
the Deep Creeks and 47 in the Bubies; but
these are somewhat smaller ranges. The
composition of the alpine floras of both of
the latter ranges, however, is very much a
Bocky Mountain-type flora. In contrast,
west of the Buby Mountains there is a dra-
matic drop in not only the total numbers of

112

GREAT BASIN NATURALIST MEMOIRS

No. 2

alpine species in the Toiyabe, the White
Mountains, and the Sierra Nevada, but also
in the number of arctic-alpine taxa present.
The arctic-alpine element consists, in these
ranges, mostly of such easily dispersed,
ubiquitous species as Oxyria digyna, Trise-
tum spicatum, Cystopteris fragilis, and An-
drosace septentrionalis. Even common spe-
cies such as Silene acaulis are missing, not
to mention relatively rare taxa such as Koe-
nigia islandica, Saxifraga flagellaris, or
Phippsia algida. Widespread arctic-alpine
species such as Silene acaulis, Polygonum
viviparum, and Salix arctica reach their
western limits in the Great Basin in the
Ruby Mountains and do not reappear in the
Sierra Nevada (Loope 1969). Since these
species are not that particular in their envi-
ronmental requirements (there are apparent
places in the Sierra Nevada where they
could grow), they either must be poorly
adapted to long-distance dispersal or there
is or has been an advantage to migrating
north or south along the old tried and true
Rocky Mountain pathway. Those arctic-
alpine species in the Sierra Nevada most
likely came down from the north in glacial
or cooler times during the Pleistocene. Mi-

gration across the Great Basin appears to
have been a chancy thing even at full-
glacial. Only species with propagules easily
dispersed by wind or birds appear to have
made it to the central Great Basin moun-
tains. Even after possible establishment,
some alpine species may have become ex-
tinct on the Great Basin peaks in dryer,
postglacial times. Axelrod (1976) also sug-
gests extinctions of moist environment sub-
alpine conifers and alpine species in the
Great Basin during these xerothermic epi-
sodes. I believe that extended post-glacial
dry periods also could have eliminated such
dry-site arctic-alpine species as Silene
acaulis and Saxifraga cespitosa in the Sierra
Nevada where they do not occur today.
This suggestion, of course, must assume that
these species migrated into the Sierra Ne-
vada, probably from the north, during fa-
vorable Pleistocene times.

Looking at the southern transect in Table
1, the same dearth of arctic species is ap-
parent in the Olancha Peak group at the
southern end of the Sierra and particularly
so on Charleston Peak in the isolated Spring
Mountains where only 9 arctic species
are known to occur. In contrast, only

Table 1. Total numbers of true alpine species and those which also occur in the Arctic (arctic-alpine) in two
transects across the Intermountain Region. Data from several sources.

Location

Total No.

of Alpine

Species

No. of

Arctic-Alpine

Species

Percent of

Arctic-Alpine

Species

Northern Transect
Beartooth Mts.

Deep Creek Mts.

Ruby Mts.

Toiyabe Range

White Mts.

(Pellisier Flats)

Piute Pass
(Sierra)

Southern Transect
Olancha Pk., etc.

(Sierra)
Spring Mts.
San Francisco Pks.

194

80

102

91

47

24

30

47

25

11

23

10

21

4

10

13

13

9

23

19

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

113

350 km to the southeast, the San Francisco
Peaks have at least 19 arctic species in spite
of their isolation. Also, there is a distinct
Rocky Mountain cast to their alpine flora.
A possible pathway for migration into this
Pleistocene volcanic mountain may have
been at full-glacial through the Arizona
White Mountains to the southeast, or the
source may have been through the Wasatch
Plateau to the north. Again, we come to the
origin versus maintenance problem: these
arctic species would not remain in the vol-
canic cinders or the glacial cirques of this
isolated mountain, no matter how they got
there, except for the summer thunder-
showers reinforcing the winter snows as a
source of water.

Much more poorly known are the endem-
ics of all of the above mountain ranges. Al-
pine endemics are much less common in the
central Rocky Mountains than in the Sierra
Nevada. This may be due to lack of isola-
tion along the Rocky Mountain system. The
Sierra Nevada as a whole has many endem-
ic alpine species but I do not know how
many of these may be narrow endemics re-
stricted to the two small regions listed in
Table 1. The White Mountains have several
alpine endemics, and since they cover so
much less area than the Sierra Nevada, the
chances of these being narrow endemics are
better than in the latter range. The same
fact holds true for the isolated ranges of the
central Great Basin. Some of these latter
endemics may also prove not to be so nar-
row, except edaphically, upon further ex-
ploration. For example, consider Primula
nevadensis (Holmgren 1967). The isolated
Spring Mountains, with an alpine flora of
only 39 species at most (some of these are
probably subalpine), has at least 4 endemic
alpine species. It seems that alpine endem-
ism tends to be more common in relation to
total alpine floras on isolated, nonvolcanic
peaks and ranges in the southwestern part
of the Intermountain Region; this includes
the southern part of the Sierra Nevada. The
real answer to the question of distribution
of endemics lies in more collecting in alpine

environments and systematic description of
the taxa.

One way of comparing alpine floras is to
use Srirensen's (1948) Index of Floristic Sim-
ilarity. This is expressed as:

IS.

X 100

1/2 (A + B)

where, ISS = Sdrensen's Index of
Similarity

c = number of species
held in common
by two alpine areas

A = total number of
alpine species
in Region A

B = total number of
alpine species
in Region B

Taking the available alpine floristic data
from each of the nine alpine areas on both
transects, I calculated SeSrensen's Indices for
each possible comparison. These indices are
presented as percentage values in Table 2.
Also, in the same table, direct distances in
km between the alpine areas are shown.
When the Indices of Similarity are plotted
against distance for each pair of alpine
areas, the result is a great deal of scatter in
the points. There is little evidence of in-
verse correlation of floristic similarity be-
tween these alpine regions with distance.
Out of 36 possible combinations, only 4
show indices above 30 percent combined
with distances below 200 km. These are the
Deep Creek— Ruby Mountains in the east-
central Great Basin and Pellisier Flats-
Olancha Peak, Pellisier Flats-Piute Pass, and
Piute Pass-Olancha Peak, all in the Sierra
Nevada- White Mountain complex at the ex-
treme southwestern edge of the Great Basin.
Two others had indices almost equally as
high but at distances between 500 and 700
km; these are the Deep Creek Mountains-
San Francisco Mountains and Beartooth-
Ruby Mountains combinations. The most
distant range, the Beartooth, has higher in-
dices of similarity with the Ruby Mountains,
Deep Creek Mountains, and San Francisco

114

GREAT BASIN NATURALIST MEMOIRS

No. 2

Mountains alpine floras at distances of 650
to 1100 km than any of these latter ranges
have with the Spring Mountains at distances
less than 500 km. The distant Rocky Moun-
tains alpine flora as exemplified by that of
the Beartooth has as great or greater an in-
fluence on those of most of the Great Basin
mountain ranges than does that of the much
closer Sierra Nevada. There is even a great-
er (two to three times) similarity between
the alpine floras of San Francisco Moun-
tains and the Deep Creek Mountains at a
distance of 531 km than there is between
the Deep Creek flora and those of the
Sierra Nevada and White Mountains, which

are equally distant from the Deep Creek
Mountains.

The rather low indices of similarity in-
dicate at least one fact rather clearly: these
Intermountain Region alpine areas are very
much like newly isolated islands. This is
well illustrated along the southern transect
from Olancha Peak to Charleston Peak to
San Francisco Peaks. With the two gaps
being only 225 and 378 km across, respec-
tively, the indices of similarity are only 13
and 14 percent. The Olancha Peak group is
clearly Sierran, Charleston Peak is unique,
and San Francisco Peaks show strong rela-
tionships to the Rocky Mountains and

Table 2. Sdrensen's indices of floristic similarity (in percent) between alpine floras for all combinations of al-
pine areas in both transects. Numbers in parentheses are distances (in km) between the alpine regions. Also, see
map in Fig. 1.

Nortl

lern Transect

Pellisier

Piute

Bear-

Deep

Ruby

Toiyabe

Flats

Pass

tooth

Creeks

Mts.

Range

(White Mts.)

(Sierra)

Beartooth

_

24%

33%

14%

14%

9%

Deep Creeks

(676)

-

39%

22%

19%

10%

Ruby Mts.

(692)

(153)

-

20%

14%

11%

Toiyabe Range

(917)

(322)

(257)

-

21%

16%

Pellisier Flats

(White Mts.)

(1102)

(451)

(418)

(145)

-

34%

Piute Pass

(Sierra)

(1167)

(515)

(483)

(217)

(72)

-

Olancha Pk. Group

(Sierra)

(1223)

(547)

(539)

(290)

(169)

(129)

Spring Mts.

(1110)

(434)

(475)

(325)

(298)

(306)

San Francisco Pks.

(1086)

(531)

(668)

(636)

(660)

(668)

South

em Transect

Olancha Pk.

Spring

San Francisco

Group (Sierra'

\

Mts

Pks.

Beartooth

14%

8%

23%

Deep Creeks

15%

13%

31%

Ruby Mts.

14%

8%

19%

Toiyabe Range

13%

18%

21%

Pellisier Flats

White Mts.)

36%

16%

19%

Piute Pass

(Sierra)

32%

10%

4%

Olancha Pk. Croup

(Sierra)

-

13%

13%

Spring Mts.

(225)

-

14%

San Francisco Pks.

(603)

(378)

-

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

115

mountain ranges far to the north. These af-
finities of the San Francisco Peaks alpine
flora have also been noted by Moore (1965).

After examining the existing floristic data
on alpine species across the Great Basin, it
is tempting to delineate some alpine phyto-
geographic boundaries. Certainly, there is a
sharp boundary just west of the Ruby
Mountains separating Rocky Mountain types
of vegetation and flora from that of the
Great Basin. Where this boundary goes to
the south, I am not prepared to say. Cer-
tainly, it would appear to be west of the
Snake Range and perhaps west of the Grant
Range; but this is tentative and probably
premature. It obviously lies between San
Francisco Peaks and the Spring Mountains.

Another alpine phytogeographic boundary
lies between the White Mountains and the
Sierra Nevada and winds north in sinuous
curves until it divides the Carson Range on
the west from the Pine Nut Range on the
east; but several species pay no attention to
such an imaginary line and go their way
from the Sierra Nevada into the altered an-
desites of the Virginia Range (Billings 1950)
and the neve cirques of the Pine Nut and
Wassuk Ranges (Billings 1954). Again, if
edaphic or moisture conditions are
right,some species can get to seemingly in-
hospitable mountain ranges and stay there
perhaps by infraspecific physiological adap-
tations (i.e., ecotypes).

Possible Origins of
Intermountain Alpine FlorAs

Alpine floras originate in all mountains
by migration and/or evolution; the moun-
tains of the Intermountain Region are no
exception. Refugia, both glacial and inter-
glacial, are also important. As Figure 1
shows, and as Simpson (1974) has demon-
strated for the paramos, successful long-
distance dispersal is greatly influenced by
the sizes and proximity of alpine islands.
Such islands were obviously larger and
closer during glacial and cooler times. From
distributional evidence, only a relatively few

arctic-alpine species appear to be able to
migrate far: Oxijria digyna, Saxifraga
cespitosa, Trisetum spicatum, and Cys-
topteris fragilis, for example. Therefore, it is
advantageous for ancestral species to be
nearby when mountains arise or new alpine
areas come into existence by climatic
causes. It also seems to be advantageous to
be on a north-south migration route such as
the old Rocky Mountains, or even the much
younger Cascade-Sierran system. The ability
to evolve rapidly by adaptive radiation is
also advantageous, as for example, in the
genus Espeletia in the Andean paramos.

Such plasticity, plus pure luck in being
near refugia, either edaphic or micro-
climatic, also helps. Interglacial refugia are
fully as important as glacial refugia. Such
interglacial refugia can be mountaintops
such as Mount Washington, New Hamp-
shire, altered volcanic rocks as in the Vir-
ginia Range of western Nevada (Billings
1950), alpine snowbanks, and the moist,
cool floors of impervious cirques (Loope
1969). Loope ascribes the presence of many
arctic species in the Ruby Mountains today
to its nonporous (moist) cirque floors as
compared to the porous, dry cirque floors
of the Jarbidge Mountains to the north and
those of the Snake and Schell Creek ranges
to the south.

There are other sources of alpine floras in
the Intermountain Region. These are the
surrounding desert floras as suggested long
ago by Went (1948) and experimentally in-
vestigated by Klikoff (1966) and by Chabot
and Billings (1972) for the Sierra Nevada.
The principal findings reported in the latter
paper would apply almost as well to the
mountain ranges of the Great Basin, provid-
ed there are suitable sites for such migra-
tion and evolution, and providing there are
species and genera of sufficient genetic
plasticity near at hand. The derivation of a
new alpine flora from a desert or semidesert
flora is aided by the following:

a. preadaptation of winter annuals and
perennials in regard to physiology and
morphology,

116

GREAT BASIN NATURALIST MEMOIRS

No. 2

b. the ability to migrate upward during
favorable climatic periods into new
but not altogether dissimilar environ-
ments,

c. selection for metabolism at low sum-
mer temperatures,

d. selection for low temperature starch
degradation and sugar translocation at
night,

e. selection of populations and ecotypes
which acclimate metabolic-ally rapidly
and ideally, and

f. selection for flowering and seed-set in
short, dry cool growing seasons.

These have been elaborated in detail by Bil-
lings (1974).

There is every reason to believe that such
upward evolution of new "alpine-desert"
taxa is taking place in the Great Basin. But
the rate may be slower than in the Sierra
Nevada due to a smaller preadapted flora
and also due to the drier and less snow-pro-
tected environments on the peaks. In such a
case, there could be a trend toward edaphic
endemism because other kinds of habitat di-
versity are in relatively short supply. Such
upward mobility and long-distance dispersal
over great distances in a north-south direc-
tion seem to be the principal sources of the
alpine floras of these isolated montane is-
lands in the central Great Basin.

Acknowledgments

This work has been aided by National
Science Foundation Grant BMS72-02356
A01 (General Ecology) for which I express
my appreciation. Many people have helped
with data and discussion; I thank particu-
larly Lloyd Loope, Brian Chabot, Juliana
Mulroy, Gaius B. Shaver, and Shirley Bil-
lings.

Literature Cited

\\mh(»1). 1). I. 1962. Post-Pliocene uplift of the
Sierra Nevada, California. Bull. Geol. Soc.
Amer. 73: 183-198.
1976. Historv of the coniferous forests, Cali-
fornia and Nevada. Univ. Calif. Pub. Bot. 70:
1-62.

Baker, R. G. 1970. Pollen sequence from Late Qua-
ternary sediments in Yellowstone Park. Science
168: 1449-1450.
Bateman, P. C, and C. Wahrhaftig. 1966. Geology
of the Sierra Nevada, pp. 105-172. In: E. H.
Bailey (ed.), Geology of northern California. Ca-
lif. Div. Mines and Geol. Bull. 190: 1-508.
Billings, W. D. 1950. Vegetation and plant growth
as affected by chemically altered rocks in the
western Great Basin. Ecology 31: 62-74.

1954. Nevada Trees, 2d ed. Univ. Nevada

Agric. Extens. Serv. Bull. 94: 1-125.
1973. Arctic and alpine vegetations: sim-
ilarities, differences, and susceptibility to dis-
turbance. BioScience 23: 697-704.

1974. Adaptations and origins of alpine

plants. Arctic and Alpine Res. 6: 129-142.
Brown, J. H. 1971. Mammals on mountaintops:
nonequilibrium insular biogeography. Amer.
Naturalist 105: 467-478.
Chabot, B. F., and W. D. Billings. 1972. Origins
and ecology of the Sierran alpine flora and veg-
etation. Ecol. Monogr. 42: 163-199.
Clokey, I. W. 1951. Flora of the Charleston Moun-
tains, Clark County, Nevada. Univ. Calif. Publ.
Bot. 24: 1-274.
Hedberg, O. 1970. Evolution of the Afroalpine

flora. Biotropica 2: 16-23.
Holmgren, N. H. 1967. A new species of primrose
from Nevada. Madrono 19: 27-29.

1972. Plant geography of the Intermountain

Region, pp. 77-161. In: A. Cronquist, A. H.
Holmgren, N. H. Holmgren, and J. L. Reveal
(eds.), Intermountain flora. Vol. 1. Hafner Publ.
Co., New York.
Howell, J. T. 1951. The arctic-alpine flora of three
peaks in the Sierra Nevada. Leafl. W. Bot. 6:
141-156.
Johnson, N. K. 1975. Controls of number of bird
species on montane islands in the Great Basin.
Evolution 29: 545-567.
Johnson, P. L., and W. D. Billings. 1962. The al-
pine vegetation of the Beartooth Plateau in re-
lation to crvopedogenic processes and patterns.
Ecol. Monogr. 32: 105-135.
Klikoff, L. G. 1966. Temperature dependence of
the oxidative rates of mitochondria in Dan-
thonia intermedia, Penstemon davidsonii, and Si-
tanion hystrix. Nature 212: 529-530.

LaMahche, V. C, and H. A. Mooney. 1967.
Altithermal timberline advance in western
United States. Nature 213: 980-982.

Linsdale, M. A., J. T. Howell, and J. M. Lins-
dale. 1952. Plants of the Toiyabe Mountains
area, Nevada. Wasmann J. Biol. 10: 129-200.

Little, E. L., Jr. 1941. Alpine flora of San Fran-
cisco Mountain, Arizona. Madrono 6: 65-96.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

117

Lloyd, R. M., and R. S. Mitchell. 1973. A flora of
the White Mountains, California and Nevada.
University of California Press, Berkeley.

Loope, L. L. 1969. Subalpine and alpine vegetation
of northeastern Nevada. Unpublished doctoral
dissertation, Duke University, Durham, North
Carolina.

MacArthur, R. H. 1972. Geographical ecology:
patterns in the distribution of species. Harper
and Row, Publications, New York.

MacArthur, R. H., and E. O. Wilson. 1967. The
theory of island biogeography. Princeton Uni-
versity Press, Princeton, New Jersey.

McMillan, C. 1948. A taxonomic and ecological
study of the flora of the Deep Creek Mountains
of central western Utah. Unpublished master's
thesis, University of Utah, Salt Lake City.

Mitchell, R. S., V. C. LaMarche, Jr., and R. M.
Lloyd. 1966. Alpine vegetation and active
frost features of Pellisier Flats, White Moun-
tains, California. Amer. Midi. Naturalist 75:
516-525.

Mooney, H. A. 1973. Plant communities and vege-
tation, pp. 7-17. In: R. M. Lloyd and R. S. Mit-
chell (eds.), A flora of the White Mountains,
California and Nevada. University of California
Press, Berkeley.

Mooney, H. A., G. St. Andre, and R. D. Wright.
1962. Alpine and subalpine vegetation patterns
in the White Mountains of California. Amer.
Midi. Naturalist 68: 257-273.

Moore, T. C. 1965. Origin and disjunction of the
alpine tundra flora on San Francisco Mountain,
Arizona, Ecology 46: 860-864.

Morrison, R. B. 1965. Quaternary geology of the
Great Basin, pp. 265-285. In: H. E. Wright, Jr.
and D. G. Frey (eds.), The Quaternary of the
United States. Princeton University Press,
Princeton, New Jersey.

Polunin, N. 1959. Circumpolar Arctic Flora. Cla-
rendon Press, Oxford.

Simpson, B. B. 1974. Glacial migrations of plants:
island biogeographical evidence. Science 185:
698-700.

Sorensen, T. 1948. A method of establishing groups
of equal amplitude in plant sociology based on
similarity of species content. Kongel. Danske
Vidensk. Selsk. Skr. (Copenhagen) 5(4): 1-34.

Vuilleumier, B. S. 1971. Pleistocene changes in the
fauna and flora of South America. Science 173:
771-780.

Vuilleumier, F. 1970. Insular biogeography in con-
tinental regions. I. The northern Andes of South
America. Amer. Naturalist 104: 373-388.

Wallace, A. R. 1880. Island life. MacMillan and
Co., London.

Wells, P. V., and R. Bercer. 1967. Late Pleisto-
cene history of coniferous woodland in the Mo-
jave Desert. Science 155: 1640-1647.

Went, F. W. 1948. Some parallels between desert
and alpine flora in California. Madrono 9:
241-249.

Willis, J. C. 1922. Age and area: a study in geo-
graphical distribution and origin of species.
Cambridge University Press, Cambridge.

PHYTOGEOGRAPHICAL VARIATION WITHIN JUNIPER-PINYON
WOODLANDS OF THE GREAT BASIN

Neil E. West1, Robin J. Tausch', Kenneth H. Rea1, and Paul T. Tueller2

Abstract.— About 22 percent of the pigmy conifer woodlands of the United States occur within the Great Ba-
sin. Only a very few reports of these woodlands exist in the literature. Available reports are either of general de-
scriptive nature or specific analysis of vegetation-environmental relationships on one mountain range. In order to
better understand basin-wide synecological patterns, a cooperative study was carried out by personnel at Utah
State University and the University of Nevada-Reno between 1972 and 1975. Vegetation, landform, geology, and
soils data obtained from 463 systematically placed stands on a randomly chosen set of 66 mountain ranges have
been used to derive patterns of latitudinal, longitudinal, and altitudinal variation in the floristic diversity in juni-
per-pinyon dominated woodlands across the Great Basin.

The latitudinal-longitudinal patterns show greatest environmental and floristic diversity on the higher mountain
ranges on the southern end of the Central Plateau portion of the study area where the Great Basin-Mojave
Desert transition occurs. This is also where the elevational breadth of the woodland belt is greatest. Juniper-
pinyon woodlands are largely lacking from northwestern Nevada. The lowest elevations for the type are found in
the Dixie Corridor centered in southwestern Utah. The general elevation of these woodlands is highest in the
west-central part of the Great Basin and declines both toward the Sierra Nevada on the west and the Wasatch
Front-High Plateaus on the east.

Use of the equilibrium theory of island biogeography gave incomplete explanations of the diversity patterns ob-
served. Certain conceptual and methodical problems forced by this overly simplistic theory are discussed. The
best correlations obtained were between species richness and an index of ecotopic diversity.

Correlation of Basin-wide patterns of woodland floristics with surficial geology, landforms, and soils is non-
discriminatory. However, broad-scale, phytogeographical variations in these woodlands are closely associated with
climatic differences. Although much direct climatic data are lacking, it seems likely that the relative contribu-
tions of the transitory frontal systems moving inland from the Pacific, continental cyclones developing over the
Great Basin, and convectional storms associated with the moist air from the Gulf of Mexico to induce precipi-
tation at different seasons are regionally important in the causation of vegetation distribution and composition.

The instability of temperature inversions is a likely determinant of the position of woodlands along the north-
ern boundary of the type. The Pacific frontal systems break the inversions most readily and are thought to be the
major cause for the lack of this vegetation in northwestern Nevada and on exposed mountain ranges along the
northern boundary of the type. Such observations provide leads for relevant ecophysiological research to support
or reject these notions.

The juniper-pinyon woodlands are a ma- siderations of other Great Basin vegetation

jor vegetation type of the Intermountain types and their distributions have been

West. West, Rea, and Tausch (1975) have made by Billings (1951), Cronquist et al.

indicated that about 325,000 km2 (125,000 (1972), Tueller (1975), and Young, Evans,

mi2) are involved. About 7.1 million ha, or and Tueller (1976). They all characterized

22 percent of this total, occur in the Great these woodlands as generally covering low

Basin. It is the phytogeography of this por- hills or as forming a belt of vegetation at

tion of the type that we wish to consider lower elevations on the higher mountains

here. with sagebrush-dominated belts both above

Brief descriptions of the Great Basin and below,

pinyon-juniper woodlands, along with con- Beeson (1974) mapped the typed and de-

•Department of Range Science, Utah State University, Logan, Utah 84322. Present address of Rea: H-8 Group Office, Environmental Studies, Los
Alamos Scientific Laboratory, University of California, Los Alamos, New Mexico 87544.
'Renewable Resources Center, University of Nevada, Reno, Nevada 89507.

119

120

GREAT BASIN NATURALIST MEMOIRS

No. 2

scribed the general longitudinal and latitu-
dinal variations from a large set of vegeta-
tional data collected on approximately one-
third of the mountain ranges found in the
Great Basin. In this paper we wish to exam-
ine further aspects of this and additional
data, expanding the discussion of longitudi-
nal, latitudinal, and elevational floristic di-
versity (beta diversity, according to Whitta-
ker 1975) and propose possible explanations
for some peculiarities of latitudinal limits
and variations in elevational extent of the
type.

Methods

Data set collection and prior analysis:
The data set available involves quantitative
data on landform, geology, soils, and vege-
tation collected between 1972 and 1975 for
463 stands systematically placed on a ran-
domly chosen set of 66 Great Basin moun-
tain ranges (Nabi 1978).

We defined woodlands as having at least
25 pinyon and/or juniper trees per hectare
(10/acre). At least one tree had to be in our
mature size-age-form class. These criteria
kept the samples from extending too far
into ecotones, yet allowed a good coverage
of the main juniper-pinyon woodland type.

Stands were sampled at regular elevation-
al intervals on all of the four major expo-
sures where the juniper-pinyon belt existed
on each mountain range. We are therefore
able to discuss the Basin-wide distribution
of these woodlands from fairly objective
bases. The full details of the sampling de-
sign, data collection, and prior analysis are
given elsewhere (Nabi 1978).

Supplementary sampling of the Shoshone
Mountains added stands in broad canyon
bottoms and on secondary slopes of faces
along major ridges. These additions were
made to sample for one area a greater
range of ecotopic heterogeneity than was
provided by the regularly placed upland
sites on exposures in cardinal directions.

In addition to presentation of broad,
longitudinal-latitudinal patterns of woodland

distribution via a map derived from field-
checked space photography, Beeson (1974)
grouped the sampled mountain ranges into
three first-order divisions, based on floristic
composition. The stress was on the major
species found on at least 25 percent of the
mountain ranges sampled. These major spe-
cies were commonly found in most of the
stands on the mountain range where they
occurred.

This approach yielded one group of
ranges with moderate floristic diversity in
the Central Plateau region that had consid-
erable similarities to the foothills and ranges
directly adjacent to the Sierra Nevada on
the west and the Wasatch Front-High
Plateaus of central Utah on the eastern
boundary of the basin. Another grouping
with low floristic richness was designated
for all lower elevation ranges occurring in
the generally lower, drier portions of the
Great Basin. The third grouping consisted
of ranges with the highest floristic diversity.
These all occurred along the southern end
of the Central Plateau where the Great Ba-
sin-Mojave desert transition occurs.

These floristic units were related to pat-
terns of surficial geology, landform, soils,
and climate. Weak correspondences were
found at this scale with all environmental
factors except climate. The widescale longi-
tudinal-latitudinal differences observed were
therefore thought to be largely related to
current patterns of precipitation and tem-
perature.

Current analysis methods: Mueller-
Dumbois and Ellenberg (1974), in their re-
cent textbook on the Aims and Methods of
Vegetation Ecology, emphasize that a com-
plete understanding of the distribution of
plant communities involves a consideration
of flora, accessibility, ecological properties
of the plants, habitat, and time. Our further
analysis here involves an in-depth exam-
ination of the total floristic diversity in our
samples of Great Basin pigmy conifer wood-
lands in an attempt to relate these floristic
diversity patterns to likely paleocological
influences and to present environmental

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

121

heterogeneity with emphasis on the present
climatic patterns through space and time.

We also examined the appropriateness of
applications of island biogeographical theo-
ry (Simberhoff 1974). These analyses fol-
lowed a strategy similar to that used by
Brown (1971) and Johnson (1975). Data
from the 18 mountain ranges sampled at
200 m elevation intervals were used (Nabi
1978). The variables accounted for were to-
tal woodland species per mountain range,
total mountain area, total mountain area oc-
cupied by woodland (taken from our pre-
viously published map), mountain height,
width of barrier, and an index of ecotopic
diversity.

The width of the barrier was calculated
similarly to Johnson's (1975) approach ex-
cept that we considered distance to the
nearest of three "continental" areas: The
Sierra Nevada, Wasatch Front-High
Plateaus of central Utah, and the northern
boundary of the Snake River Plains, all de-
fined as in Cronquist et al. (1972).

The ecotopic diversity where woodlands
occurred on each mountain was indexed by

the use of an arbitrary, numerical scale
(Table 1) taking into account those variables
on which we had direct data. The higher
the variety of factor conditions, the greater
the presumable ecotopic diversity. Untrans-
formed data for the variables assessed are
presented in Table 2. Correlation
coefficients (r) between the variables were
then calculated for both the raw and loga-
rithmically transformed data.

Additional tabular and graphical organi-
zations of the data were employed in order
to clarify our broad-scale phytogeographical
view of latitudinal, longitudinal, and altitu-
dinal variation in these woodlands. A series
of range-by-range comparisons were made
on the basis of total species richness for
each range, average stand species richness,
and altitudinal distribution of the woodland.
Each type of analysis was mapped and
compared with each other and the broad-
scale climatic patterns for the region.

Results and Discussion

Overall species richness: A total of 367
vascular plant species were found within

Table 1. Point system for ecotopic diversity score.

Range of Range of

Values Values

Possible Observed

Exposure Cardinal slopes having woodlands

1 = north only

2 = north and east, etc.

Slope Number of slope classes (10% incre-

ments) 0-10% = 1, 0-10% plus 11-20
= 2, etc.

Elevation Elevations at which woodlands were

sampled (200 m intervals) 1800-2000 m
= 1: 1800-2000 m plus 2001-2200 m =
2, etc.

Landform Five major landforms encountered

(valley, foothill, bajada, terrace,
mountain) (defined in Nabi 1978)

1-5

1-4

1-5

2-5

1-5

Geology
Soils

classified to clan (Nabi 1978)

Classified to subgroup according to Soil
Survey Staff, USDA, Soil Conservation
Service (1951, 1960, 1967)

Total possible score

1-19
1-24

H\

1-8
1-16

(10-40)

122

GREAT BASIN NATURALIST MEMOIRS

No. 2

the juniper-pinyon woodlands sampled. The
number of species in the woodlands on indi-
vidual mountain ranges varied from 164
species on the Shoshone Mountains of west-
central Nevada to under 30 species on small
isolated ranges of western Utah. Figure 1
shows that most of the mountain ranges had
fewer than 50 species in their woodlands.
Furthermore, the ranking of species num-
bers per mountain range frequencies (Fig. 2)
shows an expected log-normal distribution
over the sampled ranges (May 1975). Many
species were found at only one stand on a
given mountain range. If the species fre-
quency sequence is plotted against the num-
ber of stands in which they occur (Fig. 3)
the right hand portion of the curve also fol-
lows a log-normal form. The frequency val-
ues for the rare species (those occurring in
one to four stands) are in excess of the ex-
pected value (Fig. 4). These high frequen-
cies of rarer species are indicative of an
over-saturation of relictual species probably
remaining from the vegetation migrations
resulting from the more favorable climate
of the recent geologic past. This is a situa-
tion similar to that found by Brown (1971)

for small mammals and Johnson (1975) for
permanent resident birds in the same area.
Our estimates of such oversaturation are
conservative since sampling was restricted
to upland areas on all ranges except the
Shoshone Mountains, which included some
broad drainage bottoms and major ridge
faces. Judging from the high species density
of the Shoshone Range, inclusion of more
pronounced topographic situations such as
narrow drainage channels and searches of
atypical geological outcrops and unusual
landforms on the other ranges, in addition
to our predetermined sampling scheme,
would have yielded more plant taxa within
the woodland belt.

If the average number of species per
stand within a given mountain range is
plotted against the total number of species
found in the woodlands of that range (Fig.
5) the envelope including the data points il-
lustrates that there is a very general in-
crease in stand diversity with increasing flo-
ristic diversity of the ranges sampled. The
variation, however, is high. Except for the
ranges with greatest and least diversity,
there are few mountain ranges fitting with-

Table 2. Untransformed data used to test appropriateness of equilibrium theory of island biogeography.

Pigmy Conifer

Total Area

Total Area

Width of

Ecotopic

Woodland

of

of

Mountain

Smallest

Diversity

Mountain

Species

Mountain1

Woodland

Height

Barrier

Score

Range

Encountered

(km2)

(km2)

(m)

(km)

(See Table 1)

Burbank Hills

19

43

43

2384

.36

10

Confusion

32

145

145

2461

39

1.3

Pine Valley

37

143

143

1989

10

18

Black Pine

38

458

113

285.3

30

16

Toana

38

286

263

2117

31

19

White

40

1691

1305

3091

10

19

Goose Creek

42

1691

1305

2295

28

14

Tushar

46

1412

706

3713

10

18

Pilot

48

321

293

3265

20

24

Excelsior

48

187

187

2532

10

13

East Humboldt

49

592

.54

2456

42

18

Schell Creek

.54

152.3

1076

3353

12

22

Mineral

65

298

187

2921

10

20

Monitor

69

1908

1.564

2845

35

22

Toiyabe

76

2069

974

3292

42

24

Needle

93

1213

1083

2699

10

40

Highland

94

286

263

2865

15

27

Shoshone

164

519

4.50

3145

42

32

Mean

58.4

821.4

.564.1

2793.1

24.0

20.5

S.D.

33.2

708.9

512.1

466.0

13.2

7.3

'Area above 1800 m.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

123

in a similar range of limited average stand
diversity. The majority of the mountain
ranges within any total diversity class have
a wide variation of average stand diversity.
For example, in the interval of 80 to 95
species on a range, the average stand diver-
sity varies from 9 to over 25 species.

Latitudinal and longitudinal diversity pat-
terns: A currently fashionable way of ex-
plaining geographical patterns of diversity
would be by application of island bioge-
ography theory. The results of the correla-
tions of total species richness with total
mountain area, total area of woodland on
the mountain, width of barrier, mountain

Fig. 1. Total number of plant species in the wood-
land belts on each mountain range ordered from the
richest to the most depauperate.

15-30 30-60 60-120 120" 240

NUMBER OF SPECIES ON A RANGE
(frequency octoves)

Fig. 3. Relationships of stand frequency to species
richness. (366 stands of woodland on 66 mountain
ranges.)

Fig. 4. Number of species ordered by number of
stands in which they occurred (frequency octaves).
Line represents log-normal interpolation.

Fig. 2. Range frequency of total number of plant
species in the woodland belts organized by frequency
octaves. Line represents log-normal interpolation.

Fig. 5. Relationship of average number of species
per stand with total number of species in the wood-
lands sampled on each of 66 mountain ranges.

124

GREAT BASIN NATURALIST MEMOIRS

No. 2

height, and an ecotopic diversity score for
each mountain range are given in Table 3.

The linear correlations of woodland floris-
tic diversity with total mountain area,
woodland area on the mountain, width of
barrier, and height of mountain range are
all low and not related in a statistically sig-
nificant sense. There is, however, a signifi-
cant relationship of the index of ecotopic
diversity with floristic richness.

On a log-log basis the picture changes
considerably. The correlation between eco-
topic diversity and species density is highly

significant with little change in value, with
or without data from the Shoshone Moun-
tains. Species density is significantly corre-
lated with total area and woodland area at
the 0.05 level without the Shoshone Moun-
tains and barely insignificant with them in-
cluded.

The relationships between other factors
shows scattered significance. Ecotopic diver-
sity is significantly correlated with total
mountain area and woodland area for both
analyses. Mountain height and total moun-
tain area, but not woodland area, are signif-

Table 3. Correlation coefficients (r) between variables. Lower triangular matrix for log-transformed data, up-
per for untransformed data.

With Shoshone Mountains

Species
Density

Ecotopic
Diversity

Mountain
Height (m)

Total
Mountain
Area (km2)

Woodland
Area

(km2)

Width of
Barrier

(km2)

Species
Density

.7846° '

.3470

.1182

.1669

.1392

Ecotopic
Diversity

.8570° °

_

.3446

.2349

.3133

-.1355

Mountain
Height

.4408

.4227

_

.4676

.3390

-.1743

Total Moun-
tain Area

.4545

.4862°

.5326°

_

.9265 00

.0068

Woodland
Area

.4572

.5049°

.4620

.8510"°

-.1108

Width of
Barrier

-.3806

-.1691

-.1688

-.0743

-.2325

Species
Density

Without Shoshone Mountains

Total

Ecotopic Mountain Mountain

Diversity Height (m) Area (km2)

Woodland Width of

Area Barrier

(km2) (km2)

Species

Density

-

.8424°°

.3303

.3354

.3479

-.2279

Ecotopic

Diversity

.8521 °°

-

.2993

.3035

.3658

-.3125

Mountain

Height

.4120

.3849

-

.4995°

.3565

-.2580

Total Moun-

tain Area

.5612°

.5219°

.5422°

-

.9272°°

.0459

VV oodland
Area

.5229°

.5198°

.4596

.8524°°

-

-.0978

Width of

Barrier

-.2763

-.3135

-.2413

-.0790

-.2629

-

p £ 0.01

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

125

icantly correlated. Total mountain area and
woodland area are significantly correlated;
however, all these variables are so inter-
related that significant correlations are to
be automatically expected. The discrepancy
between the actual and perfect correlation
is probably due to historical accidents such
as fire eliminating some woodland from
where it has potential on those portions of
mountains where climatic unfavorableness
exists, as will be explained later.

We interpret this failure of width of bar-
rier to account for any of the variation ob-
served to be at least partially due to a vari-
ety of methodical problems created by some
simplistic decisions forced by the current
development of the theory. First of all, it is
impossible to choose one continent contrib-
uting flora to the woodlands. Any one
mountain range has floristic affinities with
all surrounding areas. The mixture could be
better described by the degree of affinities
to all possible continents. The Mojave
desert also contributes to the understory
composition. However, because some of the
mountain ranges are imbedded in the Mo-
jave Desert, we knew of no way to calcu-
late barrier width in such circumstances.

The overall height of the mountain range
probably yields a poor correlation because
these woodlands have both boreal-derived
and lowland-derived components. Thus,
height of connecting ridges or valleys would
have a variable effect on what species could
have migrated across the Great Basin and
have had a route of escape as environments
changed. As Johnson and Raven (1970) have
previously pointed out, height has inevitable
covariation with total environmental diver-
sity. Therefore, it is not surprising that the
index of ecotopic diversity, which reflects
combined physical variables, gives a much
better correlation coefficient.

The linear relationship of the index of
ecotopic diversity with species richness
changes if the Shoshone Mountains, with
their expanded sampling scheme, are ex-
cluded (R2 = .6156 to R2 = .7096). As can
be seen from Figure 6, the increased sam-

pling on the Shoshone Mountains did not
result in an obvious increase in ecotopic di-
versity, but it did give an obvious increase
in the number of species encountered. This
is possibly due to an increased micro-
climatic heterogeneity resulting from the
sampling scheme used. Some minor changes
also occurred for other relationships when
Shoshone Mountain data was excluded
(Table 3).

The influence of the increased sampling
scheme on the results with the Shoshone
Mountains has been minor with less effect
on the log-transformed data than on the lin-
ear data. What this appears to show is that,
within a reasonably uniform sampling
scheme, the floristic diversity encountered
within the woodlands is closely related to
physical site diversity. This seems to be as
much a function of where the plants can
survive as it is a function of how many hab-
itats exist on a particular range.

The primary contributor to the ecotopic
diversity index is the number of different

10 20 30

Ecotopic Diversity Score

Fig. 6. Regression between ecotopic diversity index
and species density in juniper-pinyon woodlands on 17
and 18 mountain ranges in the Great Basin (with and
without the Shoshone Mountains included).

126

GREAT BASIN NATURALIST MEMOIRS

No. 2

soil types encountered (Table 2). Analyses
thus far completed have shown no relation-
ship between any specific soil type and any
specific vegetation composition on a Basin-
wide basis. The diversity of soil and topo-
graphic conditions appears to indicate the
potential for diversity of vegetation compo-
sition but not necessarily which taxa may be
involved. This further complicates any at-
tempt at an island biogeographical ap-
proach.

Attempts to map latitudinal and longitu-
dinal differences in the equitability com-
ponents of diversity were found to be of
little value since juniper-pinyon woodlands
are almost uniformly dominated by one or
two trees plus one or two shrubs. One stand
had only three species altogether. Most
stands had only trace amounts of plants oth-
er than the most common five or six spe-
cies. What equitability gradients that oc-
curred are mostly elevational and
successional, as will be demonstrated later.
The species presence-absence approach is
deemed more useful at the scale being fo-
cused on here (Hume and Day 1974,
Schnell, Risser, and Hilsel 1976).

Floristic diversity (Fig. 7) is greatest on
the higher ranges across the southern end of
the Central Plateau portion of the study
area where the Great Basin-Mojave Desert
transition occurs. In addition to the rela-
tionship to the greater present environmen-
tal variation (Table 3), we feel that the high
diversity there is partially due to historical
probabilities of mixing of species with
varying abilities of range expansion and sur-
vival in the considerable migration of spe-
cies which has apparently taken place in
and since the Pleistocene, particularly hyp-
sithermal time (Antevs 1948 and 1955; Mor-
rison 1965; Birkeland 1969; Spaulding 1974;
Martin 1963; Martin and Mehringer 1965;
Fritts 1965; Cottam, Tucker, and Drobnick
1959; LaMarche 1974; Phillips and Van De-
vender 1974; Van Devender 1974; Wells
and Jorgenson 1964; Wells and Berger
1967; Elston 1976).

These previous papers and analysis of our

data favor the view that the pigmy conifer
woodlands were once continuous across the
Great Basin and northern Mojave Desert re-
gions. Subsequent aridity and consequent
plant migrations have fragmented these
woodlands into the patterns of distribution
seen today. The differences in floristic rich-
ness are probably more related to rates of
extinction (Stebbins 1974 and 1975) than to
long-distance dispersal from source biotas
on the several possible continents. The rich-
ness of the present woodland flora is thus
understandably related to the overall envi-
ronmental favorability of the various moun-
tain ranges at present and in the recent
past.

Elevational variability: Billings (1951) and
Hollermann (1973a and b) have previously
commented on the puzzling variation in the
elevational extent of Great Basin juniper-
pinyon woodlands. We also noted wide var-
iation in the Basin-wide elevational limits of
this vegetation type. This variation, how-
ever, shows a close relationship to the gen-
eral topography of the area, with climatic
modification being important in the north-
ern portions.

Our lowest sample in the juniper-pinyon

Number of spec

Fig. 7. Isolines of species richness in the woodlands
hetween the 66 mountain ranges.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

127

woodland was a stand at 1200 m on the
south slope of the Beaver Dam Mountains
of extreme southwestern Utah. This area is
also the topographic low point of the area
studied. The highest lower boundaries are
found in the White, Excelsior, and Silver
Peak ranges in the southwestern portion of
the study area, and the Toquima and Mon-
itor ranges of the central plateau, some of
the highest parts of the topography of the
whole study area. The regional distribution
of the lowest woodland boundaries are defi-
nitely along the eastern and southern
boundaries of the study area (Fig. 8).

The highest boundary of the woodland
belt is much less definite (Fig. 9), largely
due to great differences in mountain
heights. The upper extreme of these wood-
lands in our data set was a 2800 m eleva-
tion stand on the west slope of the White
Mountains of California. The geographical
distribution of upper boundaries is partially
correlated to those of lower boundaries.

The difference between the actual eleva-
tion of the lower limit and upper limit var-
ies greatly from aspect to aspect on a
mountain range and from range to range

over the Great Basin. An average woodland
belt width of about 350 m elevational ex-
tent occurred on the mountain ranges we
sampled. Further analysis of the data shows
extremes of belt width ranging from over
700 m in elevation in the White Mountains
of California, Highland Peak of Nevada,
and the Needle Range of Utah to essentially
zero at the northern boundaries of the type.
The band of woodland is also extremely
narrow in western Utah.

These elevational extremes and band
widths are graphically portrayed in Figures
10 through 13, which depict transects across
the Great Basin. These transects could be
combined to make a response envelope. The
breadth of these woodlands is generally
greatest in the areas where the extremes of
elevational extent occur. (Exceptions to this
will be explained later). This greater
breadth also approximately coincides with
and contributes to the higher floristic diver-
sity in the Great Basin-Mojave transition
discussed earlier.

Although undoubtedly some of the eleva-
tional variation observed is due to recent
historical causes, especially fire and wood

■ 2200 mele'S

ED 2100

H 2000

□ 1800

ffl 1600

E3 1200

(Hi

m

Fig. 8. Isolines of lowest woodland elevation sam-
pled. Interpolation between 66 mountain ranges.

Fig. 9. Isolines of highest woodland elevation sam-
pled.

128

GREAT BASIN NATURALIST MEMOIRS

No. 2

cutting (Lanner 1976), there are consistent
patterns of elevational extent related to to-
pography and climate. The west and east
relationships show this most clearly. Wood-
lands in the eastern Great Basin are gener-
ally lower than those to the west. The
woodland belt also narrows considerably at
the Wasatch Front. The average elevation
and then decreases, while the

increases,

Fig. 10. Map of the mountain ranges sampled in
this study. Lines connect the mountain ranges depict-
ed in the following figures.

West to east-central transect

4200-

4000-

3800
3600

I

i

3400.

%

3200-

A

3000-

2800

A

3 2600
| 2400-
- 2200

:

i

t

2000
1 1800

.

I 1600

" 1400
1200

/

1000

800
600

Fig. 11. Schematic cross section of the altitudinal
and latitudinal variations in the woodland belt and its
tree species dominance for a west-east transect (see
Figure 10) of some of the southcentral ranges sampled.
Lines across mountain indicate elevations of adjacent
vallev floors.

width of the woodland type generally de-
creases from south to north.

The relative composition of the dominant
trees and associated understory varies con-
siderably within the belt (Fig. 11-13). The
usual case is for pinyon to increase with al-
titude and juniper to dominate the lower
half of the belt. On many of the higher
mountain ranges in the Great Basin, the up-

fjA :«

W*^ 40C

T 38C

West to east— southern transect

a 90% P«yo«

4 0 00

? 4 WJW

3800

Sp

3 600-

>

3400

S *

320oJ

*" 6 f\

3000

2800-

* 2600

1

% 2400

- 2200

A

0

2000

A

:. a

-

a

2 1800

»

a

'■

*

I 1600

*

1200

\

1000

s.

BOO

■* .

1 1

1 1 1

Fig. 12. Schematic cross section of the altitudinal
and latitudinal variations in the woodland belt and its
tree species dominants for a west-east transect (see
Figure 10) of some of the southernmost ranges sam-
pled. Lines across mountains indicate elevations of ad-
jacent valley floors.

North to south transect

Fig. 13. Schematic cross section of the altitudinal
and longitudinal variations in the woodland belt and
its tree species dominants for a north-south transect
(see Figure 10) in the Great Basin. Lines across moun-
tains indicate elevations of adjacent valley floors.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

129

per part of the woodland belt is a pure pi-
nyon overstory. Juniper is more widespread,
dominating areas in the north and at lower
elevations that do not have pinyon. The
quantity of juniper appears to increase with
increasing soil moisture stress over most of
the Great Basin.

The greater proportions of pinyon don't
always appear to be purely due to increas-
ing precipitation, usual at higher altitudes.
Single leaf pinyon appears to have a wider
range of moisture tolerance than does juni-
per at the higher moisture levels. The in-
cidence of pinyon pine with double needles
and amounts of juniper increases as the pro-
portion of summer rainfall increases from
west to east across the Great Basin. The
summer dry, extreme western boundary of
the basin is the only place where some
mountain ranges have single-leaf pinyon but
lack juniper. On the west side of the White
Mountains there are belts of pure pinyon
both above and below juniper (St. Andre,
Mooney, and Wright 1965). These two cir-
cumstances suggest that although summer
soil moisture stress may largely explain the
lower elevational limits over most of the
basin, the circumstances in the southwestern
portion of the study area indicate that there
may be temperature regimes unsatisfactory
for juniper-pinyon mixes there.

St. Andre, et al. (1965) have speculated
that lack of competition from extensive for-
est vegetation above the pinyon-juniper belt
may explain the extreme width and high
upper limit of the type on the White
Mountains. The general absence of pinyon
and juniper on the east flank of the Sierra
might be partially due to Ponderosa pine
and montane chaparral providing more
competition than the sagebrush-grass types
typical of the Great Basin Ranges. Similarly,
oakbrush-dominated vegetation in the
Wasatch Mountains and those near the east-
ern boundary of the Great Basin commonly
occur above or replace juniper-pinyon
woodland vegetation. Could oakbrush pro-
vide competition limiting the extent of the

pigmy conifer woodlands there? Could the
likely upward migrations of juniper-pinyon
woodlands and the other conifer belts (La-
Marche and Mooney 1972) during the hyp-
sithermal have resulted in the elimination of
conifer forests at higher elevations of the
smaller mountain ranges of the Great Basin?
When temperature cooled and precipitation
increased in more recent times (LaMarche
1974), pinyon and juniper could have ex-
panded with less competition from other
tree forms. Are current precipitation-
temperature relationships suitable for more
competitive species? Unfortunately, little di-
rect data exists to answer these questions.

Ponderosa pine exists in a few scattered
locations on the ranges of southwestern
Utah and southeastern Nevada without
dominance occurring. In those areas ponde-
rosa pine seems to be limited to the most
favorable sites, e.g., moist sites along
streams or northeastern slopes. Pinyon and
juniper compete largely with shrubs at both
their upper and lower boundaries within the
Great Basin.

Until more definitive evidence is in from
palynological, dendroclimatological, and
woodrat midden studies, we can only go on
inferences from current vegetation. Al-
though competitive and historical factors
are part of the environmental complex, we
feel that we can largely explain the current
distribution of juniper-pinyon woodlands on
the basis of recent and current climatic pat-
terns.

The effect of mountain height and mass
on climatic patterns (the "Merriam Effect,"
Lowe 1964) has long been recognized by
plant geographers. Since its manifestations
are greatest in semiarid regions, we related
our data on width of woodland bands
against the variables of maximum mountain
height and area of the mountain mass. Nar-
rower belts are generally found on the
mountains of smaller area with the narrow-
est of the belts generally, but not always,
found on those ranges with less than 700
km2 area (Fig. 14). Mountain area above
this size has little relation to the width of

130

GREAT BASIN NATURALIST MEMOIRS

No. 2

belts. The widest belts are found on the
highest peaks and are relatively indepen-
dent of mountain area (Fig. 15). A high
peak will not, however, necessarily have a
wide belt of woodland (e.g., Wheeler Peak,
Ruby Mountains, etc.). In circumstances
where the belt is wide, the height of the
peak dominates, but the location of the
peak in the Great Basin is of much more
importance. Figure 16 makes a separation
of the Great Basin on the basis of where
woodland belt widths are either greater or
less than 400 meters. There are many peaks
in the area of narrow belt width that are as
high or higher than those found in the area
of the wide belt width, but they are all in
the region of greater orographic impact
from Pacific Frontal storm systems (Hough-
ton 1969). Because of their elevation, the
very high ranges of east central Nevada
have a climate similar to the ranges more
north and west. These climatic controls will
now be expanded on.

The lower woodland boundaries in the
northern half of the study area appear to be
related to valley bottom topography (Figs.
10-13). Wernstedt (1960) shows that little
difference occurs in summer temperatures
between the southern and northern Great
Basin, but greatly lowered winter values are
encountered as latitude increases. This trend
applies best to valley bottom stations.
Knowledge of temperature within the
woodlands is minimal because few mete-
orological stations with long-term records
exist. Most of the U.S. Weather Service me-
teorological data is from desert valley bot-
tom stations where most of the scarce hu-
man habitation occurs. Billings (1954) has
correlated the occurrence of juniper-pinyon
woodlands in northwestern Nevada to ther-
mal belts. This led us to suspect that the
breadth of the woodland belt all across the
northern portions of the Great Basin is at
least partially related to the strength of de-
velopment and persistence of thermal belts.
Analysis of what temperature data that do
exist (Fig. 17) shows a strong relationship of
fewer degree-days below freezing indexes

2500 3000

MOUNTAIN HEIGHT (metert)

Fig. 14. Average woodland belt width (in nearest
100 m increments) in relation to maximum mountain
height.

600 900

MOUNTAIN AREA

Fig. 15. Average woodland belt width (in nearest
100 m increments) in relation to mountain area above
the lower boundary of woodland.

Average woodland

Fig. 16. Division of the Great Basin on the basis of
mountain ranges having an average woodland belt
width of greater or less than 400 meters vertical ex-
tent.

1978

INTERMOUNTA1N BIOGEOGRAPHY: A SYMPOSIUM

131

for meteorological stations within the wood-
land belt, except in the southeastern Great
Basin (Pioche and Ursine, Nevada). Degree
days above freezing are much higher in
most woodland areas than those in the val-
ley bottoms where various cold-winter, sem-
idesert shrub-dominated vegetation types
occur.

Juniper-pinyon woodlands are nearly ab-
sent north of Interstate Highway U.S. 80
across northwestern Nevada. Critchfield and
Allenbaugh (1969) noted the correspondence
of the northwestern limits of Pinus mon-
ophyUu with the southern boundary of plu-
vial Lake Lahontan and speculate as to the
lake being a possible barrier to plant migra-
tion. The occurrence of pinyon pine on the
Virginia and Stillwater Ranges and the fre-
quent invasion of juniper in areas that were
once under the pluvial lakes of the Great
Basin makes this seem unlikely. The corre-
spondence of lakes to the pools of cold air
and/or lower precipitation now in these ba-
sins seems more of an etiological factor in
present distribution patterns.

Arlo Richardson (personal communi-
cation) has mapped the progression of late
winter and early spring chill-unit (Richard-
son, Seeley, and Walker 1974) accumula-
tions across the Great Basin. The valleys of
northwestern Nevada have the most rapid
progression and resultant ending of winter
plant dormancy of any area in the Great
Basin. Warming periods are, however, more
dramatically punctuated by cold periods
brought by Pacific Frontal storms there
than elsewhere. This knowledge, combined
with existing knowledge of air circulation
and storm patterns summarized by Hough-
ton (1969), leads us to speculate that a pri-
mary reason for the narrowing and absence
of the woodland belt in the north, in addi-
tion to the latitudinal decrease in solar in-
put, is due to the increased frequency of
Pacific frontal winds breaking thermal in-
versions. This pattern is thought to encour-
age earlier plant growth but greater sub-
sequent susceptibility of plants to direct
frost damage or frost drought in a manner

similar to that found for other conifers
(Hocker 1956, Newnham 1968). Supporting
evidence of the net effect on woodland
trees can be seen on space photography
where the most northerly and westerly of
the mountain ranges in the Great Basin are
devoid of pinyon and juniper. In the north
central portion of Nevada, ranges such as
the Santa Rosa's are almost devoid of con-
ifers altogether (Critchfield and Allenbaugh
1969). Ranges to the south and east of the
ranges devoid of the woodland contain one
or both tree species, implying a lee pro-
tection effect, e.g., Spruce Mountain, the
southern Ruby Mountains, and the southern
most tip of the East Humboldt Range (Fig.
18).

In the southern Great Basin the westerly
fronts have lost most of their energy; thus
their disruptive effect on the development
and maintenance of thermal belts is not as
frequent. Furthermore, most of the precipi-
tation in the southern Great Basin comes
from the moist Gulf air masses with rainfall
triggered by the lows aloft. As a con-
sequence, there are weaker adiabatic lapse
rates (decrease in temperature with eleva-
tion) and a corresponding smaller increase
in precipitation with elevation because of
reduced orographic effects. For example, on
the west slope of the White Mountains, the

Below 0° C

Fig. 17. Accumulated degree days cold stress below
0°C for selected woodland (PJ) and valley floor (V)
climatic stations across the Great Basin (west on left,
east on right). Data obtained from U.S. Weather Ser-
vice climatic summaries for years indicated.

132

GREAT BASIN NATURALIST MEMOIRS

No. 2

change in precipitation with elevation over
the entire woodland belt is only 10 cm (4
inches) (St. Andre et al. 1965).

Although west slopes are usually the most
mesic and diverse in the Basin, the wood-
lands on some of the mountain ranges in
our study areas have higher floristic richness
on the eastern and southern slopes (Table
4). This apparent reversal of the usual trend
is believed to be due to the effects of
nearby ranges creating rain shadows and
other orographic effects. For example, in
the Silver Peak Range most of the moisture
must come from lows aloft positioned to the
north and east. Johnson (1956) noted that
the woodland belt on the south and eastern
sides of the nearby Kawich Mountains had
the most breadth, highest cover, and largest
trees, providing corroborative evidence of
this situation.

In the southern two-thirds of the study
area, the lower boundary of woodlands ap-
pear to be more controlled by precipitation
than temperature. Degree day values for
Pioche and Ursine, Nevada (Fig. 17), are an
example of how heat sums do not differ be-
tween foothill, woodland, and valley stations

in the southeast. In fact, the situation may
be slightly reversed over that of the more
northerly and westerly pairs of stations.

Further evidence of the interaction of
temperature and moisture may be observed
in the variation of the width of the wood-
land belt as related to mountain symmetry.
Generally, the north-south axes of Great Ba-
sin mountain ranges are longer and the
highest peaks are also usually in the center
(Lustig 1969). When the range is protected
to the west (e.g., Table Mountain in the
Monitor Range) (Fig. 19), the widest belts
are in the wider, higher, central east-west
axis; and the belt-width decreases and aver-
age elevation increases toward the north
and south ends. This is an apparent re-
sponse to the Merriam effect of the greater
mountain mass intercepting more moisture
at a similar elevation. Extra high mountains
such as Arc Dome in the Toiyabes (Fig. 19)
appear to effect excessively cold, wet condi-
tions for woodland development and the
high, central axes have narrower, lower
belts primarily on western slopes under such
circumstances.

We have ignored here the short-term suc-

Table 4. Number of species (and number of stands) on each aspect of selected ranges in a west to east belt.

Average

Mountain Range

N

E

S

W

Total

per stand

Virginia

35(2)

17(1)

27(2)

-H

50(5)

19

« Pine Nut

32(2)

28(2)

23(2)

32(2)

57(8)

18

Clan Alpine
S Desatoya

21(2)

15(1)

-H

-H

28(3)

12

19(1)

20(1)

15(2)

28(2)

47(6)

15

<G Shoshone
■^ Toiyabe

98 (13)

106(17)

108 (22)

78(7)

159 (59)

25

19(1)

30(3)

23(2)

46(3)

75(9)

15

Monitor

43(2)

31(3)

29(3)

34(2)

68 (10)

17

Average

38.1

35.3

37.5

43.6

69.1

17.1

Upper Snake

8(1)

17(2)

10(2)

9(2)

27(7)

7

Snake

10(1)

23(2)

11(1)

25(2)

49(6)

12

H Deep Creek
* Pilot
a> Needle

31(2)

16(2)

17(2)

24(2)

49(8)

14

30(2)

21(2)

13(2)

26(3)

50(9)

12

55 (13)

53 (16)

33 (15)

43 (12)

92 (56)

9

| WahWah

20(2)

19(2)

17(2)

16(2)

45(8)

10

Mineral

23(4)

18(4)

31(5)

32(4)

62 (17)

9

Stansbury

13(1)

23(2)

12(1)

-H

37(4)

12

Tushar

32(2)

17(2)

18(2)

9(1)

48(7)

10

Average

26.3

22.9

18.7

23.3

52.0

10.1

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

133

cessional changes that are known to be op-
erating in these woodlands. If we plot by
aspect, relative tree cover, richness, and rel-
ative equitability against elevation for an
example mountain range (Fig. 20), we see
that richness is minimal in the closed wood-
lands in the middle of the belt where un-
derstory has been largely excluded by high
relative tree cover. Equitability patterns are
more complicated.

The approach to equitability used mea-
sures how evenly the existing total cover is
divided among the species present. For each
different total cover and species number a
different maximum equitability is possible.
To make an equitability comparison be-
tween areas of different cover and density,
it is necessary to convert the figure to a rel-
ative one; in this case the percent of the
maximum possible for each site. During suc-
cession (West et al. 1975) invasion by juni-
per and pinyon shifts in relative cover to-
ward more trees. However, before a
significant drop in total species number oc-

Fig. 18. Effects of exposure to westerly winds and
storm tracks on occurrence of the woodland belt at
the northern boundary of the type in northeastern Ne-
vada. Stippled area is juniper-pinyon woodland, outer
(lower) line is 2134 m (7000 ft.) contour, inner (higher)
line is 2743 m (9000 ft.) contour. Dots are major
mountain peaks: SS = Sulphur Springs Range, EH =
East Humboldt Range, Sp = Spruce Mountain, R =
Ruby Mountains. (Vegetation map derived from
LANDSAT-1 photography. Elevational contours from
U.S. Geological Survey maps.)

curs, equitability drops. As tree suppression
substantially reduces the number of under-
story species, equitability rises. Equitability
is thus highest at either end of the sere and
lowest in the intermediate serai stages. This
is contrary to the linear model for xeric-
forests proposed by Auclair and Goff (1971).

Fig. 19. Interactions of exposure to westerly storm
tracks, mountain height, and mountain mass on the
variation in woodland belt width and elevation in
west-central Nevada. Stippled area is juniper-pinyon
woodland. Outer (lower) line is 2134 m (7000 ft.) con-
tour. Inner (higher) line is 2743 m (9000 ft.) contour.
Dots are major mountain peaks: A = Arc Dome in
the Toiyabe Range, T = Table Mountain in the Mon-
itor Range. (Vegetation map derived from LANDSAT-1
photography. Elevational contours from U.S. Geologi-
cal Survey maps.)

2600

. Z400

\l

e

f¥ r

c

■

| 2200

%)

2000

VI

Number of species

0 50 100 0 50 100

% of total cover % of

represented by trees equitability

Fig. 20. Species density (floristic richness), relative
percent of total cover contributed by trees, and per-
cent of maximum equitability calculated by the Mcin-
tosh (1967) index plotted by elevation and aspect,
Needle Range, Beaver-Iron Cos., Utah.

134

GREAT BASIN NATURALIST MEMOIRS

No. 2

In general, the north slopes at any eleva-
tion have a larger number of species, a
lower percent relative cover of trees, and a
higher equitability value than the other
three aspects.

Higher average diversity values were ob-
tained in our data by including some of the
woodland that is encroaching on shrublands
and grasslands both above and below the
main woodland belts. These successional
considerations are sufficiently complex that
details have to be considered elsewhere
(Nabi 1978).

matic patterns are so different in the vari-
ous portions of the Great Basin. These
differences are striking even at the same
elevation on different portions and aspects
of the same mountain range. This is espe-
cially evident in data from the Shoshone
Range. In general, such differences are most
pronounced in the Mojave-Great Basin tran-
sition area. Stratification of landscapes into
units of some homogeneity for management
purposes must take these phytogeographical
patterns into account, if confusion is to be
avoided.

IMPLICATIONS

Our observations of these broad phyto-
geographical patterns in Great Basin
juniper-pinyon woodlands will hopefully
lead to some testable hypotheses about eco-
physiological responses of juniper and pi-
nyon. For instance, what are relative evapo-
transpirative and photosynthetic limits of
juniper and pinyon in regard to temper-
ature and moisture? Does needle number on
pinyon relate to effective soil moisture?
What are the tolerances of seedlings to tem-
perature and moisture extremes compared
to those of the mature trees? Are there late
winter-early spring low temperature suscep-
tibilities? Research to answer these ques-
tions provoked by our synecological results
will be necessary to confirm our hunches.

The distributional and phytosociological
variation observed within the type indicates
that effective environments are radically
different across the Great Basin. For the
time being we can only inferentially con-
clude from the phytogeographical evidence
that climate is a major factor influencing
the latitudinal, longitudinal, and altitudinal
extent of Great Basin juniper-pinyon wood-
lands. Climate is also probably a greater de-
terminant of internal floristic diversity than
migrational and evolutionary equilibria. Re-
garding the mountains and their woodlands
as islands in a uniform "sea" of desert leads
to dangerous basic conclusions and con-
founds management application, since cli-

ACKNOWLEDGMENTS

The data were collected with financial
support from the Intermountain Forest and
Range Experiment Station, U.S. Forest Ser-
vice. Subsequent analysis was done under
support of Mclntire-Stennis Forestry Re-
search Act monies through the Utah Agri-
cultural Experiment Station. We wish to ac-
knowledge the assistance of Robert Bayn in
data analysis and preparation of figures.

Literature Cited

Antevs, E. 1948. Climatic changes and pre-
whiteman, p. 167-191. In: The Great Basin
with emphasis on Glacial and post-Glacial
times. Bull. Univ. of Utah 38(2): 1-191.

1955. Geologic climatic data in the West.

Amer. Antiquity 20: 317-335.

Auclair, A. N., and F. G. Goff. 1971. Diversity
relations of upland forests in the western Great
Lakes area. Amer. Naturalist 105: 499-528.

Beeson, D. W. 1974. The distribution and synecol-
ogy of Great Basin pinyon-junipers. M.S. thesis,
Univ. Nevada, Reno, 95 p.

Billings, W. D. 1951. Vegetational-zonation in the
Great Basin of western North America, p.
101-122. In: Compt. Rend, du Colloque sur Les
Bases Ecologiques de la Regeneration de la veg-
etation des Zones Arides. Union Intern. Soc.
Biol., Paris.

1954. Temperature inversions in the pinyon-

juniper zone of a Nevada mountain range. But-
ler Univ. Bot. Studies 12: 112-117.

Birkeland, P. W. 1969. Quaternary paleoclimatic
implications of soil clay mineral distribution in
a Sierra Nevada-Great Basin transect. J. Geol.
77: 289-302.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

135

Brown, J. H. 1971. Mammals on mountaintops:
nonequilibrium insular biogeography. Amer.
Naturalist 105: 467-478.

Cottam, W. P., J. M. Tucker, and R. Drob-
nick. 1959. Some clues to Great Basin post-
pluvial climates provided by oak distribution.
Ecology 40: 361-377.

Critchfield, W. B., and G. L. Allen-
baugh. 1969. The distribution of Pinaceae in
and near northern Nevada. Madrono 2: 12-26.

Cronquist, A., A. H. Holmgren, N. H. Holmgren,
and J. L. Reveal. 1972. Intermountain flora,
Vol. 1, Hafner Publ. Co., New York.

Elston, R. (ed.). 1976. Holocene environmental
changes. Nevada Archeological Survey Research
Paper Series No. 6. Univ. Nevada Press, Reno.

Fritts, H. C. 1965. Tree-ring evidence for climatic
changes in western North America. Monthly
Weather Rev. 93: 421-443.

Hocker, H. W., Jr. 1956. Certain aspects of climate
as related to the distribution of loblolly pine.
Ecology 37: 824-834.

Hollermann, P. 1973a. Some aspects of the geo-
ecology of the Basin and Range Province (Cali-
fornia section). Arctic and Alpine Res. 5(3) Part
2: A85-A98.

1973b. Some reflections on the nature of

high mountains, with special reference to the
western United States. Arctic and Alpine Res.
5(3) Part 2: A 149- A 160.

Houghton, J. G. 1969. Characteristics of rainfall in
the Great Basin. Desert Research Institute.
Univ. of Nevada System, Reno.

Hume, L., and J. C. Day. 1974. The determination
of an efficient sampling intensity for studying
beta diversity in plant communities. Can. J.
Bot. 52: 189-199.

Johnson, N. K. 1956. Birds of the pihon association
of the Kawich Mountains, Nevada. Great Basin
Naturalist 1-4: 32-33.

1975. Controls of number of bird species on

montane islands in the Great Basin. Evolution
29: 545-567.

Johnson, M. P., and P. H. Raven. 1970. Natural
regulation of plant species diversity, p. 127-162.
In: T. Dobzhansky et al. Evolutionary Biology
Vol. 4. Appleton-Century-Crofts, New York.

LaMarche, V. C, Jr. 1974. Paleoclimatic infer-
ences from long tree ring records: Intersite
comparison shows climatic anomalies that may
be linked to features of the general circulation.
Science 183: 1043-1048.

LaMarche, V. C, and H. A. Mooney. 1972.
Recent climatic change and development of the
bristlecone pine (P. longaeva Bailey) krummholz
zone, Mt. Washington, Nevada. Arctic and Al-
pine Res. 4: 61-72.

Lanner, R. M. 1976. The tree that made Nevada's
silver, p. 443-444. In: S. A. Paher (ed.), Nevada
Official Bicentennial Book. Nevada Publications,
Las Vegas.

Lowe, C. H. 1964. Arizona's natural environment:
Landscapes and habitats. Univ. Arizona Press,
Tucson.

Lustic, L. K. 1969. Trend-surface analysis of the
Basin and Range Province, and some geomor-
phic implications. U.S. Geol. Surv., Prof. Pap.,
500-D: 1-70.

Martin, P. S. 1963. The last 10,000 years: A fossil
pollen record of the American southwest. Univ.
of Arizona Press, Tucson.

Martin, P. S., and P. J. Mehringer, Jr. 1965.
Pleistocene pollen analysis and biogeography of
the Southwest, p. 433-451. In: H. E. Wright,
Jr., and D. G. Frey (eds.). The Quaternary of
the United States. Princeton Univ. Press,
Princeton.

May, R. M. 1975. Patterns of species abundance
and diversity, p. 81-120. In: M. L. Cody and J.
M. Diamond (eds.). Ecology and Evolution of
Communities. Harvard Univ. Press, Cambridge.

McIntosh, R. P. 1967. An index of diversity and
its relation to certain concepts of diversity.
Ecology 48: 392-404.

Morrison, ^ B. 1965. Quaternary geology of the
Great Basin, p. 265-285. In: H. E. Wright, Jr.,
and D. G. Frey (eds.). The Quaternary of the
United States. Princeton Univ. Press, Princeton.

Mueller-Dombois, D., and H. Ellenberg. 1974.
Aims and methods of vegetation ecology. John
Wiley and Sons, New York.

Nabi, A. A. 1978. Variation in successional status of
pinyon-juniper woodlands in the Great Basin.
M.S. thesis, Utah State Univ., Logan, 127 p.

Newnham, R. M. 1968. A classification of climate by
principal component analysis and its relation-
ship to tree species distribution. For. Sci. 14:
254-264.

Phillips, A. M., and R. T. VanDevender.
1974. Late Pleistocene plant communities in
the lower Grand Canyon of Arizona. Bull. Ecol.
Soc. Amer. 51: 27 (abstr.).

Richardson, E. A., S. D. Seeley, and D. R.
Walker. 1974. A model estimating the com-
pletion of rest for "Redhaven" and "Elberta"
peach trees. HortScience 9: 331-332.

Schnell, G. D., P. G. Risser, and J. F. Helsel.
1976. Factor analysis of tree distribution pat-
terns in Oklahoma. 23rd annual meeting, South-
western Assoc. Naturalists (abstr.). p. 11.

Simberloff, D. S. 1974. Equilibrium theory of is-
land biogeography and ecology. Ann. Rev. Ecol.
and System. 5: 161-182.

Spaulding, W. G. 1974. Dynamics of late Pleisto-
cene vegetation change in southern Nevada.
Bull. Ecol. Soc. Amer. 51: 27 (abstr.).

136

GREAT BASIN NATURALIST MEMOIRS

No. 2

St. Andre, G., H. J. Mooney, and R. D.
Wright. 1965. The pinyon woodland zone in
the White Mountains of California. Amer. Midi.
Naturalist 73: 225-239.

Stebbins, G. L. 1974. Flowering plants: Evolution
above the species level. Harvard Univ. Press,
Cambridge.

1975. Shnibs as centers of adaptive radiation

and evolution, p. 120-140. In: H. C. Stutz (ed.).
Wildland Shrubs: Symposium and Workshop
Proc. U.S. Forest Service, Shrub Sciences Lab,
Brigham Young Univ. Press, Provo, Utah.

Tueller, P. T. 1975. The natural vegetation of Ne-
vada. Mentzelia 1: 3-6, 23-28.

United States Department Agriculture, Soil
Survey Staff. 1951. Soil Survey Manual.
U.S.D.A. Agric. Res. Admin. Hdbk. 18.

1960. Soil classification: A comprehensive

system, 7th Approximation. U.S. Government
Printing Office, Washington, D.C.

1967. Supplement to soil classification sys-
tems (7th Approximation). U.S. Government
Printing Office, Washington, D.C.

VanDevender, R. T. 1974. Late Wisconsin biotic
communities and climates of the Sonoran
Desert of Arizona: The fossil pack rat midden

record. Bull. Ecol. Soc. Amer. 51: 27 (abstr.).
Wells, P. V., and C. D. Jorgenson. 1964.
Pleistocene wood rat middens and climatic
change in Mojave Desert: A record of juniper
woodlands. Science 14: 1171-1174.

Wells, P. V., and R. Berger. 1967. Late Pleisto-
cene history of coniferous woodland in the Mo-
jave Desert. Science 155: 1640-1647.

Whittaker, R. H. 1975. Communities and ecosys-
tems. 2d ed. MacMillan Publ. Co., New York.

Wernstedt, F. L. 1960. Climatic fluctuations in
the Great Basin 1931-56. J. Range Manage. 13:
173-178.

West, N. E., K. H. Rea, and R. J. Tausch.

1975. Basic synecological relationships in juni-
per-pinyon woodlands, p. 41-53. In: G. F. Gif-
ford and F. E. Busby (eds.). The pinyon-juniper
ecosystem: A symposium. Utah Agric. Expt. Sta.

Young, J. A., R. A. Evans, and P. T. Tueller.

1976. Great Basin plant communities-
pristine and grazed, p. 186-215. In: R. Elston
(ed.). Holocene environmental change. Nevada
Archeological Survey Research Paper Series No.
6, Univ. Nevada Press, Reno.

PATTERNS OF AVIAN GEOGRAPHY AND SPECIATION
IN THE INTERMOUNTAIN REGION

Ned K. Johnson1

Abstract— The Intermountain Region comprises a huge arena where major avifaunas more highly developed
beyond the Great Basin come into contact. For this reason the distribution and composition of avifaunas in ri-
parian and pinyon-juniper woodlands and in coniferous forests in this region are understood most clearly if con-
sidered as part of a broader system of patterns evident in western North America. Riparian woodland samples
point to a complex meeting ground in the Intermountain Region of two major avifaunas of equivalent size, but
from opposite distributional backgrounds. These northern and southern avifaunas do not mix among the samples
represented except in western Nevada where two species of ultimate southern origin have penetrated a basically
northern avifauna. Approaching the Great Basin, from the two centers of abundance of riparian species in the
Snake and Colorado River drainages, species richness drops. Habitat depletion and, to a lesser extent, insularity
play roles in this impoverishment. In the most depauperate riparian avifaunas, six species commonly coexist, each
in a different family. Comparison of the pinyon zone avifaunas of two groups of mountain ranges, 90 km apart
along the California-Nevada border, demonstrates a striking trade off among species of northern and southern
biogeographic histories. The northern or Boreal forms, many of which are numerous in the Sierra Nevada, have
had easy access to the favorable, cool, and relatively moist pinyon forests in the adjacent spur ranges. In contrast,
the species of southern or Austral derivation prefer the warm and very arid pinyon woodland a short distance to
the south. Few species are confined to either northern or southern sites, overlap in species composition is great,
and equivalent species richness is achieved. However, despite these similarities, strong geographic differences in
abundance of most species in each pinyon avifauna and the occurrence of at least 12 specific and subspecific
range boundaries suggest the interposition between northern and southern pinyon areas of a substantial, but as
yet poorly characterized, climatic barrier. Boreal species richness declines,abmptly from high values in the south-
ern Cascades and Sierra Nevada to low levels in a zone of impoverishment across western and central Nevada.
From near 116° W Longitude in eastern Nevada, coincident with the appearance of fir and/or bristlecone pine
forests, species numbers climb gradually until the main Rocky Mountains are reached. There, species richness
compares favorably with that in the Sierra Nevada. The proportion of species favoring riparian woodlands over
coniferous forests is higher on the island mountaintops of the Great Basin than in the Boreal "continents" of the
Sierra Nevada-Cascades and Rocky Mountains. The western edge of the Great Basin richly demonstrates exam-
ples of stages in avian speciation. A full range of interactions is represented, from intergradation of poorly char-
acterized races, through abrupt zones of hybridization between strongly marked subspecies of different racial
complexes or of semispecies, to sympatry and infrequent interbreeding of closely-related, full species in recent
secondary contact. These various zones of population interaction coincide strikingly with sharp floristic and cli-
matic gradients. A major avifaunal break occurs between coastal or Sierra Nevadan forms that inhabit oak-
chaparral and/or coniferous forest and closely-related interior forms that prefer pinyon-juniper or aspen-willow
associations. In keeping with the special requirements of each species, contact zones and areas of disjunction
show general, rather than precise, coincidence in the western Great Basin. There is no Great Basin Boreal Avi-
fauna; the most distinctive interior forms occur across the entire span from eastern California to Colorado. The-
low desert trough along the east side of the Sierra Nevada that divides major mountain systems in the western
portion of the Intermountain Region is not the principal barrier dividing coastal Sierra Nevada from interior avi-
faunas. Instead, the major avifaunal transition occurs in a belt of variable width just east of the crest of the Cas-
cade Mountains and Sierra Nevada, where the precipitation shadow and continental climate begin to exert cru-
cial influence.

Although much neglected in the past be- mountain Region of western North America
cause of difficulty of access and false im- is emerging as a fascinating natural labora-
pressions of biotic sterility, the Inter- tory requiring the close attention of the

'Museum of Vertebrate Zoology and Department of Zoology, University of California, Berkeley, California 94720.

137

138

GREAT BASIN NATURALIST MEMOIRS

No. 2

biogeographer-ecologist. Islands of conifers
or aspen, cool mountain meadows, and at-
tendant biotas on scores of mountaintops lie
isolated by seas of desert from source stocks
in the Sierra Nevada and Rocky Mountains.
The unique physical setting of a multi-
plicity of more or less parallel ranges, each
differing in size, geology, and configuration
from adjacent ones, cannot be duplicated
elsewhere among continental mountain sys-
tems. And, because of low human density,
many of the montane habitats are relatively
undisturbed, thus permitting a good look at
original, man-unmodified distributions of
species. In the valleys as well, oases offer
insular environments for a variety of organ-
isms whose interrupted ranges intrigue the
biogeographer.

With increased field efforts by a number
of workers in recent decades, the bird spe-
cies breeding on many of the major basin
ranges have been tallied, and information
has reached a level allowing tests of certain
fundamental theories of insular biogeog-
raphy (Johnson 1975). However, these tests
are just the beginning. With additional "al-
pha" exploration in the future, the enlarged
data base will permit more diverse and de-
tailed kinds of analyses. But, already enough
distributional information has accumulated
to suggest several patterns that challenge
the avian geographer. In this paper I identi-
fy and discuss these patterns in a search for
common biogeographic themes. Further-
more, I specify among the birds of the In-
termountain Region distributional situations
relevant to hypotheses on the control of
range boundaries. Finally I describe the
prominent zone of active speciation in the
western part of the region and attempt to
interpret this zone in relation to coincident
physiographic, climatic, and floral dis-
continuities.

Nomenclature and sequence of species
follow the checklist of North American
birds except where superceded by the 32nd
supplement (American Ornithologists' Union
1957, 1973).

Avifaunal Transects in the
Intermountain Region

Riparian woodlands in valleys and can-
yons, pinyon woodlands on mountainsides,
and coniferous forests and aspen on moun-
taintops comprise the major habitats for
birds in the region between the axis of the
Sierra Nevada-Cascade Range and the
Rocky Mountains. Analyses of the richness
and composition of selected avifaunas
breeding in sample areas of these woodlands
and forests expose several patterns that
elucidate the fundamental organization of
avian distribution in the region. Moreover,
these samples along transects demonstrate
the complex modes by which historical as-
pects of avifaunal distribution influence
present-day patterns of species richness and
community structure. Figure 1 shows the lo-
cation of the sample areas and their basic
avifaunal relationships.

Latitudinal Variation in
Riparian Woodland Avifaunas

Bird species requiring riparian vegetation
for breeding occur discontinuously in cot-
tonwoods, willows, ashes, and associated
thickets along watercourses and at valley
oases throughout the Great Basin region.
When species of land birds breeding in such
habitats are compared (Table 1), among 11
sample areas from the Snake River drainage
in the north to the Colorado River drainage
in the south, prominent differences and
some unexpected similarities emerge in total
species numbers, composition of the avi-
faunas according to evolutionary derivation,
and structure of the avifaunas by ecologic
roles. That these three topics are inter-
related will become evident in the dis-
cussion to follow.

Total species numbers.— Of immediate in-
terest is the striking similarity of the geo-
graphically extreme stations, Central Idaho
and Colorado River, in species richness, 28
versus 25, from the pool of 43 species. But,
upon entering the Great Basin from the
Snake River drainage, species totals drop, to

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

139

21 in the Elko sample, 17 in Humboldt, 15
in Toiyabe, and to 12 in Kawich. Similarly,
moving north from the Colorado River, to-
tal species numbers fall to 13 in Pahranagat,
12 in Meadow Valley Wash, and 11 at Ash
Meadows-Pahrump. Figure 2 demonstrates
the strong and abrupt avifaunal attenuation

that occurs with increasing distance from
the lush riparian woodlands to the north
and south. The smallest tallies, for the Kaw-
ich area and Ash Meadows-Pahrump, prob-
ably reflect habitat impoverishment in view
of the limited extent and interrupted
growth of the riparian clumps there. For

Fig. 1. Sample areas of riparian woodland (squares and circles) and pinyon woodland (triangles) avifaunas in
the Intermountain Region. Shading grossly defines regions of forest and woodland. Numbers in squares or circles
correspond to those of sample areas in Table 1. Arrows indicate closest avifaunal relationships. Northern and
southern pinyon areas are denoted by solid and open triangles, respectively. Approximate boundary of northern
and southern riparian woodland avifaunas is shown by the dashed line between samples 7 and 8.

140

GREAT BASIN NATURALIST MEMOIRS

No. 2

Table 1. Species breeding in riparian woodlands at selected localities in the Intermountain Region'.

123456789 10 11

s£

c3 a-

Northern Element

Coccyzus erythropthalmus

X

Dendrocopos villosus2

X

X

X

X

X

X

X

-

-

-

-

Dendwcopos pubescens2

X

X

X

X1

-

X

-

-

-

-

-

Tyrannus tyrannus

X

-

X

— '

Iridoprocne bicolor

X

X

-

X

-

X

-

-

-

-

-

Parus atricapillus1

X

X

Troglodytes aedon

X

X

X

X

X

X

X

-

-

-

-

DumetelLi carolinensis

X

Turdus migratorius

X

X

X

X

X

X

X

-

-

-

-

Catharus fuscescens

X

X

-

-

-

X

-

-

-

-

-

Catharus ustulatus

X

X

X

X

X

X

-

-

-

-

-

Vireo gilvus

X

X

X

X

X

X

X

-

-

-

-

Vireo olivaceus

X

Vermivora celata

X

X

X

-

X

X

-

-

-

-

-

Oporornis tolmiei

X

X

X

X

X

X

X

-

-

-

-

Stetopliaga ruticilla

X

Passerella iliaca

X

X

X

X

X

X

-

-

-

-

-

Southern Element

Zenaida asiatica

-'

-'

X

Dendrocopos scalaris2

X

X

-

X

Centurus uropygialis'

X

Sayomis nigricans

X

X

-

X

Pyrocephalus rubinus2

X

-

X

X

Thryomanes bewickii2

-

-

-

X'

-

-

-

X

X

X

X

Toxostoma dorsale2

X

X

Sialia mexicana

-

-

-

X1

-

-

-

-

-

-

-

Phainopepla nitens2'

-5

-'

X

Vireo bellii

x<

X

Vermivora luciae

-5

X

-

X

Icterus cucullatus

X

X

X

Piranga rubra

-5

X

Guiraca caerulea

-5

X

X

X

X

Pipilo aberti

X

-

X

Widespread Element

Falco sparverius2

X

X

X

X

X

X

X

X

X

X

X

Coccyzus americanus

X

-

-

X

-

-

-

X

-

X

X

Otus asio2

X

-

-

X

-

-

-

X

X

-

X

Archilochus alexandri

X

X

-

X

-

X

-

X

-

-

X

Colaptes auratus2

X

X

X

X

X

X

X

-

X

-

X

Tyrannus verticalis

X

X

X

X

X

X

X

X

X

X

X

Empidonax traillii

X

X

-5

X

-5

-

-

-

-

X

X

Dendroica petechia

X

X

X

X

X

X

X

-

-

-

X

Icteria virens

X

X

X

X

X

X

X

X

-

-

X

Icterus galbula

X

X

X

X

X

X

X

X

X

X

X

Melospiza melodia2

X

X

X

X'

X

X

X

X

-

-

X

Totals

28

21

17

22

15

19

12

13

12

11

25

'Sources of data on occurrence as follows: (1) Central Idaho, a composite list from several localities in the center of the state, taken from Burleigh
(1972); (2) Elko, a composite list from field notes and specimens in the Museum of Vertebrate Zoology ( = MVZ) from the northern portion of Elko
County, and from Linsdale (1936); (3) Humboldt, from Tavlor (1912); (4) Truckee-Carson. a composite list from original data of author (= NKJ) and
from Linsdale (1936); (5) Toivabe, from Linsdale (1938); (6) Ely, data from Spring and Steptoe valleys, NKJ; (7) Kaw.ch, NKJ; (8) Pahranagat, NKJ and
MVZ; (9) Meadow Valley Wash. NKJ and Linsdale (1936); (10) Ash Meadows-Pahrump; (11) Colorado River, NKJ and Grinnell and Miller (1944). Spe-
cies that typically prefer or require riparian woodland are included.

'Permanently-resident species.

'Form shows closest affinity to populations in California, to the northwest.

'Former or casual occurrence.

'Recorded during the breeding period, but not proved to be nesting.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

141

Ash Meadows-Pahrump and Pahranagat
samples, we also may be seeing the effects
of insularity. Several Colorado River drain-
age species, such as the Abert's Towhee (a
highly sedentary, permanently resident
form), probably have had great difficulty
colonizing the isolated riparian areas to the
north. However, one could also argue that
this and other potential colonists are missing
from these remote areas because the proper
kinds of riparian habitats, as found along
the Colorado River and its tributaries, are
lacking.

The moderate size (22 species) of the
Truckee-Carson sample reflects its position
at the well-watered eastern base of the Car-
son Range.

Origin of riparian woodland avifaunas.—
The entire riparian woodland avifauna can
be divided on the basis of evolutionary

D widespread ED northern £/] southern

I. Central Idaho

39%

50

43

42

47

58

45

42 :

6?

44%

4 Truckee-Carson

2 Elko

6 Ely

3 Humboldt

5 Toiyabe

7 Kawich

10 Ash Meadows-Pahrump
9 Meadow Valley Wash

8 Pahranagat

II. Colorado River

5 15 25

Density and origin of riparian woodland

bird species

Fig. 2. Percent composition, according to geographic
derivation, of avifaunas in 11 riparian woodland
sample areas. These are arranged in order of decreas-
ing avifaunal size (species density) toward the bound-
ary of northern and southern avifaunas, between Ka-
wich and Ash Meadows-Pahrump. For example, the
Colorado River sample has 25 species, 11 (= 44 per-
cent of which are widespread and 14 ( = 56 percent)
of southern derivation.

background or derivation into (1) a northern
element consisting of species with northern
centers of distribution and affinities, (2) a
southern element comprised of species with
southern distributional relationships, and (3)
a widespread element including species oc-
curring throughout the western United
States or beyond. The species are grouped
in Table 1 according to these three ele-
ments. Note that none of the species of the
northern element occurs in any samples
south of Kawich and, with one exception
(see below), none of the species of the
southern element is found in any samples
north of Meadow Valley Wash or Pahrana-
gat. Therefore, the two major avifaunas ap-
proach, but do not contact and intermix, in
the southern Intermountain Region.

Although basically of northern derivation,
the riparian avifauna of the Truckee-Carson
sample is of special interest because it con-
tains a distinctive group of four species
with affinities to populations of northeastern
California. One of these species, the Downy
Woodpecker, is of northern affinity; two
species, the Western Bluebird and Bewick's
Wren, are of southern derivation. The Song
Sparrow is widely distributed. The western
Nevadan populations of these four species
probably originated from stocks in the Cen-
tral Valley of California, stocks that spread
into the western Great Basin via the Modoc
Plateau.

The widespread component comprises be-
tween 39 and 62 percent of each sample
(Fig. 2). No clear evidence of differential
decline of northern (or southern) versus
widespread elements is shown as the zone
of impoverishment is approached. Thus, the
diminution is general and involves approx-
imately equivalent shrinkage of the two ma-
jor components of each avifauna.

Community structure and ecologic roles.—
A surprising finding is that the widespread
species fraction seems to have stabilized, at
either seven or five species, for the five
sample areas with the smallest numbers of
species. Thus, the Humboldt, Toiyabe, and
Kawich samples have seven widespread spe-

142

GREAT BASIN NATURALIST MEMOIRS

No. 2

cies in addition to northern ones. And, the
lists of species are identical! The Ash
Meadows-Pahrump and Meadow Valley
Wash samples include southern species and
five that are widespread. However, in these
two samples the coincidence in identity of
widespread forms is incomplete; each shares
three of their five species.

Six of the 11 widespread species, Falco
sparverius, Colaptes auratus, Tyrannus verti-
cals, Icteria cirens, Icterus galbula, and
Melospiza melodia, occur in nine or more
sample areas. Because these species com-
monly coexist, they form a group of great
interest. I view them as a "standard riparian
woodland species" group analogous to the
"standard Boreal species" group I identified
in montane habitats of the western United
States (Johnson 1975). Note that each of the
six species is in a different family. There-
fore, coexistence in the riparian woodland
avifaunas examined here predominates
among species of fundamentally different
body designs.

Included in the tallies of Table 1 are five
pairs of closely related species, three pairs
of which are congeneric. Furthermore,
within each pair, each species is from a dif-
ferent distributional element. The pairs are
Coccyzus erythropthalmus, and C. american-
us, Dendrocopos pubescens and D. scalaris,
Tyrannus tyrannus and T. verticalis, Troglo-
dytes aedon and Thryomunes bewickii, and
Dendroica petechia and Vennivora luciae.
Not all of the species of each pair contact
locally, but the interactions of those that do
meet would merit close study. Indeed, a
fertile field awaits the ecomor-
phologist willing to analyze community
structure, foraging roles, and morphologic
adaptations among the coexisting species of
birds in any of the riparian woodlands dis-
cussed.

Latitudinal Variation in
Pinyon Woodland Avifaunas

Among arboreal habitats, groves of single-
leaf pinyon (Pinus monophylla) comprise
the most extensive formation in the Inter-

mountain Begion, and these woodlands
therefore represent an important habitat for
breeding birds. Typically this pine is scat-
tered over mountain slopes in open stands
or clumps of small trees with much inter-
vening brush. Occasionally, on favorable
sites, growth can be luxuriant, and true for-
est is achieved. Here trees grow to heights
of 10 m or more, and the closed canopy
shades a needle-strewn forest floor. This
site-to-site variability in growth form of pi-
nyon intrigues the ornithologist because, as
has long been known, birds clearly respond
to physical features of vegetation. A second
noteworthy aspect of pinyon is the wide
elevational range often occupied on a single
mountainside. In the White Mountains, Cal-
ifornia, for example, pinyon occurs from ap-
proximately 2000 to 2900 m, with the best
development between 2450 and 2600 m (St.
Andre et al. 1965).

In the Grapevine Mountains of southern
Nye County, Nevada, these variable charac-
teristics of pinyon are seen clearly (Miller
1946). Small, scrubby trees grow in scat-
tered stands, exceptionally fine tracts of old
trees with trunks two feet in diameter form
forests locally on the high northeast slopes,
and the pinyon zone is wide, extending to
2650 m at the top of the highest peak. The
breeding avifauna has responded to these
differences. The complement of species that
normally lives in pinyon of usual form and
spacing is present, and several kinds of birds
that typically avoid this plant formation
breed in the most luxuriant stands. Evi-
dently the latter species are attracted to
physiognomic features of the pinyons that
resemble those of the coniferous forest they
ordinarily inhabit. Furthermore, at
least three additional species that otherwise
would be absent occur in the cool upper
reaches of the zone, apparently responding
there to preferred or required summer tem-
perature regimes. I also have noted similar
occurrences of birds in unusually dominant
pinyon woodland in the Kawich Mountains
of central Nye County, 100 km north-
northeast of the Grapevines (Johnson 1956).

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

143

In recent years I have explored additional
areas of heavy pinyon that invite avifaunal
comparison with the groves in the Grape-
vines. These stands of old trees grow local-
ly, in Nevada near the California line, in
the Palmetto Mountains ( = Mount Ma-
gruder plus the Silver Peak Mountains), Es-
meralda County; the Wassuk Range, Miner-
al County; the Pine Grove Hills, Lyon
County; and in the Sweetwater Mountains,
Douglas and Lyon counties. In each of
these regions, open woodlands of small trees
and stands of intermediate height and den-
sity predominate on most slopes; the forests
are more local in distribution.

For purposes of comparison of the avi-
faunas, it is useful to consider these several
pinyon areas as comprising (1) a northern
subset occupying the spur ranges connected
to the east side of the Sierra Nevada, which
includes the Wassuk and Sweetwater moun-
tains and the Pine Grove Hills, and (2) a
southern subset of wooded islands lying to
the southeast, including the Palmetto and
Grapevine mountains. The massive White
Mountains lie between the two groups. The
northern and southern subsets (Fig. 1) per-
mit comparison of avifaunas along a north-
west to southeast gradient. Some striking
differences in avifaunal composition and nu-
merical relationships of species emerge
(Table 2). Although the northern and south-
ern localities are only 90 km apart, several
species reach their distributional limits be-
tween these mountain systems and thus
breed in only one of the two regions. Unex-
pectedly, 19 species change in average pop-
ulation density along the gradient. For some
(e.g., Empidonax ivrightii), the change is
subtle. For others (e.g., Vireo solitarius)
population densities are dramatically differ-
ent between northern and southern moun-
tains.

Several species partly or completely
switch habitats. For example, Thryomanes
bewickii and Otus asio, common and un-
common residents, respectively, in the pi-
nyon woodlands of the southern Great Basin
and northern Mojave Desert, prefer riparian

cottonwoods and willow in west-central Ne-
vada. Although both species use pinyon at
the more northerly sites, the wren is un-
common, and the owl is rare in this associ-
ation. Empidonax wrightii nests in tall
sagebrush, bitterbrush, and juniper in the
northwestern Great Basin where pinyon is
lacking but is scarce or absent in such vege-
tation in southern Nevada. In other species
the habitat change is more subtle. Parus in-
ornatus and Dendroica nigrescens inhabit pi-
nyon in both northern and southern local-
ities, but their numbers diminish in the
north where they prefer the more arid and
warmer portions of the pinyon belt. In 5 of
the 19 species, the change in density be-
tween north and south coincides with a
shift in subspecific status and, in 3 species,
with a concurrent change in habitat prefer-
ence (Table 2).

How are we to interpret these contrasts
in the avifaunas of the pinyon association of
closely adjacent geographic regions? Two
interrelated issues, (1) variation in the cli-
matic aspects of the pinyon zone environ-
ment and (2) temperature-moisture prefer-
enda and distributional histories of the bird

Table 2. Species composition and numerical rela-
tionships of pinyon woodland avifaunas of spur ranges
of the northern Sierra Nevada1 versus those of mon-
tane islands to the southeast2. Boreal (B) and Austral
(A) species are distinguished.

Species present only in north:

Glaucidium gnomu3(B)
Sphyraupius ruber3 (B)
Cyarwcitta stelleri (B)
Species present in both areas but
Aegolius acadicus (B)
Dendrocopos villosus4 (B)
Tachycineta thalassina (B)
Parus gambei45 (B)
Sirto carolinensis (B)
Species present in both areas but
Otus asio*5 (A)
Empidonax wrightii5 (A)
Aphelocoma coerulescens (A)
Psaltripurus minimus4 (A)
Species present only in south:
Regulus calendula3 (B)
Vireo vicinior (A)

Certhia familiaris3 (B)
Myadestes townsendi (B)

more numerous in north:

Sialia currucoides (B)
Piranga ludoviciana (B)
Carpodacus cassinii (B)
Chlorura chlorura (B)
Junco huemalis (B)
more numerous in south:
Thyromanes bewickii45 (A)
Polioptila caerulea5 (A)
Vireo soliarius (B)

Icterus parisorum (A)
Junco caniceps (B)

'Wassuk Range, Sweetwater Mountains, and Pine Grove Hills.

2Grapevine Mountains and Palmetto Mountains (= Mt. Magruder +
Silver Peak Mountains).

3Breeds onlv verv locally in pinyon woodland in the northern or south-
em mountains.

4Subspecies in north different from that in south

5Partial or complete habitat shift from north to south.

144

GREAT BASIN NATURALIST MEMOIRS

No. 2

species at northern and southern stations,
may offer partial explanation.

From rainfall data for the nearby White
Mountains and several other pinyon sites in
the southwest, St. Andre et al. (1965) con-
cluded that overall aridity characterizes this
plant formation. Unfortunately, no moisture
data exist for the pinyon zones in the
ranges of special interest here. Nonetheless,
pinyon certainly occupies a range of rainfall
regimes, and the more northern sites, al-
though still relatively arid, presumably have
greater average rainfall than the southern
sites. Temperature data also are scarce, ex-
cept for the White Mountains. However, I
can conclude at least that the upper part of
the zone averages considerably cooler than
the lower part, a fact readily apparent to
the researcher working at different eleva-
tions in this region. The unusual upper ex-
tension of the pinyon zone in the White
Mountains (St. Andre et al. 1965) and in the
Grapevine Mountains (Miller 1946) would
enhance temperature differences that nor-
mally exist between the different elevations.
Finally, because of the difference in lati-
tude, it is probable that the northern sites
average cooler than the southern ones.

The climatic preferences and distribution-
al backgrounds of the bird species agree
closely with the groupings of Table 2. The
5 species confined to the north and all 10
forms that are more numerous in the north
are Boreal species with distributional back-
grounds associated with northern coniferous
forests. In contrast, except for Regulus cal-
endula, Vireo solitarius, and Junco caniceps,
forms of Boreal derivation (Miller 1951), all
the species present only in the south or
reaching their greatest numbers in the
southern pinyon areas are of Austral origin,
having been associated in their distribution-
al histories with southwestern pinyon or oak
woodlands. I conclude that this major trend
in composition and density of pinyon zone
avifaunas along the Nevada-California bor-
der apparently results from the exchange
along a temperature-moisture gradient of
southern pinyon woodland species, adapted

principally to the warm and relatively arid
portions of the zone, and northern species,
adapted chiefly to the relatively cool and
moist portions of the zone.

The origin of the bird species of the
Wassuks, Sweetwaters, and Pine Grove Hills
is of particular interest because the pinyon
woods in these ranges directly contact the
coniferous forests of the Sierra Nevada. For
this reason, bird populations of the pine-fir
zone of the east side of the latter range
have had easy access to pinyon forests lying
to the east. Ready access may be one reason
why Cyanocitta stelleri and, locally, Certhia
familiaris and Sphyrapicus ruber breed in
pinyon in these areas. Otherwise these spe-
cies are not known to nest in this forma-
tion. It is noteworthy that none of these
forms inhabits the old pinyon groves in the
Desatoya Mountains of central Nevada, pre-
sumably because of the great distance of
this range from source populations.

Note that from the pool of 27 pinyon
zone species, nearly identical total numbers
of breeding forms, 23 and 24, respectively,
are recorded for northern and southern sta-
tions. Such equivalence is significant. Im-
portantly, no neat system of replacement or
of switches in abundance by species with
similar ecologic roles occurs along the
gradient (Table 2). Instead, the evolution of
community structure at various localities in
the pinyon formation probably has involved
more subtle adjustments among the species
in foraging zones, food quality, and /or habi-
tat preference. Those species of birds repre-
sented by different geographic races at
northern and southern localities would be
especially worthy of close study in this re-
gard.

Gradients in Boreal Avifaunas

I recently discussed controls of richness of
Boreal species on 11 sample areas in the
Sierra Nevada, Cascades, and Rocky Moun-
tains and on 20 mountaintop islands in the
Intermountain Region (Johnson 1975). Here
I wish to expand on portions of that analy-
sis through examination of both longitudinal
and latitudinal trends in numbers of species

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

145

occupying particular Boreal habitats along-
four transects across the Great Basin. These
transects (Fig. 3, lines I-IV) are roughly
equidistant latitudinally and pass through 20
of the sample areas described in my pre-
vious paper. Species numbers now have
been determined for eight additional areas,
and these new data are included. Transect
HI fails to cross the entire Intermountain
Region because no species lists are available
for restricted portions of the main Rocky
Mountains directly east of the Snake Range
in eastern Nevada.

Total species richness.— All transects
demonstrate the abrupt drop in total num-
bers of species in the north-south-oriented
swath that extends across much of western
and central Nevada (Fig. 4A). From high
species totals of 56 to 64 at Mt. Shasta,

Lassen, Tahoe, and Yosemite, species rich-
ness declines sharply along transect I, be-
tween Warner and Pine Forest; along trans-
ect II, between Lassen and West Humboldt;
along transect III, between Carson and Pine
Nut; and along transect IV, between Yose-
mite and Glass Mountain. Species numbers
then climb toward the east to totals of 56
and 55 at Uinta and Aquarius-Boulder, re-
spectively, at the ends of the tran-
sects. Highest totals for the Sierra Nevada-
Cascades always equal or exceed those for
the Rocky Mountains. In the region of gen-
eral impoverishment along each transect, a
mountain range moderately rich in Boreal
species interrupts the general smoothness of
each species density curve (Fig. 4A): I, Jar-
bidge; II, Ruby; III, Toiyabe-Shoshone; and
IV, White-Invo.

Fig. 3. Four transects across the Intermountain Region, passing through 28 sample areas of Boreal avifaunas.
These sample areas are named in Fig. 4. No native firs are known to occur between the two vertical dashed
lines.

146

GREAT BASIN NATURALIST MEMOIRS

No. 2

Species richness according to habitat.— In
Figure 4B I portray the percentage of Bo-
real species in each of four major habitats
(Johnson 1975: 549), i.e., Aquatic, Wet
Meadow, Riparian Woodland, and Con-
iferous Forest. This permits study of pos-
sible changes along gradients of proportions
of species preferring these habitats. For

each transect I have calculated the average
percentage of species in each of the four
habitats for Sierra Nevada-Cascades sam-
ples, Great Basin Islands samples, and
Rocky Mountains samples (Table 3).

Between-transect comparisons of propor-
tions of species in continental samples in-
dicate striking similarities. For example,

rz

II II I If

" a> S>i2 2s o
o £ .£= o -c

oIq.q.1 t- en

♦-Aquatic

Wet Meadow
• "^—Riparian
Woodland

.*^ Coniferous
Forest

A. Boreal bird species richness

B. Composition of boreal avifaunas

Fig. 4. Numbers of Boreal bird species (left) and percent avifaunal composition by four major habitats (right)
along four transects across the Intermountain Region. Continental sample areas are underlined; island sample
areas are not. Twenty sample areas are described more hilly in Johnson (1975). Species totals for eight new ones
added are as follows: A, Mt. Shasta (Grinnell and Miller 1944); I, West Humboldt Range (author's unpubl. data);
L. Wasatch (Behle and Perry 1975); M, Tahoe (Orr and Moffitt 1971); O, Pine Nut Mtns. (author's unpubl. data);
R, White Pine Mtns. (author's unpubl. data); S, Schell Creek-Egan Ranges (author's unpubl. data); and Y, Pahute
Mesa (Hayward et al. 1963).

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

147

number of species in coniferous forests var-
ies only from 62.5 to 69.6 percent of the to-
tal Boreal avifauna in the Sierra Nevada-
Cascades, and those of riparian woodland
vary only from 23.2 to 24.6 percent in the
several samples from the Rocky Mountains.
Other habitats on continents illustrate sim-
ilar uniformity. Island samples are more
complex and show differences among trans-
ects. Specifically, the island samples of
transects I and II have proportionately
more aquatic and riparian woodland spe-
cies, but fewer coniferous forest species,
than those of transects III and IV.

Within-transect comparisons (transects I,
II, and III) point to more riparian woodland
species and fewer coniferous forest species
in samples for the Great Basin Islands ver-
sus the Sierra Nevada or the Rocky Moun-
tains. However, the ecologic preferences of

birds in island samples resemble those of
birds in Rocky Mountains samples more
than those of forms in the Sierra Nevada-
Cascades. The avifaunas of transect IV do
not follow the pattern of the first three ir.
that the proportions of species in the sever-
al major habitats are similar among samples
of the Sierra Nevada, Great Basin, and
Rocky Mountains.

The composition of coniferous forest avi-
faunas.— The fraction of coniferous forest
bird species deserves special comment be-
cause some of the trends in total species
richness discussed earlier are explained part-
ly by the distribution of suitable tree spe-
cies. The following group of forest bird spe-
cies, distributed discontinuously in the
Intermountain Region, illustrates this point:
Sphyrapicus varius, Sphyrapicus thyroideus,
Empidonax hatnmondii, Empidonax difficilis,

Table 3. Average percent composition, according to habitat preferences, of species in avifaunas along trans-
ects.

Transect I

Aquatic-
Wet Meadow
Riparian Woodland
Coniferous Forest

Transect II

Aquatic
Wet Meadow
Riparian Woodland
Coniferous Forest

Transect HI

Aquatic
Wet Meadow
Riparian Woodland
Coniferous Forest

Transect IV

Aquatic-
Wet Meadow
Riparian Woodland
Coniferous Forest

Sierra

Rocky

vada-Cascades

Great Basin

Mountains

Continent

Islands

Continent

(Mt. Shasta)

(Warner through
Raft River)

(Uinta)

3.6

3.7

5.3

12.5

10.8

16.1

14.3

34.0

23.2

69.6

51.5
(W Humboldt

55.4

(Lassen)

through

(Wasatch and

Spruce-S. Pequop)

Uinta)

7.8

6.3

3.6

12.5

4.0

17.3

17.2

37.1

24.6

62.5

52.6
(Pine Nut

54.6

(Tahoe and

through

Carson)

Snake)

4.4

0.9

-

14.8

10.8

-

16.5

27.4

-

64.4

60.9

-

(Yosemite

(White-Inyo

and Glass

through

(Aquarius-

Mountain)

Highland)

Boulder)

2.3

0.6

3.7

12.3

7.0

14.5

19.0

23.4

23.6

66.4

69.0

58.2

148

GREAT BASIN NATURALIST MEMOIRS

No. 2

Sitta canadensis, Certhia familiaris, Regains
satrapa, Regulus calendula, Hesperiphona
vespertina, and Spinas pinus. Although most
of these forms occasionally breed in other
kinds of conifers or in deciduous trees, they
prefer firs (Abies concolor or Pseadotsaga
menziesii) and avoid or nest sparingly in
forests lacking these trees. Thus, the distri-
bution of firs in the Intermountain Region
is of special significance for these species.
Figure 4A indicates the presence of firs in
the various sample areas. Note the repeated
instances of increased numbers of species, of
all kinds of birds and of coniferous forest
birds, in relation to the appearance of firs.

Firs are absent from the zone of avifaunal
impoverishment across western and central
Nevada (Fig. 3). Although skimpy clumps
grow in the Sweetwater Mountains, no firs
occur in any of the other ranges along the
California-Nevada border southeast of the
Carson Range. Firs reappear in eastern Ne-
vada between 115° and 116° W Longitude.
Among the ranges considered here, sub-
stantial stands occur in the Jarbidge Moun-
tains, White Pine Mountains, and on Mount
Irish. Critchfield and Allenbaugh (1969) re-
port white fir locally in the Ruby Moun-
tains, but the avifauna at that site has not
been explored.

Rristlecone pine (Pinus longaeva) domi-
nates the upper slopes of several ranges in
eastern Nevada. This tree is important for
coniferous forest birds in part because it oc-
casionally forms closed, well-shaded stands
with strong physiognomic resemblance to
firs. Apparently in response to this feature,
several kinds of birds from the above list of-
ten occupy well-developed bristlecone pine
forests when firs are scarce or lacking. The
two species of Sphyrapicus, Empidonax dif-
ficilis, Sitta canadensis, and Regains calen-
dula exemplify this point. In Nevada, the
best stands of this conifer grow east of the
right vertical dashed line on Figure 3 that
marks the boundary of white fir distribu-
tion, and in many places the two species
occur in mixed stands.

Breeding Range Boundaries and

Interspecific Distributional

Relationships

Broad trends in avian geography are best
identified through comparison of major
components of avifaunas. Other, more de-
tailed issues can be seen most clearly by in-
spection of the distributions of single spe-
cies and their close relatives. Here I discuss
four contrasting examples, chosen from a
long list of possible cases, that illustrate
particular species-level distributional phe-
nomena occurring among birds of the Inter-
mountain Region. Three of these examples
also support the view offered earlier that
this region presently serves as an important
arena of disjunction, contact, and inter-
mixture of species representing avifaunas of
differing geographic derivation.

Grace's Warbler and ponderosa pine.— In
southeastern Nevada, as in the southern
Rocky Mountains generally, Grace's War-
bler (Dendroica graciae) breeds commonly
and exclusively in forests of ponderosa pine
(Pinus ponderosa). But near the north-
western edge of the occurrence of the
Rocky Mountain form of the ponderosa
pine (P. ponderosa var. scopulorum), for ex-
ample in the Quinn Canyon Range, the
scattered and stunted trees occupy marginal
sites, and the bird is rare. It is absent from
the few ponderosas that grow locally in the
Highland Range, on Mount Wilson, and in
the southern White Pine Mountains, and,
farther north, from the extensive ponderosa
pine habitat in the Snake Range (Fig. 5).

This case is particularly interesting be-
cause prior to 1963 the Grace's Warbler
was not known to breed anywhere in south-
ern Nevada. Since then, numerous records
have been obtained, several from places
that previously had been well-explored
(Johnson 1965, 1973, 1974). Therefore, the
evidence points to an active northwestward
range expansion in the past decade or two.
Although further extension into presently
unoccupied ponderosa pine habitat is pos-
sible, total distributional evidence suggests

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

149

that the approximate northwestern limit in
the Intermountain Region already has been
reached. However, given continued range
expansion, eventual colonization of the arid
ponderosa pine forests of southern Califor-
nia seems probable, for the species is al-
ready vagrant to that area, and a number of
other species of the interior Southwest have
become established there in recent years
(Johnson and Garrett 1974). Insofar as is
known, Grace's Warbler has never nested in
the vast ponderosa pine forests of the cen-
tral and northern Rockies, or in the Sierra
Nevada.

The warbler thus illustrates two bio-
geographic phenomena typically seen in
species with well-documented distributions:
(1) the actual range falls short of the geo-
graphic extent of seemingly appropriate
habitat and (2) the distributional boundaries
are unstable because of their complex con-

A Grace's Warbler v

0 Ponderosa Pine "•

Fig. 5. Approximate breeding distribution of Grace's
Warbler (triangles) and occurrence of ponderosa pine
(stippled areas) in southeastern Nevada and south-
western Utah. Lower elevational perimeters of wood-
lands in mountains also are indicated.

trol by a multiplicity of interacting factors
that are themselves fluctuating. I wish to
emphasize that no other species of warbler
or other pine-foliage insectivore obviously
competes with D. graciae and, for this rea-
son, has adjusted its distribution and abun-
dance as a result of recent population inter-
action with that species. Instead, the
northwestern distributional perimeter of
Grace's Warbler seems to be set by other
aspects of the environment. For example,
this warbler may be dependent upon popu-
lations of pine arthropods that fluctuate in
annual density, especially near the per-
iphery of their ranges. Or, summer temper-
ature and moisture barriers dictated by the
metabolic requirements and preferenda of
the warbler, the arthropods upon which it
feeds, or the pines they both inhabit, may
control the range border. I have no data to
support a convincing explanation.

Downy Woodpecker and Ladder-backed
Woodpecker in disjunct allopatry.— These
two congeners fail to contact in the Great
Basin. The Downy inhabits aspen groves in
the northern and eastern mountains (the
form D. p. leucurus), or cottonwoods and
willows along rivers in west-central Nevada
(the form D. p. turati), and the Ladder-back
lives in the cottonwood-willow association
in valleys or Joshua tree woodland in the
Mojave Desert. Unexpectedly, substantial
areas of apparently suitable habitat occur in
the wide zone now unoccupied by either
species. Thus, the Downy Woodpecker has
not been found in the fine aspen woodland
in the Schell Creek Range or on Mount
Wilson, among a number of seemingly ap-
propriate places. Similarly, the Ladder-back
avoids considerable Joshua tree growth in
Esmeralda and Nye Counties, Nevada. Fi-
nally, neither species breeds in the excellent
riparian woods in Owens Valley, although
the Ladder-back has been recorded at Lone
Pine, at the foot of the valley (Grinnell and
Miller 1944). I conclude that the respective
southern and northern limits of these two
species, within suitable habitats, are dic-
tated primarily by temperature and mois-

150

GREAT BASIN NATURALIST MEMOIRS

No. 2

ture. Considering the wide unoccupied
swath between their ranges in the central
Intermountain Region, distributional limits
there certainly are not controlled by inter-
specific competition.

Mountain Bluebird and Western Bluebird
in parapatry.— These two congeneric-
thrushes (Sialia currucoides and S. mexicana,
respectively) breed in contiguous allopatry,
or parapatry, in much of the southern Inter-
mountain Region. One might conclude that
such a distributional relationship is the re-
sult of interspecific competition (Fig. 6). In
at least three general areas where both spe-
cies have been recorded (Panamint Range,
Delamar Range, Zion National Park), their
spatial relationships are unclear. But, be-
cause of their well-known differences in
habitat preference (Grinnell and Miller
1944), competitive interaction probably is
infrequent or absent, given a situation of lo-
cal sympatry. Until more details are avail-
able, I conclude that the striking parapatry
results from the geographic position of a

sharp shift in average spring and summer
temperatures. The approximate coincidence
of the zone of parapatry with the northern
limits of the Mojave Desert supports this
suggestion.

Hammond's Flycatcher and Western Fly-
catcher in broad sympatry.— As in the pre-
vious examples, this case also concerns a
species of northern derivation, Hammond's
Flycatcher (Empidonax hammondii), and a
congener of southern derivation, the West-
ern Flycatcher (E. difficilis). These two
forms offer a particularly impressive ex-
ample of how the complex interplay of
phylogenetic and distributional history has
influenced present-day habitat selection and
ecologic interaction. Figure 7 shows the dis-
tributional and numerical relationships of
these species in western North America.

Hammond's Flycatcher is commonest in
cool, moist Boreal forests of the Sierra Ne-
vada, Cascades, and Rocky Mountains; its
principal range therefore surrounds the In-
termountain Region on three sides. In con-

A Mountain Bluebird
A Western Bluebird

Fig. 6. Breeding distributional relationship of the Mountain Bluebird {Sialia currucoides) and Western Bluebird
(S. mexicana bairdi) in southern Nevada and adjacent states.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

151

trast, the Western Flycatcher reaches its Mountains of Arizona and New Mexico
greatest densities in warm Pacific Coastal (large form E. d. hellmayri), where E. ham-
forests (small form E. d. difficilis) and in mondii is either absent or rare and local. In
arid interior highland forests of the Inter- areas of contact between Hammond's Fly-
mountain Region and southern Rocky catcher and the similarly sized E. d. diffi-

Empidonax difficilis
£. d. difficilis
E. d. hellmayri

Fig. 7. Breeding distributional relationship of Hammond's Flycatcher (principal range outlined) and Western
Flycatcher in the contiguous western United States. Relative abundance of three forms of Western Flycatcher is
shown by patterns as follows: E. d. difficilis, heavy slanted lines indicate very common to abundant populations,
and light slanted lines indicate common to uncommon populations; E. d. hellmayri, black indicates very common
to abundant populations, and dotted pattern indicates common to uncommon populations. E. d. insulicola is
common to abundant on the California Channel Islands, indicated in black. Scattered localities for F. d. hel-
lmayri are shown by circles; note that many of them occur within the main range of E. hammondii. A few local-
ities for the latter species, located south of the main range, are omitted.

152

GREAT BASIN NATURALIST MEMOIRS

No. 2

cilis in the Pacific Northwest and on the
west side of the Cascades, interspecific ter-
ritories are defended (Johnson 1966 and un-
publ. data). In the Rockies of southern Col-
orado the larger E. d. hellmayri evidently
overlaps in territories with E. hammondii
(Beaver and Baldwin 1975). The important
point is that these two species are largely
allopatric; considering the bulk of their
numbers, this results directly from selection
of similar, though not identical, habitats in
different major climatic zones. Where they
overlap, one species is nearly always com-
mon, the other uncommon or rare. Despite
the geographic shifts in relative abundance
and subtle differences in habitat selection,
which presumably are consequences of mil-
lenia of competitive interaction, the two
species do occur sympatrically in many
places. Obviously one species has not limit-
ed the geographic range of the other (Fig.

7).

Within the genus Empidonax, these two
species are separated phylogenetically by a
series of other forms of closer relationship.
The Hammond's Flycatcher is allied to the
Least Flycatcher (E. minimus) and to other
species of Boreal derivation (Johnson 1963).
The Western Flycatcher, in contrast, is clos-
est to Neotropical montane stocks such as
E. flavescens and is certainly of southern
derivation (N. K. Johnson, unpubl.). Appar-
ently such evolutionary backgrounds have
permitted extensive convergence in habitat
preference and foraging ecology, for in the
western United States, E. hammondii and E.
difficilis are the commonest Empidonaces of
dense coniferous forest.

Controls of Boundaries
of Species' Ranges

The previous accounts offer only the
barest introduction to the vast subject of
the determinants of borders of breeding dis-
tributions of birds. Nonetheless, it is appro-
priate to comment briefly on this topic be-
cause it lies at the heart of biogeography
and because of significant trends in the per-
tinent literature.

Grinnell (1914, 1917) and Bartholomew
(1958), among others, have emphasized
manifold controls of distributional bound-
aries of animal species, with the importance
of particular factors changing along the
range perimeter. They favored niche re-
quirements and physiologic-climatic prefer-
enda as the crucial dictates of species distri-
butions; interspecific competition either was
not mentioned or its relevance was subordi-
nated. In the more recent literature of orni-
thological ecology, the limitation of geogra-
phic range through competitive exclusion
has become a prominent theme (MacArthur
1972, Cody 1974). However, the invocation
of competition in this context has been pre-
mature.

In agreement with Salt (1952), whose
classic paper clearly describes how metabo-
lism and climate interrelate to control the
distributions of three species of finches (Car-
podacas), I see little reason to postulate
competition as an effective determinant of
range borders of species of birds in the In-
termountain Region. The question is not
whether competition occurs among the
many overlapping species here; rather, it is
what form does this competition take, and,
as such, can it influence range boundaries?
From all the evidence I have seen, includ-
ing that from the four examples discussed,
competition probably has had little in-
fluence in the determination of range
boundaries of birds. Instead, alterations in
density, such as shown for the two species
of Empidonax, are the more likely outcomes
of competitive interactions. Finally, I sug-
gest that it is unwise to conclude that
boundaries of tropical species are set by
competition (Terborgh and Weske 1975), in
the virtual absence of data on the occur-
rence of those environmental requisites cru-
cially related to the distribution of the bird
species in question.

Patterns of Avian Speciation
in the intermounta1n region

In this section I concentrate on what I
consider to be active zones of speciation in

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

153

the Great Basin region. These are illustrated
by currently interacting populations of
closely related taxa, well-differentiated at
various levels above that of the incipient
subspecies. For those concerned with more
subtle examples of racial differentiation in
this region, Behle (1963) provides a useful
review. Information on speciation zones is
of significance to the data previously pre-
sented on avifaunal gradients because of the
great probability that the same environmen-
tal discontinuities in climate and flora ulti-
mately are responsible for both.

Zones of Intergradation, Disjunction,
and Secondary Sympatry

Birds demonstrate a particularly wide
range of speciation phenomena along the
western edge of the Great Basin, in a zone
extending south-southeastward from south-
ern Oregon to southeastern California and
southern Nevada. Indeed, there is evidence
from the general distribution and speciation
of birds in western North America that the
zone just delineated is actually part of a
larger biotic separation that extends north-
ward to British Columbia and beyond. Here
I wish to discuss 16 selected pairs of taxa
(Table 4) in the western Intermountain Re-
gion that illustrate stages in the speciation
process. Figure 8 schematically portrays
zones of intergradation and hybridization
for nine of these pairs of taxa.

Several patterns emerge from study of
this wealth of material. The first concerns
the striking, although imprecise, geographic
coincidence of the several contact zones
that link (or divide) coastal and interior
populations. Secondly, a large number of
species are represented, suggesting that sig-
nificant environmental gradients or dis-
continuities in the western Great Basin per-
vasively influenced the differentiation of
birds. Indeed, few other regions in North
America can claim a comparable diversity
of speciation phenomena within such re-
stricted geographic limits.

Furthermore, the interacting taxa of
nearly every pair occupy different habitats

on each side of the contact zone (Table 5).
Habitats in the contact areas themselves are
typically intermediate between coastal and
interior situations. In a major trend illus-
trated by six pairs of taxa, the coastal repre-
sentative inhabits coniferous forest or oak

Table 4. Examples of coastal or Sierra Nevadan and
interior taxa that illustrate speciation phenomena
along the interface of the Cascades-Sierra Nevada and
Great Basin.

Gradual primary integradation; racial boundaries
weak1:

1. Sphyrapicus thyroideus thyroideus and S. t. na-
taliae. Williamson's Sapsucker.

Narrow zone(s) of primary or secondary intergradation
between representatives of strongly divergent racial
complexes:

2. Empidonax difficilis difficilis and E. d. hellmayri2.
Western Flycatcher.

3. Eremophila alpestris sierrae and E. a. lampro-
chroma. Horned Lark.

4. Aphelocoma coendescens superciliosa and A. c. ne-
vadae. Scrub Jay.

5. Psaltriparus minimus californicus and P. m. plumb-
eus or P. m. providentialis. Bush-tit.

6. Amphispiza belli canescens and A. b. nevadensis.
Sage Sparrow.

7. Passerella iliaca megarhynchus and P. i. fidva or P.
i. monoensis. Fox Sparrow.

8. Dendrocopos pubescens turati and D, p. leucurus.
Downy Woodpecker.

Disjunct allopatry between representatives of strongly
divergent racial complexes:

9. Parus inornatus inornatus or P. i. kernensis and P.
i. zaleptus. Plain Titmouse.

10. Vireo solitarius cassinii and V. s. plumbeus. Soli-
tary Vireo.

11. Vermivora celata lutescens and V. c. orestera.
Orange-crowned Warbler.

Narrow zone or sympatry and interspecific hybridiza-
tion:

12. Sphyrapicus ruber daggetti. Red-breasted Sapsucker
and S. varius nuchalis. Red-naped Sapsucker.

13. Junco hyemalis thurberi. Dark-eyed (Oregon) Junco
and /. caniceps caniceps. Gray-headed Junco.

Disjunct allopatry between species:

14. Pica nuttalli. Yellow-billed Magpie and P. pica
hudsonia. Black-billed Magpie.

15. Vermivora ruficapilla ridgwayi. Nashville Warbler
and V. virginiae. Virginia's Warbler.

16. Leucosticte tephrocotis dawsoni. Gray-crowned
Rosy Finch and L. atrata. Black Rosy Finch.

'Many additional examples of this category could be cited (see Behle
1963).

!In each pair of taxa the coastal-Sierran form precedes the interior
form.

154

GREAT BASIN NATURALIST MEMOIRS

No. 2

3

- OS

C —i

o -^ ^-i.

Dix

dale
1914

is

nell,

Lins
•th(

c £

"5 r

Grin
and
Swai

5 s

c ^

O S o s

f S:

N

Z 13 U

« U 0

zu Jai^uS

i * <u

=2 ^ c S o
E .a 5! &h £

5 M
o

U "1

c ™ *

E £ £

B «

■fit

2 <

So £

■S *

3 £ t5

rf s a

K £ 2

E=Q 8 K

0 fc

U 2

1 o S-g

8 £ « .3

3*

11 i 1

d 1

%1 g 1

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

155

oS

£ M

" o
i Z

o 2

< "2

2<u

a- o o

SS^

5 u
3 o

Js s

Zb, r>ozt,>u

I ar 5 J I

>'

<*-

-5

Z

c

_c

o

V

i&u ^

s

c

|

u

X

EB

S MU

-'

V

'B

U

>N

2 J^

a

<

"5

T3

S

—

T3

No contact
approach a
Nev. borde

o

o

-a

a

a)

C

o

-5

5

8

-5

U

as

—
>

z

3

=
o

o

Z

3

2

E
E

o i

s *

a- o
< "3

j« "3 g -d

"B •£ o a " «

§ '5 | ^ o 1

SiS o 3P ^ s s 5

§ | s c B^iX-N

C5WJOC/5>OC«&hO

a. -s,

<1

."8

II

-2 ■*

q §

o S£

8 "8 3

? * 1

-* V. -a

<d o

c -a

== .. <u

II

5 8

Q o 6 A

ssz § > ^ o ice

156

GREAT BASIN NATURALIST MEMOIRS

No. 2

woodland-chaparral, and the interior race or
species lives in pinyon-juniper woodland. In
four additional pairs of taxa, the coastal or
Sierran form occupies oak-chaparral, co-
niferous forest, or brush associated with
such forest, and the interior form switches
to riparian aspen or willows. Earlier I dis-
cussed the latter trend (Johnson 1970).

In 5 of the 16 pairs, the coastal and inte-
rior representatives are disjunct, with ranges
separated by gaps that overlap or straddle
many of the zones of intergradation of the
11 pairs of forms that do contact. Evidently
the intervening habitat separating the dis-
junct forms is unsuitable for these particular
species. Seven of the 11 pairs that meet and
interact occur less commonly in the contact
zone than in the adjacent regions. This

clearly suggests that population densities are
reduced for several species in and near their
zones of intergradation. Thus, the environ-
ments at the western border of the Inter-
mountain Region are unsuitable for certain
species and marginal for others. In any case,
the variety of distributional and speciation
phenomena represented underscores the
steep environmental gradients that influence
selection pressures on bird populations in
the restricted zone where the Cascade
Mountains and Sierra Nevada meet the
Great Basin. Munz and Keck (1968), Bald-
win (1973), and Miller (1951) describe the
significant physiographic, floristic, and cli-
matic changes ultimately responsible for the
evolution of contrasting avifaunas in this re-
gion.

Fig. 8. Left, zones of primary intergradation and hybridization (secondary contact), at the western border of
the Great Basin, in seven species of birds. Numbers correspond to those in Table 4: 2, Western Flycatcher; 3,
Horned Lark; 4, Scrub Jay (two zones); 5, Bush-tit (two zones, one of which is identical with a zone of the Scrub
Jay); 6, Sage Sparrow; 7, Fox Sparrow (two zones); and 8, Downy Woodpecker. Bight, localities of sympatry and
hybridization between Bed-breasted and Bed-naped Sapsuckers (triangles) and between Dark-eyed (Oregon) and
Gray-headed Juncos (dots).

1978 INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM 157

Physiographic Breaks and lations on the west side of the trough at lo-

Avifaunal Boundaries ealities well-removed from their principal

centers of distribution from central Nevada

A major gap in the mountain systems of eastward. These occurrences provide clear

the western United States results from the evidence for the suitability, for at least

low trough that begins in southeastern Ore- some interior species, of those montane en-

gon, courses southward through western Ne- vironments located directly adjacent to the

vada, veers southeastward between Walker southern Cascades and east side of the

Lake and the Paradise Range, then contin- Sierra Nevada. But these same ranges also

ues through the Death Valley area and host many Cascadian and Sierran forms that

beyond into the Mojave and Colorado have colonized from the west. Thus, pecu-

deserts. With such a geographic position, it Harly intermixed avifaunas breed in all four

would be easy to assume that this trough mountain systems (Table 6). One of the four

separates the Boreal avifaunas of the Cas- ranges, the Sweetwaters, has only a minor

cades and Sierra Nevada from those of the representation of interior forms; the others

interior. However, this is not the case, for a have numerous Great Basin-Rocky Mountain

host of interior races and species of birds species. All of the other mountain systems

occur westward to the Warner, Wassuk, along the Nevada-California border, the

Sweetwater, and White mountains. These Carson Range, Pine Nut Mountains, Pine

interior forms therefore maintain popu- Grove Hills, and Glass Mountain, lack inte-

Table 6. Intermixture of Cascade-Sierra Nevadan and interior components1 in Boreal avifaunas of mountain
ranges2 along the western edge of the Great Basin.

Cascade Mountain-Sierra Nevadan Warner Wassuk Sweetwater White

Forms Mountains Range Mountains Mountains

Dendragapus obscurus sierrae

X

-

X

X

Sphyrapicus ruber daggetti

X

X

X

-

Cyanocitta stelleri frontalis

X

X

X

X

Vireo soliturius cassinii

X

-

-

-

Vermivora ruficapilla ridgwayi

X

-

-

-

Wilsonia pusilla chryseola

-

-

X

-

Leucosticte tephrocotis dawsoni

-

-

X

X

Junco hyemalis thurberi

X

X

X

X

Passerella iliaca monoensis2

-

X

X

-

Great Basin-Rocky Mountain Forms

Selasphorus platycercus

-

X

-

X

Sphyrupicus varius nuchalis

X

X

X

X

Dendrocopos pubescens leucurus

X

-

-

-

Empidonax difficilis hellmayri

X

-

-

-*

Vireo soliturius plumbeus

-

X

X

X

Vermivora celata orestera

X

-

-

X

Vermivora virginiae

-

X

-

X

Wilsonia pusilla pileolata

X

-

-

X

Junco caniceps cuniceps

-

-

-

X

Passerella iliaca ftdva

X

—

—

—

Passerella iliaca canescens

-

-

-

X

'Emphasis is on forms of contrasting Sierra Nevadan versus interior distribution. Mam undifferentiated species of widespread distribution in the west-
ern United States are not considered here.

'Sources of data: Warner Mountains, Miller (1951) and Johnson (1970); Wassuk Range, original data of author and few records from Linsdale (1936);
Sweetwater Mountains, specimens in MVZ and original data of author; White Mountains, Grinnell and Miller (1944) and Miller and Russell (1956).

The avifaunal relationships of this form are uncertain.

'Miller and Russell (1956) report E. d. difficilis in late June, but the bird was not certainly nesting even though the specimen was in breeding condi-
tion.

158

GREAT BASIN NATURALIST MEMOIRS

No. 2

rior forms; their Boreal avifaunas are entire-
ly comprised of Sierra Nevadan and wide-
spread species. Because the most distinctive
interior species and racial complexes, such
as Selasphorus platycercus, Vermivora vir-
giniae, and the ivoodhonseii Group of Aph-
elocoma coerulescens, extend all the way
from eastern California to Colorado, there
is no Boreal avifauna typical only of the
Great Basin section of the interior.

The aforementioned distributional evi-
dence thus supports the principal conclusion
derived from the location of contact zones,
namely, that the most trenchant separation
of coastal-Sierran versus interior stocks oc-
curs in a zone of varying width that runs
southward just east of the crest of the Cas-
cades then southeastward along the inter-
face of the Sierra Nevada and Great Basin.
Presumably it is the beginning of the pre-
cipitation shadow and the consequent shift
from a coastal to an interior continental cli-
mate that provide the profound environ-
mental transition responsible for this zone.
The low desert trough to the east does not
represent a significant avifaunal barrier.

Epilogue.— Avian geography in the Inter-
mountain Region continues to progress, al-
beit cautiously, from the descriptive to the
analytic phase. With only crude patterns
now discernible, much remains to be
learned and interpreted. But, heeding E. O.
Wilson's (1970) reminder that ". . .biogeo-
graphy is far and away the most difficult of
all the biological sciences," I am inclined to
treat with great kindness even those tenta-
tive patterns that so far have surfaced from
the chaos.

Acknowledgments

Much of the field work on which the
present synthesis is based was supported by
the Committee on Research, University of
California, Berkeley. Mercedes S. Foster
read a draft of the manuscript and offered
valuable comments. Gene M. Christman
drafted the final versions of the illustrations.
I am indebted to all these individuals for
their kind assistance.

Literature Cited

American Ornithologists' Union. 1957. Checklist
of North American birds, fifth ed. Amer. Orni-
thol. Union, Baltimore, Maryland.

1973. Thirty-second supplement to the

American Ornithologists' Union checklist of
North American birds. Auk 90: 411-419.

Baldwin, J. L. 1973. Climates of the United States.
U.S. Dept. Commerce, Government Printing
Office, Washington, D.C.

Bartholomew, G. A., Jr. 1958. The role of phys-
iology in the distribution of terrestrial verte-
brates, p. 81-95. In: C. L. Hubbs (ed.), Zoogeo-
graphy. Amer. Assoc. Adv. Sci., Washington,
DC.

Beaver, D. L., and P. H. Baldwin.
1975. Ecological overlap and the problem of
competition and sympatry in the western and
Hammond's flycatchers. Condor 77: 1-13.

Behle, W. H. 1942. Distribution and variation of
the horned larks (Otocoris alpestris) of western
North America. Univ. Calif. Publ. Zool. 46:
205-316.

1963. Avifaunistic analysis of the Great Basin

region of North America. Proc. Inter. Ornith.
Congr. 13: 1168-1181.

Behle, W. H., and M. L. Perry. 1975. Utah birds:
checklist, seasonal and ecological occurrence
charts, and guides to bird finding. Utah Mus.
Natural Hist., Univ. Utah, Salt Lake City.

Burleigh, T. D. 1972. Birds of Idaho. The Caxton
Printers, Ltd., Caldwell, Idaho.

Cody, M. L. 1974. Competition and the structure
of bird communities. Princeton Univ. Press,
Princeton, N.J.

Critchfield, W. B., and G. L. Allen-
baugh. 1969. The distribution of Pinaceae in
and near northern Nevada. Madrono 19: 12-26.

Grinnell, J. 1914. Barriers to distribution as re-
gards birds and mammals. Amer. Naturalist 48:
248-254.

1917. Field tests of theories concerning dis-
tributional control. Amer. Naturalist 51:
115-128.

Grinnell, J., J. Dixon, and J. M.
Linsdale. 1930. Vertebrate natural history of
a section of northern California through the
Lassen Peak region. Univ. Calif. Press, Berkeley.

Grinnell, J., and A. H. Miller. 1944. The distri-
bution of the birds of California. Pacific Coast
Avif. 27: 1-608.

Hayward C. L., M. L. Killpack, and G. L.
Bichards. 1963. Birds of the Nevada Test
Site. Brigham Young Univ. Sci. Bull., Biol. Ser.
3(1): 1-27.

Howell, T. B. 1952. Natural history and differen-
tiation in the yellow-bellied sapsucker. Condor
54: 237-282.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

159

Johnson, N. K. 1956. Birds of the pinon association
of the Kawich Mountains, Nevada. Great Basin
Nat. 16: 32-33.

1963. Biosystematics of sibling species of fly-
catchers in the Empidonax haminondii-oberhol-
seri-wrightii complex. Univ. Calif. Publ. Zool.
66: 79-238.

1965. The breeding avifaunas of the Sheep

and Spring ranges in southern Nevada. Condor
67: 93-124.

1966. Morphologic stability versus adaptive

variation in the Hammond's flycatcher. Auk 83:
179-200.

1970. The affinities of the Boreal avifauna of

the Warner Mountains, California. Occas. Pap.
Biol. Soc. Nevada 22: 1-11.

1973. The distribution of Boreal avifaunas in

southeastern Nevada. Occas. Pap. Biol. Soc. Ne-
vada 36: 1-14.

1974. Montane avifaunas of southern Ne-
vada: historical change in species composition.
Condor 76: 334-337.

1975. Controls of number of bird species on

montane islands in the Great Basin. Evolution
29: 545-567.

1976. Breeding distribution of Nashville and

Virginia's warblers. Auk 93: 219-230.

Johnson, N. K., and K. L. Garrett. 1974. Interior
bird species expand breeding ranges into south-
ern California. W. Birds 5: 45-56.

Johnson, B. E. 1975. New breeding localities for
Leucosticte in the contiguous western United
States. Auk 92: 586-589.

Linsdale, J. M. 1936. The birds of Nevada. Pacific
Coast Avif. 23: 1-145.

1938. Environmental responses of vertebrates

in the Great Basin. Amer. Midland Naturalist
19: 1-206.

MacArthur, B. H. 1972. Geographical ecology.
Harper and Bow, New York.

Miller, A. H. 1941. Speciation in the avian genus
Junco. Univ. Calif. Publ. Zool. 44: 173-434.

1946. Vertebrate inhabitants of the pinon as-
sociation in the Death Valley region. Ecology
27: .54-60.

1951. An analysis of the distribution of the

birds of California. Univ. Calif. Publ. Zool. 50:
531-644.

Miller, A. H., and W. C. Bussell.
1956. Distributional data on the birds of the
White Mountains of California and Nevada.
Condor 58: 75-77.

Munz, P. A., and D. D. Keck. 1968. A California
flora. Univ. Calif. Press, Berkeley.

Orr, B. T., and J. Moffitt. 1971. Birds of the Lake
Tahoe region. California Academy of Sciences,
San Francisco, Calif.

Pitelka, F. A. 1951. Speciation and ecologic distri-
bution in American jays of the genus Aphelo-
coma. Univ. Calif. Publ. Zool. 50: 195-464.

Salt, G. W. 1952. The relation of metabolism to
climate and distribution in three finches of the
genus Carpodacus. Ecol. Monogr. 22: 121-152.

St. Andre, G., H. A. Mooney, and B. D.
Wright. 1965. The pinyon woodland zone in
the White Mountains of California. Amer. Midi.
Naturalist 73: 225-239.

Swarth, H. S. 1914. The California forms of the
genus Psaltriparus. Auk 31: 499-526.

Taylor, W. P. 1912. Field notes on amphibians,
reptiles, and birds of northern Humboldt Coun-
ty, Nevada. Univ. Calif. Publ. Zool. 7: 319-436.

Terborgh, J., and J. S. Weske. 1975. The role of
competition in the distribution of Andean birds.
Ecology 56: 562-576.

Wilson, E. O. 1970. Facts of zoogeography. Science
168: 1193-1194.

EXPLOSIVE EVOLUTION OF PERENNIAL ATRIPLEX
IN WESTERN AMERICA1

Howard C. Stutz2

.Abstract.— New habitats opened up in western North America since the recession of Lake Bonneville and
Lake Lahontan have admitted a host of new species of Atriplex. Every known evolutionary force is operating and
at accelerated paces. Autoploidy appears to be more common than in any other reported group of plants. Natu-
ral hyoridization between closely related species has provided a wealth of fertile segregants from which new
adaptive types have been and are being selected. Hybrids between more distantly related species are sterile and
some appear to have given rise to fertile alloploid derivatives.

All evidence points to a center of origin for Atriplex in northern Mexico. The numerous species which have
migrated northward into western United States and Canada were apparently able to do so because attributes ac-
quired to make them adaptive in the hot dry deserts of Mexico were characteristics which, uniquely, also pre-
adapted them for colder climates and alkaline clay soils to the north.

Woody species such as Atriplex canescens and A. con ferti folia hybridize rather easily with herbaceous pe-
rennial species such as A. cuneata, A. gardneri, A. corrugata, and A. obovata. Since most such hybrids are at least
partly fertile and produce F2 segregants of both woody and nonwoody types, the genetic basis
for the accumulation of wood is apparently rather simple.

According to Antevs (1955), Rroecker and
Kaufmann (1965), Morrison (1965), Russell
(1885), and others, much of the land surface
of western Utah and northwestern Nevada
was covered with ice and water as recently
as 10,000-12,000 years ago. During the sub-
sequent rapid disappearance of Lake Bonne-
ville and Lake Lahontan and the attending
desertizing of surrounding valleys, numerous
salt playas and alkali flats were formed.
Very few plants have been capable of sur-
viving the severe physiological drought
which characterizes such habitats. Com-
pounded by attending severe climatological
drought, which in many places is nearly ab-
solute, the number of adaptive plants and
animals is even further minimized. Indeed,
in many areas, islands of dry, saline, mud
hills and alkali playas are still completely
uninhabited. Such areas represent some of

the few last frontiers on earth yet to be ex-
ploited by living organisms.

The principal pioneers which reach out
farthest into these sterile desert islands are
plants belonging to the family Chenopo-
diaceae. The entire family appears to pos-
sess, uniquely, characteristics which permit
the accommodation of this double challenge
of physiological and climatological drought.
At the borders of every sterile, empty island
in these saline deserts some member of this
remarkable family is at the last frontier. In
bottomlands it is usually Sarcobatus or
Saaeda or Salicornia or Allenrolfea. On dryer
sites it is Graijia or Ceratoides or Atriplex or
Sa Isola.

Most chenopod genera are highly special-
ized with very few adaptive variables. Only
two species of Sarcobatus have been de-
scribed: S. vermiculatus (Hook.) Torr and S.

'This research was supported by USDA Forest Service, Intermountain Forest and Range Experiment Station, Agreement 12-11-204-31 No. 8.

'Department of Botany and Range Science, Brigham Young University, Provo, Utah 84602.

161

162

GREAT BASIN NATURALIST MEMOIRS

No. 2

baileyi Cov. There are only two species of
Grayia: G. spinosa (Hook.) Moq. and G.
brandegei Gray. There is only one species
of Ceratoides (C. lanata (Pursh) J. T. How-
ell) in North America and one of Cycloloma
(C. atriplicifolium (Spreng.) Coult.). But, in
contrast, Atriplex is awesomely genetically
rich, consisting of numerous species and va-
rieties, many of which are still unnamed.

Some Atriplex species are very narrowly
endemic, others are phenomenally wide-
spread. Atriplex navajoensis Hanson is con-
fined to a restricted area near Navajo
Bridge in northern Arizona, whereas Atri-
plex canescens (Pursh) Nutt. extends all the
way from Montana southward, deep into
Mexico. Some species such as Atriplex gar-
rettii Rydb. are remarkably uniform
throughout their entire distribution; others,
such as Atriplex confertifolia (Torr. and
Frem.) S. Wats, are highly flexible with nu-
merous ecotypes, chromosome races, and
subspecies.

So genetically rich is Atriplex in western
America and so new and variable are the
environs which it occupies that we are per-
mitted to witness the evolutionary process
at an unprecedented rate. Every known
strategy for speciation is evident, some of
which, such as autoploidy, have never been
reported to be nearly so significant in the
evolution of other plant groups.

Interspecific hybridization followed by in-
trogression, segregation, or alloploidy is
common; mutations, drift, and autoploidy
are found in nearly every species. Inter-
pretations are easy in some groups, but in
others the evidences are rather subtle and
difficult to interpret.

Evolutionary Hot Spots

Atriplex is abundant throughout all of
western America. One or more of the evo-
lutionary forces can be witnessed in nearly
every population, but there are a few par-
ticular sites where speciation is especially
rapid.

One of the richest evolutionary sites lies
in northeastern Mexico in a radius of about

200 miles around Monterey. Here there are
numerous endemic diploid and tetraploid
species, and numerous hybrid derivatives. As
pointed out by Johnston (1941), Atriplex
stewartii I. M. Johnston is an apparent de-
rivative of the hybrid Atriplex canescens x
A. acanthocarpa (Torr.) S. Wats. It has thin,
broad leaves with undulating margins, much
like the leaves of A. acanthocarpa, and
four-winged fruits which, although dimin-
utive, are much like those of A. canescens.
West of San Roberto is another segregant
from this hybrid but, in this case, with A.
canescens-type leaves and A. acanthocarpa-
type fruits. Even another derivative from
this hybrid grows in very heavy gypsum
soils southwest of Cuatro Cienegas. It is
much like A. acanthocarpa in many respects
of habit and leaf but has a strong tendency
for wings on an otherwise burrlike fruit.

Atriplex prosopidium I. M. Johnston is a
handsome, tall, four-winged shrub with a
deep blue cast which probably came from
introgression of A. obovata Moq. into A. ca-
nescens. It is narrowly endemic to a small
area around Monclova and Ocampo. A
much more widespread putative derivative
of this same origin covers thousands of acres
south and west of Cuatro Cienegas. It is a
short-statured, heavy bush, much utilized by
sheep and cattle in that area.

A second geographic center in which
hybridization has played, and is now play-
ing, a major role in the origin of new spe-
cies of Atriplex is in north central Nevada.
Valleys in this region run north and south
between series of mountain ranges. Each
valley is approximately 30 to 60 miles
across and 100 to 400 miles long and usual-
ly opens up to common basins to the north
and to the south. This peculiar phys-
iography permits species to migrate from
common populations into isolated valleys
where independent speciation can occur
and unique species can emerge. And they
have. In almost every valley unique gen-
otypes have become established. Further-
more, fortuitous contacts between north and
south immigrants as well as fortuitous in-

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

163

troductions of particular combinations of
species from either the north or the south
has provided many of the valleys with new
rich opportunities for interspecific hybridi-
zation followed by segregation, in-
trogression, and polyploidy.

Atriplex canescens has been particularly
contributive in this setting. In western Utah
and northern Nevada it has hybridized with
A. tridentata Kuntze yielding a host of new
variables. This is particularly unusual be-
cause A. canescens in this area is tetraploid
whereas A. tridentata is hexaploid. However
numerous viable seeds are produced by the
hybrid in nature, providing a host of segre-
gating progeny. Some of these have recov-
ered high fertility and chromosomal regu-
larity at the tetraploid level but are
phenotypically, and therefore genetically,
strongly modified by A. tridentata. Con-
versely, some segregants have settled down
chromosomally at 2n = 54 and are phenotypi-
cally like A. tridentata, introgressed with A.
canescens. Still others are not yet stabilized,
either genetically or chromosomally. Each
of these permutations have also been ob-
tained in the experimental nursery at Brig-
ham Young University from seeds collected
from natural hybrids.

In and around Battle Mountain, Nevada,
thousands of acres are occupied by a highly
variable population of Atriplex which ap-
pears to have been derived from this par-
entage. The plants are strongly A. tridentata
in many features but tendencies for, winged
fruits, earlier maturation, and woodiness all
suggest introgression from A. canescens.
This population appears to have become
chromosomally stabilized at 27 pairs but is
still unstabilized genetically.

In many of the valleys of west central
Nevada, Atriplex canescens appears to pos-
sess genetic attributes of A. falcata (M.E.
Jones) Standi. Since these A. canescens pop-
ulations are hexaploid, whereas most others
in western America are tetraploid (Stutz et
al. 1975), and, since A. falcata has so far
proven to be always diploid, these hex-
aploid A. canescens populations are prob-

ably allohexaploids derived from the hybrid
of tetraploid A. canescens x diploid A. fal-
cata. The principal difficulty with this in-
terpretation, however, is the abundant vari-
ation within these hexaploid populations. A
monophyletic hexaploid origin should have
provided a very uniform product. Con-
sequently, if the parentage has been cor-
rectly interpreted, then a polyphyletic ori-
gin is required, either initially or from
among partially fertile segregants of the
original hybrid or else abundant continuing
hybridization between the new allohexa-
ploid and one or both of the parents, fol-
lowed by subsequent introgression.

Another highly variable hexaploid popu-
lation occupies extensive valleys south of
Eureka, Nevada. These are much less like
A. canescens than are those described above
which occupy valleys to the west. If these,
too, are allohexaploid derivatives from the
hybridization of tetraploid A. canescens and
diploid falcata, considerable segregation
must have occurred prior to the event of
chromosome doubling because these plants
show a minor influence of A. canescens and
a major A. falcata input.

Certainty about whether or not these are
the correct interpretations must await fur-
ther study and experimental breeding. But
whatever the ancestry, the fact remains that
these partially isolated valleys in central
Nevada are hot beds of Atriplex evolution.
Nearly every valley possesses unique gen-
otypes. In many valleys the variation is still
rampant and has not yet settled down to a
predominant adaptive theme. It is an evolu-
tionary arena in which speciation is pro-
ceeding at unprecedented rates.

Eastern Utah is another region in which
both polyploidy and interspecific hybridiza-
tion have furnished many new species and
varieties. The most common interspecific-
hybrids are those derived from A. cuneata
A. Nels x A. canescens and A. cuneata x A.
confertifolia. The latter is highly sterile but
a few segregants are to be found in nature,
and I have now obtained several seedlings
from many thousands of hybrid "seeds."

164

GREAT BASIN NATURALIST MEMOIRS

No. 2

The hybrid A. canescens x cuneata, how-
ever, is, in some places, quite fertile and
produces extensive segregating populations.
Immediately west of Hanksville, Emery Co.,
Utah, about thirty acres is dominated by
hundreds of segregants from this parentage.
Since A. cuneata is usually a suffrutescent
shrub, dying back almost entirely each year
to the ground, whereas A. canescens is a
woody shrub, accumulating wood year after
year, it was interesting to find both growth
forms represented among the segregants,
suggesting a rather simple genetic basis for
this difference.

Numerous other natural interspecific-
hybrids have been found in eastern Utah,
e.g., A. confertifolia x A. garrettii, A. confer-
tifolia x A. corrugata S. Wats., and A. con-
fertifolia x A. canescens, but extensive suc-
cessful segregants have not yet been noted.

Probably the most significant products of
interspecific hybridization in all Atriplex is
that produced from hybridization of A. ca-
nescens with A. gardneri (Moq.) Dietr. In
northern Wyoming and southern Montana
these two species are often intimately
sympatric. Hybridization between them is
common, attended by a rich array of segre-
gant products. From such hybrid swarms
have come several highly adaptive com-
binations, one of which appears to be the
widespread, herbaceous form of A. canescens
which grows in the heavy clay soils on
the banks of the Missouri River and its
tributaries throughout most of Montana,
southern Alberta, northern Wyoming and
North and South Dakota. It was this form
which was collected in 1804 by Lewis and
Clark at Big Bend near Chamberlain, South
Dakota, and which is the type for A. canes-
cens. Because of the many differences
which distinguish it from the typical taller,
more woody, well-known four-wing saltbush
of the Intermountain West and northern
Mexico, a nomenclatural revision will be re-
quired.

Origins
It is difficult to be certain of the center

of origin of species in a genus as genetically
rich as Atriplex. According to Vavilov
(1926), it would be in one of the areas in
which there is abundant variation. How-
ever, in Atriplex this may not be a good in-
dex for origins. As pointed out above, most
of the evolutionary hot-spots are of very re-
cent vintage and possess abundant variation
not because of antiquity but rather because
of recent availability of new explorable hab-
itats. Also, unacceptable are theories such as
those of Matthew (1939) which suggest that
centers of origins can be detected by the
presence of advanced, progressive species.
This, too, would place the center of origin
in northern Utah and Nevada or eastern
Utah, where evolution is currently explosive
and where, but a few thousand years ago,
much of the area was covered with water.

In view of the widespread severe ecologi-
cal changes during Pleistocene and Holo-
cene in all of western America north of
Mexico, in view of the numerous affinities
between perennial Atriplex species of north-
ern Mexico and those to the north, and in
view of the numerous species in southern
states and Mexico such as A. acanthocarpa
(Torr.) S. Wats., A. obovata Moq. A. pohj-
carpa (Torr.) S. Wats., A. hipnenelytra
(Torr.) S. Wats., and A. torreyi (S. Wats.) S.
Wats, which have no homologues nor even
analogues in northern states or Canada and
also because there is a paucity of perennial
Atriplex species south of San Luis Potosi,
northeastern Mexico appears to be the most
likely area from which most contemporary
perennial, North American species of Atri-
plex arose. Adaptation to the hot dry cli-
mate and gypsiferous soils of this area con-
ceivably could have preadapted many forms
for occupation of the xeric saline areas
which became available to the north during
late Pleistocene. It may have involved bio-
logical adaptation as simple as the capacity
for accumulating and tolerating salts at an
unusually high concentration. Ostensibly the
attending high osmotic pressure of such an
adaptation could function as an "antifreeze"
in colder northern climates as well as a

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

165

mechanism for accommodating drought,
both physiological and climatological. No
other common denominator is yet apparent.

Such an origin is also fully supported by
other lines of evidence. Leaves of all pe-
rennial Atriplex species, so far examined,
have the "Kranz" type leaf anatomy which
appears to be always associated with the C4
photosynthetic pathway (Downton 1971,
Polya and Osmond 1972). Even Atriplex fal-
cata which now is confined solely to north-
ern latitudes and other diploid species as far
north as Edmonton, Alberta, still display
Kranz type anatomy. If, as is generally
maintained, C4 photosynthesis is an adapta-
tion to hot, bright desert conditions, (Down-
ton 1971, Mooney et al. 1974, Smith 1976,
and others) it is difficult to accept that
these northern species arose in situ under
existing climatic conditions. However, a
southern hot-desert origin for a species now
completely restricted to northern climates is
also difficult to accept. But if the Kranz
anatomy is as reliable an index to C4 photo-
synthesis as has been claimed, and if C4
photosynthesis is as restricted to hot climate
plants as has been suggested, then there are
really no other available explanations. And
if we accept all of that then we are forced
to accept, in either case, migration and evo-
lutionary tempos of very unusual rates. The
only apparent northerly migration lanes
were right through Utah and Nevada, and
these were mostly under water only 10 or
12 thousand years ago. If this migration
were earlier, it must have been during the
interstadial between the Bull Lake and
Pinedale pluvials, and the current distribu-
tion of northern species as far south as
northern Nevada and Utah would be a re-
cent subsequent southerly migration.

Since most of Alberta and Saskatchewan
was completely covered with ice and water
as recently as 20,000 years ago (Bayrock
1969, Christiansen 1971), Atriplex species
which are now restricted to areas in these
provinces as far north as Edmonton and
Peace Biver must have immigrated from the
south since that time. Also, since consid-

erable time would be required for the
wasting and disappearance of the ice sheets,
desert conditions conducive to Atriplex mi-
gration would not have been available until
even much later. And, since most of Utah
and Nevada were still mostly covered with
ice and water as recently as 10,000 years
ago (Bissell 1963, Morrison 1965), there are
not very many time periods left in which
the required major desert could have devel-
oped. In fact, it appears that there are only
two possibilities: (1) about 10,000 to 12,000
years ago, just prior to the last major Bon-
neville pluvial or (2) about 5,000 years ago
during the Altithermal period of post Bon-
neville. In view of the great distance which
migrating species would need to traverse to
get from southern deserts to far northern
latitudes, plus time for radiation into the
numerous existing adaptive ecotypes and in
situ synthesis of polyploid derivatives, nei-
ther of these time intervals are particularly
appealing. But since there appears to be no
available alternative it must have happened
during one or perhaps both of these periods.
The tempo of migration and evolution
would have necessarily been very rapid. But
since it is very rapid today, perhaps it is
not so unrealistic after all.

According to Morrison (1965), a distinct
period of dessication separated the Bonne-
ville and Draper pluvials about 12,000 years
ago. It was a relatively brief period during
which the Grantsville soil was deposited at
levels as low as 4250 feet. This is just 50
feet above the current level of Great Salt
Lake so would indicate desert conditions
perhaps as severe as those existing today.
Also Eardley (1962) found a huge bed of
Glauber's salt (Na2SO4«10H2O) from 15 to
25 feet below the present-day bottom of
Great Salt Lake, 9lA miles across and 32
feet thick, the surface of which dated at
11,600 ± 400 years B.P. Further evidence
of a major drought 10,000-12,000 years ago
is furnished by cores taken from Searles
Lake which show an 18-feet- thick layer of
salt of that age (Smith 1962).

Increasing frequencies of Atriplex confer-

166

GREAT BASIN NATURALIST MEMOIRS

No. 2

tifolia in wood-rat middens in the Mojave
Desert beginning about 17,000 years ago
(Wells 1976) may also be counted as evi-
dence for extant desert conditions in west-
ern North America prior to the last Bonne-
ville pluvial. Wells also furnished evidence
which suggests that the Chihuahuan Desert
may have originated less than 11,500 years
ago. Widespread desert conditions may have
lasted for only a few hundred or perhaps a
few thousand years but were probably suffi-
ciently extensive to permit Atriplex and oth-
er desert plants to migrate from the south.

Some of the Atriplex species which are
now endemic to the northern latitudes are,
uniquely, successfully competitive with
grasses and forbs with which they grow.
Elsewhere Atriplex is conspicuously a poor
competitor and is successful on harsh sites
primarily because nothing else can grow
there. Consequently the northern forms
have apparently become secondarily adapt-
ed for competitiveness in sites which are
considerably more mesic than those occu-
pied by other species of Atriplex further to
the south. Since such adaptation is likely
genetically somewhat complex, it probably
required considerable time to arise. This is
further argument that migration from the
south must have occurred at an early period
rather than during the Altithermal of only
5,000 to 6,000 years ago.

The pluvial period following the extant
drought 10,000 to 12,000 years ago appar-
ently wiped out all northern forms of Atri-
plex except those which were adaptively
competitive with other mesophytes. At least
one of these northern Atriplex species, near
Edmonton, Alberta, is diploid so probably
represents a product of primary evolution
involving principally genetic drift and the
acquisition of new adaptive mutations.

Considerable evidence is now accumulat-
ing which suggests that desert conditions
were also widespread during the Altither-
mal period. Scott (1965) reports mollusks
dated at 5500 B.P., buried 40 feet under eo-
lin sand near Denver, Colorado, and Smith
(1962) reports 40 feet of salt of altithermal

age deposited in Lake Searles in eastern
California. Whether the warm Altithermal
period developed synchronously throughout
all of western America to provide one huge
desert or whether deserts developed in dif-
ferent areas at different times is not known
but neither is it critical to Atriplex evolu-
tion and migration. A single widespread
desert would, of course, be simpler to com-
prehend but stepwise movement from one
area to another could have been just as ef-
fective. The principle conclusion in either
case is the same. Atriplex must have mi-
grated from the south, over thousands of
miles, across terrain which has subsequently
become so modified that intervening forms
have become completely extinct. Ancestral
types, if they still exist, are so very different
from those to the north which have evolved
from them that evidences of such ancestry
are not at all apparent.

Although the climate of northeastern
Mexico has apparently changed very little
during the past 30,000 to 40,000 years
(Meyer 1973) in comparison to the tu-
multuous surface changes to the north, Atri-
plex growing there would have certainly
continued to evolve. However, if the north-
ern derivatives migrated from the south
during the recent Altithermal period, many
original ancestral types would predictably
still be present. But none are to be found, a
fact which suggests that migration must
have been much earlier.

Wells (1966) has reported Atriplex canes-
cens fossils at Burro Mesa in the Chihua-
huan Desert which date at 36,000 years
B.P. However, today only polyploid forms
of this species have been found in Mexico
(Stutz et al. 1975). The only population of
diploid Atriplex canescens yet found is on
sand dunes in central Utah and may repre-
sent the only relic stand of the ancestral
form. However it is so genetically dis-
tinctive and so restricted in its distribution,
it is almost certainly a derived form, consid-
erably altered from its progenitors.

Many other diploid species of Atriplex
scattered throughout the Intermountain

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

167

West are likewise geographically disjunct
from any apparent Mexican relative. Most
of them have narrow restricted distributions
so were probably left behind as adaptations
to unique situations during and after the
migration rampage.

All of this argues strongly for greater an-
tiquity in Atriplex migration than can be af-
forded by the Altithermal period and con-
sequently supports other evidences for
widespread desert conditions prior to the
last Bonneville pluvial, of a magnitude suf-
ficient to accommodate migration and evo-
lution every bit as rapid as that which we
witness today in the wake of new ecological
sites exposed by post-Bonneville climatic
changes.

As new ecological permutations now
emerge, encroaching evolutionary fronts ex-
ploit them. Hence the entire Intermountain
West today is once again an evolutionary
arena in which new species are emerging
and migrating at phenomenal rates. Moun-
tains which are islands for mesophytes are
isolation barriers for xerophytes, and the
valleys which are isolation barriers for
mesophytes are islands for the xerophytes.
As these isolation barriers are surmounted
by chance migrants, or adaptive deviants,
new waves of evolutionary turbulence
evoke yet other genetic permutations which
accommodate even other challenges. It
would be difficult to imagine a more dy-
namic arena wherein to witness the evolu-
tionary process and it would also be diffi-
cult to find a genus better fitted for
adaptively responding to it than Atriplex.

Literature Cited

Antevs, E. 1955. Geologic climatic dating in the
west. Amer. Antiquity 20: 317-335.

Bayrock, L. A. 1969. Incomplete continental glacial
record of Alberta, pp. 99-103. In: H. E. Wright,
Jr. (ed.), Quaternary geology and climate. Na-
tional Academy of Sciences, Washington, D.C.

Bissell, H. J. 1963. Lake Bonneville: geology of
southern Utah Valley, Utah. U.S. Geol. Surv.
Prof. Pap. 257-B: 101-130.

Broecker, W. S., and A. Kaufman. 1965.
Radiocarbon chronology of Lake Lahontan and
Lake Bonneville II, Great Basin. Bull. Geol.
Soc. Amer. 76: 537-566.

Christiansen, E. A. 1971. Tills in southern Sas-
katchewan, Canada, pp. 167-183. In: R. P.
Goldthwait (ed.), Till, a symposium. Ohio State
University Press, Columbus.

Downton, W. J. S. 1971. Adaptive and evolutionary
aspects of C4 photosynthesis, pp. 3-17. In: M.
D. Hatch, C. B. Osmond, and R. A. Slayter
(eds.), Photosynthesis and respiration. Wiley,
New York.

Eardley, A. J. 1962. Glauber's salt bed west of
Promontory Point, Great Salt Lake, Utah. Geol.
Mineral. Surv. Spec. Stud. 1: 1-12.

Johnston, I. M. 1941. New phanerogams from Mex-
ico IV. J. Arnold Arbor. 22: 110-124.

Matthew, W. D. 1939. Climate and evolution. 2d
ed. New York Academy of Sciences, New York.

Meyer, E. R. 1973. Late Quaternary paleoecology
of the Cuatro Cienegas Basin, Coahuila, Mexi-
co. Ecology 54: 982-995.

Mooney, H. A., J. H. Troughton, and J. A.
Berry. 1974. Arid climates and photo-
synthetic systems. Carnegie Inst. Wash. Year
Book 73: 793-805.

Morrison, R. B. 1965. Quaternary geology of the
Great Basin, pp. 265-285. In: H. E. Wright, Jr.,
and D. G. Frey (eds.), The Quaternary of the
United States. Princeton University Press,
Princeton, New Jersey.

Polya, G. M., and C. D. Osmond. 1972.
Photophosphorylation by mesophyll and bundle
sheath chloroplasts of C4 plants. PI. Physiol.
(Lancaster) 49: 267-269.
Russell, I. C. 1885. Geological history of Lake La-
hontan. A Quaternary lake of northwestern Ne-
vada. U.S. Geol. Surv. Monogr. 11: 1-288.

Scott, G. R. 1965. Nonglacial Quaternary geology
of the southern and middle Rocky Mountains,
pp. 243-254. In: H. E. Wright, Jr., and D. G.
Frey (eds.), The Quaternary of the United
States. Princeton University Press, Princeton,
New Jersey.

Smith, B. N. 1976. Evolution of C4 photosynthesis
in response to changes in carbon and oxygen
concentrations in the atmosphere through time.
Biosystems 8: 24-32.

Smith, G. I. 1962. Subsurface stratigraphy of late
Quaternary deposits, Searles Lake, California, a
summary. U.S. Geol. Surv. Prof. Pap. 450-C: 65-
69.
Stutz, H. C, J. M. Melhy, and G. K. Livingston.
1975. Evolutionary studies of Atriplex: a relic
gigas diploid population of Atriplex canescens.
Amer. J. Bot. 62: 236-245.

168 GREAT BASIN NATURALIST MEMOIRS No. 2

Vavilov, N. I. 1926. Studies on the origin of culti- huahuan Desert. Science 153: 970-975.

vated plants. Trudy Prikl. Bot. Selekc. 16(2): 1- 1976. Macrofossil analysis of Wood Rat (Neo-

248.

toma) middens as a key to the Quaternary

vege-

Wells, P. V. 1966. Late Pleistocene vegetation and tational history of arid America. Quaternary

degree of pluvial climatic change in the Chi- Res. 6: 223-248.

DISTRIBUTION AND PHYLOGENY OF ERIOGONOIDEAE
(POLYGONACEAE)

James L. Reveal1

Abstract.— Eriogonoideae is a subfamily of the knotweed family, Polygonaceae, endemic to the New World,
and is composed of 14 genera and perhaps 320 species. It differs primarily from the other members of Polyg-
onaceae in lacking well-defined sheathing stipules or ochrea. The species of Eriogonoideae vary from tiny, fragile
annuals to herbaceous perennials, low subshrubs or shrubs to large and often arborescent shrubs. The seemingly
most primitive extant genus of the subfamily is Eriogonum (247 species), which is widespread in central North
America. A series of genera are closely related to Eriogonum, and probably have evolved directly from Eriogo-
num. These genera are Oxytheca (9 species) of the western United States, and Chile and Argentina of South
America; Dedeckera and Gilmania, both monotypic genera of the Death Valley region of California; Stenogonum
(2 species) of the Colorado Plateau and adjacent areas of the Rocky Mountain West; Goodmania and Hollisteria,
2 monotypic genera of central and southern California; and Nemacaulis, a monotypic genus of the southwestern
United States and northwestern Mexico. A second major complex of genera also probably evolved from Eriogo-
num. In this group, the most elementary genus is Chorizanthe (about 50 species), in which the extant perennial
members of the genus are perhaps evolutionarily the oldest taxa of the subfamily. These perennials are restricted
to Chile, while in the western United States and northwestern Mexico of North America, only annual species are
found. Mucronea (2 species) of California and Centrostegia (4 species) of the southwestern United States and
northwestern Mexico are clearly related to Chorizanthe. In a somewhat intermediate position between the Eriogo-
num complex and the Chorizanthe complex— but still more closely related to the latter than the former— is the
genus Lastarriaea (2 species) found in California, Baja California, and Chile. All of these genera belong to the
tribe Eriogoneae. A second tribe, Pterostegeae, contains only 2 discordant, monotypic genera: the shrubby per-
ennial genus Harfordia of Baja California and the more widespread annual, Pterostegia, of the western United
States. While time and evolution have obscured the relationships between Eriogoneae and Pterostegeae, the affi-
liations among the various genera of the tribes can be ascertained to some degree. The geographical center of
origin of the subfamily may have been in a subtropical climate, with the differentiation of modern-day genera oc-
curring in temperate, xeric regions of North America. The origin of Chorizanthe was an ancient development,
with the migration of the primitive perennial members into South America in the Tertiary. The subsequent de-
velopment of the annual habit, and migration of annual species of Eriogonoideae into South America has prob-
ably occurred in the Quaternary. The intermediate stages of evolutionary development of the genera and species
of the subfamily occurred in a habitat similar to the pinyon-juniper woodlands of the Great Basin, while evolu-
tion of the more advanced genera and species has occurred in xeric grasslands, chapparral scrub, or xerophytic
"hot desert" communities.

Introduction

In considering the action of evolutionary processes . . . mainly in the Northern Hemisphere. It con-
both extinction and extensive alterations of geographic tams m important agricultural and horti-

and ecological distribution patterns must be recognized. , , .-it.- n

(Stebbins 1974: 37). cultural species in addition to many well-

known and troublesome weeds. The vast
Polygonaceae Juss. is a large, temperate majority of the plants are small, inconspic-
or subtropical family of flowering plants uous members of the world's vascular plant
found throughout much of the world, but flora, and they can claim few positive attri-

'Department of Botany, University of Maryland, College Park 20742 and National Museum of Natural History, Smithsonian Institution, Washington,
DC. 20560.

169

170

GREAT BASIN NATURALIST MEMOIRS

No. 2

butes. The family is composed of about 40
genera and approximately 900 species (Law-
rence 1951, Melchior 1964, Airy Shaw
1973), with Polygonum L., Rumex L.,
Eriogonum Michx., Coccoloba P. Br. ex L.,
Rheum L., and Chorizanthe R. Br. ex Benth.
among the larger genera in terms of species
numbers. Domestically, the genus Fagopy-
rum Mill, is the commercial source of buck-
wheat, and leaf petioles of Rheum (rhubarb)
are frequently eaten. Antigonon Endl. is an
elegant ornamental both in the garden and
in nature, although it is more often a weed.
A few species of Polygonum, Eriogonum,
and Coccoloba are grown for their exotic-
properties.

The family is usually considered as the
only member of a monotypic order, Polyg-
onales, which is supposedly related to the
Caryophyllales (Takhtajan 1959, 1969,
Cronquist 1968, Hutchinson 1969), although
some authors still place it with the Caryo-
phyllales (Thome 1968, Benson 1974). Re-
cently, the relationship with the Caryophyl-
lales has been challenged on the basis of
pollen data (Nowicke, pers. comm.), and the
Polygonales might be better treated as an
isolated taxon with no immediate close rela-
tives.

Polygonaceae has been variously divided
into subfamilies (Bentham 1856, Bentham &
Hooker 1880, Dammer 1892, Roberty &
Vautier 1964), and the differences in opin-
ion cannot be resolved here. The one point
of near unanimity among all of these au-
thors, and others who have treated the Po-
lygonaceae, is that Eriogonoideae Benth.2 is
the most distinct subfamily of Polygonaceae
and can be readily excluded from the re-
maining subfamilies. The only serious differ-
ence, now largely resolved, has been the
relationship of the genus Koenigia L. to the
western United States genera, Hollisteria S.
Wats., Nemacaulis Nutt., and Lastarriaea

Remy in Gay of Eriogoneae Benth., and
Pterostegia Fisch. & Meyer of the tribe Pte-
rostegeae Torr. & Gray. Bentham and
Hooker (1880) proposed that Koenigia was
related to these genera, placing all of them
in a tribe termed Koenigeae. This was di-
rectly contrary to Torrey and Gray (1870)
who placed Nemacaulis and Lastarriaea in
the tribe Eriogoneae, and Pterostegia in Pte-
rostegeae; Koenigia was not even mentioned
by Torrey and Gray. Roberty and Vautier
(1964) removed Koenigia from Eriogo-
noideae, and placed this arctic and sub-
arctic genus in the Polygonoideae where it
certainly belongs.

All further comments in this paper will
be restricted to the subfamily Eriogo-
noideae.

Generic Composition of Eriogonoideae

The members of Eriogonoideae are re-
stricted to the xeric regions of the New
World, with the vast majority of species
confined to the western half of central
North America from the Tropic of Cancer
northward to the fiftieth parallel. In South
America, the few species that are known to
be native elements in the flora are found in
the deserts of northern Chile and scattered
parts of adjacent Argentina. The state of
California in the United States harbors more
species and genera of the subfamily than
any other comparable political area; the
state also has more endemic genera (five)
than any other area. Both Eriogonum and
Chorizanthe have a large number of species
in California (about 40 percent of Eriogo-
num and perhaps 70 percent of Chorizanthe
are in the state), and of all the genera of
the subfamily, only two, Stenogonum Nutt.
and Harfordia Greene & Parry, do not oc-
cur in California. Three genera occur in
both North and South America. Chorizanthe

The authorship of the subfamily name, Eriogonoideae, is here attributed to George Bentham's name "Subordo Eriogoneae" published in deCandolle's
Prodromus (14: 5. 1856) based upon Article 18 of the present Code (Stafleu et al. 1972), which states "Names intended as names of families, but pub-
lished with their rank denoted by one of the terms order {ordo) or natural order (ordo naturalis) instead of family, are treated as having been published
as names of families." Unfortunately, in Article 19, which deals with subfamily, no similar provision is stated. For this reason, some may reasonably ar-
gue that the correct authorship of the subfamily is Roberty and Vautier (Boissiera 10: 83. 1964).

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

171

is strictly an annual group in North Ameri-
ca where some 40 of the 50 species of the
genus are found, but in South America, all
but one (C. commissuralis Remy in Gay) of
the 10 or so species of the genus are per-
ennials, and no species is common to both
continents. The genus Lastarriaea is now
usually defined to include 2 species (Good-
man 1943, Hoover 1966, Munz 1974), L.
chilensis Remy in Gay of Chile, and L. co-
riacea (Goodman) Hoover of coastal Califor-
nia and northern Baja California, Mexico,
although the genus has been considered to
be monotypic with two variants (Gross
1913, Goodman 1934) or without any differ-
ences (Parry 1884, Abrams 1944, Munz &
Keck 1959). The third genus, Oxytheca
Nutt., has a single species in South America
and 8 species are restricted to North Ameri-
ca. All of the remaining genera are restrict-
ed to North America as are approximately
305 of the 320 species of the subfamily.3

Eriogonoideae is composed of fourteen
genera unequally divided into two tribes.
The large, typical tribe, Eriogoneae, con-
tains twelve genera and about 318 species,
with the majority of the species distributed
in two genera, Eriogonum (247 species) and
Chorizanthe (about 50 species). The least
advanced member of the tribe is the genus
Eriogonum. Associated with this genus are a
series of small, satellite genera which can
trace their probable origin to an extant
group within Eriogonum as it exists today.
Likewise, around Chorizanthe are related
genera which probably owe their origins to
Chorizanthe, with Chorizanthe itself likely
evolved from a now extinct portion of
Eriogonum. The second tribe, Pterostegeae,
is composed of two monotypic genera
which are only superficially related, and
whose relationship with Eriogoneae is frank-
ly lost.

Eriogonum is widespread in North Ameri-
ca, ranging from east central Alaska (Welsh
1974) southward to central Mexico, and

from the offshore islands of California and
Baja California eastward to the Appalachian
Mountains of Virginia and West Virginia
southward to Florida. In spite of its large
size in terms of species numbers, Eriogonum
has only three generic synonyms. Eucycla
Nutt. (Nuttall 1848a) and Pterogonum H.
Gross (1913) are now recognized as sub-
genera of Eriogonum (Reveal 1969a, b; Hess
& Reveal 1976), while Sanmartinia Buch-
inger (1950), a name proposed for Eriogo-
num divaricatum Hook. (Reveal & Howell
1976) when it was discovered as an in-
troduction into Argentina (Spegazzini 1902,
Moreau & Crespo 1969) and thought to
represent a distinct species of Eriogonum or
a valid genus, is now reduced to synonymy
completely.

The genus Eriogonum is currently being
monographed by myself, but past reviews
have been presented by Nuttall (1817), Ben-
tham (1836, 1856), Torrey and Gray (1870),
Watson (1877), Stokes (1904, 1936), and Re-
veal (1969a).

Most closely related to Eriogonum is Oxy-
theca. This genus of nine species has been
reviewed by those who revised the species
of Eriogonum (except Nuttall [1817], Ben-
tham [1836], and Reveal [1969a]) at least as
far as the genus in North America is con-
cerned, with both Stokes (1904, 1936) and,
indirectly, Roberty and Vautier (1964) in-
cluding the species of Oxytheca in Eriogo-
num. Critical reviews of the genus have
been presented by Jepson (1913), Abrams
(1944), and Munz and Keck (1959) in floris-
tic studies of the California species where
seven of the eight North American species
are found. Barbara J. Ertter, a graduate stu-
dent at the University of Maryland, is now
monographing the genus. One new species
has been discovered from California, and,
although recognized as unique by Stokes
(1904) and by Goodman (in herbaria annota-
tions), this San Bernardino Mountains en-
demic has not been described. Two generic

'See note added in proof at end of paper.

172

GREAT BASIN NATURALIST MEMOIRS

No. 2

segregates have been proposed for species
of Oxytheca: Brisegnoa Remy in Gay (1851),
a name actually proposed by Remy prior to
1848 when Nuttall described Oxytheca, but
whose publication was delayed, and Acan-
thoscyphus Small (1898) for a California
species of Oxytheca, O. parishii Parry, that
differs from most species of the genus in
having a multiple-awned, nonlobed in-
volucre. Oxytheca luteola Parry is now re-
ferred to Goodmania Reveal & Ertter, and
O. insignis (Curran) Goodman is placed in
Centrostegia Gray ex Benth. in DC. (Good-
man 1957). In this latter paper, Goodman
informally proposed to divide Oxytheca into
two new genera and at the same time sub-
merge a part of Oxytheca in Eriogonum.
Based upon herbarium annotations, he
would have placed O. dendroidea Nutt., O.
watsonii Torr. & Gray, O. perfoliata Torr.
& Gray, and O. parishii in Eriogonum, O.
luteola in a new genus, and referred O.
caryophyUoides Parry, O. emarginata Hall,
and O. trilohata A. Gray to a second new
genus. Goodman did not recognize the
South American form of Oxytheca as a dis-
tinct species as proposed by Miers (1851),
but retained it as a variant of the North
American species, O. dendroidea as sugges-
ted by Johnston (1929) who proposed var.
tonsiflora I. M. Johnst. He and Goodman
felt the Chilean and Argentinean plants were
a distinct form of O. dendroidea, which
they believed also occurred in other areas
of South America. Goodman never pub-
lished his proposed revision of Oxytheca,
and Ertter and I are now investigating the
genus.

As noted above, one of the species of
Oxytheca that Goodman proposed to place
in a distinct genus was O. luteola. This sug-
gestion has recently been accepted by Re-
veal and Ertter (1976b), who proposed the
genus Goodmania for this species. Good-
mania is restricted to alkaline places, dry
lake flats, and similar locations in the south-
ern end of the Central Valley of California
and elsewhere in the southern part of the
state. This monotypic genus seems to be re-

lated to both Oxytheca and Gilmania Cov.

Somewhat less closely related to Eriogo-
num, but still clearly derived from that
genus (rather than Chorizanthe), are a series
of highly restricted, endemic, western North
American genera. Dedeckera Reveal &
Howell (1976) is a large perennial shrub re-
stricted to a single known location just out-
side the northwestern edge of the Death
Valley National Monument near Eureka
Valley in Inyo County, California. This is
the only immediate relative of Eriogonian
that is perennial.

Stenogonum (Nuttall 1848a) is a genus of
two species and is restricted to the Colo-
rado Plateau and adjacent regions of
Wyoming southward through eastern Utah
and adjacent western Colorado into north-
ern Arizona and New Mexico. Until recent-
ly, this genus was included in Eriogonum,
where it had been placed by Hooker (1853)
shortly after it was proposed by Nuttall, but
it is now considered a valid genus on the
basis of its unique involucral construction
(Reveal & Howell 1976, Reveal & Ertter
1976a).

Gilmania (Coville 1936), another Death
Valley region endemic, was originally pro-
posed under the generic name of Phyllogo-
num Cov. (Coville 1893), but as this name
proved to be a later homonym, Coville re-
named it for a local Death Valley naturalist,
M. French Gilman. Stokes (1904, 1936)
maintained the genus as distinct from
Eriogonum, although Jones (1903) reduced it
to Eriogonum without comment. Roberty
and Vautier (1964) placed Gilmania in
Eriogonum too; but, unlike Jones, who re-
tained the species as distinct, they placed
the name in synonymy under Stenogonum
salsuginosum Nutt. (which they placed in
Eriogonum), an opinion that is totally in-
comprehensible.

On the Inner Coast Ranges of California
is the monotypic genus Hollisteria (Watson
1879). Jones (1908) proposed Chorizanthe
floccosa, which proved to be a synonym of
H. lanata S. Wats., but it seems unlikely
that Jones comprehended the significance of

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

173

his proposal and simply felt the plants rep-
resented a species of Chorizanthe, and did
not consider that he was reducing Hollis-
teria to Chorizanthe. Roberty and Vautier
(1964) placed the taxon in Eriogonum, but
this concept has not been followed by any-
one, and not even Stokes (1904, 1936) felt
compelled to reduce Hollisteria to Eriogo-
num, although what her opinion might have
been regarding its placement in Chorizanthe
was never expressed in print.

The genus Nemacaulis (Nuttall 1848a, b)
is rather widespread in the southwestern
United States and extreme northwestern
Mexico, with the single species, N. denu-
data Nutt, divided into two weakly defined
variants. Stokes (1904, 1936) reduced Nema-
caulis to Eriogonum, perhaps following the
ideas of Curran (1885), who noted the close
relationship between N. denudata and E.
gossypinum Curran. No one except Roberty
and Vautier (1964) has followed this reduc-
tion.

A second cluster of genera is related to
Chorizanthe. Unlike those which have just
been reviewed, the satellite genera in this
section can be traced to extant sections of
Chorizanthe.

Chorizanthe itself is a genus of perhaps
50 species, with about 40 species found in
west-central North America, and the re-
maining 10 or so restricted to northern
Chile in South America. All of the species
in North America are annuals, while all but
one of the South American species are sub-
fruticose perennials. The genus has been
monographed only by Bentham (1836,
1856), although the North American species
have been revised by Torrey and Gray
(1870), Watson (1877), Parry (1884), and
Goodman (1934). Remy (1851) and Philippi
(1864, 1873, 1895) have added species to
the South American component of Chori-
zanthe.

Like Eriogonum, the generic concept of
Chorizanthe has changed over the years,
with Chorizanthe being defined in both a
broad and a strict sense. As the genus is
outlined here— and it is done so only in a

tentative fashion— a middle-of-the-road ap-
proach is proposed. Several segregate gen-
era have been proposed from members fre-
quently placed in Chorizanthe. These genera
are Mucronea Benth., Lastarriaea, Centros-
tegia, Acanthogonum Torr., and Eriogonella
Goodman. In the following treatment,
Mucronea, Lastarriaea, and Centrostegia are
recognized as distinct from Chorizanthe,
with Acanthogonum and Eriogonella retain-
ed in Chorizanthe.

The genus Mucronea (Bentham 1836) was
described at the same time that Chorizanthe
was proposed, and it was retained as a dis-
tinct genus by Bentham (1856) in his mon-
ograph on Eriogonoideae in deCandolle's
Prodromus. Torrey and Gray (1870) reduced
Mucronea to Chorizanthe, and their opinion
was followed by Bentham and Hooker
(1880) a decade later. Goodman (1934)
reintroduced Mucronea into the literature
when he distinguished it from Chorizanthe
in his monograph on the latter genus. How-
ever, even with Goodman's study, the genus
remained suppressed (Abrams 1944, Munz
& Keck 1959) except for Hoover (1970),
who recognized the genus in a local flora.
As defined here, the genus is considered to
have two species, both of which are re-
stricted to California.

Lastarriaea was proposed by Remy in
Gay's Flora Chilena (1851), but it was not
associated with the tribe Eriogoneae (Ben-
tham 1856) until Torrey and Gray (1870)
placed the genus in the tribe. Bentham and
Hooker (1880) removed it, Hollisteria, and
Nemacaulis, along with Pterostegia, and
placed them in the tribe Koenigeae. Except
for Dammer (1892), this significant depar-
ture has not been followed to any degree.

As now defined, Lastarriaea contains two
species, one in North America and one in
South America (Goodman 1943, Hoover
1966).

The genus Centrostegia was published for
Asa Gray by Bentham (1856) and consid-
ered at the time to be a monotypic genus.
In 1870, Torrey and Gray added a second
species, but in 1877, Watson reduced the

174

GREAT BASIN NATURALIST MEMOIRS

No. 2

genus to Chorizanthe, where it remained
until Goodman's revision of Chorizanthe in
1934. In 1957, Goodman revised Centros-
tegia, bringing the number of species in the
genus to four. One species, Centrostegia in-
signis (Curran) A. A. Heller (1910), was
originally described as a species of Chori-
zanthe by Curran (1885) but placed in Oxy-
theca by Goodman (1934) without com-
ment. In short, this single, unusual species
has been batted around in three separate
genera, and it still seems out of place in
Centrostegia. As now defined, Centrostegia
occurs from Arizona and Utah westward to
California, where it is found from Monterey
and San Luis Obispo counties southward.

Of these three genera, all of which have
at one time or another been associated with
Chorizanthe, data would now seem to in-
dicate that only Mucronea is actually all
that close to Chorizanthe, with Centrostegia
occupying a position somewhat intermediate
between Eriogonum (not Oxytheca) and
Chorizanthe, and Lastarriaea well isolated
from all of the genera but still closer to the
Chorizanthe complex than the Eriogonum
complex.

Preliminary studies on the Chorizanthe
complex have revealed some major areas of
investigation for future studies. The most
important one is to determine the relation-
ship between the northern annuals and the
southern perennials in the genus Chori-
zanthe. The type of the genus is a South
American perennial, C. virgata Benth., and
these perennials differ markedly from the
annuals. Current plans call for a detailed
studv of the South American species, which
have not been revised in over 100 years. It
is hoped this group of plants will be the
subject of a doctoral dissertation. A second
area of investigation is whether or not the
genus Acanthogonum should be recognized
and, if so, what members of Chorizanthe
should be placed in it.

All of the genera discussed to this point
belong to the tribe Eriogoneae. The second
tribe of the subfamily is Pterostegeae. This
taxon may be characterized by the bisaccate

bracts which become enlarged, scarious, and
reticulate in fruit, and the consistently op-
posite leaves.

The genus Pterostegia is a monotypic
genus of low, spreading to decumbent an-
nual herbs. Described by Fischer and Meyer
(1835) from material gathered by the Rus-
sian explorers near Fort Ross in California,
the genus can be rapidly distinguished from
all other members of Polygonaceae by its
floral features and fruiting characteristics. A
major problem has been how to interpret
the bisaccate bracts. Fischer and Meyer
completely misunderstood the relationship
of the bracts of Pterostegia as they at-
tempted to relate these bracts to those of
Eriogonum. Bentham (1856) misunderstood
the bracts too, attempting to define them as
three leaves with a contiguous margin
which are expanded into a dorsal wing or
crest. Torrey and Gray (1870) stated that
the bracts were homologous with the bracts
of Nemacaulis, but even this seems most un-
likely today, although the concept expressed
by them was accepted by Bentham and
Hooker (1880). The involucral bracts of Pte-
rostegia are two-lobed, enlarged in fruit,
and are simply unlike anything found in any
genus of Eriogoneae.

Pterostegia is a rather variable species
which ranges from Oregon southward to
northern Baja California and eastward into
Utah and Arizona. It does not seem to be
divisible into infraspecific elements although
Nuttall (1848a) suggested some segregates.

The second genus of Pterostegeae, Har-
fordia, was proposed by Greene and Parry
in a paper published by Parry (1886). The
year before, Greene (1885) had described
Pterostegia galioides Greene, and, while he
placed the species in Pterostegia, it was the
first time that good specimens of P. macrop-
tera Benth. (Bentham 1844) were found.
Bentham's descriptions of his species, pub-
lished both in 1844 and 1856, lacked the
fine detail, and he was not even sure
whether the plants were annuals or per-
ennials. It is likely that the lack of adequate
material accounted for the long delay in as-

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

175

certaining the significant differences be-
tween the type species of Pterostegia, P.
drymarioides Fisch. & Mev., and P. macrop-
tera. Once Harfordia was described, it was
immediately accepted, and the genus is now
well recognized (Shreves & Wiggins 1964).
At present, H. macroptera (Benth.) Greene
& Parry is known only from the west coast
of central Baja California, Mexico.

Origin of Eriogonoideae

The origin of Eriogonoideae is unknown.
The subfamily is clearly a member of Po-
lygonaeceae, for it shares with the other
subfamilies of the family a large number of
morphological and biochemical similarities,
and the subfamily Eriogonoideae cannot be
raised to the familial level as proposed by
Meisner (1841) without violence to our un-
derstanding of families in the Magnolio-
phyta. Eriogonoideae has a single, basal, bi-
tegmic, crassinucellate ovule similar to that
of Polygonum and has the typical trinu-
cleate pollen of the family. Still, these are
features which are not only typical of Po-
lygonaceae, but of Plumbaginaceae and
nearly all of the families commonly associ-
ated with the Caryophyllales (Cronquist
1968). Eriogonoideae also shares with the
other subfamilies of Polvgonaceae the copi-
ously laden endospermous seeds and the an-
thocyanins pigmentation. The subfamily
does differ from the other subfamilies in
lacking the distinctly sheathing stipular och-
rea of the leaves (although an ochrea is
weakly present in some perennial species of
Chorizanthe), and the pollen of the sub-
family is the least specialized of all sub-
families of Polvgonaceae (Nowiche, pers.
comm.) suggesting that, as a group, Eriogo-
noideae may be a rather primitive member
of Polvgonaceae. Equally important in this
regard, it may also mean that Eriogo-
noideae, as a group, has retained many of
the least specialized features of the family
due to a lack of modification in organs
which have occurred in other taxa.

Although the critical similarities between

the subfamilies certainly associate these taxa
of Polvgonaceae into a distinct family, the
place and mode of development of Eriogo-
noideae from the rest of Polvgonaceae is
now obscured by time and compounded by
a lack of a fossil record. No one group of
genera outside of the Eriogonoideae can be
considered the exact point of origin of the
subfamily, and for this reason, the subfamily
(or tribe) has long been considered unique
within Polvgonaceae (Bentham 1836, Good-
man 1934, Boberty & Vautier 1964). Based
on preliminary pollen data from extant taxa
now available from the work of Dr. Joan
Nowicke at the Smithsonian Institution, it
seems clear that Eriogonoideae is clearly
differentiated from all but the South Ameri-
can tropical genus Triplaris Loefl. This is
the only genus which has a similar, un-
specialized pollen grain (and thus different
even from the related American tropical
genus Ruprechtia C. A. Meyer), but based
on extant data on chromosome numbers,
gross morphology, and other anatomical and
morphological features (especially in the in-
florescence), it seems most unlikely that the
tribe Triplarideae C. A. Meyer and the sub-
family Eriogonoideae have been connected
in any but the most remote fashion.

Boberty and Vautier (1964) placed
Triplarideae in the subfamily Calligonoideae
Boberty & Vautier which they defined as a
group of New and Old World genera. Dam-
mer (1892) referred the tribe to Coccolo-
boideae Dammer, a basically shrubby or ar-
borescent taxon about equally divided in the
New and Old World. Dammer's suggestion
seems more reasonable as he defined the
Coccoloboideae to include (using current
nomenclature) such genera as Coccoloba,
Muehlenbeckia Meisner, and Triplaris— all
genera with ruminated endosperm. Unfortu-
nately, all genera of Eriogonoideae have a
smooth endosperm. Meisner (1856), Ben-
tham and Hooker (1880), and Dammer
(1892) all placed Eriogonoideae in a posi-
tion in their revisions of the Polvgonaceae
which would imply that Eriogonoideae is
the least specialized of the family. Boberty

176

GREAT BASIN NATURALIST MEMOIRS

No. 2

and Vautier placed the subfamily at the end
of their treatment; most certainly Roberty
and Vautier are correct in their assessment
of the placement of the subfamily in the
family, for Eriogonoideae is the most ad-
vanced member of the extant subfamilies of
Polygonaceae and not the least specialized.
In Meisner, Bentham and Hooker, and
Dammer, interestingly, the Triplarideae was
considered the most advanced member of
the subfamily Polygonoideae (Meisner and
Bentham and Hooker) or Coccoloboideae
(Dammer). If this is indeed the case, then it
logically can follow that a possible origin of
the Eriogonoideae may have been within an
ancient taxon that, by definition, might in-
clude the basic expression from which the
Triplarideae has evolved or in fact was a
part. At no time, however, has Triplaris or
Ruprechtia played a direct role in the ori-
gin of any genus within Eriogonoideae.

It is likely that the divergency of Eriogo-
noideae from the rest of Polygonaceae has
been so fundamental, so sudden, and so suc-
cessful, that the new subfamily has com-
pletely swamped those groups (or that
group) from which it arose. If this divergen-
cy is an ancient one, as I suggested some
years ago (Reveal 1969b), and occurred at
the beginning or slightly before the start of
the Tertiary some 65 million years ago,
then the loss of such intermediate stages of
evolutionary development is to be expected.
However, if the origin of the subfamily has
been well within the Tertiary, as now seems
much more likely, then the loss of the inter-
mediate forms is a matter of the explosive
success and highly competitive nature of
the new form (in this case, the earliest
members of Eriogonoideae) as opposed to
the rather static parental type (see Stebbins
[1974] for a detailed discussion of this type
of explosive evolution above the generic-
level).

As just noted, it now seems more reason-
able to assume that Eriogonoideae arose
during the Tertiary, and probably during
the Oligocene or Miocene epochs (7 to 38
million years ago) when there was a general

drying of the climate coupled with the rap-
id development and increase of herbaceous
angiosperms (Gray 1964, Axelrod 1966, Tid-
well et al. 1972). Pollen grains, attributable
to Eriogonum, have been found in the Qua-
ternary, which began some 7 million years
ago (Leopold, pers. comm.). If this as-
sumption is correct, then perhaps the sub-
family Eriogonoideae had its origin from a
subtropical group of New World polyg-
onaceous plants near the beginning of the
drying period during the Oligocene, which
split off into a tropical complex (something
like Triplaris) and a northern temperate
complex (something like Eriogonum). This
point of origin has subsequently been lost,
with the extinct relatives of Triplaris and
Eriogonum extending and amplifying the
differences between the two extremes to a
point that now only the mere hint of rela-
tionship may be noted in a conservative
feature such as pollen morphology. If the
theories of Takhtajan (1969) and Stebbins
(1974) are correct regarding the differential
rates of specialization between tropical and
xeric temperate groups, then one may as-
sume that the relatives leading to Triplaris
have undergone less specialization and dif-
ferentiation than the relatives leading to
Eriogonum. This would seem to be the case
here, especially when one looks at the re-
duction of the inflorescence in Eriogonum
to a cluster of flowers, the reduced stature
of Eriogonum, and the great proliferation of
species in the Eriogonoideae when com-
pared with the Triplarideae.

Over the years I have vainly searched the
temperate members of Polygonaceae, and
especially those of Asia, for a hint to the
origin of Eriogonum. None has been found.
Stebbins (1974) has cautioned us to realize
that in evolutionary events such as the ori-
gin of taxa above the species rank extinc-
tion and extensive changes in the distribu-
tion and ecology of a taxon may occur
which can substantially change one's out-
look as to the possible site and point of ori-
gin for a given group. It now seems reason-
able to look to the New World tropics for a

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

177

point of origin rather than the Old World
steppes. Certainly, the pollen data just re-
cently reported to me by Nowicke has
greatly strengthened this preconceived idea.
The origins of the Eriogonoideae very likely
have revolved around the reduction of the
inflorescence from an extended one (such as
in Ruprechtia or Triplaris) to a capitate one,
the development of an involucre due to the
fusion of subtending bracts on the in-
florescence, and the reduction of stature
from a shrub or small tree to a subshrub or
low shrub. The final step, of course, has
been the development of an ability to
evolve successfully in a xeric habitat rather
than in a mesic, subtropical, or tropical
habitat. To my knowledge, none of these
steps is extant today.

It was proposed by me (Reveal 1969b)
that the probable ecological place of origin
for Eriogonum was in a xeric site, and that
the first forms of the genus were subshrubs
or low shrubs. This suggestion has been sec-
onded by Stebbins (1974), and there seems
to be little reason to alter this opinion. It is
important to note that this statement relates
to one genus, Eriogonum, and not to the
origin of the subfamily. It seems to me that
the stages of development leading from the
tropical origin of the precursors of the
Eriogonoideae to the extant genus Eriogo-
num must have taken many different direc-
tions, a great deal of time, and undergone
many different attempts before arriving at
this modern genus. It is now impossible to
close that gap, since the history^ of the
Eriogonoideae during the Tertiary is un-
known.

The most generalized form of Eriogonum
that exists today is a low, rounded shrub
with cauline leaves, cymose inflorescences,
small smooth achenes, and an unspecialized
flower with monomorphic tepals. These
shrubs occur in xeric habitats mainly in the
pinyon-juniper woodlands of the Great Ba-
sin in Utah and Nevada. Even so, these spe-
cies of Eriogonum are highly specialized as
all are tetraploids, and no diploid species
are known to exist in the genus (Stebbins

1942, Stokes & Stebbins 1955, Reveal
1969b). Therefore, Eriogonum, as it exists
today, is a highly evolved group, and no
species now exists which could point to the
initial element(s) which might have evolved
from other, more primitive, subtropical taxa
of Polygonaceae.

The assumption that Eriogonum is the
most basic genus of the subfamily seems
reasonable on the basis of morphological
considerations, especially in the makeup of
the inflorescence and involucre. However, I
suspect, that the most ancient extant mem-
bers of the subfamily are the perennial spe-
cies of Chorizanthe. As shall be discussed
below, Chorizanthe likely evolved from
Eriogonum, and not the other way around.
However, one feature found in these per-
ennial species of Chorizanthe seems to hint
at their ancientness: they have what can
only be considered as weakly defined, fi-
brous remains of ochrea. If these species of
Chorizanthe should prove to be diploids,
this would reinforce their evolutionary sig-
nificance. Based upon an examination of the
gross morphology of these plants, one must
add to the definition of the earliest mem-
bers of Eriogonoideae the presence of an
ochrea.

If the genus Eriogonum underwent its
early development in a xeric habitat domi-
nated by pinyon-juniper woodlands, then
where was such a site in the Miocene or
early Pliocene epoches when the genus was
undergoing its earliest development?

During the Miocene, the Great Basin was
dominated by extensive coniferous forest,
with the Sierra Nevada to the west about
1000 m in altitude, and thus an ineffective
rainshadow (King 1959). It is important to
note that these coniferous forests were tem-
perate in nature, with the subtropical for-
ests of the Oligocene largely pushed to the
south. Axelrod (1950) has suggested the ex-
istence of two major geofloras, with the
Arcto-Tertiary geoflora of hardwood-de-
ciduous and conifer forests dominating the
Great Basin region, and the Madro-Tertiary
geoflora of small-leaved, drought-resistant

178

GREAT BASIN NATURALIST MEMOIRS

No. 2

shrubs and trees of the southwestern United
States and northwestern Mexico. Axelrod
(1958) states that the Madro-Tertiary geo-
flora moved northward into the Great Basin
in Early Pliocene, but did not entirely re-
place the Arcto-Tertiary geoflora.

It would seem possible that Eriogonum
may have imdergone its early development
and differentiation in the Madro-Tertiary
geoflora during the Miocene and became
well established in the Arcto-Tertiary geo-
flora in at least two different expressions:
one typified by the subgenus Eucycla
(Nutt.) Kuntze in Post & Kuntze (with such
species similar to E. microthecum Nutt. or
E. corymbosum Benth. in DC.) and the oth-
er of members typical of the subgenus Oli-
gogonum Nutt. (with such species similar to
E. umbellatum Torr. or E. flavum Nutt. in
Fras.). Out of the Madro-Tertiary geoflora
possibly came such subgenera as Eriogonum
and Pterogonum (H. Gross) Reveal which
contain such species as E. longifolium Nutt.,
E. atrorubens Engelm. in Wisliz., and E.
alatum Torr. in Sitgr., or their progenitors
(Hess & Reveal 1976). Nonetheless, the bas-
ic expression of the genus would have be-
longed to the subgenus Eucycla, which is
basically a taxon of xeric, pygmy coniferous
forests. It is also likely that Chorizanthe
evolved during this period of time from the
subgenus Eucycla, probably when the sub-
genus was in the Madro-Tertiary geoflora
and before the subgenus underwent its mod-
ern-day development of species complexes
now typically found in the Great Basin.

Evolution within Eriogonoideae

If the hypothesis is correct that Eriogo-
num is the most primitive extant member
of the subfamily Eriogonoideae, then a
number of corollaries may be presented.

Within Eriogonum itself, if the basic ex-
pression of the genus was a low, spreading
subshrub or shrub with alternate leaves, cy-
mose inflorescences, and unspecialized tep-
als. then the subgenus Eucycla was the in-
itial expression within the genus. As just

noted at the end of the previous section, it
is probable that the differentiation of the
subgenera Eucycla, Eriogonum, Oligogo-
num, and Pterogonum occurred during the
Late Miocene or Early Pliocene in the
Madro-Tertiary geoflora of northern Mexico
and the southwestern United States. Three
of these subgenera of Eriogonum are fairly
distinct from one another, with no inter-
connecting forms. It is felt that while
Eriogonum and Pterogonum evolved from
Eucycla, these two did not evolve from any
extant member of Eucycla. As for Oligogo-
num, it is close to Eriogonum and more dis-
tantly related to Eucycla, and thus both
Oligogonum and Eriogonum may have de-
veloped from extinct, primitive members of
Eucycla at approximately the same time. Of
the remaining subgenera, Clastomyelon
Cov. & Morton, Micrantha (Benth.) Reveal,
Ganysma (S. Wats.) Greene, and Oregonium
(S. Wats.) Greene, all can be traced rather
directly to the subgenus Eucycla without
any major difficulties.

As for Eucycla, it has developed every
perennial habit expression of the genus
Eriogonum but one, the monocarpic habit of
E. alatum of the subgenus Pterogonum. The
subshrub or low shrub habit is widely seen
in the less specialized members of Eucycla.
Such plants are typically seen in the
pinyon-juniper (or pygmy) woodlands
throughout the western part of central
North America today, or essentially the en-
tire geographical range of the subgenus
which extends from the fiftieth parallel
southward to the Tropic of Cancer. Also
found in the pygmy woodland zone are sev-
eral different kinds of herbaceous perennial
expressions belonging to Eucvcla. Unlike the
shrubs which tend to be species of wide-
spread distribution, the herbaceous per-
ennials tend to be more restricted in their
range. Some of these species evolved within
the zone and have remained while others
have extended themselves beyond the con-
fines of the zone, and still others, in more
recent evolutionary times, have entered the
zone from other areas. From the pygmy

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

179

woodland zone, species of Eriogonum have
migrated into dry, xeric clay habitats, into
grasslands or chaparral habitats, and into al-
pine zones and off-shore islands. In all
cases, the present members of the subgenus
are tetraploid, derived species, and while
the majority of morphological expressions
are found in the pinyon-juniper belt, the
majority of explosive evolution within Erigo-
num, and in all of its related genera (except
for the most initial phases of Chorizanthe),
owe their origins to their survival in ecolog-
ical life zones other than the pygmy wood-
lands.

The origins of both of the predominantly
annual subgenera, Ganysma and Oregonium,
probably owe their origins to the subgenus
Eucycla and probably had an initial differ-
entiation in the pinyon-juniper woodlands of
the West. Once again the basic expression
of these subgenera are generally widespread
and found mainly in this habitat. However,
unlike portions of Eucycla, it seems— espe-
cially in Ganysma— that much of the initial
evolutionary development of these sub-
genera have been lost over time, because
there are widely scattered species of Ga-
nysma in other habitats in which the species
are obviously primitive but by no means an-
cient.

The temperate arid regions of the inland
portions of North America exhibit the
unique combination of selective drought
and cold temperatures, conditions which
have likely played a major role in the evo-
lution of the shrubby habit from which her-
baceous and cespitose perennial species
could have evolved (Axelrod 1966). The
shrubby and subshrubby species of Eriogo-
num and Chorizanthe have certainly been
subjected to the selective pressures of sea-
sonal cold, coupled with enough summer
moisture to sustain growth, and the ability
to occupy habitats that are protected
enough to allow for long-term survival in
extended periods of stress. By looking at a
pygmy woodlands as the original home of
Eriogonum and its first major dichotomy,
Chorizanthe, one can understand the variety

of habit and morphological expression in
this ecological habitat, and the economy in
terms of species diversity in this zone. On
the whole, the explosive evolution of the
modern-day species of Eriogonum and its
related genera, and Chorizanthe and its re-
lated genera, has been areas in of extreme
environmental stress outside the protective
(such as they are) confines of the pinyon-
juniper belt. Without a doubt, the majority
of the genera related to Eriogonum owe
their origin to their successful adaptation to
a stress condition, mostly selective drought,
accompanied by the occupation of ecologi-
cal areas on the margins of protective life
zones. However, as we shall see, the sub-
genus Eucycla has given rise only to the
other subgenera of Eriogonum, Chorizanthe,
and Dedeckera, but none of the other gen-
era. Oxytheca, Stenogonum, Gilmania,
Goodmania, Nemacaulis, and perhaps Hol-
listeria owe origin to Eriogonum subgenus
Ganysma, while Mucronea, Centrostegia,
and perhaps Lastarriaea owe their origin to
annual species complexes of Chorizanthe.
And note, all of these genera (except the
primitive members of Chorizanthe) are basi-
cally taxa of areas of extreme aridity, and
basically adapted to the annual habit (all
but Dedeckera).

The one major adaptation which dis-
tinguishes Chorizanthe from Eriogonum is a
combination of the production of an awned
involucre and the reduction in the number
of flowers per involucre. I believe the pro-
duction of an awned involucre has occurred
several times in the history of the sub-
family, much as the total loss or reduction
of an involucre has occurred several differ-
ent times and places in the taxon.

The key to understanding the origin and
evolution of Chorizanthe lies in the unstu-
died South American perennial species. An
examination of available herbarium material
seems to point the origin of these perennials
to the subgenus Eucycla of Eriogonum, a
theory which seems reasonable if Eriogonum
is, as I suspect, the basic element of the
subfamily. There are, however, some diffi-

180

GREAT BASIN NATURALIST MEMOIRS

No. 2

culties which in theory can be excused but
need to be mentioned.

If Chorizanthe evolved from Eriogonum
subgenus Eucycla, it did not evolve from
any extant group of the subgenus. One can
account for the subshrubby habit of the per-
ennial Chorizanthe as having come from Eu-
cycla. Even the hooked, awned condition of
the involucre could be traced to the sub-
genus as several extant species of Eucycla
have long, sharply acute involucral lobes
which, while not awned, could point to a
stage in the development of the awned con-
dition. The narrow, essentially basal leaves
of Chorizanthe can be traced to Eucycla, as
can the congested, cymose inflorescence.
Two major drawbacks exist. One is the
straight embryo of Chorizanthe (Goodman
1934), whereas all species of Eucycla have a
curved embryo (Reveal 1969a, b). The sec-
ond is the six-lobed involucre of Chori-
zanthe, while the majority of species in Eu-
cycla are five-lobed.

The critical hint here, I believe, is the
presence of the remains of the ochrea in
some species of South American Chori-
zanthe. As Grant (1971) has noted, a given
character may or may not be selected for or
against, and thus, while the direction of the
subfamily Eriogonoideae has been to get rid
of the ochrea, at sometime in its history of
divergency from the rest of Polygonaceae it
must have possessed this feature. If, as I sus-
pect, the perennial species of Chorizanthe
are the most ancient extant members of the
subfamily, then it would follow that these
plants would exhibit some of the more
primitive features of the subfamily and pro-
vide helpful keys to its origin. By the same
token, while I accept Eriogonum as the bas-
ic expression of the subfamily, and Chori-
zanthe as a derived element, one need not
look further than extant and derived mem-
bers of Eriogonum to find all unique fea-
tures of the South American perennials ex-
cept the ochrea. Thus, if Chorizanthe
evolved as a preliminary expression from
Eucycla, as did the subgenera Eriogonum or
Oligogonum, then suddenly we find species

of Eriogonum '•'.n a straight embryo and a
six-lobed involucre. The genus Oxutheca,
which can trace its immediate origin to
Eriogonum subgenus Ganysma, has awned
involucres, and the reduction in the number
of flowers per involucre can be seen in sev-
eral different groups of Eriogonum, although
admittedly this feature is almost entirely re-
stricted to annual species.

The next critical step in this discussion is
how did the perennial members of Chori-
zanthe get to South America while Eriogo-
num did not, and if Chorizanthe evolved
from Eriogonum subgenus Eucycla in North
America, why are there no perennial spe-
cies of Chorizanthe in North America?

The first part of this question can be eas-
ily answered. The only members of Eriogo-
noideae in South America are those which
have a distinctly awned, or hooked, in-
volucral lobe. As Stebbins (1974) has point-
ed out, such an adaptation can be a success-
ful means of long-distance dispersal. Thus
Eriogonum (with the exception of E. diva-
ricatum, an annual species which was found
as a waif in eastern Argentina) is perhaps
lacking from South America due to the ab-
sence of an awned involucre. The second
part of this question, why the perennial spe-
cies of Chorizanthe are missing from North
America, is much more difficult.

Two options exist about the existence of
perennials in South America and their lack
in North America, and a third option can
be proposed on the basis of either of the
first two options if future studies should
make such an option necessary from a tax-
onomic point-of-view.

The first two options are closely inter-
twined and deal with the actual origin of
the perennial species in South America and
the annual species in North America. As-
suming the idea that Chorizanthe evolved as
a perennial group from Eriogonum in North
America, then it had to have migrated to
South America as a perennial and become
established as a perennial. The South Amer-
ican populations, I believe, have remained
essentially unchanged since their (or its) in-

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

181

itial introduction with some speciation oc-
curring there within rather limited parame-
ters. One hint that this is so is that all of
the perennial Chorizanthe species fall within
extremely narrow limits morphologically,
and, while several species (close to 25) have
been described, the actual number of valid
species seems to be much less than that.
The one annual species in South America is
apparently a much more recent in-
troduction than the perennial species be-
cause it is closely related to the single most
widespread annual species in North Ameri-
ca (Goodman 1934).

The first option states that Chorizanthe
evolved in North America and migrated as
a perennial to South America as a single in-
troduction, with the North American per-
ennials gradually being replaced by annual
species. In South America, the perennial
species were subjected to little direct selec-
tion pressure, while in North America the
perennial members of Chorizanthe were
subject to intense pressures from the rapidly
evolving and highly competitive, closely re-
lated genus Eriogonum. In order to survive
and compete against Eriogoniim, which, I
feel, was rapidly adopting the annual habit,
Chorizanthe also had to change if this hy-
pothesis is feasible.

The second option is that the perennial
species of Chorizanthe in North America be-
came extinct, while the South American
species remained. The annual habit then de-
veloped in Chile, and only the annual spe-
cies were introduced into North America.
Raven (1963) has noted that, while the ma-
jority of species probably migrated north to
south, some certainly went from south to
north. Once in North America, the annual
species underwent active adaptive radiation
similar to that observed in such annual
groups of Eriogoniim as the subgenera Ga-
nysma and Oregonium.

The third option states that the South
American perennials represent a genus of
plants distinct from the North American
(and one South American) annuals. If this is
so, then the name Chorizanthe would be ap-

plied to the South American perennials,
while the annual species would be called
Acanthogonum, or, if that genus proves dis-
tinct, Eriogonella. This option takes on
added significance if the following scenario
should prove correct after careful system-
atic studies. If indeed Chorizanthe evolved
as a perennial and migrated southward, and
the northern element became extinct, did
the annual species evolve prior to the ex-
tinction of the perennial group or did the
annual species begin from a whole new
series of events? In option one, I have ac-
cepted the first part of this question, but if
the second were the case, then it will be
impossible to retain the North American an-
nuals in the genus Chorizanthe.

The recently discovered Dedeckera eu-
rekensis probably evolved from the sub-
genus Eucycla of Eriogonum, and most
likely from the section Corymbosa. Its ori-
gin is likely most recent. It differs from all
other members of the subfamily in having a
head of subsessile or sessile flowers, borne
on a slender peduncle and subtended by
two to five foliaceous bracts, and a single,
short-pedicellate axillary flower at the base
of each peduncle. It differs from Eriogonum
in lacking an involucral tube. This mono-
typic genus is known only from a single
site where about 200 individual plants are
found. It is likely that the genus evolved in
place within recent history and, while its
range has expanded and decreased through-
out its brief history, it is unlikely that the
plant has been beyond the restrictive eco-
logical confines of the Death Valley region
of eastern California.

The pubescence of Dedeckera is similar to
that of Eriogonum intrafractum Cov. &
Morton, another Death Valley endemic,
which is the only representative of the sub-
genus Clastomyelon and a few other mem-
bers of Eriogonum. Of all the subgenera of
Eriogonum, Clastomyelon is the most dis-
tinctive on pure morphological grounds in
that the stems are broken into a series of
ringlike segments, the numerous flowers
rupture the involucral tube into irregular

182

GREAT BASIN NATURALIST MEMOIRS

No. 2

segments, and the bractlets are foliaceous at
least in part. While it is possible to trace
the origin of E. intrafractum to the sub-
genus Eucycla, where D. eurekensis also
evolved from, both are amazingly distinct,
with D. eurekensis significantly more so
than E. intrafractum. It is interesting that in
the Death Valley area, where speciation has
been rather spectacular (Stebbins & Major
1965), Polygonaceae should be blessed with
so many different expressions. Much like
Gihnania, which will be discussed below,
Dedeckera and E. intrafractum have come
about in recent times, influenced by the en-
vironmentally profound selection pressures
of the area.

The remaining satellite genera related to
Eriogonum evolved from the subgenus Ga-
nysma.

Oxytheca is being studied currently by
Ertter and me to determine the exact make-
up of this genus. We have excluded O. lu-
teola, placing it in a new genus, Good-
mania. The remaining nine species,
however, may or may not be all related.
Oxytheca dendroidea, O. watsonii, and the
South American plants are related to
Eriogonum spergulinum A. Gray, and, based
on this close morphological similarity,
Goodman (in herbaria) has placed these
plants in Eriogonum. Oxytheca perfoliate!
probably belongs to this complex of species
(Goodman would have placed the taxon in
Eriogonum), but it is morphologically dis-
tant from the other members. As for O. par-
ishii and an undescribed taxon from the San
Bernardino Mountains of California, they
present a problem. Goodman (in herbaria)
would have placed these in Eriogonum, but
both seem more closely related to E. api-
culatum S. Wats, and E. parishii S. Wats,
than E. spergulinum. Small (1898) placed O.
parishii in a monotypic genus, Acanthos-
cyphus. It is possible that Acanthoscyphus
should be recognized if it can be shown
that O. parishii and its related taxon are
distinct from that group of Oxytheca species
typified by O. dendroidea.

This complex of species (excluding the

Oxytheca parishii complex for a moment)
seems to have developed in the pygmy
woodlands of the Great Basin and, in par-
ticular, along the western edge of the Great
Basin. This is a complex of volcanic sandy
soils that are widely scattered. I suspect the
group evolved during the Quaternary (prob-
ably the Pleistocene), with the introduction
of the South American phase in recent geo-
logical time (see Raven 1963).

Goodman (in herbaria) placed Oxytheca
caryophylloides, O. emarginata, and O. trilo-
bata in a new genus. Our preliminary stud-
ies of these species seem to indicate that
they too developed from Eriogonum api-
culatum, E. parishii complex, and perhaps
one should consider if these species too
ought not to be referred to Acanthoscyphus.
These three species differ from O. parishii
and its undescribed relative in having a
five-lobed involucre instead of the nonlobed
tube with 4 to 30 long bristled awns. All of
these plants are found in the granitic moun-
tains of southern California and northern
Baja California and occur in approximately
the same type of ecological niche.

I have come to look upon Oxytheca par-
ishii, O. caryophylloides, O. emarginata, and
O. trilohata as a group that has evolved in
the Pleistocene in the mountainous region
of southern California and adjacent Mexico
as the result of sudden and explosive evolu-
tionary changes in the gene makeup of the
rapidly developing annual species of both
Eriogonum and Chorizanthe. This is not to
say that Chorizanthe played a direct role in
the development of these species, but one
should remember that Centrostegia (and in
particular C. insignis) probably developed
at the same time, and this genus is similar
to Chorizanthe. Thus, I suspect, a whole
series of rapid changes were in the process
at this time in a small portion of Eriogo-
num, which possibly resulted in this group
of Oxytheca, Centrostegia, and perhaps (if
option three is correct) the annual species
of Chorizanthe as well. If this conclusion
should prove correct with regards to the
species now placed in Oxytheca, then Good-

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

183

man would have been proved correct, and
these species would have to be placed in a
different genus.

Stenogotium is a step-child in this group
of satellite genera. It is closely related to
Eriogonum, differing mainly in the construc-
tion of the involucral bracts. While in
Eriogonum the involucre is distinctly tubu-
lar, the involucre of Stenogonum is com-
posed of two whorls of three lobes. That
this condition is possible within a genus
clearly and closely related to Eriogonum is
an important step, because it does demon-
strate the potential for a six-lobed or three-
lobed involucre as found in the annual spe-
cies of Chorizanthe.

Stenogonum evolved from the Eriogonum
inflatum Torr. & Frem. complex of the sub-
genus Ganysma. It is thought that the origin
of this clay-inhabiting genus is relatively re-
cent and has evolved to a point about on
the par with the degree of divergence seen
in Oxytheca. Stenogonum is an annual
which has evolved from a "hot desert" com-
plex on the Colorado Plateau, which is an
area somewhat intermediate between the
Mojave Desert and the Great Basin in terms
of phvsiological stress. Eriogonum inflatum
var. inflatum is found on the Plateau, but
the more common phase is not the per-
ennial var. inflatum, but the annual var.
fusiforme (Small) Reveal. Likewise, var. in-
flatum is usually found in rocky places
above the clay hills and flats, while var.
fusiforme is typical of the clay sites. Thus it
is that the genus Stenogonum has evolved
by successfully occupying the clay habitat
that, for the most part, members of the £.
inflatum complex cannot enter.

Two genera are difficult to directly asso-
ciate with Eriogonum, and both perhaps
have recently evolved in the subfamily.
They are Goodmania and Gilmania. The
two seem to be related, as both are pros-
trate to low-spreading annuals with pub-
escent yellow flowers, cauline leaves, and
small, smooth achenes. Goodmania has in-
volucral bracts which subtend each cluster
of flowers and act as a protective involucre.

hi Gilmania, all involucral bracts are lack-
ing, but the three foliaceous leaves, when
the plants are immature (but that particular
branch is in full flower), are held close to-
gether by the shortened internodes so that
each cluster of flowers is positioned above
the lower whorl of three leaves so that the
flowers are protected both by these leaves
and the whorl of upper leaves as well. In
this condition, the long pedicels extend the
ripened flowers beyond the protective con-
fines of the three leaves so that pollination
may occur. In this fashion, the flowers of
Gilmania are better protected from the ele-
ments than those of Goodmania.

Goodmania could possibly be traced to
Oxytheca, but I think not. True, it has an
awned involucral bract, but, in fact, these
bracts are just that and they are not ar-
ranged into a distinct tube. The individual
bracts can be separated from each other
without disruption of tissue on an adjacent
bract. There is one bract that is longer than
the other four bracts which is unlike any
species of Oxytheca, but is a condition that
is seen in some species of annual Chori-
zanthe. The flowers of G. luteola are yel-
low, and no species of Oxytheca has yellow
flowers, and the plants of this species are
glabrous and bright green while those of
Oxytheca are glandular (at least in part) and
usually reddish or grayish in color. As I
look about the subfamily, I see a possible
close relationship with the subgenus Orego-
nium for this genus and Gilmania, and in
particular Eriogonum divaricatum. The sub-
genus Oregonium underwent a major up-
heaval in the hot, dry foothills of western
California, but E. divaricatum, E. pub-
erulum S. Wats., and other similar species
are more typical of the Great Basin. Thus,
while this group of species of Eriogonum
may hint as a possible place of origin for
Goodmania and Gilmania, the group seems
unsatisfactory, and no extant subfamily of
Eriogonum really reveals a logical place of
their origin.

As noted above, these two genera seem to
be recently evolved genera. Goodmania is

184

GREAT BASIN NATURALIST MEMOIRS

No. 2

usually found on the plains of old dry lake
beds in areas which were covered by water
during recent glacial periods. Gilmania oc-
curs on the lower rim of Death Valley on
alkaline soils near sea level, and thus in
areas that were covered by water less than
50,000 years ago. It is likely, therefore, that
both genera underwent their evolutionary
development at approximately the same
time, taking advantage of the same type of
opening environment niche.

The origin of Gilmania is somewhat more
difficult to postulate than that of Good-
mania. Cauline leaves in Eriogonum annuals
are infrequent, and when present are rarely
arranged in a pattern similar to that of Gil-
mania, nor are they like the leaves of Good-
mania. In Goodmania the leaves are two
and opposite, varying from laminar at the
lower nodes to acicular at the upper nodes.
In Gilmania, the leaves are in threes, with
two of the leaves opposite, and the third
opposite the next branch; all of the blades
are laminar. It seems unlikely that both Gil-
mania and Goodmania evolved from pre-
cisely the same element within Eriogonum,
but they probably did arise within the same
subgenus. I strongly suspect that the selec-
tive evolutionary pressures have been much
greater on Gilmania than Goodmania, thus
accounting for the great degree of demarca-
tion of Gilmania.

Curran (1885) was the first to call atten-
tion to the close relationship between
Eriogonum and Nemacaulis. Nemacaulis is
similar to E. gossypinum in that both have
copious bractlets and hairs surrounding and
protecting the flowers; in Eriogonum the
tubular involucre is broadly campanulate,
but in Nemacaulis the involucre is lacking
and replaced by subtending bracts. Beyond
this, the two taxa are notably distinct. Still,
it seems likely that the origin of Nemacaulis
can be traced to Eriogonum subgenus Ga-
nysma and, in particular, the section of Ga-
nysma which contains E. gossypinum.

Looking upon Nemacaulis as a recent de-
rivation from Eriogonum, it seems to have
undergone rapid development in the hot

deserts of southern California and adjacent
Mexico, occupying a position on the south-
ern geographical edge of Eriogonum section
Ganysma. I suspect the degree of difference
between Eriogonum and Nemacaulis is on
the magnitude of that exhibited by Eriogo-
num and Oxytheca.

The genus Hollisteria is a most difficult
genus to trace back to its possible point of
origin. It is a prostrate, spreading annual
with two sessile, yellow, woolly flowers sub-
tended by three slightly united involucral
bracts. In some respects, Hollisteria is inter-
mediate between Eriogonum and Chori-
zanthe. It differs from both in lacking a dis-
tinct involucral tube, but it is two-flowered
and thus similar to Centrostegia and has
acerose tips on the bracts similar to those
on Goodmania. It probably did not evolve
from an unknown perennial group as pro-
posed by Stebbins (1974) but more likely
developed from an annual complex.

I would like to say that Hollisteria could
have evolved from either Eriogonum sub-
genus Ganysma or Oregonium, but no ex-
tant group in either subgenus can really
point the way. I have tried to place the
genus near Chorizanthe, but still no one
group of that genus really is helpful. In
some respects, I have tried to fit it into a
pigeonhole between what Goodman termed
Eriogonella (C. membranacea Benth.) and
Centrostegia because here one can find a
combination of three-lobed involucres, yel-
lowish flowers, and a spreading annual
habit. Still, one compelling bit of evidence
that wrenches this entire scene is the nature
of the pollen grain. As Nowicke (pers.
comm.) has recently shown, the pollen
grains of Hollisteria and Lastarriaea are es-
sentially the same, and unlike any other
genus of Eriogonoideae. It is possible that
both Hollisteria and Lastarriaea evolved
from an extinct, independent complex of
annual species. One part of the complex
close to Eriogonum could have given rise to
Hollisteria while another part of the com-
plex close to Chorizanthe gave rise to Last-

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

185

As one might suspect from the foregoing
discussion, the genus Lastarriaea is also
somewhat intermediate between Eriogonwn
and the annual species of Cfiorizanthe, but
closer to the latter than the former. Last-
arriaea is a low, often spreading annual
without a distinctly tubular involucre, ace-
rose bracts, and whitish, glabrous, co-
riaceous tepals. Unlike all of the genera dis-
cussed to this point (with the exception of
some species of Chorizanthe and Netna-
caulis), Lastarriaea has only three anthers
per flower instead of the usual nine. And
unlike Hollisteria, which is an inland species
of the Inner Coast Ranges of California,
Lastarriaea is a coastal genus found in both
North and South America.

Goodman (1934) placed Lastarriaea in
Chorizanthe in the least specialized section
of the genus, and, while there are some su-
perficial similarities between the section
Suffrutices Benth. and Lastarriaea to the
point that perhaps it evolved from this sec-
tion, I doubt that its point of origin can be
traced to any extant section of Chorizanthe.
As noted above, a more likely situation is
that Lastarriaea developed early in the evo-
lution of the annual species of Chorizanthe
(or less likely, Eriogonum). I strongly sus-
pect that Lastarriaea became well estab-
lished in North America, and that it, the
one annual species of Chorizanthe, and Oxij-
theca all migrated to South America at ap-
proximately the same time as hitchhikers on
animals, probably during the Late Pliocene
(Raven 1963). The differences between the
North and South American elements are not
strongly expressed morphologically in these
annual species, although a strong difference
does not necessarily have to be expressed
(Grants 1967).

Mucronea is clearly derived from the an-
nual species of Chorizanthe. This genus has
a distinctly tubular involucre like Chori-
zanthe, but it and Centrostegia differ in
having three-lobed bracts instead of the typ-
ically entire bracts of Chorizanthe. The con-
spicuous bracts of Mucronea are united and
distinct, and in this feature the genus is sim-

ilar to Eriogonum and Oxytheca, especially
O. perfoliata. Mucronea is distinct from
Centrostegia and Oxytheca in having
straight cotyledons (Goodman 1934), but is
similar to Chorizanthe in this regard. I sus-
pect that Mucronea is a rather recent in-
novation within the Chorizanthe complex.

The genus Acanthogonum was recognized
as a distinct genus by Goodman (1955), but
I am still somewhat reluctant to recognize
it. When Torrey (1857) described Acan-
thogonum, he placed a single species, A. rig-
idum, in the genus. In 1858, Torrey ques-
tionably added a second species, A.
corrugatum, noting that this species was "al-
most intermediate between Acanthogonum
and Chorizanthe." Torrey and Gray (1870)
reduced both species to Chorizanthe and
added to the complex C. polygonoides and
C. watsonii. Goodman (1934) defined Acan-
thogonum to include A. rigidum and A. po-
lygonoides (Torr. & Gray) Goodman, and
these two species were retained in the
genus in his 1955 review. Basically Good-
man maintained the genus on the basis of
the curved cotyledons, but, as I am retain-
ing Eriogonella in Chorizanthe, which was
established (in part) on its curved cotyle-
dons, I cannot very well recognize Acan-
thogonum because of this feature. For now,
at least, the relationship between C. polyg-
onoides and C. corrugata, C. watsonii, and
C. orcuttiana Parry seems too close to allow
for a distinct genus to be established.

Centrostegia is a most difficult and di-
verse assemblage of species. As defined by
Goodman (1957), the genus consists of four
species, three of which, C. leptoceras A.
Gray, C. thurberi Gray ex Benth. in DC,
and C. vortriedei (Brandeg.) Goodman, form
one distinct element within the genus, but
C. insignis is decidedly aberrant although
even C. vortriedei is somewhat strange with-
in Centrostegia. I am inclined to restrict
Centrostegia to C. thurberi and C. leptoceras
but am lost when it comes to C. vortriedei,
and feel C. insignis should probably go into
a distinct genus. This latter species is cer-
tainly most closely related to Oxytheca,

186

GREAT BASIN NATURALIST MEMOIRS

No. 2

where Goodman (1934) placed it at one
time. As for C. thurberi and C. leptoceras,
one might look for an origin somewhat in-
termediate between Eriogonum and Chori-
zanthe. Until these species can be carefully
studied, especially cytologically, little can
be expressed about their relationships. I
have little faith in the one unifying charac-
ter, which is the three-lobed bract, and
would like to place more emphasis on the
involucral, floral, and vegetative features of
these plants.

Up to this point, the discussion has cen-
tered on the tribe Eriogoneae, which makes
up the vast bulk of Eriogonoideae. The oth-
er tribe of the subfamily, Pterostegeae, con-
tains only two monotypic genera. Time and
evolution have largely destroyed the inter-
connecting links between the two tribes so
that it is impossible to say what, if any, role
Eriogoneae might have played in the evolu-
tion of Pterostegeae, or the other way
around for that matter. The inflated bracts
of the fruiting specimens are unseen in
Eriogoneae, and the consistently opposite
leaves are rare. I suspect that the two tribes
are well separated now by time and events.

Until Parry's (1886) paper in which the
genus Harfordia was described, the true na-
ture of this narrowly restricted shrub was
unknown. Bentham (1844) had placed the
perennial in the genus Pterostegia not know-
ing if his species, P. macroptera, was a shrub
or not. It remained there until 1886. Only
Roberty and Vautier (1964) reduced Har-
fordia to Pterostegia. In spite of this, there
is little reason to closely associate Harfordia
with Pterostegia except in the feature of the
fruiting bracts and opposite leaves.

Hutchinson (1926, 1959, 1969) was a firm
believer in the concept that certain families
of flowering plants were fundamentally her-
baceous or woody. Polygonaceae, in his
view, was basically a herbaceous group in
which the woody, or shrubby, condition was
a secondary state. There is something to say
about this point, although it may seem con-
trary to the usual dicta (Bessey 1915). The
largest forms of Eriogonum, for example,

are highly derived forms from low, sub-
shrubby or shrubby groups. Eriogonum aus-
trinum (S. Stokes) Reveal is an annual spe-
cies that will form perennial individuals,
and this is a condition that will be seen in
other species as well. The reason to bring
this controversial subject up is Harfordia,
the perennial, versus Pterostegia, the annual.
It is possible that Harfordia represents the
residue of an ancient series of events in a
perennial line of evolution from which, at
some time in the past, the ancestral fore-
runners of Pterostegia evolved. This is the
reasonable approach. Another which cannot
be totally ignored is that Harfordia is a sec-
ondarily evolved perennial which developed
from an annual group in order to survive in
the extreme stress of long-term drought as-
sociated with the environment of central
Baja California. Pterostegia is basically a
mesic species, and I do not propose to im-
ply that Harfordia evolved from Pterostegia,
but anatomical and cytological studies may
be helpful in unraveling this question.

Summary

The subfamily Eriogonoideae is divided
into two tribes, Eriogoneae and Pteros-
tegeae, which are somewhat atypical mem-
bers of the Polygonaceae. The subfamily is
restricted to the more xeric areas of central
North America and western South America.
The basic extant expression in the subfamily
is Eriogonum, whose ancestral roots can
probably be traced to the tropical or sub-
tropical members of the family. Chorizanthe
was a major side-shoot from Eriogonum, and
from these two fundamental genera have
evolved a series of small, usually closely re-
lated genera. Eriogonum and the perennial
forms of Chorizanthe probably developed in
the pygmy woodlands of the Madro-Tertiary
geoflora, with the related genera evolving
mainly in the more xeric, hot deserts at ele-
vations lower than those in which Eriogo-
num is typically found.

The largest and most diverse genus is
Eriogonum, both in terms of numbers of

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

187

species and in expressions. Chorizanthe is
the next largest, but the degree of mor-
phological divergency in this genus is not as
great as in Eriogonum. As for the remain-
ing, smaller genera, each attempts to fill an
available morphological gap or ecological
niche, and for the most part, each is suc-
cessful.

Much work remains to be done on the
subfamily. The South American species of
Chorizanthe must be studied in the field.
The nature of the relationship between the
South American perennials and the North
American annuals of this genus must be de-
termined, and then, if the two should prove
distinct, we must decide by what name the
annuals should be called. Field studies are
now critically needed so that anatomical
and cytological material can be gathered,
and perhaps greenhouse investigations
made. A series of monographic studies are
now in progress, mainly on the genus
Eriogonum and its immediate relatives. In
time, these studies must be expanded
beyond the alpha taxonomic level where
they are now. This will be a continuing
challenge to anyone wishing to travel,
study, and investigate one of the world's
most unique groups of flowering plants.

Acknowledgments

The present paper was prepared at Brig-
ham Young University where I was on sab-
batical leave from the University of, Mary-
land the first six months of 1976. To the
various authorities at these two institutions,
and in particular Dr. Hugh D. Sisler of the
University of Maryland and Dr. Bruce N.
Smith of Brigham Young University, chair-
men of the Department of Botany at their
respective universities, I am most grateful
for the opportunity to be free for the one
semester to work on the proposed mon-
ograph on Eriogonum. The field and herba-
rium studies which have formed the basis
for this paper have been supported by Na-
tional Science Foundation grants GB-22645
and BMS75- 13063. I wish to thank my col-

league, Dr. C. Rose Broome, of the Univer-
sity of Maryland, for sharing with me her
ideas and thoughts on the complex problems
of evolution and speciation as they might
apply to Eriogonoideae.

Note Added in Proof

Since this manuscript was completed in
June 1976, the revision on Oxytheca has
been completed [Ertter, B. J. 1977. A revi-
sion of the genus Oxytheca (Polygonaceae).
Unpublished Master's Thesis, University of
Maryland Library, College Park] and a
number of minor changes must be ap-
pended. Ertter found that Oxytheca consists
of seven (not nine) species, with the "South
American form" being merely a subspecies
of O. dendroidea, and the "new species"
from the San Bernardino Mountains of Cali-
fornia a variant of O. parishii. Ertter con-
curs that O. dendroidea and O. watsonii are
related to Eriogonum spergidinum, and she
has shown conclusively that O. perfoliata is
clearly related to O. dendroidea. In fact, O.
watsonii, a rare species of west-central Ne-
vada, is intermediate between O. dendroidea
and O. perfoliata in many respects accord-
ing to Ertter.

Ertter also concurs that Oxytheca parishii
is most closely related to Erigonum apicula-
tum and E. parishii, but in doing so called
attention to an error in my own work on
Eriogonum (Reveal 1969a). In my revision
of Eriogonum I placed E. spergidinum and
E. apiculatum in widely separated sections
of the subgenus Ganysma. Ertter has shown
that these two species complexes are much
more closely related than I had thought. In
the present paper I raised the question that
if O. dendroidea and its allies arose from
the E. spergidinum complex (in one part of
Ganysma), and O. parishii arose from E.
apiculatum (in another part of Ganysma),
then perhaps the genus Acanthoscyphus
should be recognized. Such a situation now
is unnecessary.

The relationship between Oxytheca par-
ishii and the remaining members of the

188

GREAT BASIN NATURALIST MEMOIRS

No. 2

genus in southern California (O. caryophyl-
loides, O. trilobata, and O. emarginata) is
still tenuous. Ertter has shown, however,
that O. parishii is more similar to the O.
dendroidea complex than it is to the O.
caryophylloides. She has proposed to place
O. dendroidea, O. watsonii, and O. per-
foliate! in their own typical section, with O.
parishii in a monotypic section. As for the
other three species, these are going into a
third section, a taxon somewhat removed
from the other sections.

I still believe that the southern California
elements evolved as a group in the moun-
tainous regions of southern California dur-
ing the Pleistocene, but I now feel that this
development came not from isolated ele-
ments with Eriogonum but from a broadly
connected group of annual species all be-
longing to this one genus which were un-
dergoing collectively rapid evolution (see
Raven and Axelrod discussed below).

As for the troublesome Centrostegia in-
signis, recent conversations with Goodman
confirm the supposition that this species is
seriously out of place in Chorizanthe (Good-
man 1934), Oxytheca (Ertter 1977, cited
above), and even Centrostegia (Goodman,
pers. comm.) and that it most likely will
have to be placed in its own monotypic
genus. This question is now being explored.

Nowicke's pollen work discussed above
has now been published [Nowicke, J. W.,
and J. J. Skvarla. 1977. Pollen morphology
and the relationship of the Plumbaginaceae,
Polygonaceae, and Primulaceae to the order
Centrospermae. Smithsonian Contr. Bot. 37:
1-64.].

Lastly, my manuscript has been used by
Drs. Peter H. Raven and Daniel I. Axelrod
in a book entitled Origin and Relationships
of the California Flora published by the
University of California Press. Based upon
my review of their manuscript (which
should be published at approximately the
same time as this volume), their work will
nicely complement the present volume.

Literature Cited
Abrams, L. 1944. Illustrated flora of the Pacific-

states. Vol. II. Stanford University Press, Stan-
ford, California.

Airy Shaw, H. K. (Rev.) 1973. A dictionary of the
flowering plants and ferns. 8th ed. Cambridge
University Press, Cambridge.

Axelrod, D. I. 1950. Evolution of desert vegetation
in western North America. Publ. Carnegie Inst.
Wash. 590: 217-306.

1958. Evolution of the Madro-Tertiary geo-

flora. Bot. Rev. 24: 433-509.

1966. Origin of deciduous and evergreen

habits in temperate forests. Evolution 20: 1-15.

Benson, L. 1974. Caryophyllales. Encyclop. Brit-
annica 3: 973-976.

Bentham, G. 1836. On the Eriogoneae, a tribe of
the order Polygonaceae. Trans. Linn. Soc, Bot.
17: 401-420.

1844. The botany of the voyage of HMS Sul-
phur, under the command of captain Sir Ed-
ward Belcher . . . during the years 1836-1842.
Smith, Elder & Co., London.

1856. Eriogonum. In: A. deCandolle, Pro-

dromous systematis naturalis regni vegetabilis
14: 5-23.

Bentham, C, and J. D. Hooker. 1880. Genera
plantarum. Vol. Ill, part 1. Reeve and Co.,
London.

Bessey, C. E. 1915. The phylogenetic taxonomy of
flowering plants. Ann. Missouri Bot. Gard. 2:
109-164.

Buchinger, M. 1950. Sanmartinia, nuevo genero Pa-
togonico de poligonaceas. Comun. Inst. Nac. In-
vest. Ci. Nat., Ser. Ci. Bot. 1(4): 1-13.

Coville, F. V. 1893. Botany of the Death Valley
expedition. Contr. U.S. Natl. Herb. 4: 1-363.

1936. Gilmania, a new name for Phyllogo-

num. J. Wash. Acad. Sci. 26: 209-213.

Cronquist, A. 1968. The evolution and classifica-
tion of flowering plants. Houghton Mifflin Co.,
Boston.

Curran, M. K. 1885. Classification of the Erio-
goneae as affected by some connecting forms.
Bull. Calif. Acad. Sci. 1: 272-275.

Dammer, U. 1892. Polygonaceae. In: A. Engler, and
K. Prantl, Die natiirlichen Pflanzenfamilies
3(70): 1-36.

Fischer, F. C. L., and C. A. Meyer. 1835. Index
seminum, quae hortus botanicus and imperialis
Petropolitanus pro mutua commutatione offert.
St. Petersburg.

Goodman, G. J. 1934. A revision of the North
American species of the genus Chorizanthe.
Ann. Missouri Bot. Gard. 21: 1-98.

1943. Notes on Chorizanthe. Leafl. W. Bot.

3: 230.

1955. The genus Acanthogonwn, tribe Erio-
goneae. Leafl. W. Bot. 7: 234-236.

1957. The genus Centrostegia, tribe Erio-
goneae. Leafl. W. Bot. 8: 125-128.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

189

Grant, V. 1967. The selective origin of in-
compatibility barriers in the plant genus Gilia.
Amer. Naturalist 100: 99-118.

1971. Plant speciation. Columbia University

Press, New York.

Gray, J. 1964. Northwest American Tertiary pa-
lynology: the emerging picture. Pacific Sci.
Congr. Ser. 10: 21-30.

Greene, E. L. 1885. Studies in the botany of Cali-
fornia and parts adjacent. Bull. Calif. Acad. Sci.
4: 179-228, 276-282.

Gross, H. 1913. Beitrage zur Kenntnis de Polyg-
onaceen. Bot. Jahrb Syst. 49: 234-339.

Heller, A. A. 1910. Catalogue of North American
plants. 3d ed. Published by the author, Los
Gatos.

Hess, W. J., and J. L. Reveal. 1976. A revision of
Eriogonum (Polygonaceae), subgenus Pterogo-
num. Great Basin Nat. 36: 281-333.

Hooker, J. D. 1853. Catalogue of Mr. Geyer's col-
lection of plants gathered in the upper Mis-
souri. Hooker's J. Bot. Kew Gard. Misc. 5: 257-
265.

Hoover, R. F. 1966. Miscellaneous new names for
California plants. Leafl. W. Bot. 10: 337-350.

1970. The vascular plants of San Luis Obispo

County, California. University of California
Press, Berkeley.

Hutchinson, J. 1926. The families of flowering
plants. I. Dicotyledons. Macmillan & Co., Lon-
don.

1959. The families of flowering plants. Ox-
ford University Press, New York.

1969. Evolution and phylogeny of flowering

plants: Dicotyledons; fact and theory. Academic
Press, London.

Jepson, W. L. 1913. Eriogonum. A flora of Califor-
nia 1: 376-428.

Johnston, I. M. 1929. A collection of plants from
the high cordilleras of northwestern San Juan.
Physis (Buenos Aires) 9: 297-326.

Jones, M. E. 1903. Eriogonum. Contr. W. Bot. 11:
4-18.

1908. Contribution to western botany. Contr.

W. Bot. 12: 1-100.

King, P. B. 1959. The evolution of North America.
Princeton University Press, Princeton, New Jer-
sey.

Lawrence, G. H. M. 1951. Taxonomy of vascular
plants. Macmillan & Co., New York.

Meisner, C. F. 1841. Plantarum vascularium gen-
era. Fasc. 10, tab. diagn. 313-344, comm. 225-
252.

1856. Polygonaceae. In: A. deCandolle, Pro-

dromus systematis naturalist regni vegetabilis
14: 1-4, 28-186.

Melchior, H. (ed.) 1964. Engler's syllabus der
Pflanzenfamilien. 12th ed., Vol. II. Band An-
giosperm. Gebriider Borntraeger, Berlin.

Miers, J. 1851. On two genera of plants from the
cordillera of Chile. Proc. Linn. Soc. London 2:
154-157.

Moreau, R. L. P., and S. Crespo. 1969.
Rehabilitacion de Eriogonum ameglinoi Speg. y
caida en simonimia del genero Sanmartinia (Po-
lygonaceae). Darwiniana 15: 112-114.

Munz, P. A. 1974. A flora of southern California.
University of California Press, Berkeley.

Munz, P. A., and D. D. Keck. 1959. A California
flora. University of California Press, Berkeley.

Nuttall, T. 1817. Observations on the genus
Eriogonum and the natural order Polygonaceae
of Jussieu. J. Acad. Nat. Sci. Philadelphia 1: 24-
31, 33-37.

1848a. Descriptions of plants collected by

Mr. William Gambel in the Rocky Mountains
and upper California. Proc. Acad. Nat. Sci.
Philadelphia 4: 7-26.

1848b. Description of plants collected by

William Gambel, M.D., in the Rocky Mountains
and upper California. J. Acad. Nat. Sci. Phila-
delphia II, 1: 149-189.

Parry, C. C. 1884. Chorizanthe R. Brown. Revision
of the genus and rearrangement of the annual
species, with one exception, all North Ameri-
can. Proc. Davenport Acad. Nat. Sci. 4: 45-63.

1886. Harfordia Greene & Parry. A new

genus of Eriogoneae from Lower California.
Proc. Davenport Acad. Nat. Sci. 5: 26-28.

Philippi, R. A. 1864. Plantarum novarum Chilen-
sium. Linnaea 3: 225-228.

1873. Nuevas plantas en el herbario Chileno.

Anales Univ. Chile 1873: 536.

1895. Plantas nuevas Chilenas de las familias

que corresponden al tomo V de la obra de Gay.
Polygoneas. Anales Univ. Chile 91: 487-526.

Raven, P. H. 1963. Amphitropical relationships in
the floras of North and South America. Quart.
Rev. Biol. 38: 151-177.

Remy, J. 1851. Botanica [Flora Chileana]. In: C.
Gay, Historia fisica y politica de Chile 5(3):
257-284.

Reveal, J. L. 1969a. A revision of the genus
Eriogonum (Polygonaceae). Unpublished doctor-
al dissertation, Brigham Young University Li-
brary, Provo, Utah.

1969b. The subgeneric concept in Eriogonum

(Polygonaceae), pp. 229-249. In: J. Gunckel
(ed.), Current topics in plant science, Academic
Press, New York.

Reveal, J. L., and B. J. Ertter. 1976a. A reestab-
lishment of the genus Stenogonum (Polyg-
onaceae). Great Basin Nat. 36: 272-280.

1976b. Goodmania (Polygonaceae), a new

genus from California. Brittonia 28: 427-429.

Reveal, J. L., and J. T. Howell. 1976. Dedeckera

190

GREAT BASIN NATURALIST MEMOIRS

No. 2

(Polygonaceae), a new genus from California.
Brittonia 28: 245-251.

Roberty, G., and S. Vautier. 1964. Les genres de
Polygonacees. Boissiera 10: 7-128.

Shreve, F., and I. L. Wiggins. 1964. Vegetation
and flora of the Sonoran Desert. Stanford Uni-
versity Press, Stanford, California.

Small, J. K. 1898. Studies in North American Po-
lygonaceae. I. Bull. Torr. Bot. Club 25: 40-53.

Specazzini, C. 1902. Nova addenda ad Floram
Patagonicam, 3. Anales Mus. Nac. Hist. Nat.
Buenos Aires 7: 135-308.

Stafleu, F.' A., et al. (eds.). 1972. International
code of botanical nomenclature. Regnum Veg.
82: 1-426.

Stebbins, G. L. 1942. Polyploid complexes in rela-
tion to ecology and the history of floras. Amer.
Naturalist 76: 36-45.

1974. Flowering plants. Evolution above the

species level. Harvard University Press, Cam-
bridge.

Stebbins, G. L., and J. Major. 1965. Endemism
and speciation in the California flora. Ecol.
Monogr. 35: 1-35.

Stokes, S. G. 1904. The genus Eriogonum. A study
of its distribution and development in con-
nection with the climatic changes in the West
during and since the glacial period. Also a revi-
sion of its species. Unpublished master's thesis,
Stanford University Library, Stanford, Califor-
nia.

1936. The genus Eriogonum. A preliminary

study based on geographic distribution. J. H.
Neblett Pressroom, San Francisco.

Stokes, S. C, and G. L. Stebbins. 1955.
Chromosome numbers in the genus Eriogonum.
Leafl. W. Bot. 7: 228-233.

Takhtajan, A. 1959. Die Evolution der Angiosper-
men. Fischer, Jena.

1969. Flowering plants: Origin and dispersal.

Oliver and Boyd, Edinburgh.

Thorne, R. F. 1968. Synopsis of a putatively phy-
logenetic classification of the flowering plants.
Aliso 6: 57-66.

TlDWELL, W. D., S. R. Rushforth, and D.
Simper. 1972. Evolution of floras in the In-
termountain Region, pp. 19-39. In: A Cronquist,
et al., Intermountain flora: vascular plants of
the Intermountain West. Vol. 1. Hafner Publ.
Co., New York.

Torrey, J. 1857. Description of the general bot-
anical collections. In: A. W. Whipple, Reports
of explorations and surveys to ascertain the
most practicable and economical route for a
railroad from the Mississippi River to the Pacif-
ic Ocean. Vol. 4, part 4. A. O. P. Nicholson,
Printer, Washington, D.C.

1858. Description of plants collected along

the route by W. P. Blake. In: R. S. Williamson,
Reports of explorations and surveys to ascertain
the most practicable and economical route for a
railroad from the Mississippi River to the Pacif-
ic Ocean. Vol. 5, part 2. A. O. P. Nicholson,
Printer, Washington, D.C.

Torrey, J., and A. Gray. 1870. A revision of the
Eriogoneae. Proc. Amer. Acad. Arts 8: 145-200.

Watson, S. 1877. Descriptions of new species of
plants with revision of certain genera. Proc.
Amer. Acad. Arts 12: 246-278.

1879. Contribution to American botany, 2:

Description of some new species of North
American plants. Proc. Amer. Acad. Arts 14:
288-303.

Welsh, S. L. 1974. Anderson's flora of Alaska and
adjacent parts of Canada. Brigham Young Uni-
versity Press, Provo, Utah.

PROBLEMS IN PLANT ENDEMISM ON THE COLORADO PLATEAU
Stanley L. Welsh

Abstract.— The problem of distribution of plant endemics is explored, especially as regards distribution of
these plant species on elevational and substrate bases. Endemics per unit area are greater at elevations above
1980 m and on finely textured soils. A summary of data and a list of references is presented.

The Colorado Plateau has long been
known as a center for endemic plants. The
basis for that endemism most certainly in-
volves the unique features of the Plateau
Province, especially as regards climate, geo-
logical history, and position along migratory
routes. Each of these features has played a
role, and the interpretation of the nature of
endemism in the region must reflect the
part each has played. Climate is considered
as arid for the region generally, but the
mountains within and bordering the Prov-
ince stand as moist discontinuities in this
arid middle portion of the Colorado River
drainage system. Geology is, in large part,
displayed in technicolor, with Mesozoic and
Cenozoic strata dominating, and with the
thin cover of vegetation hardly obscuring
the geology. Only at higher elevations is the
mantle of vegetation sufficiently dense as to
cloak the geological substrate. Volcanic in-
trusives and extrusives are limited in area.
Sedimentary formations constitute the main
features of the plateau region.

Canyons of grand size are entrenched
into the surfaces of the uplifted plateau,
where formations are displayed along the
margins of horizontal or gently dipping
plateaus. Broad anticlines, synclines, or
gentle to steeply plunging monoclines have
been eroded to expose hogbacks or cuestas
where vast stratigraphies are exposed in rel-
atively short distances.

Main migratory routes into the plateau
involve the mountain sequences in the
Rocky Mountains along both the eastern
and western margins. The great river system
has provided a lane for movement of prop-
agules both up and down stream. Great
Plains influence appears to have been
mainly from the southwest. Hence, the flora
of the plateau is a function of the sources
of its plants; Mohavean, Chihuahuan, Great
Plains, Rocky Mountain, Great Basin, etc.

My own interest in plant endemics within
the Colorado Plateau began more than two
decades ago when I began research leading
to a thesis on the vegetation of the Utah
portion of Dinosaur National Monument
(Welsh 1957). The geological control of
vegetation at lower elevations in the
plateau is readily apparent, but nowhere so
overwhelming as along the margins of the
Split Mountain anticline at Dinosaur Na-
tional Monument. The interaction of various
substrates to low annual precipitation
through long periods of time has led to the
demonstration of edaphic differences on a
grand scale. Soil formation in a traditional
sense is unknown. The substrate surface is
often merely residual parent material only
slightly modified from that a few inches be-
low the surface. Alluvium and colluvium
function differentlv from residual or parent
materials.

Conditions for growth on the substrates

'Life Science Museum and Department of Botany and Range Science, Brigham Young University, Provo, Utah 84602

191

192

GREAT BASIN NATURALIST MEMOIRS

No. 2

available are often rigorous at best. Fre-
quently, those plants capable of estab-
lishment and reproduction are few in num-
ber. Competition is therefore limited, with
the harsh environments supporting very few
higher plant species.

This point is illustrated by a statement as-
cribed to the venerable Walter P. Cottam
concerning vegetation of the Mancos Shale
formation. While camped on that formation,
he is reported to have stated that, "there
were only four species of plants growing
here, and I don't know three-fourths of
them.'' This statement, whether true or not,
is indicative of the paucity of vegetation on
some of the formation, and of the peculiar
nature of some of the entities growing
there.

Worst possible substrates available are the
clays derived from saline shales such as
members of the Mancos Formation and its
counterparts such as the Tropic Shale (Tu-
nunk equivalent). Salt contents of 30,000
ppm are not uncommon. Additionally, the
clays perform in a special manner when
water is added. A rainstorm of some six
hours' duration supplied from 2 to 6 inches
of water in June 1972 to the Tropic Shale
east of Glen Canyon City in Kane County,
Utah. Penetration into the clay of the Trop-
ic Shale did not exceed 2 inches. Even
when wet to field capacity the clays pro-
vide very little moisture for plant growth.
Still, a few plants do grow on this substrate,
some of them continuing to flower and to
mature fruit at soil moisture levels below
the arbitrary 15 atmosphere level accepted
as the permanent wilting point.

At higher elevations, where rainfall is
greater, even these substrate types are over-
whelmed by vegetation and a soil mantle is
developed which eventually is effective in
insulating the vegetative cover from the ef-
fects of high salinity and from such sub-
stances as selenium.

Sandstones and their derivatives form the
other extreme in edaphic situations. Soluble
salts are fewer, selenium is more restricted,
water penetration is greater, and the pro-

portion of water which can be extracted
from soil at field capacity is greater. In-
creased species diversity on sands is appar-
ently a function of the greater ease of es-
tablishment and better all-around water and
soil relationships, given equal elevation,
temperature, humidity, and opportunity for
establishment.

Glaciation has apparently not played a
major role in the plateau, except in the
highlands along its fringes. Persistence of
plants through the Pleistocene has been a
possibility in at least the lower elevation
portions, and indeed in most of the Plateau
Province.

With this background, it seems both fea-
sible and necessary to attempt to quantify
the data on endemism. The data have been
generated from information recorded on
herbarium labels of specimens deposited at
Brigham Young University and at the Na-
tional Museum of Natural History, Smithso-
nian Institution. Further data have been
taken from various taxonomic treatments
(see list of references). Despite all attempts,
the information available is fragmentary,
and the conclusions derived from that infor-
mation must be regarded as tentative. A
portion of the problem involves the nature
of endemics generally. Some of them are
highly restricted and they have been col-
lected only once or a very few times. Infor-
mation on substrate, elevation, geological
formation, and other pertinent information
is not available either on the herbarium ma-
terials or in the literature. Further, most
plant taxonomists (and others who collect
plants) are not well prepared in identi-
fication of substrate types. The clays and
silts of some investigators will be the sands
and gravels of others. Despite the short-
comings of the information, there seem to
be trends which can form the basis for fur-
ther investigations.

The data summarized from the sources
investigated are presented in Table I. Only
34 families of vascular plants are known to
have endemic species in the region. The to-
tal number of endemic entities is some 340,

1978

INTERMOUNTA1N BIOGEOGRAPHY: A SYMPOSIUM

193

an average of about 10 per family. The en-
demies are far from equally divided among
the various families. The Boraginaeeae (20),
Compositae (58), Leguminosae (86), Polyg-
onaeeae (28), and Scrophulariaeeae (29) con-
tain some 221 species or about 65 percent
of all endemic taxa. The greatest number of
taxa occurs in the Legume family. The 86

entities constitute some 25 percent of all of
those in the region.

The genus Astragalus, with 62 endemic
taxa, is not only the largest single contrib-
utor to the list of endemics, but contributes
a greater number of taxa than any family
beside the legumes. Most of those taxa oc-
cur at low elevations, contributing to the

Table 1. Summary of plant endemics on the Colorado Plateau

Elevation

Family
Name

Taxa
No.

C, S, M

Substrate
Sand, G Both

NA

1980 m
<6500ft

1980m
>6500ft

Both

NA

Apocynaceae

Asclepiadaceae

Boraginaceae

4

4

20

1
3

3
2

7

1
2
10

4
2
9

1
3

4

4

Cactaceae

11

2

1

2

6

5

6

Capparidaceae

Caryophyllaceae

Chenopodiaceae

Compositae

Cruciferae

1

1

6

58

11

1
2
9
6

2

29

3

1

1
4

2

1
16

1

4

25

7

1

12
4

2

7

14

Cyperaceae

Elaeagnaceae

Ephedraceae

Euphorbiaceae

Gramineae

5

1
1
1
2

1

3
1

1

2

2

2

1
1

3
2

1

Gentianaceae

2

1

1

1

1

Geraniaceae

1

1

1

Hydrophyllaceae
Labiatae

12
2

6

3

1
2

2

9
1

1

1

2

Leguminosae
Astragalus

62

18

35

9

42

16

4

others

24

5

17

1

16

4

3

1

Liliaceae

4

4

3

1

Najadaceae
Nyctaginaceae
Onagraceae
Papaveraceae

1
2
9
1

3

1
2
6
1

2
6

1

1
2

1

Polemoniaceae

13

3

10

5

5

2

1

Polygonaceae
Eriogonum
Other

27
1

14

12

1

1

17

8

2

Portulacaceae

1

1

1

Prirnulaceae

1

1

1

Ranunculaceae

6

6

4

2

Rosaceae

1

1

1

Rubiaceae

2

2

1

1

Scrophulariaeeae

29

7

21

1

8

13

8

Selaginellaceae
Umbelliferae

1
12

3

1
9

1
5

5

2

Total

340

84

187

40

29

182

88

41

29

Percentage

24.7

55.0

11.8

8.5

53.5

25.9

12.1

8.5

Abbreviations: C, clay; S, shale, M, mud; G, gravel; NA, not available.

194

GREAT BASIN NATURALIST MEMOIRS

No. 2

idea that endemism is especially great at
low elevations. Despite this apparent abun-
dance of endemics at low elevations, only
slightly more than half of the entities are
known from below 1980 m (6500 feet) in
elevation; a quarter of them occur above
1980 m (6500 feet), 12 percent overlap, and
8.5 percent are unknown. However, the to-
tal land area above 1980 m in elevation
represents only about 30 percent of the re-
gion. When endemics are expressed in per
unit area figures there is 1 endemic per
1170 km2 for the higher elevations and 1
per 1318 km2 below that elevation. Thus,
the incidence of endemics generally is
greater at the higher elevations.

The area of the Colorado Plateau is
roughly 340,000 km2 (133,000 sq mi), which
places the density of endemics at about 1
per each 1000 km2. Almost half (55%) of
the endemics occur on sand and gravel, and
a quarter on clays, shales, or muds. Those
that grow on both types constitute only
11.8 percent, and data is missing for some
8.5 percent.

It is difficult to derive data on the pro-
portions of sands and gravels to other sub-
strate types. Certainly the sands and gravels
cover more of the land surface than do the
other substrate types. If one assumes that
only a quarter of the included region con-
sists of clays, shales, and muds, then the en-
demics on clays occur at about 1 per 1000
km2. Those on sands and gravels are present
at a density of only 1 per each 1350 km2.

Initial impressions by collectors that en-
demics are especially rich on the finely tex-
tured soils are supported by the fabricated
data. Certainly if the total area occupied by
clays is less than 25 percent, then the en-
demics per unit area might be expected to
be very great indeed. Surprising is the
greater density of endemics in the montane
sections of the plateau. Distribution of en-
demics is hardly at random, and any at-
tempt to provide averages or taxa per unit
area is likely to obscure the geographic
controls which result in the development
and placement of the taxa.

In a previous paper (Welsh, Atwood, and
Reveal 1975), the endangered, threatened,
extinct, endemic, and rare or restricted vas-
cular plants of Utah are plotted by geo-
graphic subdivision of the state. The large
part of those plants represent endemic
plants of Utah. Instructive from the plotting
is the apparent unequal distribution of those
plants in the state. Main centers of distribu-
tion include the high plateaus of south cen-
tral Utah, the Canyonlands section, and the
Uinta Basin. A trend is present, however,
which indicates that the greater numbers
occur near the southern end of Utah, with
progressively fewer taxa northward.

Since endemic plants are likely to be
considered as endangered or threatened, and
since most of the lands within the Colorado
Plateau are federally controlled, then some
management of these specialties is in-
dicated. Perhaps it is possible to establish
some predictive guidelines with regard to
location of endemics (i.e., where they might
be expected to occur). Outcrops of shale,
mudstone, and siltstone at all elevations
should be considered as suspect sources for
plant endemics. Any peculiar substrate, such
as the lacustrine limestones at higher eleva-
tions in the western margin of the plateau,
should be considered as important localities
for endemic taxa. Glaciated localities can
generally be excluded from consideration in
management practices.

References

Atuood, N. D. 1972. A revision of the Phacelia
crenulata group (Hydrophyllaceae) for North
America. Great Basin Nat. 35: 127-190.

Barneby, R. C. 1964. Atlas of North American As-
tragalus. Mem. New York Bot. Card. 13: 1-1188.

Beaman, J. H. 1957. The systematica and evolution
of Townsendia (Compositae). Contr. Gray Herb.
183: 1-151.

Cronquist, A. J. 1947. Revision of the North Amer-
ican species of Erigeron, north of Mexico. Brit-
tonia 6: 121-302.

Harrington, H. D. 1954. Manual of the plants of
Colorado. Sage Books, Denver. 666 pp.

Higgins, L. C. 1968. New species of perennial
Cryptantha from Utah. Great Basin Nat. 28:
195-198.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

195

1971. A revision of Cryptantha subgenus

Oreoearya. Brigham Young Univ. Sci. Bull.,
Biol. Ser. 13(4): 1-63.

Kearney, T. H. and B. H. Peebles. 1951. Arizona
Flora. University of California Press, Berkeley.

McDougall, W. B. 1973. Seed plants of northern
Arizona. Museum of Northern Arizona, Flag-
staff.

Tidestrom, I. 1925. Flora of Utah and Nevada.

Contr. U. S. Natl. Herb. 25: 1-665.

Welsh, S. L. 1957. An ecological survey of the veg-
etation of the Dinosaur National Monument,
Utah. Unpublished master's thesis, Brigham
Young University, Provo, Utah.

Welsh, S. L., N. D. Atwood, and J. L.
Beveal. 1976. Endangered, threatened, ex-
tinct, endemic, and rare or restricted Utah vas-
cular plants. Great Basin Nat. 35: 327-376.

SOME FACTORS GOVERNING PLANT DISTRIBUTIONS IN
THE MOJAVE-INTERMOUNTAIN TRANSITION ZONE

Susan E. Meyer'

Abstract.— The existence of a floristic transition zone can be inferred by the fact that a high proportion of in-
digenous plant species reach a distributional limit within the area. The vascular flora of Washington County,
Utah, exhibits this character to a marked degree with 53.6 percent of the native flora reaching a distributional
limit within the county. By looking at geographic distributions, ecological preferences, and range termination in-
formation for the component species, first approximations are made as to the probable factors mediating plant
distributions within the county, particularly of the range-terminating species. The high proportion of range-termi-
nating species in the flora may be accounted for mainly by limiting factors associated with abrupt shifts along
two environmental gradients. The first factor is climatic and is mediated largely by altitude; change occurs pri-
marily along a north-south gradient. The second is both climatic and edaphic and is mediated by factors other
than altitude per se; it is oriented in an essentially east-west direction. Species with narrower tolerances are
shown to be more sensitive to these environmental shifts. Some of the species distributions are better explained
by a model involving the effects of interactions between habitat mosaic and genetic homogeneity of given popu-
lations on relative migration rates in the transition area. These species may have the capacity to migrate farther,
but differences in migration rates give their distributional limits a quasi-stable aspect. These data suggest that
species cannot simply be divided into those which are environmentally limited in their present distributions and
those which are not. It seems more fruitful to regard these two conditions as extremes on a continuum which
can be expressed as migration rate.

A floristic transition zone is characterized
by the range termination of a high propor-
tion of the indigenous species. The vascular
flora of Washington County, Utah, exhibits
this character to a marked degree, with
53.6 percent of the 1,067 indigenous species
reaching a distributional limit within the
county. The objectives of this study are to
obtain detailed documentation of plant dis-
tributions within the transition area, to
identify environmental factors that might be
controlling range limits, and to elucidate
relationships between the distribution pat-
terns of groups of species and their ecologi-
cal amplitude.

Early observations on the floristic transi-
tion zone as it occurs in Washington Coun-
ty include those of Parry (1875), Merriam
(1893), and Jones (1910). The nature of the

zone as it occurs in Nye County, Nevada,
has been examined in some detail by Beat-
ley (1975), while Bradley (1967) has pub-
lished a phytogeographic survey of Clark
County, Nevada, which also lies along the
zone. Additional Washington County obser-
vations include those of Hardy (1947), Cot-
tarn et al. (1959), and Woodbury (1933).
Whittaker and Niering (1964) worked with
problems of a similar nature in southern
Arizona.

Description of the Study Area

Washington County is located in the ex-
treme southwestern corner of Utah. It is a
roughly rectangular area which spans an
east-west distance of about sixty miles and a
north-south distance of about forty miles. A

'Department of Biological Sciences, University of Nevada, Las Vegas. Present address: Department of Botany, Claremont Graduate School, Clare-
mont, California 91711.

197

198

GREAT BASIN NATURALIST MEMOIRS

No. 2

region of great environmental diversity, the
conntv contains extensive outcrops of carbo-
nate rocks, sandstones, shales, and both ex-
trusive and intrusive igneous rocks (Cook
1960). Elevations range from about 2,000
feet (600 m) near the common boundary
with Arizona and Nevada to over 10,000
feet (3100 m) at the top of the Pine Valley-
Mountains to the north. The dramatic na-
ture of the landscape is exemplified by Zion
National Park, most of which lies within the
county.

These diverse environments are not min-
gled randomly; they represent instead the
junction and interdigitation of three larger
land areas which are themselves relatively
homogeneous. Figure 1 shows the geo-
graphic position of the county relative to
these areas and to the boundaries which
separate them.

The first of these boundaries separates the
Basin and Range Province from the Colo-
rado Plateau Province. These two areas are
very different both lithologically and in the
relative proportion of high to low land. The
former is comprised of isolated mountain
ranges composed mainly of carbonate rocks
separated by broad, alluvium-filled valleys.
The latter is comprised of a contiguous
series of high plateaus composed mainly of
sandstone and shale, with lower elevations
restricted mainly to relatively narrow can-
yon floors. The climate of the Colorado
Plateau, as a whole, is more mesic than that
of the Basin and Range Province. The tran-
sition between these two physiographic-
provinces takes place within the county.

Superimposed on this boundary is another
boundary, which is designated on Figure 1
as the 4,000-foot contour line. Valley floors
to the north of this line consistently lie
above 4,000 feet and thus have a climate
which is comparatively cold. Valley floors
to the south always always lie well below
4,000 feet and thus have a climate which is
warmer. The line effectively divides the Ba-
sin and Range Province into two subregions,
the Great Basin region to the north and the
desert lowland region to the south. The fact

that part of Washington County lies south
of this line indicates an attenuation of
desert lowlands along the Virgin River
drainage into an area of generally higher
land.

Methods

The data presented here are summarized
from an annotated checklist of the vascular
plants of the study area (Meyer 1976). The
systems of vegetational and floristic classifi-
cation used here are more fully explained
there.

Each species was assigned to one of
thirty-eight floristic groups on the basis of
published distributional information. A flo-
ristic group is considered as an assemblage
of species which share a similar geographic-
distribution, without historical or ecological
implication per se.

The floristic groups are here classified
into two types of distribution patterns. An
areal distribution classification aggregates
the floristic groups into five classes on the
basis of their areal extent. The directional
distribution classification combines the flo-

Fig. 1. Location of Washington County, Utah, relative
to state and physiographic boundaries and to the 4000-
foot contour line.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

ristic groups into seven classes on the basis
of their main area of distribution relative to
the study area. Introduced plants are not in-
cluded in the tabulations.

All species which are reported to reach a
distributional limit within the county are
designated as range-terminating species.

The occurrence of each species in the
nine major community types within the
county was determined from field notes,
herbarium specimens, and monographic-
sources.

It is assumed here that the occurrence of
a species in a particular community type in-
dicates something about its ecological am-
plitude and more specifically about the cli-
matic tolerance of the species. That is, the
characteristic and the dominant plants of
each community type are considered to be
indicators of climate. It seems better to use
community type as a climate indicator
rather than elevation, because of the effect
of local modifying factors on the relation-
ship between elevation and climate. It
would be even better to monitor environ-
mental factors directly, but this was logisti-
cally impossible.

In order to measure the degree of habitat
specialization exhibited by various segments
of the flora, a fidelity index is used:

total species in segment

total community occurrences
in segment

FI ='

total species

total community occurrences

The values in this equation are derived in
the following way. Each species in the flora
occurs in one to nine possible community
types. The value of the denominator equals
the total species number (1067) divided by
the total number of community occurrences
(2182), or 0.489, indicating that the average
species in the flora occurs in approximately
two community types (0.489 w xk).

The value for the numerator is derived in
similar fashion, using total species number
divided by total community occurrence

number for any subcategory or segment of
the total flora. For example, the value of
the numerator for the endemic class is 0.606
(83 species in segment divided by 137 total
community occurrences).

The value of FI is defined as unity for
the total flora, since in that case numerator
and denominator have the same value. The
value of FI for the endemic class would be
1.24 (0.606 divided by 0.489). Since this val-
ue is greater than unity, the endemic class
shows a higher degree of habitat special-
ization than the flora as a whole. Values
smaller than unity values would indicate the
converse.

Given a value of 0.489 for the denomina-
tor, the value of FI can range from 0.23
(0.111 or one-ninth divided by 0.489) to
2.05 (unity divided by 0.489). These two ex-
tremes represent the respective cases of the
average species in the segment occurring in
all nine community types (one species/nine
community occurrences) and in only one
(one species/one community type).

Species which appear to be limited to a
single substrate are . designated as edaphi-
cally restricted species. Halophytes are not
included in the edaphically restricted cate-
gory because halomorphic soils develop on
a variety of substrates.

Factors Governing M

igration

Rates

Propagule dispersal is a stochastic pro-
cess. A plant disperses propagules in all di-
rections, regardless of whether the prop-
agule is likely to be dispersed to a site
favorable for its growth, or whether an ef-
fective dispersal constitutes a range exten-
sion for the species.

An effective dispersal is accomplished
when a propagule is dispersed to a site for
which it is preadapted in terms of genetic-
tolerance (Good 1930) and is therefore able
to grow and reproduce. When this site falls
outside the previous distributional area for
the species, the process is called migration.

Effective dispersal (and thus migration) is
a function of three sets of variables. These

200

GREAT BASIN NATURALIST MEMOIRS

No. 2

are distance, size of population supplying
propagules, and time. Their operation is
based simply on the laws of chance. The
first two variables may be related to the
third as an expression of rate. The rate at
which a species is able to migrate is a func-
tion of effective source-population size and
of average minimum dispersal distance (the
average minimum distance a propagule
must be dispersed in order to arrive at a
site for which it is preadapted). The larger
the effective source-population size and the
shorter the average minimum dispersal dis-
tance, the higher the migration rate.

Populations adjacent to the area of poten-
tial colonization are more likely to disperse
propagules into the area than those located
at greater distances. Therefore, the rate of
migration into an area is most influenced by
properties of populations immediately adja-
cent to that area.

The size of effective source-population is
conditioned by several factors: 1) absolute
size and density of the population, 2) repro-
ductive efficiency, and 3) dispersal effi-
ciency. These simply say that the chance of
propagules being dispersed to sites for
which they are preadapted is increased if
more propagules are produced and if prop-
agules can disperse more efficiently.

Minimum dispersal distance is the same
for all propagules produced by a population
only when they are all preadapted to iden-
tical sites. More often, minimum dispersal
distance is different for different propagules
because of their preadaptation to different
environments. The average minimum dis-
persal distance represents a mean value for
all the propagules produced. It is dependent
upon three sets of factors: 1) environmental
heterogeneity, 2) amount of genetic hetero-
geneity for tolerance characters present in
the population, and 3) degree of congruence
between environment and tolerance charac-
ters of the dispersed propagules.

If the environment is homogeneous and
favorable, the average minimum dispersal
distance will be very short, and migration
will be relatively rapid. If the environment

is homogeneous and unfavorable, the aver-
age minimum dispersal distance will be
very long and migration will be relatively
slow. Furthermore, these effects will in-
crease with the degree of genetic homoge-
neity of the population.

Thus a population with narrow tolerance
(low genetic heterogeneity) will be able to
migrate rapidly through an environment
which is homogeneous and favorable be-
cause of the very high probability that a
propagule will be dispersed to a site for
which it is preadapted. But its rate of mi-
gration through an environment which is
homogeneous and unfavorable will approach
zero, because of the very low probability
that a propagule will be dispersed to a site
for which it is preadapted.

A population with broad tolerances (high
genetic heterogeneity) would be able to mi-
grate less rapidly than a species with nar-
row tolerances through an environment
which is homogeneous and favorable. This is
because of the production of some prop-
agules preadapted to conditions other than
those present in the homogeneous environ-
ment. It would, however, be able to mi-
grate more rapidly through a homogeneous
and unfavorable environment. This is be-
cause it produces propagules preadapted to
a wider variety of environments, and thus
some propagules are likely to be preadapted
for survival in the potential migration area.

A heterogeneous environment is made up
of a mosaic of different environments. A
genetically heterogeneous population will
produce propagules preadapted to a wide
range of environments, and thus a high pro-
portion of the heterogeneous environment
will be "read" as favorable by the average
propagule it produces. This means that the
average minimum dispersal distance will be
relatively small and thus that migration will
be relatively rapid.

A genetically homogeneous population
will produce propagules preadapted to a
narrow range of environments; hence a low
proportion of the heterogeneous environ-
ment will be read as favorable. Average

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

201

minimum dispersal distance will be relative-
ly large, and migration will be relatively
slow.

This model is attractive in its simplicity
but is lacking in several respects. Most im-
portantly, it does not take into account the
fact that both the genetic constitution of
the population and the environment may
themselves be changing through time. In ef-
fect, it is an instantaneous rate expression; it
simply describes the options at a given
point in time. In nature the situation is
complicated by the fact that several of the
variables in the rate expression may them-
selves be changing.

A change in the environment might
change the configuration of favorable and
unfavorable sites for a population and thus
alter its migration rate. In some cases sites
already occupied might become unfavor-
able, causing a diminution in range. Such
secular climatic changes are well docu-
mented in the general area under discussion
(Mehringer 1965, Mehringer and Martin
1965). It is important to consider that a
species may have attained its modern distri-
bution under environmental conditions very
different from those now prevailing.

The genetic constitution of a population
is subject to change from several sources.
Two of these, gene flow and selection, are
particularly important in this discussion.
The former would tend to have the general
effect of increasing heterogeneity. The lat-
ter could have one of three effects. In a
homogeneous and constant environment,
normalizing selection would tend to de-
crease variability. In a homogeneous but
progressively changing environment, direc-
tional selection would tend to produce an
adaptive shift, providing that the original
population was sufficiently variable. In a
spatially heterogeneous environment, dis-
ruptive selection could operate to produce
genetically differentiated subpopulations,
particularly if gene flow between sub-
populations were minimized. These sources
of genetic change are especially important
in an area such as the one studied, which

meets the criteria of Stebbins (1974) for an
area in which active speciation may be ex-
pected to be proceeding. It is also an area
in which much of the habitat is hybrid-
inducing in the sense of Anderson (1949).
Many otherwise allopatric species have dis-
tributions which overlap in this area; they
can often be observed to introgress or inter-
grade, depending on the viewpoint as to
their origins.

In spite of the shortcomings of the model
of migration rate presented above, it is
quite useful in interpreting Washington
County data.

Discussion of Data

For interpretive purposes, two sorts of in-
formation about each species are examined.
One involves the distribution pattern for the
species as a whole. The other involves eco-
logical factors which govern its distribution
within the transition area. When these two
sets of information are examined together, a
high correlation between geographic distri-
bution and particular ecological require-
ments becomes apparent.

Table 1 shows the percent contribution of
each of seven distributional classes to the
range-terminating group as compared to the
total flora. It can be seen that the classes do
not contribute equally to these two groups.
The Mojave, Colorado Plateau, and endemic
classes contribute far more than their share
to the range-terminating group, making up
almost 70 percent of this group but com-
prising only about 40 percent of the total
flora. These differences are made up par-
tially by the northern class, which is no-
tably underrepresented in the range-
terminating group.

Column three of Table 1 shows the per-
centage of species contained within each
distributional class which are range-termi-
nating. This value is almost 100 percent for
the endemic and Mojave classes, indicating
that a high proportion of the species may
have reached a point which necessitates mi-
gration through an environment that is

202

GREAT BASIN NATURALIST MEMOIRS

No. 2

homogeneous and unfavorable. This possi-
bility is reinforced by the fact that they
show high fidelity indices, indicating a high
degree of habitat specialization and thus a
low degree of genetic heterogeneity for cli-
matic tolerance characters.

The Colorado Plateau and Great Basin
classes also show a relatively high percent
of range-terminating species, indicating a
higher sensitivity to the environmental tran-
sition between the Colorado Plateau and
the Great Basin than might have been ex-
pected. In fact, these two classes account
for almost a third of all range-terminating
species, a contribution almost as great as
that of the Mojave class.

The southern class also shows a high pro-
portion of range-terminating species, but
the converse is true of the northern group.
These apparently do not read the environ-
ment to the south of the county line as ho-
mogeneously unfavorable. One explanation
for this is the fact that the northern class is
not as specialized (FI = < 1) and thus
should be able to migrate more effectively
through a heterogeneous or even homoge-
neously unfavorable environment. Another
possibility is that these species attained
their distributions at some time in the past
when environment was more homoge-
neously favorable.

The range-terminating members of the
northern class have a much higher fidelity
index than the class as a whole. These spe-
cies would presumably have trouble mi-

grating through the heterogeneous environ-
ment of the transition zone because, due to
their genetic homogeneity, they tend to
read more of it as unfavorable. In fact, the
range-terminating members of a class con-
sistently show a higher fidelity index than
the class as a whole, lending credence to
the concept that low genetic variability for
tolerance characters makes migration across
a transition zone more difficult.

Table 2 shows the percentage contribu-
tion to each community type and to the to-
tal flora by each of the seven directional
distribution classes. The community types
are arranged roughly in altitudinal se-
quence.

The nondirectional class is slightly to
markedly overrepresented in every commu-
nity type except the Hot Desert Shrub com-
munity. In general, the species have broad
tolerances and occur in a variety of commu-
nity types, a fact also evidenced by their
low fidelity index (0.79).

About 85 percent of the Mojave species
occur in the Hot Desert Shrub community,
making up almost half of the total flora.
Perhaps more significant is the fact that an
appreciable percent of the species in two of
the higher elevation communities, Foothill
Woodland (11 percent) and Mountain Brush
(6 percent) is comprised of Mojave species.
Bearing in mind that over 98 percent of the
Mojave species are range-terminating, it is
clear that some of these species (almost 20
percent of the total Mojave class) reach a

Table 1. Range termination and fidelity indices by directional class.

% of Total

% Range-

Fidelity Index

% Total

Range-

Terminating

Fidelity Index

Range-

Directional

Native

Terminating

Species

Total

Terminating

Class

Flora

Species

in Class

Class

Species

Endemic
Nondirectional
Mojave

Colorado Plateau
(ireat Basin
Northern
Southern
Anomalous
Total Flora

7.8
25.5
17.6
14.3

6.6
20.5

5.5

2.2
100.0

14.0

<1.0

32.3

22.2

8.4

13.5

8.6

1.0

100.0

96.4
<1.0

98.4
83.0
68.6
35.2
83.1

53.6

1.24
0.79

1.32
0.86
1.26
0.96
1.04

1.24

1.34
1.12
1.46
1.22
1.09

1.23

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

203

termination of range in spite of the fact
that their tolerance characters permit some
individuals to disperse successfully to higher
elevation community types.

Most of the Mojave species have their
main area of distribution in terms of num-
bers of individuals in the Hot Desert Shrub
community. Their occasional anomalous oc-
currence in higher elevation community
types indicates a broader climatic tolerance
than that of species sharply restricted to the
hot desert. Such species should be able to
migrate through th^ ecotonal area because
they can read at least some of it as favor-
able. Therefore, there seems to be no a pri-
ori reason for them to reach a limit of
range in Washington County. Several fac-
tors might operate, however, to make their
rate of further migration very slow.

Even though hot desert populations might
be large, propagules preadapted for high
elevation community types might be pro-
duced at very low frequencies. In addition,
populations occurring in higher elevation
communities might have lowered reproduc-
tive efficiency, thus retarding the estab-
lishment of large populations which could
produce a higher frequency of preadapted
propagules. Both factors would have the ef-
fect of keeping the effective source-popu-
lation small.

The migrating species might be able to
occupy only a portion of the sites within
the higher-elevation vegetation types; the
proportion of favorable sites might decrease
with increasing distance from the transition
zone. This could be true if the species were

only marginally tolerant climatically, if it
had a narrow tolerance for some other envi-
ronmental factor, or if it faced a com-
petitive disadvantage in the presence of
some high elevation species. All of these
factors would have the effect of increasing
the average minimum dispersal distance by
increasing the proportion of unfavorable
sites. Finally, the species may actually have
migrated beyond the transition zone, but in
such small numbers as to have escaped de-
tection. In any case, the complex dynamics
of the situation make it impossible to divide
range-terminating species neatly into those
which are limited by environmental factors
and those which are not.

The community contributions of the re-
maining classes are easier to interpret. Since
Rocky Mountain species enter the county
via the Colorado Plateau, it is not surprising
that the Colorado Plateau class is well rep-
resented in montane communities, particu-
larly those of middle elevations. Many of
these montane species reach a western limit
of range, probably because of the much
lower proportion of favorable sites and the
much larger average minimum dispersal dis-
tances involved in migrating across the
Great Basin. The class as a whole is under-
represented in Hot Desert communities, and
most species which do occur there are not
range-terminating. The important role
played by edaphic factors in this class will
be discussed later.

Very few of the Great Basin species enter
the county at all; many reach southern and
eastern range limits a few miles northwest

Table 2. Percent composition of plant communities by directional class.

% Total

%

%

%

%

%

%

%

%

%

Directional

Native

Hot

Cold

Desert

Foothill

Mountain Transition Mountain Mountain

Mountain

Class

Flora

Desert

Desert

Riparian

Woodland

Brush

Forest

Forest

Riparian

Meadow

Endemic

7.8

6.6

7.0

2.9

10.1

8.2

7.8

4.7

2.7

4.6

Nondirectional

25.5

16.2

34.9

39.4

31.2

33.7

37.0

33.2

40.5

31.8

Mojave

17.6

46.0

9.5

17.3

10.7

5.8

1.0

1.0

1.4

0

Colorado Plateau

14.3

9.2

14.0

9.0

15.8

17.3

21.9

15.8

12.2

12.1

Great Basin

6.6

6.6

7.0

5.1

6.4

5.8

2.7

5.3

3.2

1.0

Northern

20.5

4.0

18.1

14.4

15.4

23.1

25.1

37.9

36.9

50.5

Southern

5.5

9.0

7.0

9.4

6.7

3.4

2.7

1.0

1.4

0

Anomalous

2.2

2.3

2.5

2.5

3.7

2.9

1.8

2.1

1.8

0

204

GREAT BASIN NATURALIST MEMOIRS

No. 2

of the county line. Those which do enter
the county show a high degree of habitat
specialization, but, oddly enough, this spe-
cialization is not to any particular commu-
nity type. Classic Great Basin salt desert
does not occur in Washington County,
which may account for the dis-
proportionately high representation of Great
Basin species in other community types.
Also, the total number of species is so small
that a difference of a few species has a rela-
tively large effect on the percent values.

As would be expected, species of the
northern class are abundantly well repre-
sented in high montane communities, rea-
sonably well represented in foothill commu-
nities, and poorly represented in desert
communities.

Values for the southern class show the
opposite relationship. They are poorly rep-
resented in montane communities and well
represented in desert communities.

Members of the endemic class show a
high degree of habitat specialization (FI =
1.24), and are overrepresented in relatively
xeric communities at low and middle eleva-
tions. They are noticeably underrepresented
in riparian communities. This may be be-
cause xeric environments tend to be more
heterogeneous than mesic habitats in re-
spect to variables other than moisture, and
heterogeneous environments tend to restrict
migration of specialized plant species, thus
keeping them endemic.

Possible relationships between the areal
extent of a distribution type and the ecolog-
ical amplitude of the component species
will now be explored. Table 3 gives fidelity
index values and percent community com-
position for each of the five areal distribu-
tion classes. These show that a distribution
type which covers a large area is more
likely to include species with broad toler-
ances than a distribution type which covers
a small area. The endemic and restricted
classes show higher habitat specialization
than the moderate class, which in turn
shows a higher value than the wide class.
This makes sense in light of the fact that
species with broad tolerances should not
only have more favorable sites available to
them, but they should also be able to mi-
grate more rapidly to occupy these sites.

An exception to this trend is exhibited by
the continental + class, which contains a
high proportion of transoceanic species. The
community composition values for species
of this class show that they are heavily con-
centrated in communities which provide
aquatic habitat, and in fact many of these
species are restricted to aquatic environ-
ments. They have probably been able to
achieve wide distributions in spite of being
habitat specialists by virtue of superior dis-
persal efficiency. It is these species which
bring the index value for the Continental +
group to a value which seems anomalously
high.

Table 3. Fidelity indices, percent edaphically restricted species, and percent composition of plant commu-
nities by areal class.

Endemic

Restricted

Moderate

Wide

Continental +

% Total Native Flora

7.8

27.5

16.8

33.0

12.8

Fidelity Index

1.24

1.25

1.00

0.83

1.01

% Edaphically Restricted

Species

36.1

39.8

15.7

7.2

0

% Hot Desert Shrub

6.6

44.5

23.4

21.4

1.7

% Cold Desert Shrub

7.0

21.3

17.8

43.5

7.9

% Desert Riparian

2.9

17.0

17.3

37.5

22.7

% Foothill Woodland

10.1

20.1

21.5

40.3

4.4

% Mountain Brush

8.2

14.9

21.2

46.6

6.3

% Transition Forest

7.8

18.3

15.1

46.1

11.0

% Mountain Forest

4.7

19.5

8.9

45.3

19.5

% Mountain Riparian

2.7

13.1

5.9

44.6

32.0

% Mountain Meadow

4.6

12.1

11.2

49.5

22.4

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

205

Plants of wide but subcontinental distri-
butions do not show this specialization, but
instead are well represented in all but the
Hot Desert Shrub community type. Plants
of moderate distribution tend to be better
represented in the lower-elevation commu-
nities. The trend continues in the group of
restricted distribution, which is relatively
underrepresented in all but the Hot Desert
Shrub community type. This trend seems to
indicate that groups of wider distribution
show higher percentages in mesic types,
while groups of narrower distribution show
higher percentages in xeric types. This
makes sense in view of the relative propor-
tion of mesic to xeric environments in the
world, on the continent, in the West, and in
the intermountain-southwestern area, re-
spectively. The values for the endemic class
have been discussed above.

The importance of edaphic factors in the
transition area will be examined next. Toler-
ance for edaphic factors, like tolerance for
other environmental factors, is a matter of
degree. This discussion takes into account
only those species which have a very nar-
row edaphic requirement. Even though
these make up only about 8 percent of the
total flora, the implication that edaphic fac-
tors play a minor role in the transition area
is not warranted. Many more species show
some degree of edaphic specialization, and
as noted above, distributional limits are not
all or nothing propositions based on pres-

Table 4. Percentage of all native species and of all
edaphically restricted species occurring in each plant
community.

% Total

% Edaphically

Native

Restricted

Plant Community

Species

Species

Hot Desert Shrub

32.4

63.9

Cold Desert Shrub

29.5

37.3

Desert Riparian

26.0

13.3

Foothill Woodland

27.9

32.5

Mountain Brush

19.5

15.7

Transition Forest

20.5

12.1

Mountain Forest

17.8

2.4

Mountain Riparian

20.8

2.4

Mountain Meadow

10.0

1.2

ence or absence of adaptive traits, but in-
stead are also a matter of degree. The pat-
terns for the sharply restricted species seem
to indicate that edaphics may be more im-
portant in the transition area than might be
supposed.

Table 4 shows the percentage of all spe-
cies versus the percentage of edaphically re-
stricted species occurring in each commu-
nity type. Though both these values tend to
decrease with an increase in altitude,
edaphically restricted species drop much
more sharply. In addition, this group is
poorly represented in the Desert Riparian
community. Clearly edaphically restricted
species are much better represented in xeric
than in mesic community types. This may
indicate that the environment is more heter-
ogeneous edaphically in dry than in wetter
environments. Mesic soils developed on dif-
ferent substrates tend to be more modified
and more alike than xeric soils developed
on those different substrates.

A consequence of such environmental
heterogeneity is that edaphically restricted
species tend to belong to more restricted
areal distribution types. This is because spe-
cies with narrow edaphic tolerances migrate
slowly through an edaphically hetero-
geneous area. It may also be because the
substrate itself is of restricted geographic
occurrence. The first and third rows of
Table 3 demonstrate this relationship.

An edaphically restricted species will be
able to migrate relatively rapidly in spite of
low genetic heterogeneity as long as the en-
vironment is homogeneous and favorable in
terms of substrate. When such a species
reaches a range limit, it is probably because
it has reached a point where the substrate
environment suddenly becomes very hetero-
geneous or homogeneous and unfavorable.
The favorable substrate may occur at wide-
ly spaced sites of small areal extent instead
of closely spaced sites of large areal extent,
or it may be completely absent from the
potential area. In any case, the average
minimum dispersal distance becomes prohi-

206

GREAT BASIN NATURALIST MEMOIRS

No. 2

bitively large, and migration slows to near
zero.

Table 5 shows the substrate preferences
of edaphically restricted species belonging
to each of the seven directional distribution
classes. The northern class contributes no
edaphicallv restricted species. This is not
surprising in view of the high preference of
this group for mesic environments, which
do not promote edaphic specialization.

The nondirectional class contributes only
one limestone-restricted and one sandstone-
restricted species, and as a group is marked-
ly underrepresented. The southern class and
the Great Basin class are also slightly under-
represented and contribute only a handful
of the edaphically restricted species. This
means that most of these species must be
contributed by the remaining three classes.

Over a quarter of the species are contrib-
uted by the Mojave class, but these are dis-
tributed very unevenly among the sub-
strates. Mojave species constitute almost
two-thirds of the limestone-restricted class,
about one-fourth of the sand-restricted class,
and less than 10 percent to the remaining
classes.

The Colorado Plateau class accounts for
almost a third of the total. But, in contrast
to the Mojave class, none of these are lime-
stone-restricted. Instead they make up al-
most two-fifths of the sandstone-restricted
species, one-fifth of the clay-restricted spe-
cies, and about two-thirds of the sand-
restricted species.

Table 5. Edaphically restricted species by substrate a

The endemic class accounts for the re-
maining third of the total. This class con-
tributes the only volcanic-restricted species,
about three-quarters of the clay-restricted
species, and two-thirds of the sandstone-
restricted species. None of the endemics are
sand restricted.

Over 90 percent of the edaphically re-
stricted species are range-terminating. The
role of substrate specialization as a factor in
range-termination seems clear. Limestone-
restricted species appear to migrate rapidly
across the southern Basin and Range Prov-
ince, but their rate of migration nears zero
when they encounter the sandstones and
shales of the Colorado Plateau. Conversely,
sandstone- and clay-restricted species seem
to migrate rapidly across the Colorado
Plateau, but their rate of migration nears
zero when they encounter the high-
carbonate soils of the Basin and Range Pro-
vince. Sand is a substrate which is wide-
spread, especially in the Colorado Plateau
Province; thus sand-restricted species do not
tend to be endemics. But there is sand in
the Mojave Desert; thus some sand-restrict-
ed species enter the county via the Mojave.
These may be prevented from migrating
further north by other than edaphic factors.

Summary

The data show that the high proportion
of range-terminating species in the county
flora is correlated with the abrupt shifts
nd directional class.

% Total

% Edaphicallv

% Volcanic-

% Sand-

% Sandstone-

% Limestone-

% Clay-

Directional

Native

Restricted

Restricted

Restricted

Restricted

Restricted

Restricted

Class

Flora

Group

Species

Species

Species

Species

Species

Endemic

Nondirectional

Mojave

Colorado Plateau

Great Basin

Northern

Southern

Anamalous

Total Species

% Total Edaphically

Restricted Species

7.8
25.5
17.6
14.3

6.6
20.5

5.5

2.2
1()0.0

36.1

2.4

24.0

28.9

3.6

0

3.6

1.2

100.0

100.0

0

0

0

0

0

0

0
100.0

1.2

0

0
25.0
54.2

8.3

0

8.3

4.2
100.0

28.9

62.5
0
0

37.5
0
0
0
0
100.0

19.3

15.8
10.6
63.2

0

5.2

0

5.2

0
100.0

22.9

0

8.7
21.7
0
0
0
0
100.0

27.7

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

207

along environmental gradients oriented in
north-south and in east-west directions. The
north-south gradient is primarily climatic,
while the east-west gradient is both climatic
and edaphic.

Species with narrower tolerances are
shown to be more sensitive to these abrupt
environmental shifts. It is also shown that
some of the range-terminating species have
the capacity to migrate beyond the transi-
tion zone under present conditions, but that
differences in relative migration rates give
their distributional limits within the county
a quasi-stable aspect.

Conclusion

Just about anything is statistically possible
in biogeography. This makes it dangerous to
conclude that any species has a migration
rate of zero. It is possible for a propagule
to be dispersed to a tiny enclave of favor-
able environment in the midst of a vast
area of homogeneous and unfavorable envi-
ronment. And it is possible that the very
rare propagule which has tolerance charac-
ters near an extreme for the species will be
dispersed to a site for which it is pre-
adapted. These events may be very unlikely,
but they are possible. They may result in
some of the seemingly aberrant "dots" often
seen on species distribution maps.

Instead of considering only two possi-
bilities, those species which are environ-
ment-limited in their current distribution
pattern and those which are not, it seems
far better to consider these two possibilities
as extremes on a" continuum. This involves
recognition of the interplay among a com-
plex series of factors which interact to pro-
duce the probability that a species will be
able to migrate at a given rate under a giv-
en set of conditions. These concepts may
prove difficult to tie down, but it is hoped
the ideas presented here will facilitate the
process— if they only begin to sharpen the
focus on this complex and interesting prob-
lem.

Literature Cited

Anderson, E. 1949. Introgressive hybridization.
Hafner Pub. Co., New York.

Beatley, J. C. 1975. Climates and vegetation pat-
tern across the Mojave/Great Basin Desert tran-
sition of southern Nevada. Amer. Midi. Natural-
ist 93: 53-70.

Bradley, W. G. 1967. A geographical analysis <>(
the flora of Clark County, Nevada. Ariz. Acad.
Sci. 4: 151-162.

Cook, E. F. 1960. Geological atlas of Utah: Wash-
ington County. Utah Geol. and Mineralog. Surv.
Bull. 70: 1-124.

Cottam, W. P., J. M. Tucker, and B. Drobnick.
1959. Some clues to Great Basin post-pluvial
climates provided by oak distributions. Ecology
40: 361-377.

Good, B. 1930. A theory of plant geography. New
Phytologist 30(11):150-171.

Hardy, B. 1947. The Beaverdam Mountains as a
barrier in plant and animal distributions: addi-
tional information. Proc. Utah Acad. Sci., Arts,
and Lett. 24: 138 (Abstract).

Jones, M. E. 1910. Origin and distribution of the
flora of the Great Plateau. Contr. W. Bot. 13:
46-68.

Martin, P. S., and P. J. Mehringer. 1965.
Pleistocene pollen analysis and the biogeo-
graphy of the Southwest, pp. 433-451. In: H. E.
Wright, Jr. and D. G. Frey (eds.), The Qua-
ternary of the United States. Princeton Univer-
sity Press, Princeton, N.J.

Mehringer, P. J. 1965. Late Pleistocene vegetation
in the Mojave Desert of southern Nevada. J.
Ariz. Sci. 3: 172-188.

Merriam, C. H. 1893. The Death Valley expedition:
a biological survey of parts of California, Ne-
vada, Arizona, and Utah. N. Amer. Fauna 7: 1-
193.

Meyer, S. E. 1976. Annotated checklist of the vas-
cular plants of Washington County, Utah. Un-
published master's thesis, University of Nevada,
Las Vegas.

Parry, C. C. 1875. Botanical observations in south-
ern Utah in 1874. Amer. Naturalist 9: 14-21,
139-146, 199-205, 267-273.

Stebbins, G. L. 1974. Flowering plants: evolution
above the species level. Belknap Press, Cam-
bridge, Mass.

Whittaker, B. H., and W. A. Niering. 1964.
Vegetation of the Santa Catalina Montains, Ari-
zona. I. Ecological classification and distribution
of species. J. Ariz. Acad. Sci. 3(1): 9-34.

Woodbury, A. M. 1933. Biotic relationships of Zion
Canyon, Utah, with special reference to succes-
sion. Ecological Monogr. 3: 146-245.

THE THEORY OF INSULAR BIOGEOGRAPHY AND
THE DISTRIBUTION OF BOREAL BIRDS AND MAMMALS

James H. Brown1

Abstract.— The present paper compares the distribution of boreal birds and mammals among the isolated
mountain ranges of the Great Basin and relates those patterns to the developing theory of insular biogeography.
The results indicate that the distribution of permanent resident bird species represents an approximate equilib-
rium between contemporary rates of colonization and extinction. A shallow slope of the species-area curve (Z =
0.165), no significant reduction in numbers of species as a function of insular isolation (distance to nearest conti-
nent), and a strong dependence of species diversity on habitat diversity all suggest that immigration rates of bo-
real birds are sufficiently high to maintain populations on almost all islands where there are appropriate habitats.
In contrast, the insular faunas of boreal mammals represent relictual populations that receive no significant con-
temporary immigration. The insular mammal faunas have been derived by extinction from a set of species that
colonized the islands when habitat bridges connected them to the continents in the late Pleistocene. A relatively
steep species-area curve (Z = 0.326), no effect of isolation on species diversity, and the absence of appropriate
species from large areas of apparently suitable habitat all support this conclusion. Measures of habitat diversity
that are closely correlated with bird species diversity do not account for much of the variation in number of
mammal species among islands. Insular area is the single variable that accounts for most of the variability in
both bird and mammal species diversity; this supports the approach of using standard parameters such as area in
comparative empirical analyses and general biogeographic theory. The results of this study suggest that extremes
of vagility among taxa and a recent history of paleoclimatic and geological changes make it unlikely that equi-
librial distributions, of the sort MacArthur and Wilson (1967) propose for the biotas of oceanic islands, are char-
acteristic of the insular distributions of terrestial and freshwater vertebrates of western North America.

Biogeography is an old science that has
made recent advances by assimilating new
concepts and data. Within the last two dec-
ades information on continental drift, pa-
leoclimatology, and ancient sea levels has
revolutionized our understanding of histori-
cal events and their effects on plant and an-
imal distribution. During the same period,
ecologists and evolutionary biologists have
learned much about the processes of popu-
lation growth, dispursal, extincton, speci-
ation, and interspecific interactions which
are the mechanisms that determine distribu-
tion. Biogeography appears to be entering
an exciting new period in which the volu-
minous data acquired by systematists and
descriptive biogeographers can be inter-
preted in terms of recently understood his-
torical events and ecological processes to

draw quantitative relationships and derive
general principles.

With the publication of their equilibrium
model of insular distribution, MacArthur
and Wilson (1963, 1967) contributed not
only a new theory but also a new approach
to biogeography. Prior to their work, most
biogeographic research had consisted of de-
scribing the distributions of particular taxa
and producing ad hoc, historical explana-
tions. MacArthur and Wilson advocated a
quantitative approach designed to build and
test general models based on ecological pro-
cesses. The specific model that they pro-
posed suggests that the number of species
inhabiting an island represents an equilib-
rium between opposing rates of extinction
and colonization, and that these processes
are functions of the size of an island and its

'Department of Ecology and Evolutionary Biology, University of Arizona, Tucson, Ariz

209

210

GREAT BASIN NATURALIST MEMOIRS

No. 2

distance from a source of colonists, respec-
tively.

MacArthur and Wilsons's choice of is-
lands for their revolutionary approach to
biogeography was not a fortuitous one. Ever
since the pioneering work of Darwin and
Wallace, islands have played a preeminent
role in the development of the science. It is
difficult to do experiments in biogeography,
but islands and insular habitats represent
natural experiments; they are small, replica-
ted systems among which species and envi-
ronmental parameters are distributed in dif-
ferent combinations. Often particular taxa
are distributed among insular habitats in
patterns that imply the operation of general
mechanisms of dispersal, extinction, and in-
terspecific interaction. The last decade has
seen several attempts to test the model of
MacArthur and Wilson, using a variety of
organisms that inhabit both true islands and
analogous isolated habitats (e.g., Barbour
and Brown 1974, Brown 1971, Brown and
Kodric-Brown in review, Culver 1970, Cul-
ver et al. 1973, Diamond 1969, 1970a,
1971, Johnson 1975, Harper et al. this sym-
posium, Schoener 1974, Seifert 1975, Sim-
berloff 1974, Simberloff and Wilson 1970,
Simpson 1974, Terborgh 1973, Vuilleumier
1970, 1973). Not all of these studies have
supported the model, but the exceptions
have contributed importantly to our under-
standing of the patterns of insular distribu-
tion and the historical events and ecological
processes that produce them.

Analysis of the distribution of vertebrates
among isolated habitats in the Inter-
mountain Begion of western North America
has contributed significantly to the devel-
opment and testing of biogeographic theory.
The dedicated field work of several gener-
ations of systematists has documented in de-
tail the species distributions of most verte-
brate groups (e.g., for fishes Miller 1948,
Hubbs and Miller 1948, Hubbs et al. 1974,
Smith, this symposium; for birds Behle
1943, 1955, 1958, this symposium, Grinnell
and Miller 1944, Johnson 1965, 1970, 1973,
1974, 1975, Linsdale 1936, Miller 1935,

1946, Miller and Bussell 1956, Van Bos-
sem 1936; for mammals Durrant 1952, Dur-
rant et al. 1955, Grinnell 1933, Hall 1946).
The paleoclimatic history of the region also
is becoming increasingly well understood
(e.g., Hubbs and Miller 1948, Martin and
Mehringer 1965, Smith this symposium,
Wells and Berger 1967, Wells and Jorgen-
sen 1964). This excellent data base has been
used in quantitative analyses of the insular
distributions of lacustrine fishes (Barbour
and Brown 1974) and montane birds (John-
son 1975) and mammals (Brown 1971).
Much more work remains to be done. Some
of the most interesting contributions can be
expected when additional kinds of organ-
isms and insular habitats are studied, so that
it is possible to make comparisons among
taxa which differ in ecological requirements
and dispersal abilities, and among habitats
that differ in environmental parameters and
history of isolation.

The present paper discusses the distribu-
tion of boreal birds and mammals among
isolated mountain ranges in the Great Basin
in relation to the theory of insular bio-
geography. It attempts to relate distribu-
tional patterns to mechanisms of dispersal
and extinction and to differentiate relictual
patterns that are the legacy of historical
events from equilibrial ones that can be at-
tributed to contemporary ecological pro-
cesses. The paper tries to develop a general
conceptual basis for analyzing and predict-
ing insular distributions. It first discusses the
current state of insular biogeographic theo-
ry and then utilizes empirical data on the
distribution of boreal birds and mammals to
test the theory and search for general mech-

The Theory of Insular Biogeography

Colonization, extinction, and speciation
are the primary processes that determine
the composition of insular biotas. An island
can be defined as a patch of suitable habitat
surrounded by unfavorable environment that
limits the dispersal of individuals. New spe-

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

211

cies colonize an island either by dispersing
across the habitat barriers or by immigrat-
ing sometime when the barriers were tem-
porarily absent. Speciation entirely within
an island potentially is another source of
new species, but there is no evidence that
this process has contributed significantly to
the diversity of boreal birds and mammals
in the Great Basin region. Extinction, which
may be caused by a variety of factors that
reduce population size, eliminates species
and reduces insular diversity. The composi-
tion of an insular biota is determined by the
interaction of these origination and extinc-
tion processes.

MacArthur and Wilson (1963, 1967) de-
veloped a simple, general model of insular
biogeography that represents the number of
species inhabiting an island as an equilib-
rium between contemporary rates of coloni-
zation and extinction (Fig. 1). Their model
makes the colonization rate a decreasing
function of distance to a source of dis-
persing species (usually the nearest conti-
nent) and extinction rate a decreasing func-

tion of island size. The model predicts: first,
that relatively constant insular species di-
versity is maintained by the continual turn-
over (extinction and colonization) of individ-
ual species; second, that the equilibrium
number of species is positively correlated
with island size and negatively correlated
with distance to the source of potential col-
onists; and third, that the equilibrium turn-
over rate is inversely related to both island
size and distance to a source of species. Al-
though the MacArthur-Wilson model was
designed specifically to account for the di-
versity of organisms on oceanic islands, it
has proven a useful heuristic device for ana-
lyzing many kinds of insular distributions
because it deals with general processes and
makes robust, testable predictions.

As the MacArthur-Wilson model has been
tested using a variety of taxa distributed
among both oceanic islands and several
kinds of insular habitats (see references
cited above), it has become increasingly
clear that in its present form it is in-
adequate to account for many empirical ex-

\ COLONIZATION

1

X

V

7

¥

VEXTINCTIOINi .>

A #^^

— \ """

NUMBER OF SPECIES

°SF °LN

NUMBER OF SPECIES

Fig. 1. Two models of equilibrium insular biogeography. Left, the MacArthur-Wilson model, which portrays
extinction and colonization rates as functions of island size and isolation respectively. Right, a modification of the
MacArthur-Wilson model by Brown and Kodric-Brown (1977), which incorporates the effect of insular isolation
on extinction rate. In both models intersections of the curves can be extrapolated to the abscissa and ordinate to
give equilibrial numbers of species (S) and turnover rates (X), respectively. Note that the two models predict the
same relative order of numbers of species but different orders of turnover rates with respect to island size and
isolation.

212

GREAT BASIN NATURALIST MEMOIRS

No. 2

amples. Two major problems have arisen in
attempts to use the model to account for
empirical patterns of animal and plant dis-
tribution. First, in some insular habitats, his-
torical episodes of immigration, speciation,
and extinction have produced numbers of
species that are significantly greater or less
than the equilibrial number expected on the
basis of contemporary rates of colonization
and extinction (Barbour and Brown 1974,
Brown 1971, Culver et al. 1973, Diamond
1970). Such historically determined distribu-
tions should be particularly common in or-
ganisms that are poor dispersers (such as the
small, nonflying vertebrates) and in geo-
graphic areas (such as western North Ameri-
ca) where paleoclimatic and geological
changes have drastically altered the barriers
that currently isolate insular habitats. Sec-
ond, it has proven difficult to test the Mac-
Arthur-Wilson model's critical predictions
about insular turnover. Although it is rela-
tively easy to census accurately the biota of
an island, it is much more difficult to assess
the natural turnover rate. As a result, most
of the purported measures of insular turn-
overs have been criticized (see Diamond
1969, 1971, Terborgh 1973, Lynch and
Johnson 1974, Simberloff 1974), and the
critical predictions of how turnover rate
varies with island size and isolation remain
untested. Recently, Brown and Kodric-
Brown (1977) obtained empirical evidence
that insular extinction rates may be strongly
dependent on the distance of an island from
a source of colonists, rather than on island
size alone as MacArthur and Wilson sug-
gest. Areas that are sources of colonizing
species often also may be sources of immi-
grant individuals of species already present
on the island; the arrival of these immi-
grants may refuce the probability that in-
sular populationd go extinct. We predicted
that, for many islands, species turnover rates
are lower for islands near sources of dis-
persing species than for more isolated ones.
Although this is the opposite of the pattern
predicted by the MacArthur-Wilson model,
the model can easily be modified to in-

corporate this influence of immigration on
extinction rate (Fig. 1).

Neither of these problems detracts from
the utility of MacArthur and Wilson's mod-
el as a heuristic device or the value of their
approach. Their work did much to stimu-
late biogeographers to develop precise hy-
potheses and test them with appropriately
analyzed quantitative data. Gradually a con-
ceptual understanding of insular bio-
geography is emerging that may not be as
simple and elegant as the MacArthur-
Wilson model, but it is hoped it will be
more realistic. In the following sections, I
hope to use the distribution of boreal birds
and mammals among isolated mountain
ranges in the Great Basin to illustrate this
theoretical approach and some of the result-
ing concepts.

Montane Islands

The mountain ranges of the Great Basin
are islands of coniferous forest and associ-
ated mesic habitats in a sea of desert. The
Great Basin is a vast interior drainage that
lies between two montane continents, the
central mountains of Utah (a part of the
Rocky Mountains) on the east and the
Sierra Nevada on the west (Fig. 2). Most of
this area consists of broad arid valleys,
which lie at an elevation of approximately
5000 feet and are sparsely covered with a
vegetation dominated by low woody shrubs
of the genera Artemisia, Chrijsothamntis,
and Atriplex. Between the valleys are a
series of mountain ranges oriented in a
north-south direction. Many of these rise to
over 10,000 feet; on their lower slopes they
are covered with juniper-pinyon woodland
and at higher elevations there are forests of
mixed conifers, stands of aspens, and some-
times wet meadows and permanent streams.

The present analysis is based on 19 is-
lands for which the boreal mammal fauna is
adequately known; lists of boreal bird spe-
cies are available for 13 of these. Islands
were defined by operational criteria applied
to topographic maps (U.S. Geological Sur-

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

213

vey maps of the states: scale 1:500,000). A
montane island was considered to be a
mountain range that contains at least one
peak higher than 9800 feet and is separated
from other highland areas by a valley at
least 5 miles across below an elevation of
7500 feet. This altitude represents the ap-
proximate lower boundary of juniper-pinyon
woodland. For each island, area above 7500
feet elevation, distance to nearest continent
(Sierra Nevada or central mountains of
Utah), and elevation of highest peak were
determined from topographic maps (Table
1). For 13 islands ornithologists (Johnson
1975, Behle, this symposium) have quan-
tified the diversity of habitats available to
boreal birds. This habitat diversity score in-
corporates the number of coniferous tree

species and the presence of riparian wood-
land, wet meadow, and aquatic habitats (see
Johnson 1975 for details). Since West (this
symposium) has shown that juniper-pinyon
woodland is absent or poorly developed in
the northern Great Basin, the islands se-
lected for this analysis lie south of the
Humboldt River and Great Salt Lake to in-
sure that they have somewhat comparable
habitats.

Boreal Birds and Mammals

For the bird and mammal species that
are restricted to juniper-pinyon woodland
and more mesic habitats of higher eleva-
tions, the mountains of the Great Basin are
truly islands. Their boreal avian and mam-

Table 1. Data for the boreal habitats used in the present analysis.

Great Basin Montane "Islands"

Area above
7,500 feet
(sq. miles)

Highest
peak

(feet)

Nearest

continent

(miles)

Habitat
diversity
score'

1 «

2 u

& 1

.2
'3

u

a

-a

3

Small Boreal

mammal

species1

1. White-Inyo

738

14,242

10

11

8

11

2. Panamint

47

11,045

52

3

5

3

3. Desatoya

83

9,814

83

2

4

7

4. Toiyabe-Shoshone

684

11,788

110

7

6

13

5. Toquima-Monitor

1,178

11,949

114

-

-

10

6. Roberts Creek

52

10,133

216

-

-

4

7. Diamond

159

10,614

190

-

-

4

8. Ruby

364

11,387

173

9

6

12

9. Spring

125

11,918

125

7

6

6

10. Sheep

54

9,912

86

5

5

3

11. Grant-Quinn Canyon

150

11,298

138

9

5

5

12. White Pine

262

11,188

150

-

-

7

13. Schell Creek-Egan

1,020

11,883

114

-

-

8

14. Spruce-South Pequop

49

10,262

156

4

4

4

15. Snake

417

13,063

89

14

9

10

16. Deep Creek

223

12,101

104

11

7

8

17. Pilot

12

10,704

114

-

-

3

18. Stansbury

56

11,031

39

8

6

6

19. Oquirrh

82

10,626

19

9

6

6

Sierra Nevada Mainland

20. Carson

284

10,788

0

17

13

22

21. Yosemite

828

13,090

0

18

15

23

Rocky Mountain Mainland

22. Paunsaugunt-Aquarius

1,008

11,124

0

14

13

16

23. Uinta

1,536

13,498

0

15

14

21

'From Johnson (1975) and Behle (this symposium).
'See Appendix 1 for documentation.
'See Appendix 2 for documentation.

214

GREAT BASIN NATURALIST MEMOIRS

No. 2

20

2 -

BIRDS

S= 2.526 A(

r =0.701

20

50

100

200

500

1,000

20

U. 5

MAMMALS

S = l 188 A0326

r =0.846

20 50 100 200 500

AREA ABOVE 7,500 FEET ELEVATION (SO MILES)

1,000

Fig. 2. Map of the Great Basin region of western North America showing the location of the isolated mountain
ranges used in the present analysis. Numbers refer to individual montane islands and continental sample areas
listed in Table 1. Note that islands of varying size and isolation lie in the "sea" of desert habitat between the
central mountains of Utah on the east and the Sierra Nevada of California on the west.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

215

malian faunas are depauperate subsamples
of the Sierra Nevada and Rocky Mountain
faunas. All of the bird and mammal species
that inhabit the isolated peaks are broadly
distributed on one or (more frequently) both
of the continental ranges to the east and
west (see appendix and references cited
there), but a significant proportion of the
continental species are not present on any
particular island and some are absent from
all islands. The continental mountain ranges
clearly are the source of the insular boreal
faunas, and the distribution of species on
the islands is the result of their abilities to
colonize and avoid subsequent extinction.

The definition of boreal bird and mam-
mal species is somewhat arbitrary, and the
lists of knowledgable specialists often do not
agree precisely (e.g., compare the lists of
boreal bird species of Johnson 1975, and
Behle this symposium). I have followed
Johnson's rather conservative designation of
boreal bird species (for a list of these spe-
cies and their distribution see Appendix 1).
I have restricted my analysis to those spe-
cies that he terms (again conservatively)
"permanent resident Boreal species," be-
cause I wanted to include in my com-
parison with mammals only those species
that maintain sedentary populations
throughout the year. This eliminates from
the analysis the large number of bird spe-
cies that breed on the montane islands but
are likely to migrate or disperse long dis-
tances between successive breeding seasons.
The definition of boreal mammal species in-
cludes those that inhabit juniper-pinyon
woodland or habitats of higher elevation,
but not the desert habitats of the Great Ba-
sin. I have excluded from the analysis large
carnivores and ungulates because their dis-
tributions have been drastically altered by
human activity and their original ranges
and habitat requirements are poorly known.
Also I have ignored bats, because their dis-
tributions are incompletely documented and
they, like the migratory birds, probably are
not permanent residents of the islands. The
list of boreal mammal species (Appendix 2)

is similar to that in Brown (1971), but it
differs in some details because I have used
slightly different criteria for designating bo-
real species and I have included a few more
records of occurrence. The resulting lists of
boreal bird and mammals contain species of
generally similar body sizes and habitat re-
quirements, although there are approx-
imately 50 percent more mammal than bird
species on both the large islands and the
continents.

Species— Area Relationships

The number of species (S) inhabiting an
island usually is positively correlated with
insular area (A); this relationship takes the
form S = CAZ, where the values of the
constant (C) and slope (Z) depend on the
characteristics of the specific taxon and
group of islands under consideration (Mac-
Arthur and Wilson 1967, Preston 1962). The
slope of this relationship can indicate the
relative importance of extinction and origi-
nation processes in determining the diver-
sity of insular biotas. (Barbour and Brown
1974, Brown 1971, Mac-Arthur and Wilson
1967). Low slopes, Z<0.20, tend to charac-
terize samples of different areas within a
continent, and the Z-value varies with envi-
ronmental heterogeneity (Harner and Har-
per, 1977). Islands usually have higher Z-
values; when the insular biota represents an
approximate equilibrium between rates of
colonization and extinction, Z-values tend to
lie in the range 0.20-0.35. When there is no
contemporary colonization to oppose extinc-
tion, islands tend to have even higher
Z-values, often > 0.40. The reason for this
pattern is straightforward. On continents
small sample areas contain a significant pro-
portion of rare species. If these habitats
were isolated and no immigration were per-
mitted, rare species would go extinct rapid-
ly on small islands and much more slowly
on large ones, producing a much steeper
species-area curve. Islands in equilibrium,
where a significant rate of colonization op-
poses extinction, represent an intermediate

216

GREAT BASIN NATURALIST MEMOIRS

No. 2

situation; species that go extinct on small is-
lands tend to be replaced by colonists, but
not at a sufficient rate to produce diversity
comparable to that on continents. There
can be exceptions to this pattern. For ex-
ample, if only a small subsample of the con-
tinental biota is able to colonize a group of
islands, the Z-value may be lower than ex-
pected because large islands will not ac-
quire as many species as they can support
(Barbour and Brown 1974).

This conceptual framework can be used
to compare the species-area curves for the
boreal birds and mammals. In my earlier
analysis (Brown 1971), I argued that boreal
mammals reached all of the islands during
periods of climatic change in the late
Pleistocene; since then there have been ex-
tinctions but no colonizations. This con-

clusion can be tested using a somewhat dif-
ferent data set for mammals and comparing
the mammalian and avian distributions. If
birds are better dispersers and are currently
crossing the desert valleys to colonize the
isolated peaks, then they should have a sig-
nificantly lower Z-value than mammals.
This is the case. The number of species of
both birds and mammals is correlated signif-
icantly with area (Fig. 3, Tables 2 and 3).
The Z-value for birds, 0.165, is even lower
than that obtained for most insular biotas
that are presumed to be in equilibrium and
approximates values for continental samples.
The Z-value for mammals, 0.326, is less
than I obtained in my earlier analysis, Z =
0.428, and lies in the upper range of those
observed for the biotas of true islands (Mac-
Arthur and Wilson 1967). Thus species-area

Fig. 3. The relationship of insular area to the number of permanent resident boreal bird species (above) and
the number of small boreal mammal species (below). Note that the slope of the least squares regression line of
the species-area curve for birds is much less steep than that for mammals. The equations for the fitted regres-
sions and the correlation coefficients (r) are indicated.

1978

INTERMOUNTAIN HKK.l'ni.HM'HY: A SYMPOSIUM

217

curves for the boreal birds and mammals
conform qualitatively to theoretical predic-
tions, but they differ slightly from the quan-
titative Z-values that might be expected if
the avian distribution represents an equilib-
rium between contemporary colonization
and extinction, and the insular mammalian
populations are Pleistocene relicts that do
not disperse across the desert valleys.

These slight deviations from expected Z-
values are not difficult to explain. The low
Z-value for birds suggests that the isolation
of the montane islands may not be a signifi-
cant barrier to avian colonization. The facts
that a) approximately 80 percent of the var-
iation in insular bird species diversity can
be accounted for by the combined effects of
area and elevation (Table 2) or by habitat
diversity (Table 3), and b) there appears to
be no impoverishment of species numbers
resulting from isolation by distance to the

nearest continent (Table 2, and see next sec-
tion) suggest that colonization rates are high
and there is little if any effect of insular
isolation. This is consistent with Johnson's
(1975) conclusion that the diversity of bo-
real birds is attributable primarily to habi-
tat; the impoverishment of the insular avi-
faunas is the result of reduced habitat
diversity on the isolated peaks and not to
any significant extent to low colonization
rates. It should be noted that it is not nec-
essary to infer from this that the desert val-
leys do not inhibit dispersal, only that colo-
nization rates remain sufficiently high that
boreal species which maintain breeding
populations on the islands rarely go extinct
and, if they do, they recolonize rapidly.

The fact that the Z-value for mammals
also is slightly lower than predicted has a
different explanation. As I shall show in the
next section, the mammals that have colo-

Table 2. Stepwise multi
birds and mammals inhabit

iple

in»

regression of the influence of three variables
montane islands in the Great Basin.

on the number

of species of boreal

Variable1

Order entered
in equation

Birds

Contribution
to R2

F-value

Significance
level

Area

Highest peak
Nearest continent

1
2
3

0.4915
0.2836
0.0051

10.633
12.607
0.201

0.008
0.005
0.658

Variable1

Order entered
in equation

Mammals

Contribution
to R2

F-value

Significance
level

Area

Highest peak
Nearest continent

1

2 ^

3

0.716
0.005
0.010

42.957
0.271
0.559

<0.001
0.610
0.466

'Data are log-transformed data from Table 1.

Table 3. Correlation coefficients (r) between variables for the 13 montane islands for which all data are avail-
able. Upper right matrix is log-transformed data; lower left matrix is computed with untransformed data. Note
that number of bird and mammal species are not closely correlated with the same variables.

Highest

Nearest

Habitat

Bird

Mammal

Area

peak

continent

diversity

species

species

Area

_

0.795

-0.079

0.623

0.701

0.891

Highest peak

0.778

-

-0.334

0.722

0.869

0.523

Nearest continent

-0.083

-0.224

-

-0.176

-0.367

-0.070

Habitat diversity

0.524

0.760

-0.102

-

0.851

0.539

Bird species

0.637

0.872

-0.310

0.898

-

0.611

Mammal species

0.876

0.817

0.090

0.542

0.592

-

218

GREAT BASIN NATURALIST MEMOIRS

No. 2

nized the montane islands are those species
of the continental fauna that occur at rela-
tively low elevations within the boreal
zone; species restricted to high elevations
are absent from all of the isolated mountain
ranges. The large islands, which also tend to
have the highest peaks (Table 3), have ex-
tensive areas of mixed coniferous forest, wet
meadow, and other high altitude habitats.
Even with post-Pleistocene extinctions these
mountains would be expected to support
some of the mammal species characteristic
of these habitats on the continental ranges.
Thus, if the islands had an unbiased sample
of the boreal mammal fauna, the larger is-
lands should have more species than they
do and the species-area curve would be
steeper. In my 1971 paper, I reported a
higher Z-value than obtained here. The rea-
son is that I included in the present analysis
some marginally boreal species (Eutamias
dorsalis, E. panamintinus, and Sylvilagus
nuttalli) that are characteristic of juniper-
pinyon woodland and are present on most
of the montane islands. This had the effect
of adding an approximately constant num-
ber of species to the fauna of all islands,
and thus lowering the slope of the exponen-
tial species-area relationship. This is the
same effect that is produced by the dis-
proportionate representation of low eleva-
tion species in the insular faunas relative to
continental faunas because of differential
colonization in the past.

Isolation and Paleoclimatic History

It is a common observation of insular
biogeography that remote islands support
fewer species than islands of comparable
size and habitat diversity that are nearer to
a continent (see MacArthur and Wilson
1967 and included references). This pattern
is attributed to the limited ability of organ-
isms to disperse so that the rate of immigra-
tion to an island declines with increasing
isolation. In the present analysis, neither
birds nor mammals demonstrate such a
negative relationship between number of

species and distance to the nearest conti-
nent (Tables 2 and 3). For mammals, this
result is consistent with my 1971 analysis
and conclusion that there is virtually no
contemporary immigration to the isolated
mountains because the desert valleys con-
stitute almost absolute barriers to dispersal.
The islands were colonized in the Pleisto-
cene when periodic climate changes re-
sulted in shifts in the altitudinal limits of
the boreal vegetation. Data from plant
macrofossils in woodrat middens suggest
that, as recently as 8,000 to 12,000 years
ago, periods of cooler, wetter climate en-
abled juniper-pinyon woodland to flourish
at least 2,000 feet (600 m) below its present
lower elevational limit (Wells and Berger
1967, Wells and Jorgensen 1964). This was
sufficient to make juniper-pinyon and asso-
ciated meadow and riparian habitats con-
tiguous across virtually all of the Great Ba-
sin, and to enable the boreal mammals
characteristic of these habitats to colonize
all of the isolated mountain ranges. With
the return of hotter, drier conditions, these
"habitat bridges" connecting the islands to
the continental ranges were eliminated and
the insular mammalian faunas have been de-
rived from the widespread Pleistocene fauna
by independent extinctions on each island.

Several lines of evidence are consistent
with this interpretation. 1) The relatively
steep species-area curve (Fig. 3) suggests
that extinction has played a major role in
determining mammalian diversity. This is
supported by the discovery of late Pleisto-
cene fossils of at least one boreal species
(Marmota flaviventri.s) from an island
(Spring Range) where it no longer occurs
(Wells and Jorgensen 1964). 2) The lack of
any correlation between number of mammal
species and distance to nearest mainland
(Tables 2 and 3) or any other likely measure
of insular isolation (Brown 1971) suggests
that there is no contemporary immigration
to the islands. 3) All of the species
known from the islands are found in
juniper-pinyon or other habitats of com-
parable elevation. It is possible to account

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

219

for the presence of the entire insular fauna
in terms of colonization across habitat
bridges that were known to have existed in
the late Pleistocene. 4) All of the species of
boreal mammals that are restricted to mixed
coniferous forest or habitats of higher eleva-
tion on the continental ranges are absent
from all of the montane islands, even
though large areas of apparently suitable
habitat are present on some of the larger is-
lands. These species include Maries ameri-
cana, Aplodontia rufa, Eutamias alpinus, E.
toivnsendi, E. speciosus, Tamiasciurus hud-
sonicus, T. douglasi, Glaucomys sabrimis,
Phenacomys intermedins, Clethrionomys
gapperi and Lepus americanus. Paleobota-
nies evidence indicates that the habitats of
these species were not connected across the
Great Basin during the late Pleistocene.
This is further evidence that small terrestri-
al mammals usually are unable to cross hab-
itat barriers only a few miles in extent and
thus colonize isolated habitats only when
bridges of appropriate habitat provide di-
rect access (Brown 1971, 1973).

The explanation for the lack of correla-
tion between number of species and dis-
tance to nearest continent in birds appears
to be the opposite of that in mammals:
birds are such good dispersers that they
have colonized virtually all islands with
suitable habitat regardless of their isolation.
Both Johnson (1975) and Behle (this sym-
posium) report a negative relationship be-
tween their measure of insular isolation (cu-
mulative width of desert barriers) and
number of permanent resident bird species,
but both authors included continental
sample areas in their analyses. The apparent
effect of isolation in their analyses can be
attributed largely to the fact that a signifi-
cantly greater number of species inhabit
continental sites than occur on the islands;
there is little or no effect of isolation by
distance when only islands are considered
(see Fig. 2 of Johnson 1975). This is con-
sistent with my own analysis of Johnson's
and Behle's data. There is an insignificant
negative correlation between number of

bird species and distance to nearest conti-
nent (Table 3), and the distance variable
does not contribute significantly to account-
ing for number of bird species using mul-
tiple regression analysis (Table 2). It re-
mains to explain the difference in species
diversity between the continents and islands.
Some of those species present on the conti-
nents but absent from most of the islands
may be limited in their distributions by
their sedentary natures or their aversion to
crossing inhospitable desert terrain. How-
ever, many of these species such as Lagopus
leucurus, Dryocopus pileatus, and Perisoreus
canadensis, have specialized habitat require-
ments and low population densities. Their
habitats are totally lacking from many of
the islands, and, even where they are pres-
ent, they often consist of small patches ob-
viously inadequate to support sustained
populations. Two other sources of evidence
suggest that rates of contemporary immigra-
tion by most boreal bird species are suffi-
ciently high to keep the habitats present on
the islands filled with an appropriate com-
plement of species. First, such boreal bird
species as Picoides tridactylus and Cyano-
sitta stelleri, which are restricted to high
elevation, well-developed coniferous forests,
are present on the islands. Since the habi-
tats of these species have not been con-
nected by bridges to the continental ranges
during the Pleistocene, it must be inferred
that at least these species are able to colo-
nize across significant barriers of unsuitable
habitat. Second, vagrant individuals of some
boreal species (e.g., Cyanositta stelleri and
Cinclus rnexicanas) infrequently are report-
ed significant distances from breeding popu-
lations (Johnson 1975, B. Bundick, pers
comm.). This evidence indicates that even
relatively sedentary permanent resident bo-
real birds are much more vagile than small
boreal mammals.

Determinants of Species Diversity and
Composition

Differences in dispersal account in large
part for differences in the species-area rela-

220

GREAT BASIN NATURALIST MEMOIRS

No. 2

tionships between birds and mammals and
in faunal composition between insular and
continental mountain ranges (at least for
mammals), but the considerable variation in
species diversity among the insular bird and
mammal faunas must be attributed primari-
ly to ecological characteristics of the is-
lands. Although a common set of both bird
and mammal species has had the opportu-
nity to colonize virtually all islands, the iso-
lated mountain ranges support different
numbers and kinds of species. Species diver-
sity of both birds and mammals is closely
correlated with insular area, but area is
only a correlate of factors such as the quan-
tity and diversity of habitat and food re-
sources that ultimately determine species di-
versity. Power (1972) and Johnson (1975)
have shown for birds that someone familiar
with a group of organisms can devise quan-
titative measures of habitat diversity that
account for significantly more of the vari-
ance in insular species diversity than area.
This approach has great utility for elucidat-
ing the environmental factors that influence
the diversity and distribution of particular
taxa, but it may not be useful for devel-
oping general biogeographic or ecological
theory.

There are two arguments in favor of bas-
ing biogeographic theory on simple, stan-
dard parameters such as insular area. The
first is practical. Parameters such as area,
elevation (another correlate of habitat diver-
sity), and distance to the nearest continent
are easy to determine from maps. By ob-
taining these measurements and species lists
from the literature, it is possible to describe
clear patterns of plant and animal distribu-
tion without doing all the original fieldwork
required to quantify accurately more direct
environmental variables. Much of the recent
synthetic work in insular biogeography has
been done this way, and even those authors
that have attempted to quantify variables
such as habitat diversity often have used
data available from the literature or topo-
graphic maps (e.g., Power 1972, Johnson
1975). The second argument is that area

may be the best parameter for constructing
general theory that can be applied to di-
verse taxa and various kinds of islands. In
the development of theory some sacrifice of
precise explanation of specific cases usually
must be made in order to obtain a desirable
degree of generality. It is questionable
whether it would be practical or profitable
to base biogeographic theory on organism-
specific parameters such as habitat diversity,
food availability, or carrying capacity. Cer-
tainly at present we have no accepted, stan-
dardized techniques for measuring these
variables in the field that are suitable for a
variety of taxa and habitats.

Some test of the merit of this approach
can be made by comparing the correlates of
species diversity for boreal birds and mam-
mals. Johnson (1975) developed a quan-
titative "habitat diversity score" that ac-
counted for most of the variability in the
number of boreal bird species in his analy-
sis. This appears to measure accurately the
requirements of boreal birds in western
North America, because Behle (this sym-
posium) tested it and obtained gratifyingly
similar results. Since boreal mammals utilize
the same forest, meadow, and freshwater
habitats as boreal birds, it would be encour-
aging if Johnson's habitat diversity score ac-
counted for similarly large proportions of
the variability in insular species diversity
for both taxa. Unfortunately this is not the
case (Table 2). The number of boreal bird
and mammal species inhabiting the same is-
lands are not very closely correlated (r2 <
0.37); whereas 81 percent of the variability
in bird species diversity can be attributed to
habitat diversity score, the comparable fig-
ure for mammals is only 29 percent. Area is
by far the best correlate of species diversity
for both birds and mammals. When the ap-
propriate log-transformed data are used,
area accounts for 49 and 79 percent of the
variability in bird and mammal species
numbers, respectively (Table 3). Inter-
estingly, elevation of highest peak (another
correlate of habitat diversity readily obtain-
able from topographic maps) and area taken

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

221

together in multiple regression analysis
(Table 2) account for almost as much varia-
bility in bird species diversity (r2 = 0.76)
as the habitat diversity score (r2 = 0.81).

Two explanations can be offered for the
lack of good correlation between Johnson's
habitat diversity score and mammalian spe-
cies diversity. First, the habitat diversity
score is highly dependent on the species di-
versity of coniferous trees. Particular tree
species and forest types probably are impor-
tant to the boreal birds, but they are much
less likely to influence the diversity of bo-
real mammals. As I have shown, boreal
mammals that depend on the well-developed
coniferous forests of high elevations have
been unable to colonize any of the islands,
and those mammals that have colonized, if
they require conifers at all, are primarily
species of the juniper-pinyon woodlands
that are well developed on almost all of the
islands. Second, Johnson's habitat diversity
score is based primarily on the presence or
absence of particular tree species or habitat
types. This qualitative approach appears to
work well for birds, because they have suffi-
ciently high immigration rates to maintain
populations on almost all islands where suit-
able habitats are present. In contrast, mam-
mals, which have had to maintain insular
populations for thousands of years in the
absence of any significant immigration, may
be much more dependent on the quantity of
boreal habitats (particularly those character-
istic of low elevations that were connected
in the Pleistocene) which may be more
closely correlated with insular area than
with Johnson's more qualitative index.
Whether one or both of these nonexclusive
explanations turn out to be correct, it is
clear that different factors determine the di-
versity of insular species of birds and mam-
mals. These results suggest that parameters
such as area, even though their effects are
indirect and imprecise, may be the variables
best suited for constructing and testing gen-
eral biogeographic theory.

However, specific factors that influence
the insular species diversity and composition

of particular taxa can contribute to theory
by providing information on the mecha-
nisms and effects of the underlying process-
es of colonization and extinction. The great
importance of habitat in determining insular
bird species diversity, coupled with a low
slope of the species-area curve and the lack
of any significant effect of isolation by dis-
tance, suggests that boreal birds are able to
disperse at a sufficient rate to maintain
populations on most islands with suitable
habitats. The distribution of boreal mam-
mals presents a dramatic contrast. Mammal
species are absent from many islands that
have suitable habitats, either because they
have never colonized or because they have
gone extinct since the last episode of
Pleistocene colonization. The interacting ef-
fects of habitat requirements and paleocli-
matic changes in determining the set of spe-
cies that colonized the islands already have
been discussed. Of equal interest are the
factors that have resulted in extinctions
among the original colonists to produce the
present insular faunas.

The mammalian faunas of the montane is-
lands have been derived by extinction from
a common set of 14 functional species that
were widely distributed across the Great
Basin during the late Pleistocene when their
habitats were connected by bridges to the
continental mountain ranges (Table 4). After
the islands were isolated by the contraction
of boreal habitats to approximately their
present position, extinctions reduced the
faunas of each island and thus played a ma-
jor role in determining species composition.
Five large islands have retained at least 10
of their original 14 species, but five small
islands have lost all but 3 or 4 of the origi-
nal set of species in the period of approx-
imately 10,000 years that the boreal habi-
tats have been isolated. The distribution of
extinctions among species is highly non-
random and appears to be related primarily
to population size, as MacArthur and Wil-
son (1967) predicted. Herbivorous species of
generalized habitat requirements and small
to intermediate body size have persisted on

222

GREAT BASIN NATURALIST MEMOIRS

No. 2

most of the islands (Table 4). In contrast,
herbivores of large body size and/or spe-
cialized habitat requirements and carnivores
have had higher extinction rates and persist
on only a small proportion of the 19 islands.
The frequency of occurrence of boreal spe-
cies on the islands (Table 4) corresponds
very closely to the relative abundance of
these species where they occur together on
large islands or continental mountain ranges
(personal observations). The dependence of
extinction on population size probably is
even more precise than is apparent here,
because population size is influenced by the
presence of particular habitats and com-
peting species which vary among islands.
For example, one of the few mountain
ranges where a juniper-pinyon chipmunk
(Eutamias dorsalis or E. panamintinus) does
not occur is the Ruby Mountains, where
juniper-pinyon woodland is very poorly de-

veloped. Similarly, the chipmunk character-
istic of higher elevations (Eutamias um-
brinus) is absent from the Pilot Range,
which has only small stands of mixed con-
ifers on its single tall peak. On this moun-
tain, in the absence of E. umbrinus, E. dor-
salis has extended its altitudinal range
upward into the mixed conifers on the
peak.

Conclusions, Predictions, and
Unanswered Questions

The insular distributions of boreal birds
and mammals in the Great Basin differ in
ways that are consistent with both current
biogeographic theory and independent evi-
dence of ecological and historical factors
that have affected colonization and extinc-
tion. All data seem consistent with the in-
terpretation that insular bird populations
represent an equilibrium between contem-

Table 4. Characteristics of the boreal mammal species that inhabit the montane islands of the Great Basin.
Species are listed in decreasing order of frequency of occurrence.

Body

Number

weight

of islands

Species

(grams)

Diet

Habitat

inhabited

Eutamias umbrinus

60

mostly seeds

generalist: forests, wood-
lands, talus

17

Neotoma cinerea

300

green vegetation

generalist: rock outcrops
or talus in all habitats

17

Eutamias dorsalis, E.

55

mostly seeds

generalist: primarily

16

panamintinus

juniper-pinyon woodlands

Spermophilus lateralis

170

green vegetation
seeds

generalist: open forest,
meadow

14

Microtus longicaudus

45

green vegetation

generalist: meadow, open
forest, streams

13

Sylvilagus nuttallii

800

green vegetation

generalist: all habitats
except dense forest

12

Marmota fhviventris

3,000

green vegetation

generalist: open forest,
meadow

10

Sorex vagrans, S.

7

invertebrates

generalist: forest, meadow.

8

tenellus'

streams

Sorex palustris

14

invertebrates

permanent streams

6

Oehotona princeps

120

green vegetation

talus adjacent to meadow

5

Zapus princeps

25

seeds, green
vegetation

wet meadow, streams

4

Mustela erminea

.50

small vertebrates

generalist: meadow, open
forest

4

Spermophilus beldingi

300

green vegetation

wet meadows

3

Lepus townsendii

3,000

green vegetation

large open meadows

1

Congeners listed on the same line are ecological and geographic replacements.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

223

porary rates of colonization and extinction;
immigration rates are sufficiently high that
boreal bird populations inhabit almost all is-
lands where there are suitable habitats, re-
gardless of their isolation from continents.
In contrast, the boreal mammal populations
represent relicts of species that were wide-
spread in the Great Basin during the late
Pleistocene when habitat bridges connected
the islands and continents; the mammal
faunas of the islands have been derived
from a common set of Pleistocene colonists
by subsequent extinctions in the absence of
immigration.

Neither the avian nor the mammalian dis-
tributions fit the equilibrium model pro-
posed by Mac Arthur and Wilson (1967) to
account for diversity on oceanic islands.
The insular mammal faunas clearly are not
in equilibrium; in the absence of immigra-
tion they are gradually relaxing toward an
equilibrium of zero species at rates in-
versely related to island size. The distribu-
tion of birds does represent a sort of equi-
librium between contemporary colonization
and extinction, but the immigration rates
are so high that the islands are virtually sat-
urated with species for which the appropri-
ate habitats are present. The avifaunas of
the islands differ from the MacArthur-Wil-
son model in that there is no significant ef-
fect of isolation by distance, species compo-
sition is quite precisely determined by
habitat, and rates of faunal turnover prob-
ably are very low (see Brown and Kodric-
Brown 1977).

The information available on the distribu-
tion of other vertebrates among insular hab-
itats suggests that the kinds of patterns de-
scribed here for birds and mammals may be
widespread. Migratory birds, bats, and very
large carnivorous and herbivorous mammals
probably are at least as vagile as permanent
resident boreal birds. They would be ex-
pected to colonize most suitable habitats
and show little effect of insular extinction
or isolation by distance. On the other hand,
amphibians and reptiles probably have dis-
persal capacities similar to mammals. These

taxa are unlikely to cross habitat barriers of
even modest extent, and it is well known
that fishes can colonize new areas only
when suitable habitat bridges are present
(Barbour and Brown 1974, Hubbs and Mill-
er 1948, Hubbs et al. 1974, Smith, this sym-
posium). The distributions of these taxa
should be extremely sensitive to paleocli-
matic and geological events; where insular
habitats have been connected and reisolated
they should show relictual distributions in
which the identity of the colonists and the
effects of subsequent extinctions produce
patterns of diversity comparable to those of
the small boreal mammals. It is interesting
to speculate that the extremes of vagility
which characterize most of the vertebrates
and the extensive climatic and geological
changes that have drastically changed the
landscape of western North America within
the last million years make it unlikely that
land and fresh water vertebrates in this re-
gion will show the sort of equilibrium distri-
butions that have been demonstrated for the
biotas of oceanic islands with a long history
of isolation and relative environmental sta-
bility (see MacArthur and Wilson 1967,
Simberloff 1974).

In insular biogeography many important
questions remain to be answered, and much
theoretical and empirical work is yet to be
done before this promising and vigorous
young science can afford to rest on its lau-
rels. The vertebrates that are distributed
among the numerous isolated habitats in
western North America continue to offer
great potential as systems for testing theory
and developing general concepts. For those
organisms that appear to demonstrate equi-
librial distributions it is particularly impor-
tant to measure turnover rates accurately,
and to incorporate the results into an ap-
propriate conceptual framework. For those
organisms that exhibit relictual distributions
it is important to elucidate patterns of colo-
nization and extinction and to relate those
to theory. Three generations of icthyologists
(Hubbs and Miller 1948, Hubbs et al. 1974,
Smith, this symposium) have carefully dis-

224

GREAT BASIN NATURALIST MEMOIRS

No. 2

sected historical changes in Great Basin
drainages and have related them to the con-
temporary distribution of fishes. As yet
there has been little attempt to relate these
patterns to biogeographic theory, but the
possibilities for doing so seem great and it is
hoped it will be attempted soon. The distri-
butions of many organisms are patchy on
scales smaller than the gross biogeographic
one considered here. Smith (1974a, b) has
obtained some encouraging results by using
the theory of insular biogeography and the
dynamics of opposing colonization and ex-
tinction rates to account for the distribution
of a small boreal mammal (Ochotona prin-
ceps) among the isolated patches of its spe-
cialized rockslide habitat in the Sierra Ne-
vada. It will be interesting to see to what
extent this approach continues to prove use-
ful for understanding insular distributions on
various scales.

Acknowledgments

This paper could not have been written
without the field work and published spe-
cies lists of numerous collectors, system-
atists, and biogeographers who have worked
to document the distributions of birds and
mammals in the Great Basin. I am particu-
larly grateful to N. K. Johnson for providing
most of the data on birds and to W. H.
Behle for making available information
presented in this symposium. My wife, A.
Kodric-Brown, and my students have done
much to encourage my work in insular
biogeography. I would like to dedicate this
paper to the memory of Stephen D. Dur-
rant, whose knowledge of the Great Basin
and its mammals was matched only by his
enthusiasm for the outdoor life and his abil-
ity as a teacher and storyteller.

Appendix 1. Records of occurrence ot permanent resident boreal bird species.'

Sample Area
Species 1 2 3 4 8 9 10 11 14 15 16

18 19 20 21

22 23

Dendragapus obscurus
Bonasa umbellus
Lagopus leucurus
Oreortyz picta
Glaucidium gnotna
Strix occidentals
S. ncltulosa
Drijocopus pileatus
Dendrocopus villostis
D. pubescens
D. albolaruatus
Picoides arcticus
P. tridactylus
Perisorcus canadensis
Cyanositta stellch
Varus gambeli
Sitta carolinensis
S. pi/gmaea
Cinclus mexicanus
Pinicola cnucleator

Total

5 4 6 6 6 5

6 13 15

'Data from Johnson (1975) and Behle (this symposium).

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

225

» 2 = 5 »-~

= ;£ j: •§ E § | J =~' ~ ~

8. £ = -- ^

2 >

a. c £

\%\ llllli | 111 ill 111! lillj ill fill iff

6 | |
^11

226

GREAT BASIN NATURALIST MEMOIRS

No. 2

Literature Cited

H. Brown. 1974. Fish spe-
lakes. Amer. Naturalist 108:

Barbour, C. D., and J.
cies diversity in
473-489.

Behle, W. H. 1943. Birds of Pine Valley Mountain
Range, southwestern Utah. Univ. Utah Biol. Ser.
7(5): 1-85.
1955. The birds of the Deep Creek Moun-
tains of central western Utah. Univ. Utah Biol.
Ser. 11(4): 1-34.

1958. The birds of the Raft River Mountains,

northwestern Utah. Univ. Utah Biol. Ser. 11(6):
1-40.

1978. Avian biogeography of the Great Basin

and Intermountain Region. Great Basin Nat.
Mem. 2: 55-80.

Brown, J. H. 1971. Mammals on mountaintops:
nonequilibrium insular biogeography. Amer.
Naturalist 105: 467-478.

1973. Species diversity of seed-eating desert

rodents in sand dune habitats. Ecology 54: 775-
787.

Brown, J. H., and A. Kodric-Brown. 1977.
Turnover rates in insular biogeography: effect of
immigration on extinction.' Ecology 58: 445-449.

Culver, D. C. 1970. Analysis of simple cave com-
munities. I. Caves as islands. Evolution 29: 463-
474.

Culver, D. J. R. Holsinger, and R. Baroody.
1973. Toward a predictive cave biogeography:
the Greenbrier Valley as a case study. Evolution
2: 689-695.

Diamond, J. M. 1969. Avifaunal equilibria and spe-
cies turnover rates on the Channel Islands of
California. Proc. Nat. Acad. Sci. 64: 57-63.

1970. Ecological consequences of island colo-
nization by Southwest Pacific birds. II. The ef-
fect of species diversity on total population den-
sity. Proc. Nat. Acad. Sci. 67: 1715-1721.

1971. Comparison of faunal equilibrium

turnover rates on a tropical and a temperate is-
land. Proc. Nat. Acad. Sci. 68: 2742-2745.

Durrant, S. D. 1952. Mammals of Utah, taxonomy
and distribution. Univ. Kansas Pnbl. Mus. Natu-
ral Hist. 6: 1-549.

Durrant, S. D., M. R. Lee, and R. M. Hansen.
1955. Additional records and extensions of
known ranges of mammals from Utah. Univ.
Kansas Publ. Mus. Natural Hist. 9: 69-80.

Grinnell, J. 1933. Review of the recent mammal
fauna of California. Univ. California Publ. Zool.
40: 71-234.

Grinnell, J., and A. H. Miller. 1944. The distri-
bution of the birds of California. Pac. Coast.
Avif. 27: 1-608.

Grinnell, J., and T. I. Storer. 1929. Animal life in
the Yosemite. Univ. of California Press, Berke-
ley.

Hall, E. R. 1946. Mammals of Nevada. Univ. Cali-
fornia Press, Berkeley.

Harner, R. F., and K. T. Harper. 1977. The role
of area, heterogeneity, and favorably in plant
species diversity of pinyon-juniper ecosystems.
Ecology 57: 1254-1263.

Harper, K. T., D. C. Freeman, W. K. Ostler, and L.
G. Klikoff. 1978. The flora of Great Basin
mountain ranges: diversity, sources, and dis-
persal ecology. Great Basin Nat. Mem. 2: 81-
103.

Hubbs, C. L., and R. R. Miller. 1948. Correlation
between fish distribution and hydrographic his-
tory in the desert basins of western United
States, p. 17-166. In: The Great Basin with em-
phasis on glacial and postglacial times. Bull.
Univ. Utah 38(2): 1-191.

Hubbs, C. L., R. R. Miller, and L. C.
Hubbs. 1974. Hydrographic history and relics
fishes of the north central Great Basin. Memoirs
Calif. Acad. Sci. 7: 1-259.

Johnson, N. K. 1965. The breeding avifaunas of the
Sheep and Spring ranges in southern Nevada.
Condor 67: 93-124.

1970. The affinities of the Boreal avifauna of

the Warner Mountains, California. Occas. Pap.
Biol. Soc. Nevada 22: 1-11.

1973. The distribution of Boreal avifaunas in

southern Nevada. Occas. Pap. Biol. Soc. Nevada
36: 1-14.

1974. Montane avifaunas of southeastern Ne-
vada: historical change in species distribution.
Condor 76: 334-337.

1975. Controls of number of bird species on

montane islands in the Great Basin. Evolution
29: 545-657.

Linsdale, J. M. 1936. The birds of Nevada. Pac.
Coast Avif. 23: 1-145.

Lynch, J. F., and N. K. Johnson. 1974. Turnover
and equilibria in insular avifaunas, with particu-
lar reference to the California Channel Islands.
Condor 76: 370-384.

MacArthur, R. H., and E. O. Wilson. 1963. An
equilibrium theory of insular biogeography. Ev-
olution 17: 373-387.

1967. The theory of island biogeography.

Princeton Univ. Press, Princeton, N. J.

Martin, P. S., and P. J. Mehringer.
1965. Pleistocene pollen analysis and biogeo-
graphy of the Southwest, p. 433-451. In: H. E.
Wright and D. G. Frey (eds.), The Quaternary
of the United States, Princeton Univ. Press,
Princeton, N.J.

Miller, A. H. 1935. Some breeding birds of the Pine
Forest Mountains, Nevada. Auk 52: 467-468.

1946. Vertebrate inhabitants of the pinyon

association in the Death Valley region. Ecology
27: 54-60.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

227

Miller, A. H. and W. C. Russell. 1956.
Distributional data on the birds of the White
Mountains of California and Nevada. Condor
58: 75-77.

Miller, R. R. 1948. The cyprinodont fishes of the
Death Valley system of eastern California and
southwestern Nevada. Misc. Publ. Univ. Mich-
igan Mus. Zool. 68: 1-155.

Power, D. M. 1972. Numbers of bird species on the
California islands. Evolution 26: 451-463.

Preston, F. W. 1962. The canonical distribution of
commonness and rarity: Part I. Ecology 43: 185-
215.

Schoener, A. 1974. Experimental zoogeography:
colonization of marine mini-islands. Amer. Nat-
uralist 108: 715-738.

Seifert, R. P. 1975. Clumps of Heliconia in-
florescences as ecological islands. Ecology 56:
1416-1422.

Simberloff, D. S. 1974. Equilibrium theory of is-
land biogeography and ecology. Ann. Rev. Ecol.
Syst. 5: 161-182.

Simberloff, D. S., and E. O. Wilson. 1969.
Experimental zoogeography of islands. The colo-
nization of empty islands. Ecology 50: 278-296.

Simpson, R. R. 1974. Glacial migrations of plants:
island biogeography evidence. Science 185: 689-
700.

Smith, A. T. 1974a. The distribution and dispersal

of pikas: consequences of insular population
structure. Ecology 55: 1112-1119.

1974b. The distribution and dispersal of

pikas: influence of behavior and climate. Ecolo-
gy 55: 1368-1376.

Smith, G. R. 1978. Riogeography of Intermountain
fishes. Great Rasin Nat. Mem. 2: 17-42.

Terborgh, J. 1973. Chance, habitat, and dispersal
in the distribution of birds in the West Indies.
Evolution 27: 338-349.

Van Rossem, A. J. 1936. Rirds of the Charleston
Mountains, Nevada. Pac. Coast Avif. 24: 1-65.

Vuilleumier, F. 1970. Insular biogeography in conti-
nental regions. I. The northern Andes of South
America. Amer. Naturalist 104: 373-388.

1973. Insular biogeography in continental re-
gions. II. Cave faunas from Tesin, southern
Switzerland. Syst. Zool. 22: 64-76.

Wells, P. V., and R. Rerger. 1967. Late Pleisto-
cene history of coniferous woodland in the Mo-
jave Desert. Science 155: 1640-1647.

Wells, P. V., and C. D. Jorgensen. 1964.
Pleistocene wood rat middens and climatic
change in the Mojave Desert: a record of juni-
per woodlands. Science 143: 1171-1173.

West, N., R. J. Tausch, K. H. Rea, and P. T.
Tueller. 1978. Phytogeographical variation
within juniper-pinyon woodlands of the Great
Rasin. Great Rasin Nat. Mem. 2: 119-136.

BIOGEOGRAPHY AND MANAGEMENT OF NATIVE WESTERN SHRUBS:
A CASE STUDY, SECTION TRIDENTATAE OF ARTEMISIA

E. Durant McArthur and A. Perry Plummer1

Abstract.— Biogeographical considerations are important in the management of western shrublands. Seedings
on rangelands have a higher probability of success when tried and tested principles are followed. It is usually
best to seed mixtures that include adapted shrubs and herbs. Shrubs generally are well adapted to the environ-
mental extremes of western ranges. Throughout the year, they provide nutrients for herbivores that are only sea-
sonably available in herbs.

Section Tridentatae of Artemisia is endemic to western North America and distinct from the analogous Eura-
sian section (or subgenus) Seriphidium. The groups have separate, distinguishable centers of diversity; the two
groups seem to be connected in the geologic past by way of the more primitive subgenus Artemisia. Preliminary
karyotypic evidence suggests different but advanced karyotypes for both groups; chemotaxonomic and morpholo-
gical data indicate differences in the groups. The Tridentatae likely evolved in North America during late Ter-
tiary or early Quaternary times under the stimulus of cycles of aridity.

The Tridentatae (sagebrushes) are morphologically variable. Different accessions are differentially adapted.
Management practices for various taxa should take into consideration the individual taxon's characteristics. Effort
should be made to seed adapted taxa and accessions. Sagebrush management requires maintenance, seeding, or
thinning, depending upon the circumstances.

Aside from serving as a review of the lit-
erature, this paper presents original data
and touches upon unpublished material and
reports being prepared for publication.

Shrubs are an important component of
the vegetation of the American West, as
well as in other parts of the world. In fact,
much of the West is classified as shrubland
because the dominant plant species are
shrubs (Kiichler 1964, McGinnies 1972).
Plummer (1974) recognized six major shrub-
land types in the American West.

Since the arrival of the white man, shrub
ecosystems have been exploited and des-
poiled by his livestock and land manage-
ment practices (Cottan 1961, Heady 1975).
The distribution and composition of many
plant and animal communities have been
substantially altered. The relatively new sci-
ence of range management was born, in
part, to systematize efforts to make shrub-
lands more productive. The value of shrubs
in their own right has long been under-

estimated (McKell 1975b). In fact, much ef-
fort, time, and money have been spent to
eradicate and control shrubs in order to fa-
cilitate the establishment of exotic grass
monocultures.

Shrubs have many current and potential
uses. In the proceedings of an international
symposium on useful wildland shrubs
(McKell et al. 1972), a section on the pres-
ent and possible uses of shrubs included
chapters devoted to browse and cover for
wildlife, low-maintenance landscaping, soil
cover and stabilization, fire relations, medic-
inal values, and industrial raw materials.
This list was not exhaustive; for example, no
reference is made to the use of shrubs as
food for livestock.

Our work on shrubs began with a need to
improve critical winter game ranges (Plum-
mer et al. 1968) and has continued with
emphasis shifting to improving all types of
disturbed sites (McArthur et al. 1974, Mon-
sen 1975, Plummer, in press). Numerous

'Research geneticist and project leader, Intermountain Forest and Range Experiment Station, Forest Service, U.S. Department of Agriculture, Ogden,
Utah 84401; located at the Intermountain Station's Shrub Sciences Laboratory, Frovo, Utah 84601.

229

230

GREAT BASIN NATURALIST MEMOIRS

No. 2

species and ecotype adaptation plots have
been set out both in and out of the natural
ranges of particular shrub taxa. Plummer (in
press) listed 85 native and 14 exotic shrubs
that have proved to be adapted to disturbed
sites in one or more vegetative communities
of the Intermountain Region. Some shrubs
can be established only vegetatively or by
transplanting and do not reproduce by seed.
Even these, however, have value since they
cover and stabilize disturbed areas and
serve as a nurse crop on critical sites for
plants that establish more slowly.

A principal thrust of range management
has been in the modification of vegetation—
often disclimax vegetation (Heady 1975).
Mechanical, chemical, and burning methods
have been used widely to control unwanted
plants— mostly woody ones— on rangelands.
These treatments are generally followed by
seeding. Practical trials and experiments
have succeeded and failed often enough for
cause and effect relationships to be ana-
lyzed and to serve as a set of principles.
Recommendations relative to the seeding of
many range sites can be made with a high
probability of success. Plummer et al. (1968)
and Heady (1975) outlined procedures and
site criteria necessary for successful range
seedings. Except for weather conditions,
these are controllable by the range man-
ager. We wish to emphasize, as did Plum-
mer et al. (1968) and Heady (1975), the im-
portance of planting mixtures of adapted
taxa because:

(1) Many seedings are on variable terrain
that includes diverse microhabitats.

(2) A mixed diet is usually more palatable
to and nutritious for herbivores.

(3) Periods of growth vary for different
plant taxa and classes (shrubs, forbs and
grasses); so succulent forage is provided for
a longer time.

(4) Some plants benefit others by provid-
ing habitat and nutrients.

(5) Diseases and insect pests do not at-
tack all species equally. Shrubs should be
included in most seedings since they are
well adapted to drought, salinity, acidity,
wide temperature fluctuations, and other
environmental extremes of western ranges.
In addition, they provide habitat for ani-

mals that other classes of plants do not
(McKell 1975a).

This paper examines the biogeography
and the management implications of a par-
ticular group of shrubs— the sagebrushes,
section Tridentatae of Artemisia. The sage-
brushes are perhaps the most common shrub
in western North America. In order to get a
better idea of how to better manage and
utilize this resource, we have tried to catch
a glimmer of its evolutionary past. Such in-
formation should benefit the management of
sagebrush ranges by providing bases for bet-
ter understanding species adaptation and
distribution and plant improvement pro-
grams through hybridization and selection.

M

ETHODS AND MATERIALS

M

Species Distributions.— General distribu-
tion of species of the Tridentatae was ob-
tained from Ward (1953), Reetle (1960), and
by examining the Artemisia collections of
the herbaria of Brigham Young University
(BRY) and the Intermountain Station's
Shrub Sciences Laboratory (SSLP). We have
followed Beetle's (1960) and Beetle and
Young's (1965) nomenclature for the Triden-
tatae (Fig. 1). Distribution of the Seriphi-
dium species that occurs in the Soviet Union
was taken from Polyakov (1961). For spe-
cies that occur outside the Soviet Union, de
Candolle (1838) and Boissier (1875) were
consulted (Fig. 2). We reviewed other re-
gional and national floras peripheral to the
Soviet Union but found little information to
add to the aforementioned references.

Karyotyping.— The karyotypes of section
Tridentatae were obtained from a prelimi-
nary analysis of data for a later report
(McArthur, Pope, and Plummer, in prepara-
tion). A sample cell was selected from each
of 13 diploid and 14 tetraploid populations
representing 9 diploid and 9 tetraploid taxa.
Squashed root tips of seedlings were micro-
scopically analyzed. The slides were pre-
pared by fixing colchicine-pretreated root
tips in 1:3 acetic alcohol and squashing
them in acetocarmine. The idiogram of sec-
tion Tridentatae (Fig. 3) was prepared from
measurements made on 20 X 25 cm photo-
micrographs prints at a magnification of
3120X. The idiogram for the two species of

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

231

126 120

Fig. 1. Distribution ranges of the species of section Tridentatae.

232

GREAT BASIN NATURALIST MEMOIRS

No. 2

Seriphidium (Fig. 3) was prepared from
data presented by Filatova (1974). We used
the shorter of two lengths Filatova gives for
chromosome 3 of A. juncea. Chromosome
sizes (L = large, M = medium, S = small)
were determined by pairing the chromo-
somes according to relative length and cen-
tromere position, dividing the genome into
relative lengths, averaging the length of
each pair, and proportionalizing the pair so
the genome had a length of 100 arbitrary
units. Thus, L > 12.4, M = 9.6 - 12.4, S
< 9.6 for diploids and L > 6.2, M = 4.8
- 6.2, S < 4.8 for tetraploids. The centro-
mere positions (M = metacentric, SM =
submetacentric, and ST = subterminal)
were determined by the ratio of the length
of the short arm to the length of the long
arm. Thus, M > 75 percent, SM = 50 -
75 percent, ST < 50 percent (Tables 1 and
2).

Species Adaptation.— Shrub wildings have

been collected in the spring (March- June)
and fall (October-November) and trans-
planted to uniform gardens and smaller spe-
cies adaptation plots. Periodic ratings are
made as to their height, crown, vigor, her-
bage yield, reproduction, and survival.
These data are on file principally at the
Great Basin Experimental Area in Ephraim,
Utah, but also at the Shrub Sciences Labo-
ratory in Provo, Utah. Collection numbers,
prefixed by a U, have been assigned to ac-
cessions within each taxon (Tables 1, 2, 3).
Voucher herbarium specimens are deposited
at the Shrub Sciences Laboratory (SSLP).

Section Tridentatae of Artemisia

The genus Artemisia is a group of about
200 plant species which belong to the tribe
Anthemideae of Compositae (de Candolle
1837, Clapham 1962). Good (1974) consid-
ers Artemisia as belonging to his temperate
genera, group 1, subgroup 2, which includes

Fig. 2. Distribution of the section (subgenus) Seriphidium in Eurasia and North Africa. The numbers indicate
maximum number of species in each area.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

233

"genera found throughout the northern ex-
tratropical latitudes with some extension
southwards in all directions, usually to cer-
tain tropical mountains only."

Besser (1829) began the first com-
prehensive monograph on the genus early in
the 19th century. He recognized four sec-
tions— Abrotanum, Absinthium, Dracunculus,
and Seriphidium. His monograph was not
completed before he died, but de Candolle's
(1837) "Prodomus Systematis Naturalis" and
Hooker's (1840) "Flora Boreali-Americana"
include significant portions of Besser's work.
Besser's sections are still considered the
principal divisions of the genus, but in some
treatments they have been elevated to the
rank of subgenus and additional subordinate
sections created (Rydberg 1916, Polyakov
1961). The comprehensive treatment of Ar-
temisia in the "Flora of the USSR." (Poly-
akov 1961) unites Absinthium and Abrota-
num into the single large subgenus
Artemisia. Hall and Clements (1923) recog-
nized the close affinity of these two groups,
which were separated in Besser's scheme

only by the hairiness of the floral recep-
tacle. Members of Seriphidium are distin-
guished by having only perfect disk flowers
whereas members of the other subgenera
have fertile or sterile ray flowers and vari-
ous degrees of disk flower fertility. Rydberg
(1916), while recognizing the subgenus Seri-
phidium, created the section Tridentatae for
most North American members of Seriphi-
dium. He also created monotypic sections
for A. rigicki and A. pygmaea as well as a
new genus for the largely herbaceous A.
palmeri. More recent monographers (Hall
and Clements 1923, Ward 1953, and Beetle
1960) have been more conservative than
Rydberg, but have been divided on the re-
tention of Tridentatae as a section. Beetle
(1960) included all woody, North American,
homogamous Artemisia species in the sec-
tion Tridentatae along with A. bigelovii,
which often has a ray flower or two on oth-
erwise discoid heads. Beetle's Tridentatae
form a group of 11 unequal species endemic
to western North America (Fig. 1). The spe-
cies distribution ranges vary in size from

SERIPHIDIUM

TRIDENTATAE

Fig. 3. Idiograms of two species of section Seriphidium, A. juncea, and A. leucodes (after Filatova 1974) and a
generalized idiogram of section Tridentatae. Note secondary constriction in long arm of the first chromosome in
the Tridentatae. The arrows indicate a chromosome substitution in some taxa (Table 1).

234

GREAT BASIN NATURALIST MEMOIRS

No. 2

-r >n

o o —

ID IC

05 C ~

A £

4*1

Z A w

C ~ - — 3)

■3 °

3 U

'' $ - £ -9 J ^

«p g« 4 -S «A ° "9 •

a 3

* - c -

< < <!

Q $j 9

S ^ 2 -n

nil

ill?

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

235

those that cover most of the range of sec-
tion Tridentatae (A. tridentata and A. cana)
to relatively small ranges covering about 10
percent of the section's range (A. pygtnaea
and A. longiloba) to the restricted type lo-
cation range of A. argilosa. A. rothrockii is
shown with a disjunct range (Fig. 1). Ward
(1953) and Cronquist (1974) maintain that
A. rothrockii is restricted to the Sierra Ne-
vada and its outliers. Beetle (1960), on the
other hand, records the species from the
Rocky Mountains. The California material is
tetraploid and hexaploid (Ward 1953),
whereas our work indicates that Rockv

Mountain material (from Utah's Wasatch
Plateau) is diploid (McArthur, Pope, and
Plummer, in preparation). The Rocky
Mountain material may be an undescribed
taxon. However, in this paper, we are treat-
ing it as A. rothrockii.

Evidence for Separating Tridentatae
from seriphidium

Regardless of the rank (subgenus or sec-
tion) that one attaches to Tridentatae or Se-
riphidium, the two taxa should be sepa-
rated. Hall and Clements (1923), Ward

Table 2. Chromosome constitution of representative tetraploid Tridentatae.

Collection data

Ch

omosome

pair typesa

Taxon

LSM

LM

MM

MSM

MST SM

SSM

SST

Artemisia

U3-Antimony, Garfield

3

4

2

3

1 1

2

2

arbuscula

Co., Utah

A. bigelovii

U9-Flagstaff, Coconino
Co., Arizona

2

2

5

3

1 2

1

2

A. bigelovii

U13-Chinle, Apache Co.,
Arizona

3

2

5

3

1 2

2

A. cana

U3-Strawberry Reservoir,
Wasatch Co., Utah

2

4

4

2

1 1

2

2

A. nova

U16-John's Valley,
Garfield Co., Utah

3

1

6

2

3

2

A. nova

U22-Steinaker Reservoir,
Uintah Co., Utah

3

1

5

3

2 2

1

1

A. pygmaea

Ul-John's Valley,
Garfield Co., Utah

2

2

6

4

1

1

2

A. rigida

U5-Elder Flat, Monow
Co., Oregon

2

1

8

3

2

2

A. tridentata

U75-Wingate Mesa,

2

4

4

2

2

2

2

tridentata

San Juan Co., Utah

A. tridentata

U89-Steptoe Valley,

2

4

4

2

2

2

2

tridentata

White Pine Co., Nevada

A. tridentata

U14-Colton, Utah Co.,

2

2

6

3

2

1

2

vaseyana

Utah

A. tridentata

U23-Pinto Canyon, Wash-

2

2

6

3

2

1

2

vaseyana

ington Co., Utah

A. tridentata

U61-Milford, Beaver

2

2

8

3

1

1

1

wyomingensisCo., Utah

A. tridentata

U80-Jaeob's Lake,

2

3

6

3

2

2

wyomingensis Coconino Co., Arizona

Average

2.3

2.4

5.4

2.8

.9 1.2

1.2

1.8

Expected

2.0

2.0

6.0

4.0

0 2.0

0

2.0

aChromosome types: The first letter stands for chromosome size; L = large. M = medium, S = small. The other letters stand for centromere posi-
tion; M = metacentric, SM = submetacentric, ST = subterminal. Some of the LSM have an obvious secondary constriction, others do not.

236

GREAT BASIN NATURALIST MEMOIRS

No. 2

Table 3.

Establishment and survival of Artemisia tridentata wildi

ng transplants

at the Snow Field Station, Eph-

raim, Utah.

Number of plants

Percent

Initial

Established

Survivors

Repro-

Subspecies

Collection data

(1969-1970)

(1971)

(1976)

1976/1971

Vigor3

ductionb

Tridentata

U44-Loa, Wayne
Co., Utah

33

17

11

64.7

4

7

U74-Dove Creek,

40

31

25

80.6

9

10

Dolores Co., Colorado

U75-Wingate Mesa,

47

36

22

61.1

7

4

San Juan Co., Utah

U76-Clear Creek Can-

44

33

21

63.6

6

5

yon, Sevier Co., Utah

U79-Dog Valley,

46

46

44

95.7

10

10

Juab Co., Utah

U82-Brush Creek,

25

21

14

66.7

7

7

Uintah Co., Utah

Total
Mean

Vaseyana

Total

Mean

Wyomin-
gensis

Total
Mean

184

U13-Indian Peaks
Beaver Co., Utah
U14-Colton, Utah Co.,
Utah

U15-Sardine Canyon,
Cache Co., Utah
U19-Salina Canyon,
Sevier Co., Utah
U23-Pinto Canyon,
Washington Co., Utah
U24-Clear Creek
Canyon, Sevier Co., Utah
U31-Durkee Springs,
Sevier Co., Utah
U52-Alton, Kane Co.,
Utah

U72-Petty Bishop's
Log, Sanpete Co., Utah

U17-Milford,
Beaver Co., Utah
U37-Evanston, 23
Uinta Co., Wyoming
U58-Trough Springs,
Humboldt Co., Nevada
U80-Jacob's Lake,
Coconino Co., Arizona

41

36

32

124

74.4

6.1

7.2

41

24

58.6

6

7

65

45

69.2

8

3

24

20

83.3

5

7

9

7

77.8

3

4

28

26

92.9

7

7

.34

12

35.3

4

3

28

14

50.0

7

2

7

3

42.9

3

6

425

335

239

71.3

5.7

5.2

41

15

10

66.7

7

4

23

8

4

50.0

8

2

63

26

14

53.8

5

3

42

22

17

77.3

8

4

169

71

45

7.0

3.2

"Vigor rating is an ocular estimate of the accession's general condition 1 =
with little or no dieback.

""Reproduction rating is an ocular estimate of the accessions reproduction:
dant and of several age classes.

5 = average vigor with some dieback; 10 =
:; 5 = seedlings common but even aged; 10 •

exceptional vigor
= seedlings abun-

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

237

(1953), and Beetle (1960) all recognized the
distinctness of the North American forms,
although each treated them in a different
manner. We believe there are several com1
pelling reasons for separation, which we
will discuss.

Members of Seriphidium are Eurasian
(and North African). Their center of distri-
bution is Central Asia (Fig. 2). The present
distribution suggests an evolutionary radi-
ation from the areas of greatest species di-
versity. On the basis of karyotypic, anato-
mical, and morphological evidence, Filatova
(1974) supported the view that the most
primitive members of the Seriphidium occur
in arid areas in Kazakhstan, USSR and adja-
cent areas in Central Asia. She suggested an
annual /biennial herbaceous species (A. leuc-
odes) and a perennial shrub (A. junceus) as
being ancient primitive types. Musaev
(1965) gave detailed distribution maps of
five species of Seriphidium, which indicate
that the northern limit for these species is
about 52° -53° N. latitude. Musaev con-
tended that the Turan flora, composed in
part of these Seriphidium members, is
southern and Mediterranean and not strong-
ly connected to the northern (grass-forb)
steppes. This body of evidence suggests that
Eurasian Seriphidium originated in Central
Asia and radiated outward. Few members of
the Seriphidium are found in northerly lati-
tudes. The most widespread northern Old
World species, A. martima (sensn lato)
(Polyakov 1961, Clapham 1962), may have
been able to migrate north in milder coastal
habitats. The North American Tridentatae
and at least two other Artemisia species re-
main problematical— A. pahneri and A. men-
dozana. Artemisia palmeri is largely her-
baceous, is endemic to coastal badlands on
both sides of the California-Baja California
border, and little resembles the Tridentatae.
Rydberg (1916) created a new genus for it—
Artemisiastrum. Artemisia mendozana, pre-
sumably from west central Argentina, is
little known, but Artemisia, in general, is
not prominent in South America (Bonpland
et al. 1820, Harvard Univ. 1968, Heusser
1971, Good 1974). It was listed as a mem-
ber of the Seriphidium by de Candolle
(1837); Kawatani and Ohno (1964) indicated

that it was an octoploid (2n = 72).

Then, the fact that the Seriphidium and
Tridentatae are advanced taxa, occurring in
widely disjunct areas with no discovered
pollen record in Artemisia profiles of the
Beringia (Northeast Siberia-Alaska) area,
further supports their taxonomic separation.
Artemisia is a phylogenetically advanced
genus in an advanced family (Compositae)
(Stebbins 1974). Present consensus (Carlquist

1966, Cronquist 1968, Stebbins 1974) sup-
ports a woody ancestral prototype for the
Compositae. However, woodiness in Arte-
misia is probably secondary. Most woody
Artemisia species are bunched in Seriphi-
dium and Tridentatae, which are specialized
(reduced) in floral characteristics (Hall and
Clements 1923). Anatomical characteristics
of Tridentatae stems suggest an herbaceous
ancestry (Diettert 1938, Moss 1940, Stebbins
1974). The Seriphidium do not presently oc-
cur in northeastern Siberia nor the Triden-
tatae occur in Alaska or adjacent areas of
Canada (Fig. 1, 2). On the other hand,
members of both the subgenera, Dracun-
culus and particularly Artemisia (sensu lato),
grow in those areas and span, or nearly
span, Beringia (Rydberg 1916, Mizushima
1972, Korobkov 1972, Yurtsev 1972). No
palynological evidence has been found to
indicate that Seriphidium or Tridentatae were
ever in Beringia (Hopkins 1967, Graham
1972). Most pollen identification of Ber-
ingian Artemisia is only to the genus
level; those samples that are identified to
species are from the subgenus Artemisia.
Furthermore, paleobotanists infer from pres-
ent distributions and climates that Artemisia
pollen from Quaternary deposits came from
herbaceous species of the subgenus Arte-
misia (Colinvaux 1967, Johnson and Packer

1967, Kind 1967). Although Artemisia is
fairly common in Beringia, it is apparently
not as prevalent as it was earlier during
Quaternary time and reflects a harsher pres-
ent day climate (Colinvaux 1967). Woody
Chenopodiaceae of the same genera, but
different species, often grow in association
with the Tridentatae in North America and
the Seriphidium in Central Asia (Musaev
1965, Blauer et al. 1976). Stebbins (1974)

238

GREAT BASIN NATURALIST MEMOIRS

No. 2

believes that both chenopod shrubs and Ar-
temisia shrubs of the Tridentatae and Seri-
phidium evolved from herbaceous ancestors
on xeric "islands" of the Arcto-Tertiary For-
est. In the case of Artemisia, the principally
herbaceous, florally primitive subgenus Ar-
temisia probably served as parental stock.

The limited data so far available indicate
that the basic karyotypes of Tridentatae and
Seriphidium are significantly different. Fila-
tova (1974) presented idiograms of two spe-
cies of Seriphidium which she considered to
be primitive (Fig. 3), but Polyakov (1961)
listed these taxa as being advanced. Cur-
rently, those two idiograms are the only
ones available for diploid Seriphidium. We
are preparing and will publish karyotypes
of several Tridentatae taxa (McArthur,
Pope, and Plummer, in preparation). Infor-
mation presented here is tentative, coming
as it does from small samples. Although
there is variation throughout the section
(Table 1), it appears that one basic genome
is present which (with small changes) ac-
counts for the chromosomal variation in the
Tridentatae. Further evidence for one basic
genome in the Tridentatae is the high in-
cidence of multivalent formations in poly-
ploid populations (McArthur and Pope
1975, McArthur, Pope, and Plummer, in
preparation). Generally, diploid Tridentatae
have a pair of large submetacentric chromo-
somes which may have a secondary con-
striction in the long arm (Fig. 3), a pair of
large metacentric chromosomes, three pair
of medium metacentric chromosomes, two
pair of medium submetacentric chromo-
somes, a pair of small metacentric chromo-
somes, and a pair of small subterminal
chromosomes. In some cases (Table 1) a
pair of medium subterminal chromosomes
replaces the small metacentric chromo-
somes. The karyotype of tetraploid Triden-
tatae would appear to be an approximate
doubling of the diploid karyotype (Table 2).

Karyotype evolution in several Compos-
itae taxa often leads to: (1) shortening of
absolute chromosome length, (2) less ho-
mogeneity in chromosome lengths within
the complement, and (3) rearrangements of
chromosomal materials resulting in terminal
centromere positions (Stebbins 1953, Huzi-

wara 1962, Dejong 1965, Anderson 1966,
1970). Based on these criteria, both the Se-
riphidium and Tridentatae have advanced
karyotypes as their morphologies suggest.
Unpublished data (McArthur, Shrub Sci-
ences Laboratory) indicate some American
members of the subgenera Dracunculus and
Artemisia have simpler karyotypes.

Chemotaxonomic studies have shown that
section Tridentatae is a more or less natural
group (Hanks et al. 1973, Kelsey et al.
1973, Geissman and Irwin 1974). Seriphi-
dium members are probably less homo-
geneous owing to their greater numbers and
wider distribution. However, a particular
sesquiterpene lactone, the anthelminic san-
tonin, occurs widely in the old world Seri-
phidium (Polyakov 1961), but santonin has
not been discovered in any of the Triden-
tatae (Geissman and Irwin 1974).

The Tridentatae are all woody shrubs,
whereas about 15 percent of the Seriphi-
dium are herbaceous and others are semi-
shrubs with woody bases and herbaceous
tops (Hall and Clements 1923, Polvakov
1961).

Origin of the Tridentatae

Artemisia is much more richly developed
in Eurasia than it is in the Western Hemi-
sphere. Consensus places the origin of the
genus in Eurasia (Hall and Clements 1923,
Beetle 1960, Stebbins 1974). Some workers,
for example, Beetle (1960), believe that the
section Tridentatae migrated to North
America in a differentiated form. We be-
lieve, however, that the North American
Tridentatae differentiated in situ from her-
baceous ancestors in the subgenus Artemisia.
This subgenus is circumpolar at present, and
evidence indicates that it was formerly even
more prevalent in the northern latitudes
(Colinvaux 1967, Estes 1968). Hall and Cle-
ments (1923) pointed out that A. bigelovii
shared traits found both in Abrotanum (Ar-
temisia) and Seriphidium (Tridentatae).
They thought A. bigelovii or A. bigelovii-
like stock gave rise to the Tridentatae in
the Southwest during the Pleistocene. They
believed that with subsequent warming and
drying, Tridentatae ancestors migrated
north and developed into present-day taxa.

1978

INTERMOUNTAIN BIOGEOCRAPHY: A SYMPOSIUM

239

A study on the distribution of chromatogra-
phic patterns of phenolic constituents by
Hanks et al. (1973) supports this view.

In either case, the Tridentatae have only
been a prominent part of western American
flora since the late Tertiary or early Qua-
ternary periods (Axelrod 1950). The first Ar-
temisia pollen recorded in the Inter-
mountain Region is from Miocene Epoch
deposits (Tidwell et al. 1972). Tridentatae
progenitors probably were present on xerie
"islands" in or near the North American
Tertiary floras (Axelrod 1950). Alternating
moist and dry climates during the Pleisto-
cene Epoch favored the repeated evolution-
ary cycles that provided our present-day
Tridentatae. Aridity is a powerful evolution-
ary stimulus (Axelrod 1972, Stebbins 1974).

It is interesting that arid, interior conti-
nental climates may have produced inde-
pendently in western North America and
Central Asia analogous series of woody Ar-
temisia—the Seriphidium and the Triden-
tatae. An apparent similar example of con-
vergent evolution is found in woody
chenopod shrubs in the same two areas and
in Australia.

Management Implications in Section
Tridentatae

Although Tridentatae are recognized as a
natural group, gross differences among sec-
tion members have long been recognized.
Black sagebrush (A. nova) has been re-
garded as a desirable, palatable species, but
big sagebrush (A. tridentata) has been con-
sidered to be an unwanted range weed
(Cottam 1961). We now know there is
much variation within these taxa and that
palatable and unpalatable forms occur in
both (Stevens and McArthur 1974). In addi-
tion to the gross differences mentioned,
there are many differences of smaller orders
of magnitude. These last have led to differ-
ences of opinion among taxonomists.

Excluding A. palmeri, Rydberg (1916) rec-
ognized 13 species of Tridentatae, Hall and
Clements (1923) 4 species, Ward (1953) 7,
and Beetle (1960) 11. Each monographer
(except Rydberg) also recognized subspecific
taxa. Beetle (1960) and Beetle and Young
(1965) divided the Tridentatae into 11 spe-

cies, 9 subspecies, and 2 forms. Winward
(1970) has suggested an additional form. Al-
though we have learned much about sage-
brush in the past 60 years, we have much
to learn. Its ubiquitous western presence, its
value as a soil stabilizer, its habitat and
landscape value, and its attraction to the
public will demand attention for a long
time.

Range managers will benefit by learning
as much as they can about variation in
sagebrush. Characteristics of the subspecies
of big sagebrush listed in Table 4 illustrate
some differences in a group of closely re-
lated plants. It is difficult to separate these
taxa by any one character, but taken in
combination, each taxon is clearly distinct.

Such differences as adaptation, palatabi-
lity, and height in these as well as in other
sagebrush taxa should be taken into account
in management decisions. This is not to say
a manager should not make decisions on the
basis of readily available information. We
think that he should, but he should also re-
alize that the knowledge base is expanding.

Some broad aspects of shrub management
were discussed in the introduction. We
pointed out that some taxa are widely adap-
tive. In looking at shrub adaptation in a
narrower sense, it can be important for us
to consider sources of particular taxa indi-
vidually. On specific sites, different acces-
sions of the same plant taxon may perform
differently as do the subspecies of big sage-
brush (A. tridentata) in a uniform garden
(Table 3). Marchand et al. (1966) also ob-
served differential performance of big sage-
brush taxa in a uniform garden. As one
might expect, the same accessions may per-
form differently on different sites. Some
seem to have broader amplitudes of adapta-
tion than others. In a paired performance
(establishment, production, and reproduc-
tion) several accessions of Tridentatae in the
taxa, A. tridentata tridentata, A. nova, A.
cana, and A. bigelovii, showed differential
adaptation. Nine accessions performed
equally well at two sites in central Utah
(Snow Field Station near Ephraim and Gor-
don Creek Winter Game Range near Help-
er), but seven performed markedly better at
one or the other of the sites. When consid-
ering seed sources for a seeding, it is best to

240

GREAT BASIN NATURALIST MEMOIRS

No. 2

Table 4. Characteristics of subspecies of Artemisia tridentata.

Subspecies

Characteristic

tridentata

vaseyana

wyomingensis

References

Habitat and
range

Foothills and
valley floors.
610-2140 m.
British Columbia
and Montana to
Baja California
and New Mexico.

Foothills and
mountains.
915-3030 m.
British Columbia
and Alberta to
California and
New Mexico.

Foothills and
valley floors.
1520-2150 m.
Montana to
Arizona.

This paper;
Beetle and
Young 1965;
McArthur et al.
1974.

Smell

Bitter pungent

Pleasant

Bitter pungent

McArthur et al.
1974.

Leaf shape

Narrowly cuneate
to oblanceolate

Cuneate

Cuneate

Marchand et al.
1966; McArthur
et al. 1974;
McDonough et al.

1975.

Height

1.0-4.0 m

0.7-1.2 m

0.5-0.8 m

McArthur et al.
1974.

Sesquiterpene
lactone compounds

4-7

3-6

2

Kelsey et al.

1973.

Ultraviolet

visible

coumarins

Trace

Abundant

Trace

Shafizadeh and
Melinkoff 1970;
Stevens and
McArthur 1974;
Brown et al.
1975.

Tendency to
layer

None

Strong

None

Beetle and
Young 1965.

Palatability
to deer and
sheep

Low

High

Intermediate

Hanks et al. 1973;
Sheehy and
Winward 1976.

Protein content

High

Low

Low

Sheehy and
Winward 1976;
Welch et al. 1977.

Seed germination

High

Low

Intermediate

Harniss and
McDonough 1976.

2n

18,36

18,36

36

Ward 1953; Taylor
et al. 1964;
Winward 1970; Kel-
sey et al. 1975;

McArthur, Pope, and
Plummer (in prep.)

Flower and seed
phenology

Late

Early

Late-
Intermediate

Marchand et al.
1966; Winward
1970; Hanks et al.
1973.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

241

obtain seed from a site as similar and as
close as possible to the site to be seeded. If
seed must be brought in from some dis-
tance, it is best to get it from a harsher cli-
mate or more northerly location (Plummer
et al. 1968, Plummer 1977). For ex-
ample, accessions of A. bigelovii from
nearby Emery County, Utah, perform much
better at the Snow Field Station and the
Gordon Creek Winter Range than those
from the more distant and southern San
Juan County, Utah, and Coconino County,
Arizona. Van Epps (1975) presented data
supporting a similar conclusion for fourwing
saltbush (Atriplex canescens). Although the
foregoing discussion is based on wilding
transplant experiments, we have noted dif-
ferential success of accessions of big sage-
brush on seeded areas.

The sagebrushes are a valuable resource,
which can be more valuable if managed
wisely. Wise management entails maintain-
ing productive natural populations, seeding
to improve depleted or disturbed ranges,
and thinning closed stands. When mixed with
grasses, forbs, and other shrubs sagebrush
adds beauty to the range, food and cover
for animals, and stability to the soil. More-
over, sagebrush variation should permit se-
lection of material for particular purposes,
such as palatable, high-protein forms for
mule deer winter ranges. We think that
such an opportunity exists, since putative
natural hybrids have been discovered (Ward
1953, Beetle 1960), and we have cytological
and artificial pollination data that hybridi-
zation occurs (McArthur and Pope 1975,
McArthur, Pope, and Plummer, in prepara-
tion). We predict that in the years ahead
the much-maligned sagebrush will be re-
garded with increasing favor by land man-
agers.

Acknowledgments

Federal funds for wildlife restoration
were provided through Project W-82-R. We
thank A. C. Blauer, B. C. Giunta, R. C.
Holmgren, G. L. Jorgensen, K. R. Jorgensen,
J. C. Pederson, C. L. Pope, R. Stevens, and
B. L. Welch for technical advice and/or as-
sistance. Ann Lund made translations from
Russian of portions of Polyakov (1961) and
Filatova (1974).

Note Added in Proof

Since this paper went to press, we
learned of a monographic study on the
cytotaxonomy of the Artemisia maritimia
complex of section Seriphidium (Persson, K.
1974. Biosystematic studies in the Artemisia
maritima complex in Europe. Opera Bot-
anica 35:1-188). Persson's karyotypes of
diploid A. santonicum are apparently more
primitive than those shown by Filatova
(1974). However, the A. santonicum karyo-
types are unlike the Tridentatae karyotype.

Literature Cited

Anderson, L. C. 1966. Cytotaxonomic studies in
Chrysothamnus (Astereae, Compositae). Amer. J.
Bot. 53: 204-212.

1970. The karyotype of Chrysothamnus par-

ryi ssp. parryi and its implications. Trans. Kan-
sas Acad. Sci. 72: 399-401.

Axelrod, D. I. 1950. The evolution of desert vege-
tation in western North America. Publ. Car-
negie Inst. Wash. 590: 215-306.

1972. Edaphic aridity as a factor in angio-

sperm evolution. Amer. Naturalist 106: 311-320.

Beetle, A. A. 1960. A study of sagebrush, the sec-
tion Tridentatae of Artemisia. Bull. Univ.
Wyoming Exp. Sta. 368: 1-83.

Beetle, A. A., and A. Young. 1965. A third sub-
species in the Artemisia tridentata complex.
Bhodora 67: 405-406.

Besser, W. S. 1829. Synopsis absinthiorum. Bull.
Soc. Imp. Naturalistes Moscou 1: 219-265. Cited
by Hall and Clements (1923), see below.

Blauer, A. C, A. P. Plummer, E. D. McArthur, R.
Stevens, and B. C. Giunta. 1976. Char-
acteristics and hybridization of important inter-
mountain shrubs. II. Chenopod family. USDA
Forest Service Res. Pap. INT-177: 1-42.

Boissier, E. 1875. Flora orientalis. Vol. 3, Georg
Bibliopolam, Geneva.

Brown, D., R. O. Asplund, and V. A. McMahon.
1975. Phenolic constituents of Artemisia tri-
dentata ssp. vaseyana. Phytochemistry 14: 1083-
1084.

Candolle, A. P. de. 1838. Compositae. Prodromus
6: 1-687.

Carlquist, S. 1966. Wood anatomy of Compositae:
A summary with comments on factors con-
trolling wood evolution. Aliso 6: 25-44.

Clapham, A. R. 1962. Artemisia, pp. 858-862. In: A.
R. Clapham, T. G. Tutin, and E. F. Warburg,
Flora of the British Isles. 2d ed. Cambridge
University Press, Cambridge.

242

GREAT BASIN NATURALIST MEMOIRS

No. 2

Colinvaux, P. A. 1967. Quaternary vegetational his-
tory of arctic Alaska, pp. 207-231. In: D. M.
Hopkins (ed.), The Bering land bridge. Stanford
University Press, Stanford.

Cottam, W. P. 1961. Our renewable wildlands— a
challenge. University of Utah Press, Salt Lake
City.

Cbonquist, A. 1968. The evolution and classifica-
tion of flowering plants. Houghton Mifflin, Bos-
ton.
_. 1974. Discussion, pp. 140-143. In: G. Bendz,
and J. Santesson (eds.), Chemistry in botanical
classification. Proceedings of the twenty-fifth
Nobel Symposium. Academic Press, New York.

DeJonc, D. C. D. 1965. A systematic study of the
genus Astranthium (Compositae, Astereae).
Publ. Mus. Michigan State Univ., Biol. Ser. 2:
433-528.

Diettert, R. A. 1938. The morphology of Artemisia
tridentata Nutt. Lloydia 1: 3-74.

Estes, J. R. 1968. Cytotaxonomic studies in the Ar-
temisia ludoviciana polyploid complex of the
Pacific Northwest. Unpublished doctoral dis-
sertation, Oregon State University Library, Cor-
vallis.

Filatova, N. S. 1974. Kariosistematika dvukh vidov
polynei podroda Seriphidium (Bess.) Rouy.
[Karyotaxonomy of two species of wormwood of
the subgenus Seriphidium (Bess.) Rouy.] Izv.
Akad. Kaz. SSSR, Ser. Biol. 1: 16-20.

Geissman, T. A., and M. A. Irwin. 1974. Chemical
constitution and botanical affinity in Artemisia,
pp. 135-143. In: G. Bendz, and J. Santesson
(eds.), Chemistry in botanical classification. Pro-
ceedings of the twenty-fifth Nobel Symposium.
Academic Press, New York.

Good, R. 1974. The geography of flowering plants.
4th ed. Longman Group Limited, London.

Graham, A. 1972. Floristics and paleofloristics of
Vsia and eastern North America. Elsevier Pub-
lishing Co., Amsterdam, Netherlands.

Hall, H. M., and F. E. Clements. 1923. The phy-
logenetic method in taxonomy. The North
American species of Artemisia, Chrtjsothamnus.
and Atriplex. Publ. Carnegie Inst. Wash. 326: 1-
.355.

MxsKs. D. L., E. D. McArthur, R. Stevens, and A. P.
Pi i mmer. 1973. Chromatographic character-
istics and phylogenetic relationships of Arte-
misia section Tridentatae. USDA Forest Service
Res. Pap. INT-141: 1-24.

Harniss, R. O., and W. T. McDonough. 1976.
Yearly variation in germination in three sub-
species of big sagebrush. J. Range Managem.
29: 167-168.

Harvard University. 1968. Artemisia. Gray Her-
barium Index. 1: 605-616.

Heady, H. F. 1975. Rangeland management.
McGraw-Hill Book Co., New York.

Heusser, C. J. 1971. Pollen and spores in Chile.
University of Arizona Press, Tucson.

Hooker, W. J. 1840. Flora boreali-americana 1: 1-
351.

Hopkins, D. M. 1967. The Bering land bridge. Stan-
ford University Press, Stanford.

Humboldt, F. H. A. von, A. J. Bonpland, and C. S.
Kunth. 1820. Compositae. Voyage aux re-
gions equinoctiales du Nouveau Continent, fait
en 1799-1804, partie 6, Botanique. Sect. 3, Nova
genera et species plantarum 4: 1-312.

Huziwara, Y. 1962. Karyotype analysis in some
genera of Compositae. VIII. Further studies on
the chromosomes of Aster. Amer. J. Bot. 49:
116-119.

Johnson, A. W., and J. G. Packer. 1967.
Distribution, ecology, and cytology of the
Ogotoruk Creek Flora and the history of Ber-
ingia, pp. 245-265. In: D. M. Hopkins (ed.), The
Bering land bridge. Stanford University Press,
Stanford.

Kawatani, T., and T. Ohno. 1964. Chromosome
numbers in Artemisia. Bull. Natl. Inst. Hyg. Sci.
(Tokyo) 82: 183-193.

Kelsey, R. G., M. S. Morris, N. R. Bhadane, and F.
Shafizadeh. 1973. Sesquiterpene lactones of
Artemisia: TLC analysis and taxonomic signifi-
cance. Phytochemistry 12: 1345-1350.

Kelsey, R. G., J. W. Thomas, T. J. Watson, and F.
Shafizadeh. 1975. Population studies in Ar-
temisia tridentata ssp. vasetjana: Chromosome
numbers and sesquiterpene lactone races. Bio-
chem. Syst. Ecol. 3: 209-213.

Kind, N. V. 1967. Radiocarbon chronology in Si-
beria, pp. 173-192. In: D. M. Hopkins (ed.), The
Bering land bridge. Stanford University Press,
Stanford.

Korobkov, A. A. 1972. Ktsitotaksonomicheskoi kha-
rakteristike nekotorykh vidov roda Artemisia L.
Severo-vostoka SSSR. [Cytotaxonomical charac-
teristics of some species of the genus Artemisia
L. in northeastern USSR.] Bot. Zurn. (Moscow
& Leningrad) 57: 1316-1327. (Abstr. 64806 in
Biol. Abstr. 56(12). 1975).

Kuchler, A. W. 1964. Manual to accompany the
map, Potential natural vegetation of the con-
terminous United States. Amer. Geogr. Soc.
Spec. Publ. 36: 1-116.

Marchand, L. S., A. McLean, and E. W. Tisdale.
1966. Uniform garden studies on the Artemisia
tridentata Nutt. complex in interior British Co-
lumbia. Canad. J. Bot. 44: 1623-1632.

McArthur, E. D., B. C. Giunta, and A. P.
Plummer. 1974. Shrubs for restoration of
depleted ranges and disturbed areas. Utah Sci.
35: 28-33.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

243

McArthur, E. D., and Lorenzo Pope. 1975.
Genetic studies in section Tridentatae of Arte-
misia, pp. 164. In: H. C. Strutz (ed.), Wildland
shrubs: Proceedings of the symposium and
workshop. Brigham Young University Press, Pro-
ve

McDonough, W. T., R. O. Harniss, and R. B.
Campbell. 1975. Morphology of ephemeral
and persistent leaves of three subspecies of big
sagebrush grown in a uniform environment.
Great Basin Nat. 35: 325-326.

McGinnies, W. G. 1972. North America, pp. 55-66.
In: C. M. McKell, J. P. Blaisdell, and J. R. Goo-
din (eds.), Wildland shrubs— their biology and
utilization. USDA Forest Service Gen. Tech.
Rep. INT-1: 1-494.

McKell, C. M. 1975a. Shrubs and forbs for improve-
ment of rangelands, pp. 62-75. In: R. S. Camp-
bell, and C. H. Herbel (eds.), Improved range
plants. Society of Range Management, Denver.

1975b. Shrubs— a neglected resource of arid

lands. Science 187: 803-809.

McKell, C. M., J. P. Blaisdell, and J. R.
Goodin. 1972. Wildland shrubs— their biolo-
gy and utilization. USDA Forest Service Gen.
Tech. Rep. INT-1: 1-494.

Mizushima, M. 1972. Taxonomic comparison of vas-
cular plants found in western North America
and Japan, pp. 83-91. In: A. Graham (ed.), Flo-
ristics and paleofloristics of Asia and eastern
North America. Elsevier Publishing Co., Am-
sterdam, Netherlands.

Monsen, S. B. 1975. Selecting plants to rehabilitate
disturbed areas, pp. 76-90. In: R. S. Campbell,
and C. H. Herbel (eds.), Improved range plants.
Society of Range Management, Denver.

Moss, E. H. 1940. Interxylary cork in Artemisia
with reference to its taxonomic significance.
Amer. J. Bot. 27: 762-768.

Musaev, N. F. 1965. Northern distribution limits of
characteristic components of the Turan Desert
flora, pp. 84-129. In: A. I. Tolmachev (ed.), Dis-
tribution of the flora of the USSR (Translated
from the Russian in 1968 by the Israel Program
for Scientific Translations.)

Plummer, A. P. 1974. Morphogenesis and manage-
ment of woody perennials in the United States,
pp. 72-80. In: K. W. Kreitlow, and R. H. Hert
(eds.), Plant morphogenesis as the basis for sci-
entific management of range resources. USDA
Agric. Res. Serv. Misc. Publ. 1271: 1-232.

Revegetation of disturbed intermountain area

sites. In: J. L. Thames (ed.), Symposium: Re-
clamation of disturbed land in the southwest,
p. 302-339. University of Arizona Press, Tucson.

Plummer, A. P., D. R. Christensen, and S. B.
Monsen. 1968. Restoring big game range in

Utah. Utah Division of Fish and Game Publ.
68-3: 1-183.

Polyakov, P. P. 1961. Artemisia. In: V. L. Komarov
(ed.), Flora USSR 26: 425-631.

Rydberc, P. A. 1916. Artemisia and Artemisiastrum.
N. Amer. Fl. 34: 244-285.

Shafizadeh, F., and A. B. Melinkoff. 1970.
Coumarins of Artemisia tridentata ssp. vaseyana.
Phytochemistry 9: 1311-1316.

Sheehy, D. P., and A. H. Winward. 1976. Relative
palatability of seven Artemisia taxa to mule
deer. In: Abstracts of papers, 29th anuual meet-
ing of the Society for Range Management, pp.
60.

Stebbins, G. L. 1953. Chromosomes and phytogeny
in Compositae-Cichorieae. Univ. Calif. Publ.
Bot. 26: 401-429.

1974. Flowering plants: Evolution above the

species level. Belknap Press of Harvard Univer-
sity Press, Cambridge.

Stevens, R., and E. D. McArthur. 1974. A simple
field technique for identification of some sage-
brush taxa. J. Range Managem. 27: 325-326.

Taylor, R. L., L. S. Marchand, and C. W.
Crompton. 1964. Cytological observations on
the Artemisia tridentata (Compositae) complex
in British Columbia. Canad. J. Genet. Cyto. 6:
42-45.

TlDWELL, W. D., S. R. RUSHFORTH, AND D.

Simper. 1972. Evolution of floras in the In-
termountain Region, pp. 19-39. In: A.
Cronquist, A. H. Holmgren, N. H. Holmgren,
and J. L. Reveal, Intermountain Flora: vascular
plants of the Intermountain West. Vol. 1. Haf-
ner Publ. Co., New York.

Van Epps, G. A. 1975. Winter injury to fourwing
saltbush. J. Range Managem. 28: 157-158.

Ward, G. H. 1953. Artemisia, section Seriphidium
in North America. Contr. Dudley Herb. 4: 155-
205.

Welch, B. L., E. D. McArthur, and B. C.
Giunta. 1977. Variation of crude protein
among sources of Artemisia tridentata grown in
uniform gardens. In: Abstracts of papers, 30th
annual meeting of the Society for Range Man-
agement, p. 11-12.

Winward, A. H. 1970. Taxonomic and ecological
relationships of the big sagebrush complex in
Idaho. Unpublished doctoral dissertation, Uni-
versity of Idaho Library, Moscow.

Yurtsev, B. A. 1972. Phytogeography of north-
eastern Asia and the problem of transberingian
floristic interrelations, pp. 19-54. In: A. Graham
(ed.), Floristics and paleofloristics of Asia and
eastern North America. Elsevier Publishing Co.,
Amsterdam, Netherlands.

APPLYING BIOGEOGRAPHIC PRINCIPLES TO RESOURCE MANAGEMENT:
A CASE STUDY EVALUATING HOLDRIDGE'S LIFE ZONE MODEL

James A. MacMahon1 and Thomas F. Wieboldt1-2

Abstract.— Conservation schemes require some portioning of the world into a discrete number of subunits
which might benefit from similar management practices. The Holdridge Life Zone scheme, proposed for such
purposes previously, is applied to Utah as an initial test of its usefulness. A hypothetical life zone map was devel-
oped entirely on the basis of weather station data and data derived from these values. This predictive map com-
pared favorably to data on patterns of vegetation, mammal, reptile, and amphibian distribution in Utah.

Since the publication of MacArthur and
Wilson's (1967) analytical approach to the
biogeography of islands, several authors
have attempted to apply island bio-
geography theory to the problem of opti-
mum size and number of nature reserves.

Several aspects of theory have been in-
voked including species/ island area relations
(Diamond 1975, Sullivan and Shaffer 1975,
Wilson and Willis 1975) and extinction rates
(Terborgh 1974). Not all workers agree that
this approach is correct or robust (Sim-
berloff and Abele 1976). [See also rebuttals
to Simberloff and Abele, e.g., Diamond
(1976)]. Simultaneously the theories of is-
land biogeography are being changed or re-
stated in verifiable formats (Simberloff
1976).

Currently economics (Myers 1976) and
even international cooperation (Franklin
1977) are being included as factors jn na-
ture reserve establishment. Despite all of
the interest in the application of theory to
practical problems there are few attempts
to provide a reasonable plan to test the effi-
cacy and appropriateness of the implied
theory /practice merger.

Recently, without judging whether or not
theory was adequate or correct, MacMahon

(1977a) suggested a series of concrete steps
one might opt to test the various hypoth-
eses. It was hoped that this would be appli-
cable on a world scale.

In summary the suggestion was to deter-
mine the most sensitive species to be pre-
served (i.e., rarest, largest, highest trophic-
level, most vagile, etc.). This species deter-
mines the minimum size of the reserve area.
One might also, for a community, use the
species area relationship (MacArthur and
Wilson 1967). That relationship is

S = CAZ

where S = number of species, A = island
area, C = parameter characteristic of a
ticular taxon and archipelago and z = pow-
er derived by assuming species abundances
are distributed in a log normal manner
(Preston 1962). The value of C is a commu-
nity coefficient and not known for most
communities. One must also invoke consid-
eration of extinction rates for determining
the number of replicate areas necessary
(Terborgh 1974).

MacMahon pointed out that if the total
world variety of communities was reduced
into some small grouping to determine
functions like C and determine whether or
not there is similar value of C for most ex-

'Department of Biology and Ecology Center, Utah State University, Logan, Utah 84322.
'WieDoldt reduced some data for this paper. He did not prepare the manuscript and should

be held accountable for the content.

245

246

GREAT BASIN NATURALIST MEMOIRS

No. 2

amples of communities within one of these
groups, then the conservation process could
proceed more rapidly and be applied more
broadly.

The suggestion was made that a possible
system of classification, often used but sel-
dom tested, was that of Holdridge (1947,
1967). Holdridge's scheme uses mean annual
biotemperature (the sum of mean monthly
temperature above 0 C divided by 12),
mean annual precipitation and potential
evapotranspiration to place localities on a
chart divided into hexagons representing
vegetation types (Fig. 1). Corrections are
made for altitudinal differences at similar
latitudes (Holdridge 1967, Sawyer and Lind-
say 1963, Tosi 1964) (Fig. 2).

The hexagonal areas are termed life
zones. When a specific site is located on
the Holdridge model, a trinomial designa-
tion corresponding to its latitudinal position,
altitudinal belt, and humidity province is
applied. If the station lies in the basal al-

titudinal belt, the latitudinal region alone is
used with the humidity province. An ex-
ample will illustrate this procedure. Logan,
Utah, lies at an elevation of 4,790 ft (1460
m), has a mean annual biotemperature of
10.0 C and a mean total annual precipi-
tation of 16.64 inches (423 mm). Using the
standard lapse rate of 6 C per thousand me-
ters, a mean annual sea level biotempera-
ture of 18.76 C is calculated corresponding
to the subtropical latitudinal region (Fig. 2).
The actual biotemperature of 10.0 C corre-
sponds to the montane altitudinal belt (Fig.
2). The lines for a 10.0 C biotemperature
and 423 millimeters precipitation intersect
within the steppe life zone (Fig. 1). The tri-
nomial life zone designation of Logan,
Utah, would be subtropical montane steppe.
The question is, do all communities of the
world falling into a particular life zone,
e.g., subtropical montane steppe, have sim-
ilar values of C and can they be treated in
the conservation sense in a similar way?

LATITUDINAL REGIONS
POLAR

£ \ :■'. i a i -*

ALTITUDINAL BELTS

\ \

64 00 32 00

\ \

SEMIPARCHED SUPERARID PERARID

\ \

PERHUMID SUPERHUMID

HUMIDITY PROVINCES

Fig. 1. Graphical model to assign life zone designations to localities based on three abiotic measurements (after
Holdridge 1967).

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

247

Note that we are aware that the evolution
of species, major taxa differences, and the
specific species-species interactions in the
various examples of particular life zones
will all dictate detailed differences from
place to place. Despite this, the question re-
mains—is there a least common denominator
of similarity such that not every community
in the world will have to be studied before
it can be conserved? If ecological laws
really exist this least common denominator
should be identifiable at some level of reso-
lution, albeit generalized.

The first step in answering the specific
question at hand is to ascertain whether or
not the Holdridge model is applicable si-
multaneously to plants and animals (commu-
nities), outside of the tropics and in an area
where sufficient climatic and distribution
data are available.

Since the few tests of the Holdridge mod-
el by people outside the Holdridge group

(see review Holdridge et al. 1971) suggest
its plausibility for vegetation (Lindsey and
Sawyer 1970, Sawyer and Lindsey 1963,
1971, Thompson 1966), tropical birds (Slud
1960, 1976), and tropical reptiles and am-
phibians (Heyer 1967), a more detailed
faunal and vegetational test seems appropri-
ate. We report here on an attempt to an-
swer the question whether life zones pre-
dicted on Holdridge's model do coincide
with the patterns of distribution of animals
and plants in nature to a significant degree.

Methods

Data from 68 weather stations, selected
from throughout Utah and having long cli-
matic records (i.e., a minimum of 30 year
averages of monthly temperatures and an-
nual precipitation based on 1931-1960
records) were the main data base to gener-
ate the "theoretical life zones" of Utah. Re-

BASAL BELT POSITION

0" 1.5" 3

fO L A^O LAr|_ _BOR EAL_ I _COO L TEMPE R_ATU R E_ _ _[ _W A RM_T E MP ERAJ U R E _ j_ _ _ SUB T R OjMCAL _TR°f ICAL

LATITUDINAL REGIONS (MEAN ANNUAL BIOTEMPERATURE AT SEA LEVEL. uO)

Fig. 2. Approximate guideline positions of latitudinal and altitudinal corrections to the Holdridge life zone sys-
tem. Data based on an average adiabatic lapse rate of 6°C/1000 m (after Holdridge 1967).

248

GREAT BASIN NATURALIST MEMOIRS

No. 2

cords of longer duration are available for
certain stations. In these cases averages
based on the total vears of record are used.

Records of shorter duration are used wher-
ever climatological data gaps occurred in
critical geographic areas.

Subtropical Lower
Montane Desert Scrub

["""I Subtropical Montane

Dry Scrub

Warm Temperate Montane
Moist Forest

Subtropical Lower
Montane Steppe

Subtropical Montane
Steppe

Warm Temperate Montane
Wet Forest

□ Subtropical Montane
Desert

Subtropical Montane
Moist Forest

Warm Temperate Subalpine
Moist Forest

F!^1 Subtropical Montane
Li^J Desert Scrub

Subtropical Montane
Wet Forest

□ Warm Temperate Subalpine
Wet Forest

Fig. 3. Predicted potential life zone map of Utah. Assignment of localities to Holdridge life zones derived en-
tirely from abiotic data based on United States Weather Bureau data corrected as described in the text.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

249

Temperature and precipitation data pro-
vide only a rough approximation of the life
zones because of topographic and phys-
iognomic complexity. Maps of available
heat, potential evapotranspiration, and nor-
mal annual precipitation from Jeppson et al.
(1968) provided information to correct for
these irregularities. Additionally, we had ac-
cess to precipitation data from the United
States Forest Service, Dixie National Forest,
snow survey data from the Soil Con-
servation Service, and data reported in the
literature (Price and Evans 1937). All of
these data were used to derive estimates of
values of the three factors necessary to de-
fine the hypothetical life zones of a particu-
lar locality.

The development of certain regression
relationships (Figs. 9, 10) permitted us to
take point data and derive yearly values
and thus estimate life zone boundaries
where few weather stations were available.
This was particularly valuable because Utah
weather stations are generally in the valleys.
The station data or calculated points were
plotted on a life zone chart (Fig. 1). Names
were applied to life zones using the altitu-
dinal and latitudinal corrections (Fig. 2).
Life zone designations were then used to
generate a theoretical life zone map (Fig.
3).

To determine whether or not the hypo-
thetical life zones coincided with existing or
potential life zones, several maps were gen-
erated or used.

The recent vegetation map of Cronquist
et al. (1972), interpreted in light of
Kiichler's (1964) map, was used as the vege-
tation standard (Fig. 4). Additionally maps
were prepared of conifer (Fig. 5) and some
semiarid shrub (Fig. 6) distributions. Conifer
distributions are taken from Little (1971).
Shrub distributions are based on field data,
unpublished dissertations, and analysis of
herbarium specimens at Brigham Young
University and Utah State University.

Maps of mammal biotic provinces (Fig. 7)
were adapted from Durrant (1952), while a
series of reptile and amphibian maps were

drawn (Fig. 8) based on data in Stebbins
(1966). These two groups of vertebrates
were used because they represent taxa of
relatively high habitat fidelity, relatively
low vagility (compared to birds), and well-
known distributions, and they represent a
range of physiologies (e.g., ectotherms vs.
endotherms).

Results and Discussion

The "hypothetical" life zone map derived
entirely from estimates of abiotic factors ac-
cording to the Holdridge (1967) scheme is
presented in Fig. 3.

A comparison of Fig. 3 to the Utah vege-
tation map (Fig. 4) reveals interesting sim-
ilarities and differences. The vegetation map
is more detailed in the Great Basin desert
(western one-half of the state) area and the
Canyonlands floristic section (southeastern
one-fourth) than the hypothetical map. In
mountainous regions (Montane Zone, Fig. 4)
the hypothetical map is more detailed.
These differences are merely matters of
scale in some cases and map basis in others.
Much of the Great Basin desert mosaic is
undoubtedly due to edaphic factors in the
Cronquist et al. (1972) presentation. Since
the vegetation map relies on species distri-
butions, shadscale (Atriplex confertifolia) vs.
sagebrush (Artemisia tridentata) commu-
nities are differentiated even though they
may be in similar climate zones.

In the montane areas the Cronquist map,
because of scale, recognizes one community
while the hypothetical map includes topo-
graphically related climatic changes and
predicts community differences. Similar rea-
soning explains the Canyonland area dis-
crepancies between the two maps.

An integrated interpretation of the maps
suggests a high degree of coincidence. The
creosote bush vegetation zone clearly coin-
cides with an isolated part of the Sub-
tropical Lower Montane Desert Scrub. The
Subtropical Montane Desert Scrub plus Sub-
tropical Montane Steppe clearly coincides
with the sagebrush plus shadscale vegetation

250

GREAT BASIN NATURALIST MEMOIRS

No. 2

I-" j

L I Absolute desert

Shadscale Zone

Montane
Zone

D

Water

Creosote
Bush Zone

Sagebrush Zone

Pinyon-Juniper
Zone

Fig. 4. Vegetation map of Utah redrawn from Cronquist et al. (1972) and modified slightly with reference to
held notes and the map of Kuchler (1964).

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

251

zones. The Canyonlands floristic section
coincides well with large sections of the
Subtropical Lower Montane Desert Scrub.

The Absolute Desert vegetation zone
coincides with Subtropical Montane Desert.

Finally there is high overlap of the aggre-
gated montane hypothetical life zones and
the Montane vegetation zone.

Analysis of conifer distributions (Fig. 5),
mainly montane forms, in conjunction with

Fig. 5. Distribution outlines of 10 conifers in Utah (adapted from Little 1971). Species include Abies concolor,
A. lasiocarpa, Juniperus communis, J. scopulorum, Picea engelmannii, P. pungens, Pinus contorta, P. flexilis, P.
ponderosa, Pseudotsuga menziesii.

252

GREAT BASIN NATURALIST MEMOIRS

No. 2

the hypothetical map (Fig. 3) shows surpris-
ingly close fits of species boundaries and
hypothetical life zone boundaries. This fit is
biologically more meaningful than the rec-

ognition of a single montane vegetation for
the purpose of predicting management
practices. It is well known that steep topo-
graphic gradients cause community bound-

Fig. 6. The distribution limits of 13, mainly semiarid woody plants with limits in southern Utah. Data derived
from unpublished theses, dissertations, and analysis of herbarium specimens and field notes. Species include: Ber-
beris fremontii, Encelia frutescens, Fraxinus anomala, Yucca augustissima, Yucca baccata, Lycium pallidum, L.
torreyi, Coleogyne ramosissima, Quercus undulata, Dalea fremontii, D. lanata, D. polyadenia, and D. thompsonae.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

253

aries to become discrete when they other-
wise form continua (Beals 1969). The
vegetation map is generalized beyond man-
agement usefulness in the area of moun-
tains.

The non-Great Basin desert semiarid
woody plants mapped (Fig. 6) approximate
the distribution of the Subtropical Lower

Montane Steppe plus Subtropical Lower
Montane Desert Scrub hypothetical life
zones. The Mojave desert section of Utah
(extreme southwestern corner) is not differ-
entiated in the hypothetical map from the
larger Canyonlands area of Subtropical
Lower Montane Desert Scrub.

Interestingly, the woody plant, mammal,

□

Middle Rocky
Mtn.

Southern Rocky
Mtn.

Colorado Plateau

Columbia Plateau

Great Basin

Northern Great
Plains

Fig. 7. Faunal areas of Utah based on mammal distributions. (Durrant 1952). Lines within a faunal area are
boundaries of one of Durrant's provinces or subcenters.

254

GREAT BASIN NATURALIST MEMOIRS

No. 2

reptile, and amphibian data (Figs. 6, 7, 8)
suggest no absolute differences, but rather a
gradual diminution of species numbers in
the case of animals and replacement of
shrub species. This is not to say that the use

of animal and plant species distributions
cannot be used to delimit the Mojave
desert. MacMahon (1977b) defines bound-
aries of all North American deserts based on
exactly such data.

Fig. 8. Distribution limits of reptiles and amphibians in Utah based on Stebbins (1966). A-boundaries of am-
phibians. B-boundaries of lizards. C-boundaries of snakes. D-analysis of the percentage of the entire Utah her-
petofauna occurring in any one area. Analysis based on a grid of 10 x 10 mi (16 x 16 km) cells.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

255

The point is that one must pick certain
species as indicators to rigidly define the
Mojave desert, but an analysis of the distri-
butions of all species in the desert commu-
nities shows gradual loss or replacement
over a broader area. The same is true for
the Great Basin desert. The gradual species
shifts and assignment of indicators is de-
tailed in MacMahon (1977b).

The Holdridge scheme separates the
Great Basin desert from the other semiarid
areas of the state by segregating western
Utah into Subtropical Montane Desert
Scrub plus Subtropical Montane Steppe. We
believe this differentiation is probably im-
portant, because western Utah communities
subjectively appear to us to be quite differ-
ent from those of the southeast corner de-
spite floral overlap. Thus the vegetation
map implies a higher degree of similarity
than we believe exists, particularly for man-
agement purposes. Again, animal boundaries
(Figs. 7, 8) are strongly coincident with sep-
aration of the Canyonlands area from the
western deserts.

All data suggest that the Holdridge

8186 - 0 5058X » 0 0578 >
D 5685 - 1 .0496 X <

scheme effectively predicts plant- and
animal-based life zone boundaries, if one
can ignore a few "indicator" species. These
data reinforce the general findings of others
mentioned previously that the Holdridge
scheme shows promise, but are more per-
suasive in that they use plants and animals
in a topographically complex area where
distributions are well known.

An interesting offshoot of these analyses
are shown in Figures 9 and 10. It is usually
assumed that long-term weather records
would be necessary to derive Holdridge's
climatic variables. Our data suggest, for
Utah, that mean annual biotemperature can
be predicted from mean annual temperature
(Fig. 9). This is not surprising, but it means
that one can take mean annual temperature
data from a weather station in a valley cor-
rected for altitude, etc., to get a corrected
mean annual temperature for a place dis-
tant from that station. Mean biotemperature
may be read directly from the curve or cal-
culated from the formulae. Additionally, for
Utah sites, a single measurement of snow
accumulation on one day a year (1 April,
Snow Water Equivalents) predicts annual
precipitation (Fig. 10). For Utah this high

Fig. 9. Relationship between mean annual biotem-
perature and mean annual temperature for Utah
weather stations.

Fig. 10. Relationship between the 1 April snow wa-
ter equivalent (SWE) and the annual precipitation for
localities in the Manti-LaSal National Forest having
snow courses studied by the United States Soil Con-
servation Service.

256

GREAT BASIN NATURALIST MEMOIRS

No. 2

correlation is due in part to the high per-
centage of precipitation (60 percent or
more) that comes as snowfall. A similar
relationship may be of no value in other
areas, but these results suggest that a person
might be able to develop various other cor-
rections specific to particular geographic
areas. The prospect of using simple data ex-
tends the potential use of the Holdridge
scheme. Additionally, since one only needs
values for two variables to assign an area to
a life zone hexagon, one may actually read
the approximate value of the unknown third
parameter off of the life zone model
(Thompson 1966).

The Holdridge life zone model does pre-
dict plant and animal aggregations for Utah
with reasonable accuracy. The next step to
reach the final goal of using this scheme to
portion the earth into a reasonable number
of subsections for management purposes is
to apply the scheme to another area con-
taining organisms not closely related to
those of Utah and then to determine wheth-
er or not a similar value of C obtains for
the unrelated areas falling into the same life
zone. That analysis must be left for another
place and time.

Acknowledgments

This work was made possible by the
US/IBP Desert Biome funded by the Na-
tional Science Foundation (Grant GB32139).
Donna Baranowski and Bette Peitersen
aided in manuscript preparation, and Linda
Finchum typed the manuscript. The United
States Forest Service, Dixie National Forest,
and Soil Conservation Service, Manti-LaSal
Snow Course Project, kindly permitted use
of their data.

Literature Cited

Reals, E. W. 1969. Vegetational change along al-
titndinal gradients. Science 165: 981-985.

Cronquist, A., A. H. Holmgren, N. H. Holmgren,
and J. L. Reveal. 1972. Intermountain
Flora: Vascular plants of the Intermountain
West. Vol. 1 Hafner Publ. Co., New York.

Diamond, J. M. 1975. The island dilemma: lessons
of modern biogeographic studies for the design
of natural reserves. Biol. Conserv. 7: 129-146.

1976. Island biogeography and conservation:

strategy and limitations. Science 193: 1027-1029.

Durrant, S. D. 1952. Mammals of Utah. Univ. Kan-
sas Publ. Mus. Natl. Hist. 6: 1-549.

Franklin, J. F. 1977. The biosphere reserve pro-
gram in the United States. Science 195: 262-
267.

Heyer, W. R. 1967. A herpetofaunal study of an
ecological transect through the Cordillera de
Tilaran, Costa Rica. Copeia 1967: 259-271.

Holdridge, L. R. 1947. Determination of world
plant formations from simple climatic data. Sci-
ence 105: 367-368.

1967. Life zone ecology. Rev. ed. Tropical

Science Center, San Jose, Costa Rica.

Holdridge, L. R., W. C. Grenke, W. H. Hatheway,
T. Liang, and J. A. Tosi. 1971. Forest envi-
ronments in tropical life zones— a pilot study.
Pergamon, New York.

Jeppson, R. W., G. L. Ashcroft, A. L. Hurer, G. V.
Skogerboe, and J. M. Bagley. 1968.
Hydrologic atlas of Utah. Utah Water Research
Laboratory, Utah State University, Logan.

Kuchler, A. W. 1964. Manual to accompany the
map. Potential natural vegetation of the
counterminous United States. Amer. Geogr. Soc.
Spec. Publ. 36: 1-116.

Lindsey, A. A., and J. O. Sawyer. 1970.
Vegetation-climate relationship in the eastern
United States. Proc. Indiana Acad. Sci. 80: 210-
214.

Little, E. L. 1971. Atlas of United States trees.
Vol. 1, conifers and important hardwoods.
USDA Forest Service Misc. Publ. 1146.

MacArthur, R. H., and E. O. Wilson. 1967. The
theory of island biogeography. Princeton
Monogr. Popul. Biol. 1: 1-203.

MacMahon, J. A. In press-a. Thoughts on the opti-
mum size of natural reserves based on ecologi-
cal principles. In: J. F. Franklin (ed.), Biospher-
ic reserves. USDA.

MacMahon, J. A. In press-b. North American
deserts. In: R. Perry (ed.), Deserts of the world.
Cambridge University Press, Cambridge.

Myers, N. 1976. An expanded approach to the
problem of disappearing species. Science 193:
198-202.

Preston, F. W. 1962. The canonical distribution of
commonness and rarity. Ecology 43: 185-215.

Price, R., and R. B. Evans. 1937. Climate of the
west front of the Wasatch Plateau in central
Utah. Monthly Weather Rev. 65: 291-301.

Sawyer, J. O., and A. A. Lindsey. 1963. The Hold-
ridge bioclimatic formations of the eastern and
central United States. Proc. Indiana Acad. Sci.
73: 105-112.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

257

1971. Vegetation of the life zones in Costa

Rica. Indiana Acad. Sci. Monogr. 2: 1-214.

Simberloff, D. S. 1976. Species turnover and equi-
librium island biogeography. Science 194: 572-
578.

Simberloff, D. S., and L. G. Abele. 1976. Island
biogeography theory and conservation practice.
Science 194: 285-286.

Slud, P. 1960. The birds of finca "La Selva", Costa
Rica: a tropical wet forest locality. Bull. Amer.
Mus. Natl. Hist. 121: 51-148.

1976. Geographic and climatic relationships

of avifaunas with special references to com-
parative distribution in the Neotropics. Smithso-
nian Contr. Zool. 212: 1-149.

Sullivan, A. L., and M. L. Shaffer. 1975.
Biogeography of the megazoo. Science 189: 12-
17.

Stebbins, R. L. 1966. A field guide to western rep-
tiles and amphibians. Houghton-Mifflin, Boston.

Terborgh, J. 1974. Preservation of natural diversity:
the problem of extinction prone species. Bio-
Science 24: 715-722.

Thompson, P. T. 1966. A test of the Holdridge
model in midlatitude mountains. Prof. Geogr.
18: 286-292.

Tosi, J. A., Jr. 1964. Climatic control of terrestrial
ecosystems: a report on the Holdridge model.
Econ. Geogr. 40: 173-181.

Wilson, E. O., and E. O. Willis. 1975. Applied
biogeography, pp. 522-534. In: M. L. Cody, and
J. M. Diamond (eds.), Ecology and evolution of
communities. Harvard University Press, Cam-
bridge.

1978

INTERMOl NTAIN BIOGEOGRAPHY: A SYMPOSIUM

259

SUBJECT AND SPECIES INDICES FOR
INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

Abajo Mountains, p. 61, 77.

Albion Mountains, p. 83.

Allopatric ranges, p. 71, 150, 153, 201.

Alpine biota, p. 11, 67, 72, 105, 109, 112.

Altitude, p. 29, 245.

Amphibians, p. 43, 245.

Angiosperms, p. 5, 6, 7.

Annual plants, p. 85, 87, 115, 170.

Aquarius Plateau, p. 65.

Arctic, p. 11, 105, 111, 112.

Arcto-Tertiary flora, p. 8, 9, 10, 238.

Area, p. 90, 91, 122, 209, 220, 245.

Arizona, p. 4, 43, 68, 105.

Ash Meadows, p. 140.

Asia, p. 7, 8, 237.

Aspen forest, p. 137.

Australia, p. 8.

Barriers, p. 17, 20, 28, 29, 30, 31, 32, 38,

47, 61, 63, 66, 71, 83, 84, 88, 107, 121,

122, 152, 167, 197, 217.
Baseline data, p. 72.
Beartooth Mountains, p. 105, 110, 111.
Beaver Dam Mountains, p. 50, 58, 73.
Bees, p. 13.

Beringia, p. 10, 13, 237.
Biogeoraphie regions, p. 106, 115, 119, 120,

134, 137, 202, 245.
Biogeography, p., 158, 209, 245.
Biotemperature, p. 255.
Bird dispersed plants, p. 96, 97.
Birds, p. 5, 12, 55, 89, 98, 137, 209, 220.
Bison, p. 12.

Black Pine Mountains, p. 122.
Blue Lake, p. 101.
Body size, p. 68, 69, 99, 215, 221.
Boreal biota, p. 31, 55, 59, 63, 71, 77, 137,

157, 209.
Bryce Canyon National Park, p. 83, 88, 93.
Burbank Hills, p. 122.
Burro, p. 12.

California, p. 13, 43, 55, 56, 57, 82.
Camel, p. 12.
Canada, p., 11, 161.
Carnivores, p. 222.
Carson Range, p. 212.

Cassia Mountains, p. 83.

Center of distribution, p. 7, 8, 237.

Charleston Peak, p. 105.

Chemotaxonomy, p. 238.

Chihuahua, p. 46.

Chile, p. 8.

Circumboreal flora, p. 11.

Clark Mountain, p. 98.

Climate, p. 3, 85, 88, 99, 108, 119, 130,

134, 144, 197, 245.
Clinal variation, p. 55, 64, 65, 66, 67, 68.
Coahuila, p. 46.
Colonization, p. 17, 31.
Color, p. 68.

Colorado, p. 4, 6, 44, 56, 57, 67, 68.
Colorado Plateau, p. 4, 43, 47, 57, 192, 198,

206, 255.
Colorado River, p. 17, 20, 27, 31, 33, 34,

38, 67, 139, 140.
Community structure, p. 141.
Competition, p. 37, 129, 152.
Confusion Range, p. 122.
Coniferous forest, p. 11, 137, 147, 221.
Continental climate, p. 3, 4, 7, 99.
Continental drift, p. 4, 18.
Craters of the Moon, p. 5.
Cretaceous, p. 4, 5, 6, 7, 9, 12, 105.
Death Valley, p. 20, 26, 32.
Deciduous trees, p. 8.
Deep Creek Mountains, p. 61, 65, 68, 72,

73, 77, 83, 92, 93, 111, 212.
Desatoya Range, p. 212.
Diamond Mountains, p. 212.
Dicotyledons, p. 6, 7.
Dinosaurs, p. 12.
Diptera, p. 13.

Disjunct ranges, p. 44, 81, 147, 153, 237.
Dispersal, p. 5, 17, 27, 71, 81, 93, 94, 96,

98, 111, 197, 199, 212.
Distance, p. 89, 90, 91, 93, 114, 200, 209.
Drainage basins, p. 18, 19, 20, 21.
East Humboldt Mountains, p. 108, 122, 133.
East Tintic Mountains, p. 83, 93.
Ecological role, p. 141.
Edaphic factors, p. 10, 89, 99, 109, 110,

260

GREAT BASIN NATURALIST MEMOIRS

No. 2

115, 116, 161, 164, 191, 197, 205, 207.
Elephant, p. 12.
Elevation, p. 11, 83, 85, 89, 90, 91, 122,

126, 220.
Elk Ridge, p. 61.
Elko, p. 140.
Ely, p. 140.

Endangered species, p. 24, 25, 37, 72.
Endemics, p. 13, 17, 22, 35, 36, 43, 55, 57,

81, 87, 91, 97, 105, 110, 111, 113, 171,

192, 204, 229.
Eocene, p. 4, 5, 6, 7, 12, 18.
Eurasia, p. 7, 8, 9, 12.
Europe, p. 13.
Evolution, p. 3, 8, 13, 36, 115, 161, 170,

229, 238.
Excelsior Mountains, p. 122.
Extinct species, p. 22, 24, 25, 98, 217, 221.
Extinction, p. 3, 12, 13, 17, 28, 31 32, 37,

39, 73, 98, 99, 209.
Fauna, p. 3.
Fidelity index, p. 199.
Fish, p. 17, 210, 223.
Fleshy fruits, p. 84, 94, 96.
Flora, p. 3, 81, 86.
Forbs, p. 96.

Fossils, p. 17, 25, 26, 28, 44, 55, 56, 166.
Frenchman Flat, p. 43.
Frisco Mountains, p. 61, 77.
Geologic history, p. 3, 4, 71.
Glaciers, p. 10, 20, 44, 69, 71, 106, 108,

161, 192.
Gondwanaland, p. 6, 9, 13.
Goose Creek Mountains, p. 122.
Goose Lake, p. 20, 33, 34.
Graminoides, p. 87, 96.
Grand Canyon, p. 17, 33.
Grant Range, p. 212.
Grasses, p. 3, 87, 96.
Grassland, p. 10, 73, 170.
Great Rasin, p. 3, 17, 19, 43, 45, 47, 49, 55,

57, 81, 119, 198, 209, 255.
Great Plains, p. 44, 55.
Great Salt Lake, p. 55.
Green River, p. 27.
Green River flora, p. 6.
Gymnosperms, p. 5.

Habitat diversity, p. 59, 63, 89, 119, 122,
125, 147, 197, 209, 215, 220.

Harney Rasin, p. 34.

Henry Mountains, p. 61, 77.

Herbaceous plants, p. 3, 7, 8, 98, 170.

Herbivores, p. 222.

Highland Mountains, p. 122.

Holdridge's Life Zones, p. 245, 246, 247.

Horse, p. 12.

House Range, p. 61, 77.

Humboldt River, p. 139, 140.

Hybridization, p. 8, 10, 11, 55, 58, 69, 153,

156, 161, 241.
Hymenoptera, p. 13.
Hypsithermal (Altithermal) period, p. 11,

20, 109, 111, 129, 167.
Idaho, p. 4, 43, 44, 56, 82, 140.
Immigration, p. 209.
Independence Mountains, p. 108.
Insects, p. 13.

Intercalary meristem, p. 12.
Introductions, p. 17, 37, 39, 72, 98.
"Islands," p. 5, 17, 55, 59, 72, 73, 81, 82,

88, 90, 105, 106, 161, 209.
Jarbidge Mountains, p. 72, 83, 97 108, 115.
Jemez Mountains, p. 44.
Jurassic, p. 5, 12, 13.
Kaibab Plateau, p. 83.
Karyotype, p. 161, 229, 230, 231, 238.
Kawich, p. 140.

Klamath River, p. 17, 20, 26, 33, 34.
Lake Ronneville, p. 20, 26, 32, 34, 71, 161.
Lake Idaho, p. 19, 24.
Lake Lahontan, p. 20, 26, 32, 34, 55, 71,

131, 161.
La Sal Mountains, p. 61, 77.
Lassen Volcanic National Park, p. 83.
Latitudinal effects, p. 126, 137, 142.
Laurasia, p. 6, 9, 13.
Lava, p. 4.

Lifeform of plants, p. 84, 87.
Longevity, p. 99.
Longitudinal effects, p. 126.
Madro-Tertiary flora, p. 8, 9, 10.
Mainlands (continents), p. 88, 90, 121, 215.
Mammals, p. 5, 11, 31, 55, 62, 89, 98, 209,

220, 245.
Management, p. 1, 2, 17, 39, 53, 72, 100,

101, 134, 229, 230, 239, 245.
Markagunt Plateau, p. 65.
Meadow Valley Wash, p. 38, 58, 139, 140.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

261

Mediterranean region, p. 8.

Merriam effect, p. 129.

Mexico, p. 7, 8, 9, 43, 46, 161.

Mineral Mountains, p. 61, 77, 122.

Miocene, p. 3, 4, 7, 10, 12.

Mojave biota, p. 202.

Mojave Desert, p. 7, 13, 38, 55, 58, 105,

125, 253.
Molluscs, p. 55.
Monitor Range, p. 122, 133.
Monocotyledons, p. 7.
Montana, p. 7, 68.
Montane biota, p. 11, 63, 203, 212.
Mount Timpanogos, p. 83, 93.
Nature reserves, p. 245.
Navajo Mountains, p. 61, 77.
Needle Range, p. 61, 77, 122.
Nevada, p. 4, 18, 43, 45, 55, 56, 65, 67.
New Mexico, p. 4, 43, 44, 68.
New World, p. 9.
Niche breadth, p. 28, 61, 96, 100, 200, 207,

222.
Normalizing selection, p. 201.
North America, p. 7, 8, 9, 12.
Oak woodland, p. 10, 11, 129, 137.
Olancha Peak, p. 105, 111.
Old World, p. 7, 8.
Oligocene, p. 4, 6, 7, 18, 105.
Oquirrh Mountains, p. 61, 65, 66, 68, 77,

212.
Oregon, p. 33, 34.
Overgrazing, p. 53.
Overkill, p. 3.
Owens River, p. 20, 26, 32.
Pack rat middens, p. 43.
Pahranagat, p. 140.
Pahrump, p. 140.
Paleocene, p. 6, 7, 105.
Palm, p. 5, 6.
Panamint Range, p. 212.
Parasites, p. 99.

Paunsaugunt Plateau, p. 65, 212.
Pavant Mountains, p. 65.
Permian, p. 13, 105.
Physiological adaptations, p. 116.
Phytogeography, p. 105.
Pilot Peak, p. 122, 212.
Pine Forest Mountains, p. 97, 101.
Pine Valley Mountains, p. 61, 77, 83, 93,

122, 198.
Pinyon-juniper woodland, p. 10, 43, 55, 57,

73, 100, 119, 137, 142, 213.
Pit River, p. 17.
Plant succession, p. 133.
Pleistocene, p. 3, 12, 44, 48, 55, 63, 71, 81,

221, 223.
Pleistocene pluvials, p. 17, 19, 43, 105.
Pliocene, p. 4, 7, 17, 38, 105.
Pluvials, p. 31, 71, 166.
Pollen record, p. 5, 240.
Pollinating insects, p. 3, 13.
Polyploidy, p. 10, 109, 240.
Population size p. 221.
Precipitation, p. 255.
Precipitation/evaporation ratio, p. 5, 10, 11,

28, 31, 71.
Predation, p. 37.
Primitive man, p. 3, 55.
Pyramid Lake, p. 55.
Quaternary, p. 56, 170.
Raft River Mountains, p. 61, 67, 68, 77, 83.
Range limits, p. 197.
Red Butte Canyon, p. 83, 93.
Refugia, p. 43, 48, 115.
Reptiles, p. 43, 245.
Riparian woodland, p. 137, 147, 205.
Roberts Creek Mountains, p. 212.
Rocky Mountains, p. 4, 11, 13, 17, 55, 99,

105, 108, 215.
Ruby Mountains, p. 20, 33, 72, 83, 87, 92,

93, 97, 105, 108, 110, 111, 115, 133, 212.
Sacramento Mountains, p. 44.
Sagebrush steppe, p. 55, 56, 73, 129.
Sagehen Creek, p. 83.
San Francisco Peak, p. 105, 108, 111, 113,

115.
Santa Rosa Mountains, p. 83, 97, 108.
Saprophytes, p. 99.
Schell Creek Mountains, p. 122, 212.
Sequoia-Kings Canyon National Park, p. 83.
Sevier River, p. 53.
Sheep Mountains, p. 212.
Shoshone Mountains, p. 122.
Shrubs, p. 8, 87, 96, 170, 229.
Sierra Nevada Mountains, p. 4, 17, 31, 55,

57, 67, 81, 96, 99, 105, 108, 111, 212,

215.
Silene acaulis, p. 110, 112.

262

GREAT BASIN NATURALIST MEMOIRS

No. 2

Similarity indices, p. 17, 21, 26, 27, 95, 105,

113.
Snake Range, p. 72, 109, 212.
Snake River, p. 17, 20, 26, 31, 32, 33, 139.
Snake River Plain, p. 4, 17, 18, 38, 44.
Sonora, p. 46.
Source areas, p. 81, 90, 91, 92, 93, 115, 141,

164.
South Africa, p. 8.
South America, p. 8, 9, 12.
Southwest, p. 57.
Speciation, p. 138, 152.
Species-area curves, p. 17, 31 61, 81 88, 97,

122, 215, 245.
Species diversity, p. 31, 37, 56, 59, 81, 82,

87, 96, 101, 105, 122, 124, 125, 138, 139,

144, 220.
Species equilibrium, p. 37, 81, 98, 99, 134,

209, 211.
Spring Range, p. 83, 87, 91, 108, 111, 212.
Spruce Mountain, p. 133, 212.
Stansbury Mountains, p. 61, 77, 212.
Sticktights, p. 84, 94, 96.
Stream Capture, p. 27.
Subalpine biota, p. 67.
Sulphur Springs Range, p. 133.
Sympatric species, p. 36, 71, 150, 153.
Synecology, p. 119.
Tavaputs Plateau, p. 65.
Temperature inversions, p. 119, 131.
Tertiary, p. 4, 6, 7, 8, 9, 12.
Teton Mountains, p. 81.
Therapsid reptiles, p. 12.
Threatened species, p. 17, 23, 25.
Timberline, p. 11, 106.
Toana Mountains, p. 122.
Toiyabe Mountains, p. 83, 87, 93, 108, 110,

111, 122, 133, 140, 212.
Toquima Range, p. 110, 212.
Transition zone, p. 197.
Trees, p. 7, 87, 96, 99.
Triassic, p. 12, 105.
Trophic dynamics, p. 81.
Truckee River, p. 140.
Turnover rates, p. 223.
Turtle Range, p. 98.
Tushar Mountains, p. 65, 67, 122.
Uinta Mountains, p. 44, 61, 64, 66, 68, 83,

212.

Ungulates, p. 12.

Utah, p. 4, 6, 43, 50, 51, 55, 56, 67, 82,

245.
Utah Lake, p. 31.
Vagility, p. 209, 223.
Vegetation, p. 3, 72, 81, 119, 245.
Vicariant species, p. 17, 27, 36.
Virgin River, p. 33, 38, 47, 58, 68, 198.
Volcanic activity, p. 17, 18.
Vulnerable species, p. 17, 23, 24, 25.
Wah Wah Mountains, p. 61, 77.
Warner Mountains, p. 71, 83.
Wasatch Mountains, p. 4, 17, 19, 31, 38, 44,

55, 61, 64, 66, 68, 77, 81, 83, 93, 96.
Washington Co., Utah, p. 197.
Wheeler Peak, p. 72, 83, 87, 93.
White Mountains, p. 83, 108, 111, 113, 122,

212.
White Pine Mountains, p. 212.
White River, p. 33, 38.
Wind dispersal, p. 84, 94, 96.
Wyoming, p. 4, 43, 82.
Yellowstone region, p. 18, 83.
Yosemite National Park, p. 212.
Z-values, p. 17, 31, 89, 209, 216.
Zion Canyon, p. 58.

Abies bahamea, p. 11.
Abies concolor, p. 97, 98, 100, 148.
Abies lasiocarpa, p. 11, 86, 100.
Accipiter cooperi, p. 77.
Accipiter gentilis, p. 11.
Accipiter striatus, p. 77.
Acer grandidentatum, p. 86, 100.
Acipenser transmontanus, p. 22.
Acipenseridae, p. 22.
Aconitum columbianum, p. 86.
Acrocheilus alutaceus, 23.
Aegolius acadicus, p. 77.
Aeroncmtes saxatlis, p. 77.
Agastacke pallidiflora, p. 87.
Agelaius phoeniceus, p. 67.
Agropyron pringlei, p. 86.
Agropyron spicatum, p. 92.
AUenrolfca, p. 161.
Allium, p. 11, 86.
Amelanchier alnifolia, p. 92.
Amelanchicr utahensis, p. 92.
Androsace septentrionalis, p. 112.

1978

INTKHMOl NTAIN BIOOKOCHAPHY: A SYMPOSIUM

263

Aneides Jiardyi, p. 44.

Angelica seabrida, p. 87.

Antennaria marginata, p. 87.

Antcnnaria rosea, p. 111.

Antennaria soliceps, p. 87.

Anthemidae, p. 7.

Anthus spinoletta, p. 79.

Aphelecora eaerulescens, p. 66.

Aphelocoma coerulescens, p. 158.

Aplodontia rufa, p. 225.

Aquila chrysaetos, p. 77.

Aquilegia triternata, p. 87.

Archilochus alexandri, p. 78, 140.

Archoplites, p. 26.

Arctium, p. 85.

Arctostaphylos patula, p. 94.

Arenaria, p. 11.

Arenaria confusa, p. 87.

Arizona elegans, p. 48.

Artemisia, p. 7, 229.

Artemisia arbuscula, p. 92.

Artemisia douglasiana, p. 86.

Artemisia (section Seriphidium), p. 229.

Artemisia (section Tridentatae), p. 229.

Artemisia argilosa, p. 235.

Artemisia bigelovii, p. 233, 238, 241.

Artemisia cana, p. 235, 239.

Artemisia junceus, p. 237.

Artemisia leueodes, p. 237.

Artemisia hngiloba, p. 235.

Artemisia maritima, p. 237.

Artemisia mendozana, p. 237.

Artemisia nova, p. 239.

Artemisia palmeri, p. 237, 239.

Artemisia pygmaea, p. 233.

Artemisia rigida, p. 233.

Artemisia rothrockii, p. 235.

Artemisia santonicum, p. 241.

Artemisia tridentata, p. 7, 92, 100, 235, 239,

249.
Asio otus, p. 77.
Asteraceae, p. 7.
A.stereae, p. 7, 8.
Astragalus, p. 10.
Atriplex, p. 8, 161.
Atriplex acanthocarpa, p. 164.
Atriplex canescens, p. 161, 163, 164, 241.
Atriplex con ferti folia, p. 161, 163, 249.
Atriplex corrugata, p. 161, 164.

Atriplex cuneata, p. 161, 163.
A t rip lex fa lea ta, p . 163.
Atriplex gardneri, p. 161, 164
Atriplex garretH, p. 162, 164.
Atriplex hymenelytra, p. 164.
Atriplex navapensis, p. 162.
Atriplex obovata, p. 161, 164.
Atriplex polycarpa, p. 164.
Atriplex torreyi, p. 164.
Atriplex tridentata, p. 162.
Balsamorhiza, p. 7, 86.
Balsamorhiza macrophylla, p. 86.
Balsamorhiza sagittata, p. 86.
Bidens, p. 85.
Bonasa umbeuus, p. 77.
Boraginaceae, p. 8.
Brassicaceae, p. 8.
Bromus anomalies, p. 93.
Bromus breviaristatus, p. 86.
Bubo virginianus, p. 66, 77.
Ba/o fcoreas, p. 49, 50, 51.
Bufo microseaphus, p. 45.
Bafo punctatus, p. 45.
Bwfo ivoodhousei, p. 45.
Bufeo jamaicensis, 66, 77.
Calamagrostis scopulorum, p. 86.
Callisaurus draconoides, p. 46.
Caltha leptosepala, p. 93, 110.
Calyptridium umbellatum, p. 111.
Cardamine, p. 8.
Carduelis pinus, p. 79.
Care.r amplifolia, p. 86.
Carex aurea, p. 86.
Carex elynoides, p. 110.
Carex helleri, p. 111.
Carex lanuginosa, p. 86.
Carex pulvinata, p. 110.
Carex scopulorum, p. 110.
Carex tahoensis, p. 86.
Carpodaeus cassinii, p. 79.
Castilleja clokeyi, p. 87.
Castilleja linoides, p. 87.
Castilleja viscidula, p. 87.
Catharus fuscescens, p. 79, 140.
Catharus ustulatus, p. 79, 140.
Catostomidae, p. 22.
Catostomus ardens, p. 24, 36.
Catostomus catostomus, p. 24.
Catostomus clarki, p. 24, 36.

264

GREAT BASIN NATURALIST MEMOIRS

No. 2

Catostomus columbianus, p. 24.
Catostomus discobolus, p. 24, 36.
Catostomus fumeiventris, p. 24, 36.
Catostomus insignis, p. 24.
Catostomus latipinnis, p. 24.
Catostomus luxatus, p. 25.
Catostomus macrocheilus, p. 24, 36.
Catostomus occidentalis, p. 24.
Catostomus platyhrynchus, p. 17, 25, 27.
Catostomus sp., p. 24.
Catostomus tahoensis, p. 24, 36.
Catostomus warnerensis, p. 24, 36.
Ceanothus martini, p. 86, 93.
Ceanothus velutinus, p. 93.
Centurus uropygialis, p 140.
Ceratodes, p. 10, 161.
Certhia familiaris, p. 65, 78, 144, 148.
Chaenactis, p. 7.

Chamacbatiaria millefolium, p. 100.
Charina bottae, p. 49, 50, 51.
Chasmistes brevirostris, p. 25, 33.
Chasmistes cirius, p. 25.
Chasmistes liorus, p. 25.
Chasmistes sp. p. 25.
Cheilanthes gracillima, p. 86.
Chenopodiaceae, p. 8, 161.
Chordeiles minor, p. 67.
Chlorocrambe hastata, p. 86.
Chorizanthe, p. 170.
Chrysothamnus, p. 10.
Circaea, p. 85.
Cirsium clokeyi, p. 87.
Clematis columbiana, p. 86.
Clethrionomys gapperi, p. 225.
Cnemidophorus tigris, p. 48.
Coccyzus americanus, p. 140.
Coccyzus erythropthalmus, p. 140.
Colaptes auratus, p. 67, 140.
Colaptes cafer, p. 69.
Colaptes chrysoides, p. 67.
Coleogyne, p. 43.
Coleonyx varigatus, p. 46.
Coleoptera, p. 13.
Coluber constritor, p. 49.
Columba fasciata, p. 77.
Contopus borealis, p. 78.
Contopus sordidulus, p. 78.
Corallorhiza, p. 99.

Cottidae, p. 22.

Cottus bairdn, p. 17, 25, 27, 29, 36.

Cottus beldingi, p. 25.

Cottus eonfusus, p. 25.

Cottus echinatus, p. 25, 36.

Cottus extensus, p. 25, 36.

Cottus greenei, p. 26.

Cottus klamathensis, p. 26.

Cottus leiopomus, p. 26.

Cottus pitensis, p. 26.

Cottus princeps, p. 26.

Cottus rhotheus, p. 26.

Cottus tenuis, p. 26.

Crenichthys baileyi, p. 25, 36.

Crenichthys nevadae, p. 25, 36.

Crotalus cerastes, p. 48, 50.

Crotalus michelli, p. 48, 50.

Crotalus scutulatus, p. 48, 50.

Crotaphytus wislizeni, p. 46.

Cryptantha, p. 8.

Cryptantha mohavensis, p. 86.

Cyanocitta stelleri, p. 66, 78, 144, 157.

Cycloloma atripticifolium, p. 162.

Cymopterus nivalis, p. 87.

Cyprinidae, p. 22.

Cyprinodon breviradis, p. 25, 36.

Cyprinodon diabolis, p. 25.

Cyprinodon nevadensis, p. 25.

Cyprinodon radiosus, p. 25.

Cyprinodontidae, p. 22.

Cystopteris fragilis, p. 112, 115.

Dedeckera, p. 170.

Delphinium occidcntale, p. 93.

Dendragnpus obscurus, p. 65, 77, 157.

Dendrocopos pubescens, p. 140, 149, 157.

Dendrocopos scalaris, p. 140.

Dendrocopos villosus, p. 140.

Dendroica coronata, p. 79.

Dendroica gracial, p. 79, 148.

Deschampsia caespitosa, p. 110.

Dipsosaurus dorsalis, p. 46.

Dodecatheon jeffreyi, p. 111.

Dw/w. p. 110.

Draba arida, p. 87.

Draba lemmonii, p. 111.

Dryocopus pileatus, p. 78.

Dumetella carolinensis, p. 140.

Eleocharis montana, p. 87.

Ernpetrichthys lotos, p. 25, 36.

1978

INTERMOUNTAIN BIOGEOCRAPHY: A SYMPOSIUM

265

Empetrichthys merriami, p. 25, 36.

Empidonax difficilis, p. 78, 147, 150, 157.

Empidonax hammondii, p. 78, 147, 150.

Empidonax minimus, p. 152.

Empidonax oberholseri, p. 78.

Enceliopsis, p. 7.

Epilobium angustifolium, p. 86.

Eremichthys acros, p. 23.

Eremophila alpestris, p. 67, 78.

Erigeron, p. 8, 11.

Erigeron peregrinus, p. 110.

Erigeron ursinus, p. 86.

Erigeron ivatsoni, p. 87.

Eriogoneae, p. 186.

Eriogonoideae, p. 170.

Eriogonum, p. 10, 11, 110, 170.

Eriogonum gracilipes, p. 110.

Eriogonum hohngrenii, p. 87.

Eriogonum kingii, p. 87.

Eriophyllum, p. 7.

Eumeces skiltonianus, p. 50.

Eutamias alpinus, p. 225.

Eutamias amoenus, p. 225.

Eutamias dorsalis, p. 100, 221, 225.

Eutamias panamintinus, p. 225.

Eutamias quadrimaculatus, p. 225.

Eutamias quadrivittatus, p. 225.

Eutamias speciosus, p. 225.

Eutamias townsendii, p. 225.

Eutamias umbrinus, p. 100, 221, 225.

Ffl/co sparverius, p. 77, 140.

Festuca arizonica, p. 87.

Fritillaria atropurpurea, p. 86.

Fundulus spp., p. 25.

Gentiana, p. 11.

Geothlypis tohniei, p. 79.

Geothlypis trichas, p. 67.

Geranium fremontii, p. 93.

Geum macrophijllum, p. 86.

Geum rossii, p. 86, 110.

Gi/fl alvordensis, p. 23, 36.

Gifa atraria, p. 23, 36.

Gifo fcico/or, p. 17, 23, 27, 28,33, 36.

Gi/a coerulea, p. 23.

Gi/a copei, p. 23.

Gt7a cj/p/ia, p. 23, 33.

Gi/a elegans, p. 23, 33.

Gi'/ia, p. 8.

Gilmania, p. 170.

Glaucidim gnoma, p. 77.

Glaucomys sabrinus, p. 225.

Glyceria elata, p. 86.

Goodmania, p. 170.

Gopherus agassizi, p. 46.

Grazia, p. 161.

Guiraca caerulae, p. 140.

Hackelia floribunda, p. 86.

Haplopappus, p. 7, 10.

Harfordia, p. 170.

Heliantheae, p. 7, 10.

Hesperiphona vespertina, p. 148.

Hesperoleucus symmetricus, p. 23.

Hierochloe odorata, p. 86.

Holbrookia maculata, p. 47.

Hollisteria, p. 170.

Holodiscus dumosus, p. 93.

Hu/sea, p. 11, 86.

Hidsea algida, p. 111.

Hydrophyllaceae, p. 8.

Ht//a arenicolor, p. 45.

Hymenoxys, p. 11.

Icterus bullockii, p. 69.

Icterus cucullatus, p. 140.

Icterus galbula, p. 69, 140.

Icterus galbula, p. 140.

Iotichthys phlegethontis, p. 23.

Iridoprocne bicolor, p. 140.

Ivesia pygmaea, p. 111.

Junco caniceps, p. 70, 80, 157.

Junco hycmalis, p. 70, 80, 157.

Juniperus, p. 7, 85.

Juniperus osteosperma, p. 43, 93, 98, 119.

Kalmia polifolia, p. 94.

Koeleria cristata, p. 110.

Koenigia islandica, p. 112.

Lampetra lethophaga, p. 22.

Lampetra minima, p. 22.

Lampetra tridentata, p. 22.

Lampropeltis getulus, p. 48, 49.

La Rivers, I., p. 18, 22, 23.

Larrea, p. 4, 43.

Lastarriea, p. 170.

Lathyrus pauciflorus, p. 93.

Ledum glandidosum, p. 94.

Lepidium, p. 8.

Lepidomeda albivallis, p. 24, 36.

Lepidomeda altivelis, p. 24, 36.

266

GREAT BASIN NATURALIST MEMOIRS

No. 2

Lepidomeda mollispinis, p. 24, 36.
Leipidomeda nittata, p. 24.
Leptotyphlops humilis, p. 48.
Lepus americanus, p. 225.
Lepus townsendii, p. 221, 225.
Lesquerella, p. 8.
Leucosticte atrata, p. 79.
Leucosticte tephrocotis, p. 157.
Lewisia pygmaea, p. 111.
Lewisia rediviva, p. 93.
Libocedrus decurrens, p. 86.
Lithospermum, p. 8.
Lomatium, p. 11.
Lonicera involucrata, p. 86.
Lo.ri'fl curvirostra, p. 79.
Lupinus breweri, p. 111.
Marmota flaviventris, p. 221, 225.
Maries americana, p. 225.
Masticophis flagellum, p. 48.
Melospiza melodia, p. 140.
Mertensia arizonica, p. 86.
Mertensia toyabensis, p. 87.
Microtus longicaudus, p. 221, 225.
Microtus richardsoni, p. 225.
Mimulus torreyi, p. 86.
Moapa coriacea, p. 24.
Moldavica parviflora, p. 86.
Molothrus atcr, p. 66.
Mucronca, p. 170.
Muhlenbergia wrightii, p. 87.
Mustela erminea, p. 221, 225.
Myadestes townsendi, p. 79.
Mylocheilus caurinus, p. 24.
Mylopharodon, p. 26.
Myosotis, p. 11.
Myoxocephalus, p. 26.
Nemacaulis, p. 170.
Neotoma cinerea, p. 221, 225.
Scotoma lepida, p. 43.
Nicotiana attenuata, p. 94.
Nucifraga columbiana, p. 78, 97, 98.
Ochotona princeps, p. 221, 224, 225.
Oenothera caespitosa, p. 93.
Oncorhyncus tshawytscha, p. 22.
Opheodrys oemalis, p. 49.
Ophisaurus attenuatus, p. 44.
Opuntia charlestonensis, p. 87.
Oporonis tolmiei, p. 140.
( hobanche, p. 99.

Orthocarpus tolmiei, p. 86.

Orthodon, p. 26.

Oryzopsis kingii, p. 86.

Osmorhiza chilensis, p. 86.

Otws as/o, p. 67, 140.

Ofus flammeohis, p. 77.

Oxynrt rfigi/na, p. 110, 111, 112, 115.

Oxytheca, p. 170.

Pachistima myrsinites, p. 93.

Farws atricapillus, p. 66, 78, 140.

P«n/.s gambeli, p. 65, 78.

Passerella iliaca, p. 71, 140, 153.

Passerina amoena, p. 69.

Passerenia cyanea, p. 69.

Pedicularis, p. 11.

Penstemon, p. 10, 11.

Penstemon keckii, p. 87.

Peraphyllum ramosissirrium, p. 94.

Perisoreus canadensis, p. 78.

Petrochelidon pyrrhonota, p. 66.

Phacelia, p. 8.

Phainopepla nitens, p. 140.

Phenacomys intermedins, p. 225.

Plicucticus melanocephalus, p. 79.

Phippsia algida, p. 112.

P/i/o.v caespitosa, p. 111.

P/i/o.t covillei, p. 110.

P/i/o.v longifolia, p. 93.

Phrynosoma douglassi, p. 50.

Phijsaria, p. 8.

Picea engelmannii, p. 11, 97, 100.

Picea glauca, p. 11.

P/refl pungens, p. 11, 86, 100.

Picoides tridactylus, p. 78.

Picoides villosus, p. 66, 78.

Pinicola enucleator, p. 79.

Pinus albicaulis, p. 94, 97.

PinMS aristata, p. 94, 97, 98.

Finns banksiana, p. 11.

Pm?/.s contorta, p. 11.

P/m/.s edufcs, p. 86, 97.

Pinus flcxilis, p. 97, 98.

Pinus jefferyi, p, 86, 98.

Pnn/.s longaeoa, p. 109, 148.

/'/M//.s monophylla, p. 98, 129.

/'/m/.s ponderosa, p, 86, 94, 98, 100, 129,

148.
/'//)//(> flfoertt, p. 140.
Pipilo chlorura, p. 79.

1978

INTERMOUNTAIN BIOGEOGRAPHY: A SYMPOSIUM

267

Pipilo erythrophthalmus, p. 79.

Piranga ludoviciana, p. 79.

Piranga rubra, p. 140.

Plagiobothrys, p. 8.

Plagopterus argentissimus, p. 24.

Plethodon neomexicanus, p. 44.

Plethodontidae, p. 44.

Polemoniaceae, p. 8.

Polemonium eximium, p. 111.

Polygonaceae, p. 170.

Polygonum bistortoides, p. 110.

Polygonum lapathifolium, p. 98.

Polygonum minimum, p. 111.

Polygonum pensylvonicum, p. 98.

Polygonum viviparum, p. 112.

Pooecetes gramineus, p. 79.

Populus tremuloides, p. 86.

Populus trichocarpa, p. 86.

Potentilla heanii, p. 87.

Primula capillaris, p. 87.

Primula parryi, p. 86, 93.

Progne subis, p. 78.

Prosopium abyssicola, p. 23.

Prosopium gemmiferum, p. 23.

Prosopium prolixus, p. 23.

Prosopium spilonotus, p. 23.

Prosopium williamsoni, p. 17, 23, 27, 36.

Prunus emarginata, p. 86.

Pseudoacris triseriata, p. 49.

Pseudtsuga menziesii, p. 11, 100, 148.

Pterostegeal, p. 186.

Pterostcgia, p. 170.

Ptychocheilus lucius, p. 23.

Ptychochcilus oregonensis, p. 23, 33.

Purshia tridentata, p. 86.

Pyrocephalus rubinus, p. 140.

()uercu.s gambelii, p. 11, 86, 98, 100.

()j/erc?/.s' turbinella, p. 11.

fian<7 pretiosa, p. 49, 51.

Ranunculus, p. 11.

Ranunculus jovis, p. 93.

Rcgulus calendula, p. 79, 148.

Regulus satrapa, p. 79, 148.

Rclictus solitaries, p. 24, 27.

Rhinichthys cataractac, p. 17, 24, 27.

Rhinichthys falcatus, p. 24.

Rhinichthys oscidus, p. 17, 24, 27, 28, 36.

Rhinichthys sp. p. 24.

Rhniocheilus lecontei, p. 53.

fi//K'.v cereum, p. 86.

ftj'/w wolfii, p. 86.

Richardsonius balteatus, p. 17, 23, 27, 36.

Richardsonius cgrcgius, p. 23, 36.

Rorripa, p. 8.

Rubus pair i floras, p. 94.

Salicomia, p. 161.

Salix arctica, p. 110, 112.

Salmo apache, p. 22.

SaZmo cZarfci, p. 17, 22, 27, 28, 36.

Salmo gairdneri, p. 22.

Salmo sp., p. 22.

Sahnonidas, p. 22.

Salpinctes obsoletus, p. 79.

Salsola, p. 161.

Salvadora hexdepis, p. 48.

Salvelinus malma, p. 22.

Sarcobatus, p. 161.

Saxifraga, p. 11.

Saxifraga cespitosa, p. 112, 115.

Saxifraga flagellaris, p. 112.

Sayornis nigricans, p. 140.

Sclcroporus elongatus, p. 51.

Sclcroporus graciosus, p. 50.

Sclcroporus magister, p. 47.

Sedum rosea, p. 111.

Selaginella selaginoides, p. 87.

Selasphorus platycercus, p. 78, 157.

Senedo midtilobatus, p. 10.

Senecio streptanthifolius, p. 10.

Sequoiadendron giganteum, p. 86.

SiaZta currucoides, p. 79, 150.

Sifl/io mexieana, p. 79, 140, 150.

Sibbaldia procumbens, p. 110.

Silene clokcyi, p. 87.

Sitanion hystrix, p. 86.

Sitta canadensis, p. 78, 148.

Sitta carolinensis, p. 78.

Sifta pygmaea, p. 78.

Sonom semianulata, p. 48, 49.

Sorex /[/c//i, p. 225.

Sorex palustris, p. 221, 225.

Somv tenellus, p. 221, 225.

Sorex trowbridgei, p. 225.

Sorex vagrans, p. 221, 225.

Spermophilus armatus, p. 225.

Spermophilus beldingi, p. 221, 225.

Spermophilus lateralis, p. 221, 225.

Sphyrapicus ruber, p. 144, 157.

268

GREAT BASIN NATURALIST MEMOIRS

No. 2

Sphyrapicus thyroideus, p. 78, 147.
Sphyrapicus varius, p. 78, 147, 157.
Spinus pinus, p. 148.
Spizella passerina, p. 80.
Stanleya, p. 8.
Stellula calliope, p. 78.
Stenogonum, p. 170.
Stetophaga ruticilla, p. 140.
Stipa californica, p. 86.
Streptanthus, p. 8.
Sfn'x occidentalis, p. 77.
Suaeda, p. 161.

Sylvilagus nuttallii, p. 221, 225.
Synthyris ranunculina, p. 87.
Tachycineta bicolor, p. 78.
Tachycineta thalassina, p. 78.
Tamiasciurus douglasii, p. 225.
Tamiasciurus hudsonicus, p. 225.
Tanacetum compaction, p. 87.
Taxus brevifolia, p. 86.
Tegeticula, p. 3.

Thamnophis cyrtopsis, p. 48, 49.
Thalictntm fendleri, p. 86.
Thamnophis elegans, p. 49.
Thelypodium, p. 8.
Thermopsis montana, p. 86.
Thryomanes bewickii, p. 140.
Toxostoma dorsale, p. 140.
Trifolium andersonii, p. 86.
Tri folium monoense, p. 110.

Trisetum spicatum, p. 110, 112, 115.
Troglodytes aedon, p. 78, 140.
Tsuga mertensiana, p. 86.
Tardus migatorius, p. 79, 140.
Tyrannus tyrannus, p. 140.
Tyrannus verticalis, p. 140.
Urosaurus ornata, p. 47.
t/ta stansburiana, p. 47.
Valeriana occidentalis, p. 93.
Vermivora celata, p. 79, 140, 157.
Vermivora luciae, p. 140.
Vermivora ruficapilla, p. 157.
Vermivora virginiae, p. 79. 157.
Viguiera, p. 7.
Vio/a adunca, p. 86.
Virgo beUu, p. 140.
Virgo gilvns, p. 79, 140.
Virgo olivaceus, p. 140.
Virgo solitarius, p. 79, 157.
Wi/sonia pusilla, p. 79, 157.
Wyethia, p. 7.
Xanthium, p. 85.
Xyrauchen texanus, p. 25, 33.
Yucca, p. 3.
Zapus princeps, p. 225.
Zenaida asiatica, p. 140.
Zonotrichia iliaca, p. 80.
Zonotrichia leucophrys, p. 80.
Zonotrichia lincolnii, p. 80.
Zonotrichia melodia, p. 66, 67, 80.

NOTICE TO CONTRIBUTORS

Original manuscripts in English pertaining to the biological natural history of western
North America and intended for publication in the Great Basin Naturalist should be di-
rected to Brigham Young University, Stephen L. Wood, Editor, Great Basin Naturalist,
Provo, Utah 84602. Those intended for the Great Basin Naturalist Memoirs should be sim-
ilarly directed, but these manuscripts are not encumbered by a geographical restriction.

Manuscripts. Two copies of manuscripts are required. They should be typewritten,
double spaced throughout on one side of the paper, with margins of at least one inch on
all sides. Use a recent issue of either journal as a format, and the Council of Biology Edi-
tors Style Manual, Third Edition (AIBS 1972) in preparing the manuscript. An abstract,
about 3 percent as long as the text, but not exceeding 200 words, written in accordance
with Biological Abstracts guidelines, should precede the introductory paragraph of each ar-
ticle. Authors may recommend one or two reviewers for their article. All manuscripts re-
ceive a critical peer review by specialists in the subject area of the manuscript under con-
sideration.

Manuscripts that are accepted and that are less than 100 printed pages in length will
automatically be assigned to the Great Basin Naturalist. Those manuscripts larger than 100
printed pages in length will be considered for the Memoirs series.

Illustrations and Tables. All illustrations and tables should be made with a view toward
having them appear within the limits of the printed page. Illustrations that form part of an
article should accompany the manuscript. Illustrations should be prepared for reduction by
the printer to either a single-column (2 5/8 inches) or double-column (5 1/2 inches) width,
with the length not exceeding 7 1/2 inches.

Costs Borne by Contributor. Contributors to the Great Basin Naturalist should be pre-
pared to donate from $10 to $30 per printed page toward publication of their article (in
addition to reprint costs outlined in the schedule below). Authors publishing in the Great
Basin Naturalist Memoirs may be expected to contribute $35 per printed page in addition
to the cost of printed copies they purchase. No printed copies are furnished free of charge.
A price list for reprints and an order form are sent with the galley proof to contributors.

Reprint Schedule of the Great Basin Naturalist

2 pp.

4 pp.

6 pp.

8 pp.

10 pp.

12 pp.

Each

100 copies

$20

$24

$28

$32

$36

$40

additional

200 copies

28

32

36

40

44

48

2 pp.

300 copies

36

40

44

48

52

56

$4

Great Basin Naturalist Memoirs

No. 1 The birds of Utah. By C. L. Hayward, C. Cottam, A. M. Woodburrv, H. H. Frost.

$10.
No. 2 Intermountain biogeography: A symposium. By K. T. Harper, J. L. Reveal, et al. $15.

tfg^v

cflcme

Bookbinding Co., Inc.

100 Cambridge St.
Charlestown. MA 02129

3 2044 072 231 384

Date Due