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 In general, only previously unpublished manuscripts of less than 100 printed pages in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. Reveal 2

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. Smith 1

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 -

mi 2 xlO 3 .5
km2 x l03

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

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. Tanner 1

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 R 2 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 R 2 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,

N 2

N 3

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; N 2 = number of these permanently resident; N 3 = 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,

N 2

N 3

AR

WB

DM

EHP

LPH

HDS

1.000

N,

.001

.933

1.000

N 2

.001
.970

.001
.816

1.000

N 3

.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

R 2

R 2 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. R 2 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 Freeman 1 , W. Kent Ostler 1 , and Lionel G. Klikoff 2

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
(R 2 ). 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 (R 2 )

.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 (R 2 )

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 (R 2 )

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

r s = .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 r w -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 Mountains 2 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 km 2 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, IS S = 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. West 1 , Robin J. Tausch', Kenneth H. Rea 1 , and Paul T. Tueller 2

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 km 2 (125,000 (1972), Tueller (1975), and Young, Evans,

mi 2 ) 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

Mountain 1

Woodland

Height

Barrier

Score

Range

Encountered

(km 2 )

(km 2 )

(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 (km 2 )

Woodland
Area

(km 2 )

Width of
Barrier

(km 2 )

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 (km 2 )

Woodland Width of

Area Barrier

(km 2 ) (km 2 )

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 (R 2 = .6156 to R 2 = .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 00

? 4 WJW

3800

Sp

3 600-

>

3400

S *

320oJ

*" 6 f\

3000

2800-

* 2600

1

% 2400

- 2200

A

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
km 2 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

50 100 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. Johnson 1

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 villosus 2

X

X

X

X

X

X

X

-

-

-

-

Dendwcopos pubescens 2

X

X

X

X 1

-

X

-

-

-

-

-

Tyrannus tyrannus

X

-

X

— '

Iridoprocne bicolor

X

X

-

X

-

X

-

-

-

-

-

Parus atricapillus 1

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 scalaris 2

X

X

-

X

Centurus uropygialis'

X

Sayomis nigricans

X

X

-

X

Pyrocephalus rubinus 2

X

-

X

X

Thryomanes bewickii 2

-

-

-

X'

-

-

-

X

X

X

X

Toxostoma dorsale 2

X

X

Sialia mexicana

-

-

-

X 1

-

-

-

-

-

-

-

Phainopepla nitens 2 '

- 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 sparverius 2

X

X

X

X

X

X

X

X

X

X

X

Coccyzus americanus

X

-

-

X

-

-

-

X

-

X

X

Otus asio 2

X

-

-

X

-

-

-

X

X

-

X

Archilochus alexandri

X

X

-

X

-

X

-

X

-

-

X

Colaptes auratus 2

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 melodia 2

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 Nevada 1 versus those of mon-
tane islands to the southeast 2 . Boreal (B) and Austral
(A) species are distinguished.

Species present only in north:

Glaucidium gnomu 3 (B)
Sphyraupius ruber 3 (B)
Cyarwcitta stelleri (B)
Species present in both areas but
Aegolius acadicus (B)
Dendrocopos villosus 4 (B)
Tachycineta thalassina (B)
Parus gambei 45 (B)
Sirto carolinensis (B)
Species present in both areas but
Otus asio* 5 (A)
Empidonax wrightii 5 (A)
Aphelocoma coerulescens (A)
Psaltripurus minimus 4 (A)
Species present only in south:
Regulus calendula 3 (B)
Vireo vicinior (A)

Certhia familiaris 3 (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 bewickii 45 (A)
Polioptila caerulea 5 (A)
Vireo soliarius (B)

Icterus parisorum (A)
Junco caniceps (B)

'Wassuk Range, Sweetwater Mountains, and Pine Grove Hills.

2 Grapevine Mountains and Palmetto Mountains (= Mt. Magruder +
Silver Peak Mountains).

3 Breeds onlv verv locally in pinyon woodland in the northern or south-
em mountains.

4 Subspecies in north different from that in south

5 Partial 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 2 s 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

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
weak 1 :

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. hellmayri 2 .
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

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

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

S S ^

5 u
3 o

J s s

Zb, r>ozt,>u

I a r 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 components 1 in Boreal avifaunas of mountain
ranges 2 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 monoensis 2

-

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 AMERICA 1

Howard C. Stutz 2

.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 F 2 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 i