Full text of "Importance, preservation and management of riparian habitat: a symposium, Tucson, Arizona, July 9, 1977"

See other formats

Historic, Archive Document

Do not assume content reflects current
scientific knowledge, policies, or practices.

Rocky Mountain Forest and
Range Experiment Station
Forest Service

U.S. Department of Agriculture
Fort Collins, Colorado 80521

USDA Forest Service

General Technical Report RM-43

Importance, Preservation and Management

of Riparian Habitat:

A Symposium

Tucson, Arizona

July 9, 1977

Foreword

The material offered in this symposium is
both urgently needed and late in coming. Only
in very recent years have scientists and mana-
gers begun to shift their perspective on
riverine systems away from localized, single-
practice values toward a broader, more coopera-
tive and ecological set of approaches.

The dearth of relative and detailed his-
toric data on riparian habitats is lamentable.
We begin investigations on a complex and
closely interwoven ecosystem without a
reliable baseline. Much of our knowledge to
date focuses on the aftermath of decades of
abuse. But armed with this hindsight, we are
beginning to collect, organize and apply hard
data to difficult problems. What percentage
of breeding birds are dependent on riparian
habitat? How much grazing is too much? How
many species of plants occur only in riverine
ecosystems?

We have begun the long and expensive task
of quantifying our generalities about the
riparian habitat, so that we can offer valid
alternatives to managers in their attempts to
preserve the scant remnants of what was once
a vast network of thriving and varied habitats.

Agencies and individuals that favored
cutting, grazing, damming, and channelizing
were opposed by those who favored preserving
intact areas that supported unique plant and
animal species, or were of high recreational
or esthetic value. We are beginning to
reevaluate both these polarized positions,
realizing that entire riverine systems cannot
be solely maintained for a single interest,
whether it be water salvage or the remnant
population of a species endangered by our
activities .

This symposium stresses the continuity
and interrelationships of riparian ecosystems,
their wildlife and vegetation, historic and
current uses. We have designed this proceedings
to bring together material that represents the
current state of knowledge, and to point out

directions for the next critical steps in the
interlocking problems of research and manage-
ment .

This symposium and its proceedings evolved
after much planning. Some individuals on the
steering committee have discussed a symposium
such as this for several years. However, with
riparian research and management still in its
infancy, only recently has a good symposium
seemed possible. Even now many of the prominant
researchers and managers whom we contacted
during the preparation of the symposium felt
that their grasp of the workings of riparian
ecosystems was still inadequate. We found
research projects and management plans devel-
oping as rapidly as the state of the art allows
in the western United States, especially the
arid Southwest. And although subjects such as
river recreation are developing rapidly in the
eastern United States (see Proceedings, River
recreation management and research symposium —
January 24-27, 1977, General Technical Report
NC-28, North Central For. Exp. Stn. , Minneapolis,
Minn.) one is hard pressed to find riparian
projects east of the Rocky Mountains, except for
those associated with the continuing destruction
of even more riverine ecosystems.

We predict that by the early 1980 's research
projects currently being planned and undertaken
will lead to a vast expansion in the knowledge
necessary to properly manage riparian ecosystems.
In 3 to 5 years, a full-scale workshop on
sophisticated means of research, preservation,
and management of riparian habitat seems possible.
Today we offer a relatively primitive state of
the art in the "importance, preservation, and
management of riparian habitat." If our predic-
tions are correct, this science will have
developed from a newly born discipline into pro-
gressing maturity within two decades. Far-
sighted men like the late Douglas C. Morrison,
to whose memory this symposium is dedicated, will
have played a major role in the birth of scien-
tific management of riparian habitat.

R. Roy Johnson

USDA Forest Service

General Technical Report RM-43

Importance, Preservation and
Management of Riparian Habitat:

A Symposium

Tucson, Arizona
July 9, 1977

Technical Coordinators:

R. Roy Johnson
National Park Service
and

Dale A. Jones
USDA Forest Service

Cosponsored by:

Arizona Game and Fish Department

Arizona-New Mexico Section, The Wildlife Society

Arizona State University

Bureau of Land Management

Museum of Northern Arizona

National Park Service

U.S. Fish and Wildlife Service

USDA Forest Service

Acknowledgment

We wish to thank the authors of the papers
in this proceedings for their enthusiasm and
cooperation. The one day available for the
symposium provided insufficient time to present
all of the fine papers brought to our attention.
Thus, we are pleased to have several excellent
contributed papers published in the same volume
with the presented papers. The authors of con-
tributed as well as presented papers submitted
camera-ready manuscripts to expedite their
publication. The Bureau of Land Management
provided generous financial support for the
publication of these papers.

Finally, we acknowledge those who assisted
with the mechanics of organizing and conducting
a symposium of this scope. The steering com-
mittee, composed of representatives from the
sponsoring agencies listed, assisted in organ-
izing and publicizing the symposium. And the
fact that you are able to read these papers
so shortly after the symposium is due largely
to the combined efforts of the authors and the
Publications Group, Rocky Mountain Forest and
Range Experiment Station, Fort Collins, Colo-
rado. To all of these we owe our sincere
thanks .

R. Roy Johnson

Senior Research Scientist

National Park Service

U.S. Department of the Interior

Grand Canyon, Arizona 86023

Dale A. Jones

Director of Wildlife Management
Forest Service

U.S. Department of Agriculture
Washington, D.C. 20013

Abstract

Twelve presented and 15 contributed papers highlight what
is known about this unique, diminishing vegetative type:
characteristics, classification systems, associated fauna, use
conflicts, management alternatives, and research needs.
Speakers stressed the continuity and interrelationships of
riparian ecosystems, their wildlife and vegetation, historic
and current uses.

Contents

Page

SESSION I . 1

Discussion Leader: Dale Jones

Importance, Preservation and Management of Riparian Habitat:
An Overview 2

Steven W. Carothers

Classification of Riparian Habitat in the Southwest 5

Charles P. Pase, and Earle F. Layser

Inventory of Riparian Habitats 10

David E. Brown, Neil B. Carmony, and Raymond M. Turner

Importance of Riparian Ecosystems: Biotic Considerations l*t

John P. Hubbard

Importance of Riparian Ecosystems: Economic Considerations .... 19

Kel M. Fox

Vegetation Structure and Bird Use in the Lower Colorado

River Valley 23

Bertin W. Anderson, and Robert D. Ohmart

A Riparian Case History: The Colorado River 35

Robert D. Ohmart, Wayne 0. Deason, and Constance Burke

SESSION II i»8

Discussion Leader: Robert Jantzen

Wildlife Conflicts in Riparian Management: Grazing kS

Charles R. Ames

Wildlife Conflicts in Riparian Management: Water 52

Charles E. Kennedy

Management Alternatives for the Riparian Habitat in the

Southwest 59

Gary A. Davis

Endangered Species vs. Endangered Habitats: A Concept 68

R. Roy Johnson, Lois T. Haight, and James M. Simpson

Riparian Research Needs 80

David R. Patton

Riparian Habitat Symposium: Closing Remarks 83

Milo J. Hassell

CONTRIBUTED PAPERS

Classification of Riparian Vegetation 85

William A. Dick-Peddie, and John P. Hubbard

Fishes Inhabiting the Rio Grande, Texas and Mexico,

Between El Paso and the Pecos Confluence 91

Clark Hubbs, Robert R. Miller, Robert J. Edwards,
Kenneth W. Thompson, Edie Marsh, Gary P. Garrett,
Gary L. Powell, D. J. Morris, and Robert W. Zerr

An Overview of Riparian Forests in California: Their

Ecology and Conservation 98

Anne Sands, and Greg Howe

Regeneration and Distribution of Sycamore and Cottonwood

Trees Along Sonoita Creek, Santa Cruz County, Arizona 116

Richard L. Glinsky

The Development and Perpetuation of the Permanent Tamarisk

Type in the Phreatophyte Zone of the Southwest \2h

Jerome S. Horton

Avian Use of Saltcedar Communities in the Lower Colorado

River Val ley 128

Bertin W. Anderson, Alton E. Higgins, and Robert D. Ohmart

Influences of Riparian Vegetation on Aquatic Ecosystems
with Particular Reference to Salmonid Fishes and Their
Food Supply 137

William R. Meehan, Frederick J. Swanson, and James R. Sedell

Ecological Study of Southwestern Riparian Habitat:

Techniques and Data Applicability 1^6

Bertin W. Anderson, Ronald W. Engel-Wilson,
Douglas Wells, and Robert D. Ohmart

The Importance of Riparian Habitat to Migrating Birds 156

Laurence Stevens, Bryan T. Brown, James M. Simpson,
and R. Roy Johnson

Significance of Rio Grande Riparian Systems Upon the

Avifauna 165

Roland H. Wauer

Some Effects of a Campground on Breeding Birds in Arizona 175

Stewart W. Aitchison

Population Fluctuations in Nocturnal Rodents in the Lower

Colorado River Valley 183

Bertin W. Anderson, Jeff F. Drake, Jr., and Robert D. Ohmart

CI imatological and Physical Characteristics Affecting Avian
Population Estimates in Southwestern Riparian Communities
Using Transect Counts 193

Bertin W. Anderson, and Robert D. Ohmart

Southwestern Riparian Communities: Their Biotic Importance

and Management in Arizona 201

David E. Brown, Charles H. Lowe, and Janet F. Hausler

Terrestial Mammals of the Riparian Corridor in Big Bend

National Park 212

William J. Boeer, and David J. Schmidly

Session I

Discussion Leader: Dale Jones
Director, Wildlife Management
USDA Forest Service, Washington,

Importance, Preservation,
and Management
of Riparian Habitats: An Overview1

Steven W. Carothers

In the early 19th century, huge and
numerous squawfish were being taken along the
lower Gila; there was a commercial fishery of
the humpback sucker in the San Pedro; wild
turkeys hatched out of the waist-high grasses
of the Auga Fria; and grizzly bears and
mountain lions were encountered with alarming
frequency in the riparian woodlands associated
with some of these drainages. The historical
literature that documents the early exploration
and settlement of the southwestern United States
is replete with similar accounts of the original
condition of the area's rivers, streams and
springs (see Lacey et al., 1975; Hastings and
Turner, 1965, for original references).

The river valleys of the arid Southwest
have undergone significant physical and biolog-
ical changes since the early 1800' s. This
change has involved almost exclusively a
deterioration of the natural resources. Drain-
ages like the Gila and San Pedro Rivers that
once supported pristine riparian communities
are now, in many sections, dry and devoid of
native trees and shrubs. Concomitant with
the deterioration of the riparian habitats,
the range habitats throughout the same area
also reflect a decline in the production of
palatable forage plants and an increase in
topsoil erosion.

We can, and have, argued about the causes
of deteriorating range quality. But there is,
in my opinion, no controversy concerning the
causes of the irrevocable consumption of the
riparian habitat that has occurred in less
than 150 years. The imminent demise of the
riparian woodland can be most assuredly linked
to the land utilization practices of man.

The title of this symposium indicates
that we will address the "importance, preserva-
tion and management" of riparian habitats.
The word "importance" here is meant to reflect
the relative contribution of riparian habitats

Paper presented at the Symposium on
Importance, Preservation and Management of
the Riparian Habitat, Tucson, Arizona,

July 9 , 1977.

Head of Biology Department, Museum of
Northern Arizona, Flagstaff, Arizona 86001.

in a natural ecosystem; "preservation and
management" refer to implementing land management
practices that will forestall the possible extinc-
tion of these habitats. We are tempted to
approach these issues in purely scientific
definitions. But the issues extend beyond
the realm of biology. Man has contributed,
in large part, to these problems. It is only
by examining how he views the riparian habitat
as important to his economic base, and thereby
consumes this resource, that we can understand
why it is in danger of extirpation.

The settlement patterns of the native
American Indians clearly reflect the initial
consumptive use of riparian habitats . They
first settled the river valleys, needing water
for themselves, and subsequently drinking water
for livestock and irrigation water for crops.
This pattern of consumption accelerated greatly
in the early 1820' s, with the settlement of
the early Anglo-Americans in the Southwestern
river valleys. Prospectors, farmers, and
ranchers all found uses for the limited water
and its associated vegetation. The area was
rapidly settled and in 1896 the Governor of
Arizona Territory could report, "In Arizona
by 1883-84 every running stream and spring was
settled upon, ranch houses built, and adjacent
ranges stocked." (Report of Governor, 1896:21
fide Hastings and Turner, 1965) .

The settlers cleared large expanses of
native vegetation, using some for building
materials; but for the most part, they did not
view the woodlands as a valuable resource , and
removed them so that the soil of the alluvial
bottom could be put into "production" for
agricultural and domestic livestock grazing
purposes. Eventually, farming and ranching
became thriving concerns; river water was
channeled into irrigation canals, wells were
excavated and in time the water tables began
to drop. Responding to changing water regimes,
damaging floods, or in simple attempts to increase
the yield of the land, dams were finally con-
structed, inundating and destroying even more
riparian woodland and free-flowing streams and
rivers. By the late 1920' s, America had
shifted from a rural to a predominantly urban
population. The beginning of an urban-industrial
civilization in the arid Southwest required the

utilization of many innovative technological
advances in developing veritable oases where
once only parched deserts prevailed. As pop-
ulation centers experienced rapid and prolific
expansion, terms such as water production,
water management, and water salvage became
very meaningful.

As recently as the late 1960 ' s , belts of
native riparian woodland along the river valleys
of central and southern Arizona were still being
actively removed by water salvage and flood
control agencies. These "phreatophyte control"
and channelization projects were easily justi-
fied when based on the standard cost/benefit
ratio that was used for project evaluation at
the time. The important parameters of these
evaluations were: 1) streamside vegetation
requires substantial amounts of water, water
that is lost to the atmosphere through evapo-
transpiration; and 2) streamside vegetation
impedes the rapid transport of flood waters
and increases the apparent severity of floods
by temporarily and partially damming channels,
thus forcing high water into the adjacent
f loodplain lands .

The question of how much water is gained
and to what degree floods are prevented by
phreatophyte control and channelization has
been a battle of the minds almost since
vegetation removal was first suggested (see
Lacey et al., 1975; Paylor, 1974, for references).
Still, we have not seen the last of phreatophyte
control and streambed channelization in the
Southwest, and we know for certain that addition-
al dams will consume even more of the still
extant riparian areas . But the most insiduous
threat to the riparian habitat type today is
domestic livestock grazing. Many riparian areas
appear to be in good health; on closer examina-
tion, we find that while the mature vegetation
approaches senescence, grazing pressures have
prevented the establishment of seedlings. We
are very concerned that when many of these
mature stands of trees die of natural causes,
there will be no young forms to take their
place. Heavy grazing pressures can and do
produce even-aged, non-reproducing vegetative
communities. Our concern for this habitat's
survival can only mount until this situation
is remedied.

For more than a century, then, the riparian
habitats of the Southwest were viewed only in
terms of their consumptive value, while their
values for non- consumptive purposes — aesthetics,
recreation, wildlife, and so on — were largely
ignored. It was not until the mid-1960 's that
various agencies and individuals, particularly
in the Arizona Game and Fish Department (see
Bristow, 1969; Gallizioli , 1965) and the United
States Forest Service (pers. communication,
Dale Jones and Douglas C. Morrison) began to
point out that substantial numbers of both
game and non-game wildlife species were

dependent upon riparian vegetation. And it
was only in the fall of 1968 that efforts to
quantify the impact of streamside vegetation
removal on wildlife were first undertaken. It
is for his efforts in this regard that this
symposium is dedicated to the late Douglas C.
Morrison.

Working through the wildlife staff on the
Coconino National Forest, in 1968 Mr. Morrison
participated in the design of a research project
that would quantify the effects of phreatophyte
control on breeding birds in the native riparian
woodland of the Verde River. A Forest Service
contract was awarded to the Department of Biology
at the Museum of Northern Arizona in the spring
of 1969. Through Mr. Morrison's efforts, the
study was funded for two years by the Forest
Service. After that time, the Arizona Game
and Fish Department supported the project for
an addi tonal three years.

The results of that study (see Carothers
et al. , 1974) substantiated, for the first
time , two facts that had long been suspected
by many wildlife biologists: 1) that vegetation
manipulation in native riparian communites was
extremely detrimental to breeding bird popula-
tions, the extent of the impact being significant-
ly correlated with the degree to which phreato-
phytes were removed, and 2) that for a given
number of acres of habitat, the riparian type
supports higher population densities than any
other forest habitat type. Indeed, the surpris-
ing discovery resulting from these studies is
that the homogeneous cottonwood riparian type
of the Verde River contains some of the highest
avian population densities per unit area that
have been recorded in the continental United
States. Recently, other investigators working
in the river valleys of the Gila (Hubbard, 1971) ,
the Colorado (Anderson, Ohmart et al. , this
symposium; Carothers and Sharber, 1976) and the
Salt (Johnson and Simpson, 1971) have demonstrat-
ed the remarkabley high wildlife potential of
riparian habitat types.

The influence of the riparian type on
wildlife is not limited to those animal species
that are restricted in distribution to the
streamside vegetation. Preliminary investigations
conducted by us (see Stevens et al., this sympo-
sium) , in both river valley and mountain riparian
types, demonstrate that the population densities
of birds in habitats adjacent to the riparian
type are influenced by the presence of a riparian
area. Our present interpretation of these
preliminary data is that when a riparian habitat
is removed or severely manipulated, not only are
the riparian species of the area adversely
influenced, but wildlife productivity in the
adjacent habitat is also depressed. The actual
width the zone of influence riparian habitats
have on adjacent habitat wildlife productivity
may, for some animal species, extend several
hundred meters beyond the edge of the stream-

side vegetation. Under the auspices of the
Forest Service, we are presently attempting to
determine this for a variety of riparian types
in Arizona and New Mexico.

Thus, the history of man's use of the
riparian habitats in the Southwest indicate
that it has been and continues to be an
important and valuable asset to the settlement
and progress of this country. On the other
hand, ecological research on this habitat type
has conclusively demonstrated that riparian
areas are integral and indispensable components
of desert and mountain ecosystems. Past
riparian habitat management practices have
resulted in widespread destruction of these
areas. That they are non-renewable resources
as suggested by Lacey et al. (1975) is a
frightening possibility. And even though there
are many Southwestern drainages still forested
by riparian vegetation, current land use
practices still threaten the future existence
of these native communities.

We should not look back on the land
management practices of the past with too
much remorse and certainly with no blame.
A summary of man's activities in and the
destruction of woodlands, streams, and rivers
simply reflects man's successful settlement of
this arid land, allowing those of us who ,live
and work in the Southwest the lifestyle we now
enjoy. Land management practices of the past
should, in fact, be a foundation for learning
and understanding how to cautiously move
forward in our interactions with the environ-
ment .

We are here today to exchange information.
The time is at hand for the ecologist, economist,
engineer, environmentalist and land manager to
strike a compromise ... a compromise that will
provide a future for native Southwestern riparian
habitat types. Accepting and assessing the
environmental mistakes of the past, becoming
aware of the intricate needs and associations
of man and the environment can lead to the
implementation of land management practices that
will achieve this end.

LITERATURE CITED

Bristow, Bud. 1969. Land and water projects
investigations, investigation of proposed
projects. Ariz. Game and Fish Dept., Proj .
FW-16-R-8, Jl Completion Report. Pp. 2 3-27.

Carothers, S.W. , R.R. Johnson, and S.W. Aitchison
1974. Population structure and social
organization of southwestern riparian birds.
Amer. Zool. 14:97-108.

Carothers, S.W. and N.J. Sharber. 1976. Birds
of the Colorado River. Iii An Ecological
Survey of the Riparian Zone of the Colorado
River between Lees Ferry and the Grand Wash
Cliffs, Arizona. Final Research Report,
National Park Service.

Gallizioli, S. 1965. Phreatophytes and wildlife
In Pacific Southwest Interagency Committee,
Phreatophyte Subcommittee. Minutes of
Phreatophyte Subcommittee Meetings. Vol. 2,
Jan. 1960 through Dec. 1966. Pp. 317-319.

Hastings, J.R. and R.M. Turner. 1965. The
Changing Mile. University of Arizona
Press, Tucson. 317 pp.

Hubbard, J. P. 1971. The summer birds of the
Gila Valley, New Mexico. Nemouria, 1-35.
Occ. Pap. Delaware Mus . Natur. Hist.

Johnson, R.R. and J.M. Simpson. 1971. Impor-
tant birds from Blue Point cottonwoods ,
Maricopa County, Arizona. Condor 73:379-
380.

Lacey, J.R., P.R. Ogden and K.E. Foster. 1975.

Southern Arizona Riparian Habitat: Spatial

Distribution and Analysis. Univ. of Ariz.,

Tucson. 148 pp.
Pay lore, P. (ed.) 1974. Phreatophytes:

A Bibliography. Water Res. Sci. Info.

Center, U.S. Dept. of the Interior. 277 pp.

Classification of Riparian Habitat in the Southwest1

Charles P. \Pase, and Earle F. '^Layser

4/
2

Abstract. — The riparian areas in Arizona and New Mexico
are uniquely productive wildlife habitats. A tentative classi-
fication based on the work of Brown and Lowe is proposed as
a working model. Six biomes, nine series and 23 associations
are tentatively recognized. Additional research is proposed to
further refine the classification. The classification of ripar-
ian vegetation can provide a strong management tool.

INTRODUCTION

Riparian habitats in arid and semiarid
environments are unique reservoirs of plant
and animal diversity. Breeding bird densities
are likely to be high, especially in the most
productive Fremont cottonwood (Populus f remontii)
stands, with as many as 1,000 pairs or more
per 100 acres (Carothers and Johnson 1975).
Also, the small but highly productive riparian
communities support the greatest variety of
birds (and probably other vertebrates) in the
Southwest.

"Riparian" type habitats are streamside
or riverside communities, stretching from high
forest to low desert. Soil moisture is seldom
a limiting factor, at least for successfully
established perennials, although surface water
may be lacking at times in marginal areas. The
wide array of habitats thus included sustains
an equally wide array of plant and animal com-
munities .

FLORISTIC HISTORY

The southwestern mountains and valleys con-
tain diverse floristic elements, which have
mixed and adapted to provide a unique flora.

■'-Paper presented at the Importance, Pres-
ervation and Management of the Riparian Habitat
Symposium, held at Tucson, Arizona, July 9, 1977.

Principal Plant Ecologist, USDA Forest
Service, Rocky Mountain Forest and Range Experi-
ment Station, at the Station's Research Work
Unit at Arizona State University, Tempe. Cen-
tral headquarters is maintained at Fort Collins
in cooperation with Colorado State University.

Resources Analyst, Land Use Planning, USDA
Forest Service, Southwestern Region, Albuquerque,
New Mexico „

Madro-tertiary elements such as Celtis,
Juglans , Prosopis , Platanus , and Sapindus
along the Mexican cordillera are representative
of families with strong subtropical affinities.
During the drier postpluvial period, increasing
dryness forced their retreat to the riparian
zones, where today they form characteristic
deciduous forests and woodlands. According to
Martin (1963), many of the Mexican flora and
fauna probably were isolated in the southern
Arizona mountains during the mid-postglacial,
perhaps 4,000 to 8,000 years ago. If so, modern
communities were relatively recently composed.

During the Pliocene-Pleistocene era,
Arcto-tertiary Geoflora extended southward
into the mountains of Arizona and New Mexico,
establishing such northern-affinity genera as
Alnus , Salix, Populus, and Betula. The same
xerothermal conditions — the so-called altither-
mal — that isolated the mexican plateau elements
also trapped these genera in moist, riparian
habitats (Lowe 1964) .

Development of modern riparian plant and
animal communities, and the isolation of many
species into highly restricted habitats, happened
relatively recently. Further restriction of
these small, sensitive areas by agriculture,
recreational, and other developments may pose
a serious threat to species largely or wholly
dependent on this habitat.

THE RIPARIAN ZONES

Despite the species diversity and produc-
tivity of the riparian zone, it is relatively
small. The total riparian area in Arizona is
some 279,600 acres (Babcock 1968), of which
100,700 acres are along the Gila River. The
areas within New Mexico may be comparable or
slightly larger, with the inclusion of substan-
tial areas of mesquite, Fremont cottonwood,
and salt cedar (Tamarix spp.) along the main
stem of the Rio Grande.

The most productive Fremont cottonwood
areas are surprisingly small — about 6,000 to
8,000 acres in Arizona, according to Barger and
Ffolliott (1971). Unfortunately, estimates of
area by plant associations are not yet available
for other types. However, National Forests in
USDA Forest Service Region 3 plan to inventory
riparian habitat, because of its high wildlife
value, as part of forest land management plan-
ning.

Riparian communities generally exhibit a
predictable vertical zonation in relation to
each other (fig. 1), although absolute upper
and lower limits may vary within the Region
(Freeman and Dick-Peddie 1970). Certain species,
such as box elder (Acer negundo), are found
throughout the area, and may occur as scattered
minor components in a number of related asso-
ciations. Others such as Salix irrorata, have
much lower ecological amplitudes.

WHY A CLASSIFICATION SYSTEM?

2. Utilizing a classification system would
allow for more uniform identification
of the different riparian situations,
thus providing a means to more accur-
ately assess the distribution and the
relative amounts of the different com-
munities that may exist.

3. Development of the classification would
provide an inventory of the major plant;
that exist in the different communities.
Inventory of fauna associated with the
different plant communities could also
be done.

4. Successional roles of the different
species would be better determined
which would allow more accurate pre-
diction of results of management prac-
tices .

5. It would provide a framework for addi-
tional research and reporting of re-
search results.

While there is substantial intergradation

between adjacent riparian units, certain species The development of a site-based vegetation class-

or combinations of species tend to dominate ification system for riparian habitats is not an

standSo Recognition of these natural ecological end in itself. The purpose of the classif icatioi

units may help the land manager as follows: is to provide land and resource managers with a

management tool. As pointed out above, this

1. It would assist the identification, would enable managers and research to better

description, and communication about deal with problems involved in the management

riparian habitats. of riparian habitats in the Southwest.

STATUS OF CLASSIFICATION

While there have been notable studies of ripar-
ian vegetation in the Southwest (Campbell and
Green 1968; Lacey, et al. 1975; Lowe 1961; Free-
man and Dick - Peddie 1970; Horton 1960), there
has been only little work to date towards de-
velopment of a systematic classification of ri-
parian vegetation or sites. Such classification
involves methodology as described by Pfister and
Arno (1977). In generalized terms this means:

1. Sampling of stands or communities in
respect to ecological and floristic
characteristics .

2. Analyzing the above data by various
techniques .

3. Considering various groupings (classes)
based on the data and analysis; and
selecting groupings appropriate to

the purpose.

4. Defining the classes as simply and pre-
cisely as possible.

The taxonomic classification is then used to
apply the appropriate class name to communities
as they are encountered in the field. This is
called "identification", and differs from the
actual development of the classification (Bai-
ley, et al. in press).

A TENTATIVE CLASSIFICATION OF RIPARIAN
COMMUNITIES
(generally after the system of Brown
and Lowe 1974)

Boreal Riparian Mixed Forest Biome
Spruce — Mixed Shrub Series

Picea pungens - Alnus tenuf olia
Association
Temperate Riparian Deciduous Forest Biome
Mixed Broadleaf Series

Mixed Broadleaf Associations (fig. 2)
Acer negundo Associations
Alnus oblongif olia Associations

(fig. 3)

Platanus wrightii Associations

(fig. 4)

Fraxinus velutina Associations

(fig. 5)

Juglans major Associations
Cottonwood — Willow Series

Populus f remontii — Mixed Broadleaf

Associations
Populus f remontii Associations (fig. 6)
Salix bonplandiana Associations
Populus f remontii — Salix gooddingii

Associations
Salix gooddingii Associations
Populus angustif olia Associations

Subtropical Riparian Evergreen Forest Biome
Palm Series

Washingtonia f ilif era Associations
Boreal Riparian Woodland Biome
Willow Series

Salix bebbiana Associations
Salix irrorata Associations
Salix mixed Associations (fig. 7)
Alder Series

Alnus tenuif olia Associations
Alnus tenuif olia — Salix Associations

(fig. 8)

Temperate Riparian Deciduous Woodland Biome
Willow Series

Salix exigua Associations
Subtropical Riparian Deciduous Woodland Biome
Mesquite Bosque Series

Prosopis julif lora Associations
Prosopis julif lora — Mixed narrowleaf
(Tamarix, Chilopsis, Celtis)

Associations

Tamarix Disclimax Series

Tamarix chinensis Associations

The above list is not to be considered complete.
Lacey et al. (1975) suggests that a number of
these should be subdivided further. The list-
ing, however, generally displays the current
"classification" of riparian vegetation, and
represents a tentative identification of types
which are a first approximation toward the de-
velopment of a classification.

Figure 2. — Mixed broadleaf Association with
Platanus, Populus , Fraxinus , Juglans , Prunus ,
and Sapindus. This is primary habitat for
rose- throated becard and Apache fox squirrel.
Arroyo Cajon Bonito, Sonora, Mexico.

Figure 3. — Alnus oblongif olia Association on
the upper Gila River. This type occur only on
upper middle elevation living streams, where
roots can reach the water table during much of
the year (Freeman and Dick-Peddie 1970) .

Figure 4. — Platanus wrightii Association along
Turkey Creek, Rincon Mountains, Coronado National
Forest. Limited understory and absence of repro-
dection charaterizes these grazed channel
bottoms .

Figure 5. — A nearly pure stand of Fraxinus
velutina along Ash Creek, Rincon Mountains.
Heavy past cattle use will be restricted under
new management plans in this area.

Figure 6. — Narrow bands of Fremont cottonwood,
sycamore, velvet ash and Goodding willow pro-
vide sharp contrast to dry grassland communi-
ties on the Prescott National Forest.

Figure 7. — Salix Mixed Association, upper Black
River, Apache-Sitgreaves National Forest.
J3. pseudocordata and j>. lasiolepis provide
shade and cover along one of Arizona's productive
trout streams. The area is valuable elk summer
range, with little livestock use.

Figure 8. — A lush Alnus tenuif olia - Salix
Association flanks Los Pinos Creek on the Carson
National Forest. The narrow, overhanging ripar-
ian woodland community greatly improves trout
habitat.

LITERATURE CITED

Babcock, H.M. 1968. The phreatophyte problem
in Arizona. Ariz. Watershed Symp. Proc.
12:34-36.

Bailey R.G., R.D. Pfister, and J. A. Henderson
(in press) . The nature of land and re-
source classification. Jour, of For. (draft
Ms.) 28 p.

Barger, Roland L. and Peter F. Ffolliott. 1971.
Prospects for cottonwood utilization in
Arizona. Progr. Agric. in Ariz. 23(3):
14-16.

Brown, David E., and Charles H. Lowe. 1974. A

digitized computer-compatible classification
for natural and potential vegetation in the
Southwest with particular reference to
Arizona. J. Ariz. Acad. Sci. 9, Suppl. 2.
Campbell, C.J., and Win Green. 1968. Perpetual
succession of stream-channel vegetation in
a semi-arid region. J. Ariz. Acad. Sci.
5:86-98.

Carothers, Steven W. , and R. Roy Johnson. 1975.
Water management practices and their effects
on nongame birds in range habitats.
p0 210-222. In Proc. Symp. on Manage, of
For. and Range Habitats for Nongame Birds,
Dixie R. Smith, tech. coord. USDA For.
Serv. Gen. Tech. Rep. WO-1, 343 p.

Freeman, C.E., and W.A. Dick-Peddie. 1970.

Woody riparian vegetation in the Black and
Sacramento mountain ranges, southern New
Mexico. Southwest Nat. 15:145-164.

Horton, J.S., R.C. Mounts, and J.M. Kraft. 1960.
Seed germination and seedling establishment
of phreatophyte species. U.S. Dep. Agric,
For. Serv. Rocky Mt. For. and Range Exp.
Stn., Stn. Pap. 48, 26 p.

Lacey, John R., Phil R. Ogden, and Kenneth E.
Foster. 19751 Southern Arizona riparian
habitat: spatial distribution and analysis,
OALS Bull, 8. School of Renewable Natural
Resources and Office of Arid Land Studies,
Univ. of Ariz., Tucson.
Lowe, C.H. 1961. Biotic communities in the sub-
mogollon region of the inland Southwest.
Jour. Ariz. Acad. Sci. 2:40-49.

Lowe, Charles H. 1964. The vertebrates of
Arizona, Part I, Arizona landscapes and
habitats. The Univ. of Ariz. Press,
Tucson.

Martin, Paul S, 1963. The last 10,000 years„
A fossil pollen record of the American
Southwest. The Univ. of Ariz. Press,
Tucson.

Pfister, R.D. and S.F. Arno 1977. Forest hab-
itat type classification methodology.
USDA Forest Service Intermtn. For. & Rg.
Expt. Sta. Mimeo. 29 p.

Inventory of Riparian Habitats'

David E. Brown
Neil B. Carmony
Raymond M. Turner

Abstract. — A recently published map of Arizona's peren-
nial streams and important wetlands was presented and dis-
cussed. Perennial streams are illustrated rather than ripar-
ian vegetation because the streams are of more direct biotic
significance and are more readily identifiable. Inventory
procedures used in preparing the map were outlined and the
categories of streams and wetlands described. A planned re-
vision of this map will incorporate selected altitude contours,
base-flow classes, and an inventory of seasonal streams.

The map distributed to members of the
symposium presents Arizona's perennial streams
and important wetlands; it is a synthesis of
information compiled by numerous investigators
(Brown, Carmony, and Turner, 1977). While the
map implies the distribution of riparian and
marshland vegetation, it does not show riparian
vegetation directly: rather, it shows the dis-
tribution of live streams and wetlands — the
physical basis upon which riparian and wetland
biotic communities depend.

Efforts and plans to modify and even
eliminate many of the Southwest 's remaining
riparian communities have stimulated interest
in the inventory and preservation of riparian
habitats. One of the results of this interest
was the formation in 1972 of a "Riparian
Recovery Committee" within the New Mexico-
Arizona Section of the Wildlife Society. This
committee has promoted the investigation of
riparian habitats, and its efforts resulted in
this symposium sponsored by the U.S. Forest
Service. It soon became apparent to the
committee and to other workers involved with
riparian vegetation and fauna that the clas-

Paper presented at the Riparian Habitat
Symposium, Tucson, Arizona, July 9, 1977.

David E. Brown, Arizona Game and Fish
Department, Phoenix.

Neil B. Carmony, U.S. Geological Survey,
Tucson.

Raymond M. Turner, U.S. Geological
Survey, Tucson.

sification and inventory of the various ripar-
ian habitats is a necessary first step in under-
standing the problems associated with ripar-
ian habitats.

Because of the immense biological impor-
tance of these habitats, it was originally
planned to prepare a map of Arizona's natural
and potential riparian deciduous forests and
woodlands as described by Lowe (1964, p. 60-62)
and as classified by Brown and Lowe (1974a,
1974b). An informal program of collaboration
between the Arizona Game and Fish Department
and the U.S. Geological Survey was initiated
to map riparian vegetation from low-level
aerial photographs of sub-Mogollon Arizona
(photographs were taken in June 1973, and later
checked at a limited number of sites for
ground truth) . Although the determination of
established deciduous forests proved feasible,
it soon became apparent that the inventory of
riparian forest vegetation in itself left much
to be desired for the purposes of biotic
assessment. Two major deficiencies of this
approach were:

1. Extreme variations in the biota within
the riparian deciduous forests were not
measured. Deciduous forests can and do
exist along near-perennial as well as
perennial streams. Habitats along inter-
mittent or nearly perennial streams do
not, of course, support fish and other
important aquatic forms that are found in
perennial streams. Therefore, the
presence or absence of a potential ripar-
ian aquatic biota and its predators, e.g.

black hawks (Buteogallus anthraoinus ) ,
ospreys (Pandion haliaetus), bald eagles
(Haliaeetus leucocephalus) , otters (Lutra
canadensis), and kingfishers (Megaceryle
alcyon, Chlovoceryle amerioana) 3 would not
be indicated by an inventory of riparian
deciduous forests alone.

2. The dynamic nature of riparian commu-
nities is such that potential riparian
deciduous forests could not always be
determined. It soon became apparent that
floods, intensity of livestock grazing,
geologic and hydrologic conditions, and
other factors could continually or tempo-
rarily affect the presence of several
deciduous-forest species and thereby
determine the presence of the forest
community itself. These factors can
result in the presence of successional
stages from seedlings to decadent stands,
or in the absence of trees altogether.
These conditions may prevail over a
relatively short term or may be as long
as several human generations.

For these reasons we decided to determine
the feasibility of modifying existing map(s) of
live (perennial) streams and rivers to show wet-
lands and potential riparian vegetation. To our
surprise no comprehensive inventory of perennial-
flow reaches existed, and the only map of "perma-
nent streams" available for Arizona was a small-
scale drainage map by Miller (1954). 5 This map
was presumably based on fish collections and,
while serviceable, was very much in need of
revision. Perennial-stream data were trans-
ferred and extrapolated from Miller's map to a
1:500, 000-scale U.S. Geological Survey map of
Arizona. This became the base work map. U.S.
Geological Survey minimum flow records were used,
where available, to identify the nature of flow
in streams. Those streams for which no stream-
flow records were available became the subject
for interviews with wildlife managers, U.S. Forest
Service personnel, back packers, fishermen, and
others who had visited them. Fortunately, these
efforts coincided with a statewide stream inventory
conducted by the Arizona Game and Fish Department
(Silvey, 1977) and with a statewide map inventory
of biotic resources, including perennial streams,
by regional personnel of the Arizona Game and Fish
Department. Personal knowledge and investigation
by the authors supplemented these data and allow-
ed for further comparison and correction. A
major but not sole criterion used in determining

Perennial-flow reaches are indicated on U.S.
Geological Survey topographic and other maps, but
in many cases these streams do not appear to have
been adequately field checked and are in error.

the permanence of flow in an ungaged stream
was the presence or absence of fish. We
considered as perennial those streams which
supported a fish fauna regardless of minimum
flow. Perennial streams, as classified here,
may have surface water only in pools during
times of extreme drought or other low-flow
periods.

It soon became obvious that there are
three major categories of perennial streams
based on the biota actually or potentially
present. These categories and their character-
istics are:

1. Unregulated perennial streams. These
streams usually possess a native and/or
introduced aquatic fauna. Those that are
now devoid of native fishes possess the
potential for restoration if the native
species formerly present are still
available. Native riparian communities
and their associated biota are actually
or potentially present, and the riparian
exotic saltcedar (Tamarix ahinensis) can
be expected to be poorly represented.
Some of these streams may be partially
regulated by relatively small upstream
storage or diversion.

2. Regulated perennial rivers and streams.
These rivers and streams are characterized
by a totally unnatural streamflow pattern.
They lack normal seasonal high water, and
on occasion base flows may be artificially
reduced or eliminated. There may also be
major daily changes in streamflow that
negatively affect the life cycles of many
fish species. Water-temperature regimes
are also greatly altered. This factor,
and the lack of normally occurring high
flows, tend to inhibit natural repro-
duction of those plants and animals
which evolved with a natural flowing
system. These 'managed" streams are now
often largely populated by introduced
fishes, reptiles (for example, the
soft-shelled turtle (Trionyx spinifera)) ,
and amphibians (for example, the bullfrog
(Rana catesbeiana)) , and the native biota
are reduced and some species are extir-
pated. Exotic plants occupy many exten-
sive riparian habitats. Numerous species
of birds characteristic of riparian
communities usually persist, however.

3. Stream reaches containing only
effluent or waste-water dishcarge. These
stream reaches are characterized by an
almost completely introduced aquatic fauna
adaptable to polluted waters, and include
mosquito fish (Gambusia af finis)* sailfin

mollies (Poeoilia latipinna) , carp (Cyprinus
carpio), and crayfish (Astaoidae) . The
riparian vegetation is usually a mixture
of native and exotic species. The status of
these reaches is expected to change sig-
nificantly in the future as consumptive uses
for these waters are developed.

Important wetlands, the other major category
shown, are separated into two size categories,
classified by regulatory criteria similar to
those for perennial streams, and plotted on the
base map in the appropriate size and color.
Wetlands are here defined as poorly drained lands,
seasonally or periodically submerged lands, and
shallow bodies of water supporting emergent
vegetation. Important wetlands are those that,
because of their size and/or location, are much
used by nesting, migrating, and wintering water-
fowl or that provide habitats for rare, unusual
or interesting wetlands species of flora and
fauna. We were highly dependent on Game and Fish
personnel, especially Richard L. Todd and Thomas
K. Britt, for the locations of many of these
major wetlands.

All these data were transferred from the
1:500,000 work map to a U.S. Geological Survey
Drainage Map, scale 1:1,000,000. This map- is now
out of print but approximately 1,800 copies are in
circulation. Its rapid dissemination was due to
the interest of fishermen, hikers, realtors, gold
miners, and others as well as those interested in
natural-resource inventory. Because of this
broad appeal we intend to publish a second edition,
possibly in 1978. In an effort to make the map
more useful and meaningful to biologists, we
propose to incorporate additional data and crite-
ria, in addition to making those corrections
called to our attention. These new items include:

A. The addition of altitude contours 1,000 ft,
3,500 ft, 5,500 ft, and 7,750 ft above mean
sea level. These contours, based on the
occurrence of riparian species, were select-
ed after consultation with biologists
possessing knowledge of Arizona's endemic
fishes (R. R. Miller; W. L. Minckley;

W. Silvey, personal communications, 1977).
These contours will roughly partition the
state into arctic-boreal, cool-temperate,
warm-temperate, and subtropical zones. In
addition to providing information as to which
species of riparian flora and fauna can be
expected, these contours will allow for
rapid identification of cold montane waters
and associated fishes, waters dominated by
cool-water fish, waters dominated by warm-
water fish, and waters populated exclusively
by warm-water species.

B. The division of unregulated streams into size
categories on the basis of base-flow data.

Streams would be categorized as those
possessing a 7-day minimum flow of >50
cubjic feet per second (ft Is), >10-<50
ft /s, and <10 ft Is. These divisions
would roughly separate Arizona's natural
flowing streams as follows:

1. Base flow >50 ft3/s

Unregulated perennial streams with a base
flow greater than 50 ft-fys potentially
support a varied aquatic fauna of both
native and introduced species. Some fish
species present can be expected to reach
sizes of 20 lbs or more. Consequently,
these rivers are potential habitat for
some of the larger piscivorous animals,
e.g. otter, black bear (Ursus amerioanus),
bald eagle, osprey, etc. While these
rivers potentially support all native
species of riparian vegetation common to
their life zone, periodic floods and
seasonal fluctuations in flow regimes
usually prevent the establishment of
"gallery" forests, and long stretches of
river are often characterized by wide,
barren flood plains. Examples are few
and include the Lower Little Colorado
River below Blue Springs, the Verde River
from below Perkinsville to Clarkdale, the
Verde River from West Clear Creek to
Horseshoe Reservoir, and the Salt River
above Roosevelt Lake. Formation of a
perennial river from effluent discharge
is a recent phenomenon in Arizona.
Rivers in this class are formed by sewage
releases from growing cities and from
irrigation return flow. The aquatic
fauna is almost entirely introduced, as
is the dominant riparian plant, saltcedar
(Tamarix ohinensis). Native vegetation
is profuse, however, and appears to be
increasing.

2. Base flow >10-<50 ft3/s

Unregulated perennial streams in this
category can also be expected to support
a varied native and introduced aquatic
fauna. Select habitats can be expected
to support species of fish over 1 lb.
These streams are potential habitat for
most piscivorous species and normally
possess well-developed native riparian
communities and their animal associates.
Riparian deciduous forests are common
features below 6,500 ft. Examples of
streams in Arizona with base flows in
this range are the Virgin River, Bright
Angel Creek, the Verde River above
Perkinsville, Oak Creek, the lower
portions of West Clear Creek, Black River,
north fork of the White River, and the
San Francisco River.

3. Base flow <10 ft /s

Unregulated perennial streams of this size
usually support only the smaller fishes
and their predators. The native fish
fauna may be rich in species, and a sur-
prising number of amphibians and aquatic
reptiles can be expected. Riparian
deciduous species frequently form "gallery"
forests, and excellent examples are
present at many places.

C. We plan also to add seasonal streams.
These streams, termed semi-perennial by
Zimmerman (1969), flow during the winter
and early spring; they do not have a
sufficient base-flow component to maintain
a surface flow during warm dry months when
evapotranspiration losses are high, but
can be expected to flow from mid-winter
to about April. These streams potentially
support well-developed riparian vegetation,
including deciduous forests. The aquatic
biota is, of necessity, limited to those
forms adapted to withstand periods of
desiccation, and fishes are lacking.
Nonetheless, avian and amphibian species
characteristic of, and associated with,
riparian forests may be well represented,
with the exception of piscivorous species.
Such non-piscivorous animals include in the
appropriate environments the zone-tailed
hawk (Buteo albonotatus ) , grey hawk (Buteo
nitidus), summer tanager (Piranga rubra),
blue grosbeak (Guiraoa oaerulea) 3 Wood-
house's toad (Bufo woodhousei) t the
canyon tree frog (Hyla areniaolor) 3 etc.
These streams will be differentiated from
those of an ephemeral nature.

We believe that these inventory procedures
are applicable to other Southwestern States and
will prove useful in the delineation of their
aquatic, riparian, and wetland biotic resources.

We suggest that future mapping be at the scale of
1:1,000,000, which would be an effective
standard for the illustration of biotic resources
because of its easy metric conversions (1 mm =
1 km), adaptability for demonstration purposes,
and utility for field use.

LITERATURE CITED

Brown, D. E. , Carmony, N. B. , and Turner, R. M.
(compilers), 1977, Drainage map of Arizona
showing perennial streams and some important
wetlands: [Published by] Arizona Game and
Fish Dept., Phoenix, scale 1:1,000,000.

Brown, D. E. , and Lowe, C. H. , 1974a, A digitized
computer-compatible classification for
natural and potential vegetation in the
Southwest with particular reference to
Arizona: Ariz. Acad. Sci. Jour., v. 9,
supp. 2, 11 p.

1974b, The Arizona system for

natural and potential vegetation:
Illustrated summary through the fifth
digit for the North American Southwest:
Ariz. Acad. Sci. Jour., v. 9, supp. 3,
56 p.

Lowe, C. H., 1964, Arizona's natural environ-
ment, landscape, and habitats:
Arizona University Press, Tucson, 136 p.

Miller, R. R. , 1954, A drainage map of

Arizona: Syst. Zool., v. 3, p. 80-81.

Silvey, W. , 1977, Drainage systems of Arizona:
Ariz. Game & Fish Dept. mimecrept, 18 p.

U.S. Geological Survey, 1974, State of Arizona,
base map: Scale 1:1,000,000.

Zimmerman, R. C., 1969, Plant ecology of an
arid basin, Tres Alamo s-Redington area,
southeastern Arizona: U.S. Geol. Survey
Prof. Paper 485-D, 51 p.

Importance of Riparian

Ecosystems:
Biotic Considerations1

John P. Hubbard2

By biotic considerations I am referring
to flora and fauna, and specifically I would
like to probe the question of the importance
that riparian ecosystems play in sustaining
the rich biotas of the Southwest, i.e.
Arizona and New Mexico. To begin, these two
states are among the richest of any in the
United States as far as their diversity is
concerned in species of plants, terrestrial
vertebrates, and many invertebrates. This
biotic richness stems from several factors,
including the great environmental variety of
the region and the fact that several major
biotic areas impinge on the area, i.e. the
Great Basin, Rocky Mountains, Great Plains,
Mexican Plateau, and the Southern (Chihuahuan
and Sonoran) Deserts.

New Mexico is the fourth and Arizona the
fifth largest of the United States, with areas
of 121,666 and 113,909 square miles, respec-
tively. In size these states are thus on a
par with such well-known entities as the
British Isles, Italy, and the Philippines.
In elevation New Mexico ranges from 2800 to
13,161 feet above sea level, while Arizona
ranges from near sea level to 12,670 feet.
Although often through of as "deserts", both
states support extensive montane forests, and
New Mexico especially is crowned with alpine
tundra in the north. On the other hand,
aridity is a dominant climatic feature of the
region, and particularly at elevations below
6000 feet surface water is scarce and natu-
rally restricted to a few thousand miles of
generally narrow drainageways in the two
states .

Floristic diversity is revealed by the
fact that New Mexico supports 3500 to 3600
species of higher native plants within its
borders (Wagner, 1977), while the latest
summary for' Arizona lists 3438 (Kearney and
Peebles, 1960). For the continental United
States and Canada as a whole, an estimated

Ipaper presented in Importance, Preser-
vation, and Management of the Riparian Habitat,
Tucson, Az, July 9, 1977.

2Endangered Species Program Supervisor,
New Mexico Department of Game and Fish,
Santa Fe, N. Mex. 87503.

40,000 to 50,000 species of higher plants have
been recorded. Thus, the floras of New Mexico
and Arizona comprise about 7 to 9 percent of
the total flora of what might be termed tem-
perate North America.

Among terrestrial vertebrates one finds
that even higher percentages of the overall
temperate North American faunas are recorded
in these two states (Table 1) .

Table 1. Vertebrate Fauna of the Southwest,
Including Species Totals and as Per-
centages of the Total Fauna of North
America North of Mexico.-^

Mammals

species
percent

Birds

species
percent
species
percent

Reptiles

species
percent

Amphibians

species
percent

(all)

(breeding)

Arizona

134
41.6

431

62.0
245

38.0

35.2

13.5

New Mexico

139
43.2

413

59.4
245

38.0

30.3

14.2

As one can see, except for amphibians, Arizona
and New Mexico harbor disproportionate portions
of the terrestrial vertebrates of temperate
North America, with figures ranging from about
one-third to almost two-thirds among mammals,
birds, and reptiles. Amphibians, which mainly
depend on water for reproduction, in the two
states constitute about one-sixth of the North
American fauna.

^Data sources include Findley et al.,
1975; Hubbard, 1970; Lowe, 1964; Phillips
et al., 1964; Stebbins, 1966.

Fishes, although they face an overall
scarcity of habitats in the Southwest, are
nonetheless well-represented in the faunas.
Arizona has 32 native species (Minckley,
1973), while New Mexico has 59 species record-
ed within its boundaries (Koster, 1957). The
latter area supports a richer fauna by virtue
of its location in both the Atlantic and
Pacific drainages of the continent. In fact,
several species from the very rich
Mississippian ichthyofauna reach western lim-
its in New Mexico, including the blue sucker
(Cycleptus elongatus) . Even with their lim-
ited faunas, these two states still host — or
hosted — reasonably rich percentages of the
overall U.S. fish fauna in their boundaries,
i.e. 5.3% in Arizona and 9.3% in New Mexico.

From the above it should be apparent
that Arizona and New Mexico are truly diverse
in their floras and faunas, even when one
largely restricts the discussion of animals
to vertebrates. Thousands of species of in-
vertebrates also occur in the two states,
including especially terrestrial arthropods.
For example, Howe (1975) lists almost 700
species of butterflies from temperate North
America, and of these about one- third are re-
corded from New Mexico and somewhat higher
figure from Arizona.

In evaluating the biotic importance of a
region, one approach is through the considera-
tion of endemism, i.e. the degree to which
species are restricted to an area in question.
Both Arizona and New Mexico are host to endem-
ic plants and animals, including vertebrates
as well as invertebrates. Although I know
of no compendium of such species, several
examples illustrate some of the endemism.
For example, among vertebrates New Mexico
hosts the only known populations of such spe-
cies as the White Sands pupfish (Cyprinodon
tularosa) , Jemez Mountain salamander
(Plethodon neomexicanus ) , and Sacramento
Mountain salamander (Aneides hardii) . Both
states boast endemic plants as well, while
together they share a number of other endemics
that occur nowhere outside the Southwest, in-
cluding the minnow genera, Tiaroga and Meda,
in the Gila Basin.

Although endemism is an important means
of evaluating the biotic importance of an
area, other considerations also pertain. For
example, the kinds of assemblages of plants
and animals are important, and in these two
states virtually unique associations have
arisen because of the interdigitation and/or
mingling of diverse biotas. Such associations
are interesting and important form evolution-
ary, ecological, and other biological points
of view. Unique or unusual assemblages of

plants and animals provide scientists and
others the extended opportunity to understand
better our ecosystems and life itself. An
example of a notable biological assemblage is
the breeding avifauna of the lower Gila Valley
of New Mexico, where species characteristic
of the Sonoran, Mexican Plateau, and
Holarctic avifaunas occur side-by-side (Hubbard,
1971) . That fauna has been compared to another
in the ecologically similar San Juan Valley,
250 miles to the north and in the same drain-
age basin (i.e. Colorado River). Both avi-
faunas have similar numbers of species (i.e.
105 versus 112 in the Gila), but they differ
importantly; for example, only 58.7% of the
Gila species breed in the San Juan, while
only 64.8% of the species in the latter area
breed in the Gila (Schmitt, 1976).

The essence of the above comparisons is
that not only are Arizona and New Mexico
biotically diverse and host to certain endem-
ics, but they also show significant and im-
portant area-to-area differences in the com-
position of biotas occupying similar situa-
tions. Each river valley, mountain range, hot
spring, or alkaline playa is apt to differ from
those occurring nearby, and this fact alone
underscores even more the biotic importance of
these two states. This is not to imply that
other regions are lacking in biotic importance,
for such is not the case. However, Arizona
and New Mexico stand apart from most other
states in having both very rich floras and
faunas and in having many factors that promote
ecological departures from the "norm", i.e.
disjunct or limited habitats, varied biotic
sources, and so on.

Having established the credentials of the
Southwest in terms of richness and importance
of its floras and faunas, let us turn to the
question of how riparian ecosystems may be
important in perpetuation of these features.
In terms of any one group for which such ri-
parian ecosystems must be regarded as essential,
certainly no question exists that the most im-
portant would be fishes. I have already men-
tioned that Arizona hosts — or hosted — 32 native
species and New Mexico 59. Together these
total 75 species when combined, no fewer than
6 of which are federally endangered, i.e.
Colorado River squawfish (Ptychocheilus lucius),
humpback chub (Gila cypha) , woundf in (Plagopterus
argentissimus) , Gila trout (Salmo gilae) , Gila
topminnow (Poeciliopsis occidentalis) , and
Pecos gambusia (Gambusia nobilis) , plus one
species that is threatened, the Apache trout
(S. apache). In addition, the New Mexico De-
partment of Game and Fish lists 30 species of
native fishes as endangered in the state, in-
cluding the squawfish, Gila trout, topminnow,
and gambusia mentioned above. On a percentage

basis, about half of New Mexico's ichthyo-
fauna is regarded as endangered at the state
level, whereas 8 percent of the overall south-
western fauna is federally endangered.

It is obvious that riparian ecosystems
are of paramount importance in the survival
of native fishes in the Southwest, where the
vast majority of the species are riparian
(versus lacustrine) in their habitat occu-
pancy. The major threat to the survival of
these fishes involves degradation of the
required habitats, including lowering of the
water table, construction of dams, diversions,
and reservoirs, vegetation clearing, pollu-
tion, roads, grazing, and the introduction
of exotics. This degradation will no doubt
continue, for it is partly an outgrowth of
man's quest for water and the environments
that it fosters. There is little that the
dependent biota can do to stem this quest,
and man continues to take the aqueous spoils
and leave the biota high and dry. Obviously,
this approach cannot continue if the ichthy-
ological portion of the rich and important
biota of the Southwest is to persist.

Next to fishes, there is no single large
group of southwestern vertebrates so depen-
dent for survival on water, that essential
and basic element of riparian ecosystems.
Yet, there are aquatic plants and inverte-
brate animals that are just as dependent,
including invertebrates. Among the latter
are certain mollusks and arthropods, such as
Exosphaeroma thermophilum — an endemic crus-
tacean confined to a warm spring run near
Socorro, New Mexico. Some animal and plant
species are seasonally dependent of riparian
ecosystems, such as many amphibians which
breed in water. The exact numbers of non-
fish species dependent on aquatic habitats
in the area has not been determined, but it
is significant.

So far, the emphasis on the importance
of riparian ecosystems to the biota of the
Southwest has concentrated mainly on the
question of surface water, as in the cases
of fishes and of certain other animals and
plants. However, there are other riparian
features involved that should also be men-
tioned, and among the most important is the
vegetation characteristic of these ecosystems.
A great variety of plants utilize stream
courses in the Southwest, including both
obligate and facultative species. Typical
of the obligates are cottonwoods (Populus
spp.), willows (Salix spp.), alders (Alnus
spp.), and other broadleaf trees. Faculta-
tive species are those that invade stream
courses from other habitats, but which may
survive without riparian systems. Over 100

kinds of woody plants occur regularly in
floodplains in New Mexico, of which about 40%
are obligates (Hubbard, ms . ) .

Riparian plants are biologically impor-
tant from a number of standpoints. One
aspect of their importance is an individual
species, for some are restricted in range,
numbers, or both. For such species, degrada-
tion of the riparian ecosystem could be
especially detrimental, even critical to
survival. Conversely, for some such species
the continued availability of acceptable ri-
parian ecosystems is essential if survival
is to continue. Another aspect of impor-
tance is at the level of plant assemblages,
such as vegetational communities. The matter
of communities is especially important, for
a great deal of diversity exists among ri-
parian communities in the Southwest (Hubbard,
ms . ) and this deserves perpetuation. In
addition, the assemblage concept is impor-
tant from the standpoint of revealing
evolutionary, ecological, and other biological
information, such as any divergence among
fragmented populations. There is even a
historic (or prehistoric) consideration, in
that we may view the broadleaf assemblages of
trees and shrubs along many southwestern
streams as the major remnant of the ancient
Arctotertiary Flora that was dominant in
North America 50 to 100 million years ago.

Besides assemblages of plants, aggrega-
tions of considerable biological importance
are those involving animals as well. Perhaps
the aggregation that has attracted most
attention recently involves riparian vegeta-
tional communities and their attendant bird-
life. Although virtually unstudied until
recent decades, this biotic aspect of the
Southwest has now become better known, and
studies have included such streams as the
Verde (e.g. Carothers and Johnson, 1973) and
Colorado (Ohmart, mss.) in Arizona and the
San Juan (White and Behle, 1961; Schmitt,
1976) and Gila (Hubbard, 1971) in New Mexico.
All of these systems are extremely rich in
breeding birds; for example these two
New Mexico river valleys support 16-17% of
the entire breeding avifauna of temperate
North America over the course of only a few
score of miles.

The requirements of these avifaunas in-
volve both the aquatic and the vegetational
aspects of riparian ecosystems, but the
greater, direct dependence is on the plant
communities. Actually, on both the San
Juan and the Gila, aquatic habitats other
than the river per se are limited, and thus
few aquatic species are present. Considering
both aquatic and vegetational aspects together

as constituting together riparian habitats,
one finds that in the Gila Valley some 25.0%
of the 112 breeding bird species are restrict-
ed to them, while 24.1% occur in them primar-
ily (Hubbard, 1971). Neither group of bird
species, totalling 49.1% of the breeding
avifauna, would probably occur in the area
in the absence of these riparian habitats.
The figures for the 105 breeding species in
the San Juan Valley are similar, i.e. 26.5%
and 19.4%, or a combined total of 45.9%
showing riparian dependence (Schmitt, 1976).
In addition, 22.3% of the Gila species and
28.6% of the San Juan species also show some
to much utilization of riparian habitats,
and several species achieve maximal numbers
in them. Clearly, in these two areas the
presence of riparian habitats is extremely
important, and in essence they double the
avian diversity that might otherwise be
present. The same degree of importance no
doubt pertains elsewhere in the Southwest,
and is apparent that riparian ecosystems
play a key role in maximizing avian diversity
in the region.

Other riparian faunal-plant assemblages
seem to have been little studied, but there
is no doubt that others will show a strong
relationship between biotic diversity and
the presence of riparian ecosystems. For
example, although there appear to be fewer
southwestern mammals than birds with a strong
riparian dependence, nonetheless there are
certainly some species that do show this,
e.g. water shrew (Sorex palustris), Arizona
gray squirrel (Sciurus arizonensis) , beaver
(Castor canadensis) , meadow vole (Microtus
pennsylvanicus ) , muskrat (Ondatra zibethica) ,
raccoon (Procyon lotor) , mink (Mustela vison) ,
and otter (Lontra canadensis) . The same can
be said of reptiles, such as various turtles
(e.g. Kinosternon spp., Trionyx spp . ) , green
snakes (Opheodrys spp . ) , water snake (Natrix
erythrogaster) , and garter snakes (Thamnophis
spp.). On the other hand, amphibians show a
pronounced dependence on riparian — or at
least aquatic — ecosystems, because of the
general need of water for reproduction, e.g.
in various toads and frogs.

At this point, I believe that it has
become readily apparent that riparian eco-
systems are of paramount importance in
producing and maintaining a large degree of
the biotic diversity of the southwestern
United States. Although this importance is
perhaps most apparent in fishes and best
quantified in birds, it is clear that, for
many plants and animals, riparian ecosystems
are critical for them to flourish or even
survive in the region. I am hopeful that
more studies will be done to quantify this

importance, particularly with reference to
the degrees of dependency that exist among
biotic elements on these ecosystems and to
the niches that are occupied. It goes with-
out saying that the better we understand
these aspects, the better we can anticipate
the needs of the biota and manage for its
preservation. We have already witnessed
extremely widespread destruction and modifica-
tion of riparian ecosystems in the Southwest,
mainly as the result of man's activities over
the last several decades. As population
pressures and the demands on the riparian
ecosystem grow, we will be hard-pressed to
preserve what is left of the southwestern
riparian biota. Yet, if we do not meet the
challenge and achieve better preservation,
we will have allowed one of the richest of
all of the world's temperate floras and
faunas to have been diminished.

The time to obtain data and take positive
management steps is all too short, but at the
same time it is not too late to act. For
example, several important examples of ri-
parian ecosystems remain in the Southwest,
such as in the lower San Francisco Valley
in southwestern New Mexico and southeastern
Arizona. This particular tract lies in U.S.
National Forest, and with more enlightened
management it could provide along over 30
river miles of public land for the mainte-
nance of the very rich lowland riparian biota.
At the present time, grazing and off-road
vehicles are causing much damage to the tract,
which embodies everything about a wilderness
or wild river except in terms of management.
At higher elevations, more extensive ri-
parian ecosystems lie on public land and are
available for preservation, although manage-
ment again is frequently not accomplishing
this ,

The sad fact is that even public lands
have priorities' upon them that are not in
the best interest of preserving riparian eco-
systems, and changing this outlook for even
limited areas is often difficult. On private
lands the situation is generally worse, al-
though here and there some degree of preserva-
tion has been obtained for some tracts. There
is a critical need for a better education of
managers of both public and private lands
supporting riparian ecosystems as to their
importance and values, which range from eso-
teric to the practical. For example, points
of practical importance and value include the
role of vegetation in soil retention, effect
on climate, and in the harboring species that
provide both consumptive and non-comsumptive
recreation. These practical uses combine
with esoteric considerations to provide a
telling argument in favor of better preservation

of our native riparian ecosystems, fragmented
and misused as they have become. Hopefully,
individuals and agencies will soon join
forces to ensure such preservation, which is
long overdue and which cannot be delayed much
longer .

BIBLIOGRAPHY

Burt, W. H. and R. P. Grossenheider . 1964.
A field guide to the mammals. Houghton-
Mifflin Co., Boston, Mass.

Carothers, S. W. and R. R. Johnson. 1973.
A summary of the Verde Valley breeding
bird survey, 1971. Arizona Game and Fish
Department Land and Water Projects Inves-
tigations. Progress Report 7-1-76 to
6-30-72.

Conant, R. 1975. A field guide to reptiles
and amphibians of eastern and central
North America. Houghton-Mifflin Co. Boston,
Mass .

Eddy, S. 1957. How to know the freshwater
fishes. W. C. Brown, Dubuque, Iowa.

Findley, J. S., A. H. Harris, D. E. Wilson,
and C. Jones. 1975. Mammals of New Mexico.
Univ. New Mexico Press, Albuquerque.

Howe, W. H. 1975. The butterflies of North
America. Doubleday and Co. Garden City
N.Y.

Hubbard, J. P. 1970. Check-list of the
birds of New Mexico. New Mexico Orn.
Soc. Publ. 3.

Hubbard, J. P. 1971. The summer birds of
the Gila Valley, New Mexico. Nemouria
No. 2.

Kearney, T. H. and R. H. Peebles. 1970.

Arizona Flora, Univ. Arizona Press, Tucson.

Koster, W. J. 1957. Guide to the fishes
of New Mexico. Univ. New Mexico Press,
Albuquerque .

Lowe, C. H. (editor). 1964. The vertebrates
of Arizona. Univ. Arizona Press, Tucson.

Minckley, W. L. 1973. Fishes of Arizona.
Arizona Game and Fish Dept., Phoenix.

Phillips, A. R. , J. Marshall, and G. Monson.
1964. Birds of Arizona. Univ. Arizona
Press, Tucson.

Robbins, C. S., B. Bruun, and H. S. Zim.
1966. Birds of North America. Golden
Press, New York.

Schmitt, C. G. 1976. Summer birds of the
San Juan Valley, New Mexico. N. Mex. Orn.
Soc. Publ. No. 4.

Stebbins, R. C. 1966. A field guide to
western amphibians and reptiles.
Houghton-Mifflin Co., Boston.

Wagner, W. L. 1977. Floristic affinities
of Animas Mountains, southwestern New Mexic
Unpub. M.S. Thesis, Univ. New Mexico,
Albuquerque. XI+180 pp.

White, C. M. and W. H. Behle. 1961. Birds
of the Navajo Reservoir basin in Colorado
and New Mexico, 1960. Univ. Utah Anthro.
Paper 55:129-154.

Importance of
Riparian Ecosystems:
Economic Considerations1

I 2/
Kel Fox-

Efforts to preserve riparian habitat must
recognize man's growing demands to put this area
to other uses. Economic pressures, at conflict
with environmental concerns, pose an inevitable
threat to vegetation and wildlife. A compromise
in the balance of preservation and development
must be maintained.

Everyone here today feels the ri-
parian zone is important, or we would
not be holding this symposium. Most
of you want to preserve it. A few of

; you will admit, grudgingly, that it
needs to be managed. We all can agree
that huge areas of riparian habitat
have been removed by development-
oriented pressures. But few will ad-

■ mit , in this age of environmental con-
cern, that the same economic realities
of past decades still pose a threat to
much of the remaining habitat. These
threats are direct, as, for example,
through conversion to farming or
residential use, or indirect, as
through lowered groundwater levels
due to pumping.

These are the general conclusions
reached by the staff of the Arizona
Water Commission, who helped in the
I preparation of this paper. Now let's
examine them in detail.

When the white man first came to
this country, the low deserts were
broken by oases along the major rivers.
To farm he had to irrigate and the
bottom lands were the closest to water.
Mesquite bosques and cottonwood thickets

— Paper presented at the sym-
posium on Importance, Preservation
and Management of Riparian Habitat,
Tucson, Arizona, July 9, 1977.

— Kel Fox, Chairman, Arizona
Water Commission, 222 North Central
Avenue, Phoenix, Arizona 8500^

were cleared and put into production.
Southern Arizona hasn't been the same
since, but that's not necessarily bad.
The major metropolitan areas, wouldn't
be here but for the pioneering efforts
of those early farmers.

Clearing the riparian— zone
for agriculture is not a historic
phenomenon; it still goes on. The
bulk of the clearing is of phreato-
phytes, and most of that is along the
Colorado River, below Davis Dam on
Indian lands. The Indian reservations
along the river in Arizona and Calif-
ornia have decreed rights that allow
the consumption of 5^6,000 acre-feet
primarily for irrigation. In 1961
there were 80,000 acres of prime arable
land on these reservations that were
covered with phreatophytes . Clearing
will no doubt continue as the tribes
develop lands to use their water.

If Senator Kennedy has his way
with the Indian water rights bill,
250,000 acres of land on the -Gila
River Indian Reservation will be irri-
gated. It is a safe bet that some
of the new development will be on areas
now in phreatophytes.

— In this paper, riparian is used
in reference to vegetation which occurs
in or adjacent to drainage ways or their
floodplains which may be perennial or
ephemeral. Phreatophytes are those
riparian plants which habitually obtain
their water supply either directly or
through capillary fringe, from the
zone of saturation.

There are concentrations of private
land along all of the desert rivers.
Much of this land is suitable for ag-
riculture and is subject to clearing.

Farther up the watershed, the
higher elevation riparian zone tends
to be immune from these pressures.
Much of it is locked up in federal
administration. It probably will
remain unchanged. That small fraction
that is in private hands is often too
rough, too stony or otherwise unsuited
for agriculture. But that doesn't
mean it's safe from development pres-
sures .

It's prime country for subdivisions.
Pick up a map of a National Forest in
Arizona. Those white spots along the
major headwater streams are private
land, turn-of-the-century homesteads.
When the ranchers retire, those places
will be up for sale. In a land where
small lots in the pines sell for
thousands of dollars to greenery-starved
desert dwellers, lots in and along a
creek bottom, with real running water,
are pure gold. And cost as much.
Pave the road to the homestead and
it's good-by habitat, hello homesite.

Most of the headwater streams
are protected from this pressure by
ownership and topography. The mid-
elevation streams, however, are as a
rule, totally exposed. Privately
owned bottom lands along Oak Creek,
Tonto Creek, the Verde River and the
upper San Pedro are now actively being
offered for sale and converted into
retirement homes.

And what phreatophytes the farmers
didn't clear along the Colorado River
are subject to clearing or encroachment
by the subdivisions sprouting up there
for winter visitors.

Water yield improvement projects
have taken most of the blame for
removal of riparian habitat. However,
if you were to analyze the areas
cleared, you would find that the large
majority of the clearing was a result
of direct development pressures; it
was the land that was needed for other
purposes, rather than the water.
Clearing and/or conversion for down-
stream water yields was much in vogue
in the 50 's and 60's, however.

The Los Angeles report of the
Phreatophyte Subcommittee of the Pacific
Southwest Inter-Agency Committee of
August 1969 listed nearly two dozen
major clearing projects in Arizona.
A multitude of benefits were listed,
but the common thread was water. The
biggest of these was the 42,000 acre
project of the Bureau of Reclamation
along the Colorado River from Davis
Dam to the Mexican boundry.

The project was designed to
salvage water to enhance the supply
available to the C.A.P. The Bureau
surveyed the floodplain in this reach
of the river in 196l, identifying
155,000 acres of phreatophytes.
Environmental concerns have stalled
the project, as it has most of the
others, and it is not now being
actively pursued. That doesn't mean
that clearing of the floodplain won't
be accomplished however. Of the
nearly 65 percent of surveyed habitat,
98,000 acres, was on private, state,
or Indian lands. About half of the
proposed clearing program, 20,000
acres was on nonarable Indian lands.
All of this is subject to clearing
for irrigation or homesites. Only
6,000 acres were in wildlife refuges.

The term "water salvage" has
fallen from repute in describing the
benefits of these projects, but the
intent remains; to make more water
available for man's direct use. The
U.S. Geological Survey, in a study
reported in the Proceedings of the
1968 Arizona Watershed Symposium,
estimated annual evapotranspiration
of all phreatophytes in Arizona to
be about 940,000 acre-feet annually.
The monumental study of Ffolliott
and Thorud in 1974 indicates that
water yield in the riparian habitat
in Arizona might be increased as much
as 600,000 acre-feet per annum by
clearing and conversion. The Compre-
hensive Framework Study for the Lower
Colorado Region estimated 435,000
acre-feet could be salvaged each year
under a feasible management program
from the phreatophyte habitat. No
matter whose estimate you use — it's
a great deal of water, a potential
increase from 15 to 35 percent of
the state's dependable supplies.

Increasing the dependable supplies
will become more important as energy
costs for pumping and groundwater
levels increase. Changes in ground-
water law, mandated during this session
of the legislature, will no doubt be
aimed at reducing groundwater overdraft.
This will reduce local supplies still
further, which will intensify the demand
for increased dependable supplies from
other sources. Under these sorts of
pressures, environmental considerations
will have less importance.

Flood control, or more properly,
flood damage reduction, will become
more important with time as development
increases in or near riparian areas.

Salt cedar, with its prolific
growth, will rapidly take over a bare,
well watered site. Frequently, that's
in the main stem of the channel, and
as the stand develops, the channel's
conveyance is reduced. Not that high
flows, or flood waters, won't get
downstream, they do, but at the expense
of increased depth of flow and con-
sequent enlargement of the inundated
area.

The flows will eventually sweep
out the choking vegetation but only
after building up to sufficient head
and unnecessary levels of damage.
Clearing or maintaining a channel thus
duplicates Mother Nature's handiwork,
and avoids the incremental flood damage.

About a third of those two dozen
clearing projects mentioned earlier
claimed flood control as one of the
project's benefits. The clearing of
cottonwoods in the Verde Valley some
years back was designed to aid in flood
hazard reduction. The phenominal
growth of bottom land subdivisions in
that Valley may well work to reinstitute
such a program.

Environmental pressures have so
far worked to set aside the Corps of
Engineer clearing projects on the
Gila River downstream of Phoenix.
Sportsmen and hunter groups joined in
that cause, and as a result of their
efforts to save dove nesting and
roosting cover, they have been denied
hunting access to the privately owned
portions of the bird's habitat. As
more and more hunting areas are denied
them, sportsmen may well choose to
side with the Corps to work out a
compromise solution.

Conservationists are getting more
sophisticated with time. They had to,
as increasingly they've lost habitat
to esoteric causes. Mesquite bosques
and salt cedar thickets have died from
unseen causes throughout Southern
Arizona. Not fire, flood, or pestilence.
Not from developers or bulldozers or other
direct threats. Still the habitat died.

The cause — a decline in water
levels in response to pumpage under
developed areas miles distant. Ground-
water basins in southern Arizona underlie
the valleys from mountain to mountain.
The aquifers are quite productive, but
recharge is low. This means the cones
of influence from a pumping well, or
more commonly, hundreds of pumping wells,
spread wide, reaching seemingly protected
riparian habitat.

The famed San Xavier bosque south
of Tucson so died, as have large areas
in the thickets along the lower Santa
Cruz on the Gila Indian Reservation.
It's not done yet, either.

The riparian habitat of the upper
San Pedro, site of the proposed
Charleston Dam, so successfully opposed
by conservationists, is threatened
nonetheless. The expansion of Fort
Huachuca's mission has brought a
subdivision boom to the area. Water
level declines are accelerating, the
cone of depression rapidly expanding
toward the San Pedro.

As I said, conservationists have
become more sophisticated, but they've
yet to really embrace technology in
their fights. The impact of the Fort's
expansion in the riparian areas was
predicted by our Water Commission in
a computer model study of the Fort's
groundwater supplies. No one used
the information in assessing the
proposed expansion. Not that it would
have necessarily helped. The move had
strong local support. But at least
all impacts would have been discussed.

II' the march of progress can't be
averted, perhaps technology can be used
to facilitate the necessary compromises.
Recently the Maricopa County Flood
Control District acquired a small tract
of privately owned riparian habitat
along the lower Salt River for mandated
mitigation of the impacts of the county
flood control program. Will that parcel
eventually dry up and die as groundwater

withdrawals continue in the Valley?
No one knows — the eventuality wasn't
even considered. The Arizona Game and
Fish Department intends similar purchases
in the area. Should they utilize exist-
ing computer models to select tracts?
Most assuredly, if they wish to preserve
the habitat.

Recreation ranks right next to
mom and apple pie as typifying the
American. No where else is the
pressure on the riparian habitat, and
the wildlife it supports, more aptly
described by the character from Pogo :
"We have met the enemy, and he is us."
The physical presence of people is the
problem.

Those that sought to protect the
eagles at the Orme Dam site enlisted
the aid of the river's tubers in the
fight. But the tubers themselves are
a threat to the eagles, and no doubt
one day these bedfellows will part.

The riparian zone is especially
attractive to recreationists in this
water short land. Demand for this type
of recreation exceeds the supply,

causing continual pressure on the
developed sites, and insuring the
certainty of loss of more habitat.

So what does the future hold for
riparian habitat when faced with the
economic realities of life? Without
a doubt, more of the habitat will be
lost — there are too many pressures
for it to be entirely preserved.
Most of the loss will be on private
and Indian lands. Here the pressures
are felt most keenly, and the manage-
ment goal is not preservation.
Federal lands will probably be preserved
to a large extent, although demands
for flood control, grazing and recrea-
tion will require some concessions.

Society will come to realize that
our standard of living cannot be main-
tained without utilization of all of
our resources — today's topic of concern
is no exception. Our current preoccu-
pation with wildlands is a luxury we
can afford only because the wilderness
was subjugated. We must strike a
balance today, if only because society
did not do so in the past .

Vegetation Structure and Bird Use
in the Lower Colorado River Valley1

i I 2

Bertin W. Anderson and Robert D.lOhmart /

Abstract. — Data from riparian communities along the
lower Colorado River are used in discussing relationships
between the avifauna and the structure of plant communities.
Correlations between bird population parameters and vege-
tation structural characteristics were found to vary season-
ally. The mean habitat breadth of all species is narrowest
with respect to vegetative structure in winter and broadest
in summer; permanent residents occupy the structural types
more evenly than visitors. The habitat breadth of various
species is greater in summer than winter. Narrower habitat
breadths are accompanied by reduced habitat overlap among
the species in winter, suggesting that winter is potentially
the time of greatest stress. Permanent residents tend to be
less specialized with respect to structure than visitors.
These facts suggest that since winter requirements are
different from but equally as important as breeding require-
ments, they should receive at least equal attention. The
requirements of wintering visitors should receive special
attention because they showed a higher degree of habitat
specialization than permanent residents.

Since MacArthur and MacArthur (1961) first
reported the relationship between breeding bird
species diversity (BSD) and foliage height
diversity (FHD) much research effort has been
expended and quantitative data gathered in an
effort to explain relationships between various
bird population parameters, such as BSD, and
various features in the landscape (e.g. Balda
1969, Cody 1968, Karr 1968, Karr and Roth 1971,
Willson 1974). Although of great heuristic value
to the theoretician and manager alike, the
approach, as Balda (1975) stated, is too imper-
sonal; that is, it does not consider the biology
of individual species. This problem has been
successfully attacked in part by analyzing the
vegetation in the vicinity of singing males
during the breeding season (Anderson and
Shugart 1974, Conner and Adkisson 1977, James
1971, Whitmore 1975 a, b) . These investigators
have successfully applied multivariate analyses

1/ Paper presented at the symposium on
Importance, Preservation and Management of
Riparian Habitat, Tucson, Arizona 9 July 1977.

2/ Respectively, Faculty Research Associate
and Associate Professor of Zoology, Arizona
State University, Dept. Zoology and Center for
Environmental Studies, Tempe, Arizona 85281.

to the problem of habitat selection. Such
analyses have an advantage of reducing problems
associated with data interpretation, although
multivariate axes are often difficult to
precisely verbalize (Shugart et al. 1975). While
studies of the vegetation around individual
singing males undoubtedly provide valuable
information concerning the characteristics of
the breeding environment, it is only a small
step toward understanding species' habitat
requirements and there are definite problems
associated with the technique. Anderson (1974)
found that sub-adult male grosbeaks (Pheucticus
ludovicianus and P_. melanocephalus) tended to
establish territories in suboptimal portions
of the habitat. Furthermore, several of these
sub-adults failed to acquire mates although they
sang almost constantly. The suboptimal
portions of the habitat often included parks
and roadsides, and therefore, these birds were
much more easily found and observed than the
adult males which tended to inhabit denser
thickets. Inclusion of these males in greater
proportion than their occurrence in the popu-
lation could result in misleading conclusions
concerning their breeding habitat requirements.
Misleading conclusions could also be reached in
spatially restricted studies for many other spe-
cies, especially polygynous species such as the
Long-billed Marsh Wren (Cistothorus platensis)

(Verner and Engelsen 1970) , Yellow-headed
Blackbird (Xanthocephalus xanthocephalus)
(Willson 1966) , Red-winged Blackbird (Agelaius
phoeniceus) (Holm 1973, Linsdale 1938), and
Dickcissel (Spiza americana) (Harmeson 1974,
Martin 1971, Zimmerman 1966). In the south-
west desert there is an additional problem —
several species breed in a number of vegetative
types over a relatively long period of time
(four or five months). Obviously, analysis
should include males proportionate to their
occurrence in all vegetative types throughout
the breeding season. Breaking the analysis
into several spatial and temporal components
perhaps would be most appropriate.

Seasons other than the breeding period
have received little attention even though
they may be equally or more important than
the breeding season. Fretwell (1972) provided
cogent arguments to the effect that quality of
the wintering habitat is critically important
to survival of the breeding population and
that populations will decrease despite abundant
high quality breeding habitat when wintering
habitat is poor. Shugart et al. (1975)
presented two discriminant functions for seven
species from data gathered in fall and winter,
but such studies are rare.

Another factor which may affect the way
a habitat is utilized by a species, but which
is often neglected, is population size. Popu-
lation sizes, in turn, are affected by a number
of biotic and climatic factors. Some measure
of the population size should be related to
statements concerning species' requirements.
All of these factors can best be studied and
understood by censusing large areas several
times a year for a period of years. Unfortu-
nately, such studies require a relatively
large staff and are, therefore, costly; but
if quality data are the result, such studies
may prove to be the most efficient and inex-
pensive in the long run.

In this report we evaluate the importance
of vegetative structure at the habitat level
to birds in the valley of the Colorado River
from Davis Dam to the Mexican boundary, about
443 km. The relationship of entire avian
communities, guilds, and individual species
to structure of the vegetation will be discussed
on a seasonal basis. The importance of season
(especially winter), climate, and population
sizes will be presented. Although our data
are suited for multivariate analyses, we are
not prepared to present the results of such
analyses as they relate solely to structure
at this time. Such analyses are underway and
preliminary results are consistent with our
comments here.

In defining habitat we concur with
Whittakeret al. (1973) that terms such as
niche and habitat should be stabilized, and
we follow their recommendation in considering
the niche as an intracommunity variable and
the habitat as a broader concept usually encom-
passing more than one community. Communities,
in this report, refer to fairly homogeneous
areas with respect to dominant vegetation and
structure and range in size from 10 to 40 ha.
Niche requirements, as opposed to habitat
requirements, are not presented here because
of space limitations. Similarly, data concerning
dietary preference and feeding behavior must
be omitted from this report but will be available
at a later date.

CORRELATIONS

The relationships between bird population
parameters and the various vegetative parameters
were examined with regard to the spatial and
temporal dimensions of the environment. Further,
the diversity measures are of all birds and not
just breeding birds and are used to discuss the
value of vegetative structure in evaluating the
wildlife use values of an area. Each year was
divided into five seasons: winter included
December, January and February; spring included
March and April; summer included May, June and
July; late summer included August and September;
and fall included October and November.

Our study area encompasses about 4,828 km
between Davis Dam, located on the Nevada-Arizona
border, and the Mexican boundary south of Yuma,
Arizona. The riparian vegetation was divided
into six community types based on the dominant
plant species and into six structural types
based on the vertical profile of each community.
Methodology employed for determining structure
and for censusing birds is discussed elsewhere
in these proceedings (Anderson, Engel-Wilson,
Wells and Ohmart) .

Vegetation Parameters

Since the data did not fit a normal curve,
correlations were determined using the non-
parametric Kendall rank correlation (Sokal and
Rohlf 1969). Correlations between structural
variables are significant and positive in four
cases and significant and negative in one
instance out of ten comparisons (Table 1) .

Bird Density and Vegetation
Structural Parameters

Bird density in winter correlated with
vegetation at the 1.5 to 3.0 m level (Table 2).

Table 1. — Correlations between vegetative structural characteristics in the riparian vegetation
along the lower Colorado River Valley.

Height

0-0.6 m 1.5-3.0

>4.5 m

Total
Relative
Density

0-0.6 m
1.5-2.0 m
>4.5 m
Total
FHD

-0.076

-0.42**
0.49**

0.02

0.67**

0.54**

FHD

-0.12
0.11
0.35*
0.02

* and ** significant at p<0.05 and <0.005 respectively.

Table 2. — Correlations between bird population parameters and vegetation parameters at five seasons
in the lower Colorado River Valley.

Relative
Density of
Vegetation at

Dec 1974
Jan 1975
Feb

March
April

Season

May

June

July

Aug
Sep

Oct
Nov

BIRD DENSITY

0.6 m
1.5-3.0 m
>4.5 m
Total
FHD

BSD

0.6 m
1.5-3.0 m
>4.5 m
Total
FHD

BSD WITH 10% DOVES
0.6 m
1.5-3.0 m
>4.5 m
Total
FHD

SPECIES

0.6 m
1.5-3.0 m
>4.5 m
Total
FHD

-0.02

0.10
-0.16

0.16

0.04

-0.40*
-0.10
-0.14
-0.19
0.29*

-0.16
-0.02

0.23
-0.07

0.38*

-0.32*
-0.37*

0.00
-0.44**

0.32*

-0.25
0.05
0.14
0.04
0.13

-0.23
-0.39*
-0.13
-0.32*
0.11

-0.21
-0.25

0.15
-0.18

0.09

-0.30*
-0.34*
-0.17
-0.35*
0.09

-0.36*
0.42*
0.52*
0.39*
0.02

0.48*
-0.18
-0.12
-0.04

0.09

0.17
0.03
0.11
-0.04
0.43*

-0.24
0.20
0.17
0.22
0.31*

-0.22

0.31

0.18

0.26
-0.18

0.01

0.51**
-0.59**
-0.74**

0.03

-0.02

-0.35*

-0.52*

-0.35*

-0.25

0.04
-0.14
-0.28
-0.20
-0.13

0.18
0.04
0.12
0.03
0.19

-0.10
0.25
0.43*
0.16
0.63**

-0.01
0.14
0.23
0.09
0.45**

-0.12
-0.02

0.22
-0.07

0.58**

* and ** significant at p<0.05 and <0.005 respectively.

This is at least partly because honey mesquite
stands with volume at this level are correlated
with mistletoe. The presence of mistletoe adds
another dimension to the habitat and supports
several species which are nearly absent else-

where. In spring, bird densities were relatively
high in several structural types, thus density
was not correlated with any structural parameter .
Spring is a period of transition — winter resi-
dents are still present and summer residents

are returning. In the summer, after winter
residents have departed, the greatest numbers
of birds were found in areas with the greatest
total vegetation. Late summer is another
transition period; many summer residents depart
and wintering species arrive. In fall the
greatest numbers of birds were found in relativ-
ely open areas (low total density of vegetation)
as well as in areas with volume at the inter-
mediate levels. Explanations of this are too
numerous and conjectural to present here.

BSD and Vegetation Structure

In winter there was a significant negative
correlation between volume of 0 to 0.6 m and
BSD (Table 2). Areas with greatest vegetative
volume at the lower levels tended to be dominated
numerically by White-crowned Sparrows
(Zonotr ichia leucophrys) . There was a positive
correlation in winter with FHD. In the spring
BSD was significantly negatively correlated
with the densest areas with relatively dense
vegetation at 1.5 to 3.0 m. In the summer BSD
was significantly correlated with density of
vegetation in the lower layer (0 to 0.6 m) .
Doves in summer in the more lush areas have
such overwhelming numerical dominance that BSD
in such areas is suppressed. In fall, after
many doves had migrated, BSD and FHD were
positively correlated. Removing 90 percent
of the doves from the data revealed a corre-
lation between FHD and BSD in winter, summer,
and fall but not in spring or late summer.
The number of species is correlated with FHD
in winter, summer, and fall. Number of species
and BSD are correlated with each other at all
seasons (Anderson and Ohmart, unpubl. data) .

Comparison of Years

Correlations between the bird parameters
and structural features were nearly the same
each year (Table 3) . The greatest deviation
was found in summer 1976 when total bird density
was significantly correlated with FHD, contrary
to findings in 1975. This may have occurred
because doves, which are most numerous in
areas of low or moderate structural diversity,
were at least -40 percent fewer in numbers in
most areas in 1976.

In summary, it would
vegetation is more import
early summer than at othe
This is generally true, b
to 1500 doves per 40 ha t
the importance of dense v
because of large numbers
Sparrows and Phainopeplas
the reaction of many spec
is obscured in the winter
Sparrows are primarily gr
found in relatively open

appear that dense
ant to birds in the
r times of the year,
ut the presence of up
ends to exaggerate
egetation. Similarly,
of White-crowned

(Phainopepla nitens) ,
ies to sparse vegetation

White-crowned
anivorous and are
areas, whereas

Phainopeplas are found wherever there is mistle-
toe, which is most abundant in areas of moderate
density .

BSD was not correlated with FHD in summer,
but this is misleading for the excess of doves
in summer masks the real diversity of birds
found in areas where doves predominate. This
relationship is revealed by using only 10 per-
cent of the doves in calculating BSD or by
simply considering the number of species
involved. By using either of the above alter-
natives, the data suggest a correlation in
summer between BSD and FHD.

The correlation between BSD and FHD in
this study does not provide as good a fit to a
regression line as in a number of other studies,
thus indicating that along the lower Colorado
River only a relatively small part of the
observed diversity is due to structural complex-
ity. That it is significant in three of five
seasons, however, suggests that structural
complexity does have important management
implications. Obviously, BSD must be evaluated
within the context of other population parameters ,
such as the number of species and the density
of vegetation.

HABITAT BREADTH BY STRUCTURE OF VEGETATION

The extent to which each species occupies
the various structural types of vegetation is
referred to here as a species' habitat breadth
by structure (HB ) . This is calculated by
HB = -Ep.log p. where p. is the proportion of

g i e i i
ividuals found in the ith structural type.

This parameter is independent of the distribution

in which we designate habitat breadth for

vegetative type (HB ). For example, a species

equally abundant in all six structural types

in cottonwood-willow communities would have

the same HB as one found in equal numbers in

all dominant vegetation of all structural types.

The former would, of course, have HB of 0.0

while HB of the latter would be log of 6 or
1.8 (based on six dominant community types in
riparian vegetation) . HB and HB are calcu-
lated for each species ani the means for all
species occurring at densities of at least
1/40 ha each month and for each season are
calculated .

Mean HB

Mean HBg for each season (fig. 1) reveals
four things of potential ecological importance
as related to seasons. First, high summertime
values in mean HBS are followed by a lower
mean in the cooler time of year. Second, the
smallest mean HBS occurred in the winter of
1974-75. Third, visiting species tend to have
lower mean HBg than permanent residents.

Table 3. — Correlations between bird population parameters and structural features in summer 1975
and 1976 along the lower Colorado River Valley.

Height Total

Relative

0-0.6 m 1.5-3.0 m >4.5 m Density FHD

DENSITY

1975 -0.36* 0.42** 0.52** 0.39** 0.02

1976 -0.55** 0.44** 0.67** 0.47** 0.36**

BSD

1975 0.48** -0.18 -0.12 -0.04 0.09

1976 0.49** -0.22 -0.39* -0.15 -0.15

BSD WITH 10% DOVES

1975 0.17 0.03 0.11 -0.04 0.43**

1976 0.04 0.09 0.31 0.17 0.34*

SPECIES

1975 0.24 0.20 0.17 0.22 0.31*

1976 0.15 0.20 0.28* 0.24 0.24

* and ** significant at p<0.05 and <0.005 respectively.

ALL RESIDENTS VISITORS

Finally, HBg was not reduced in fall 1976; in
fact, it increased for permanent residents.

The relatively low mean in cooler seasons
coincides with the time of year when produc-
tivity is lowest. As insects decline in number,

they probably become more restricted in their
distribution (Raitt and Pimm 1976); this disjunct
distribution is probably mirrored by the
restricted distribution of insectivorous birds
which tended to have narrower HBS. Similarly,
the patchy distribution of seeds in the sparser

j I

areas is reflected by the relatively narrow
HB^ of wintering seed eating birds.

The low HBS in winter 1974-75 appears to
be accounted for in that the winter was signif-
icantly colder with more days of frost and was
windier than subsequent winters (Anderson and
Ohmart, MSjV). We predict that habitat breadth
will increase in a given season only if
resources become so abundant that consumption
is limited more by the birds' ability to harvest
than by competition during the season or if
regulation during that season is by predation
(Anderson and Ohmart, MSjV). Insects were
probably less abundant (Raitt and Pimm 1976)
and this, in addition to the fact that the
colder windier conditions required more energy
per day per bird, appears to have restricted
HBS during that period. The subsequent winter
was milder than average and mean HBS was
correspondingly higher. Visitors tend to have
narrower mean HBS than do permanent residents,
indicating that they are more sensitive to
features of the vegetation structure on the
average than permanent residents (see below).

Unusually heavy rainfall in September
followed by above average fall temperatures
may have resulted in greater insect productivity
than normal and allowed expanded HBS in the
fall of 1976.

In summary, these data indicate that the
use of structure is not static. Not only is
there variation in mean HBS from species to
species, but from season to season and year to
year, reflecting seasonal and annual differences
in climate and other factors. In addition,
permanent residents and visitors adapt to
vegetation structure in different ways. HBS
and HBV and their ecological significance are
discussed at greater length, especially with
respect to the theory of competition in
Anderson and Ohmart (MS-^/).

FORAGING GUILDS AND VEGETATION STRUCTURE

(Myiarchus tyrannulus) was most numerous in
type I. The mean HBS for this group in the
summer was 1.435.

In winter the Say Phoebe (Sayornis saya)
was the only flycatcher occurring in densities
of at least 0.5 per 40 ha. Its HBS was 1.363
which is 5.7 percent smaller than the summer
average for the group.

Medium-sized Insectivores

In summer a group of six medium-sized
insectivores were present at densities of at
least 0.5 per ha. Among them, the Blue
Grosbeak (Guiraca caerulea) , Cactus Wren
(Campy lorhynchus brunneicapillus) , and Northern
Oriole (Icterus galbula) had relatively large
HBS (Table 4) . The Yellow-breasted Chat
( Icteria virens) was intermediate and the
Summer Tanager (Piranga rubra) and Yellow-
billed Cuckoo (Coccyzus amer icanus) were
specialists, being found most extensively in
structural type I. The average HBS for the
group was 1.413.

The Cactus Wren is the only member of
this group present in winter. Its winter HBS,
1.695, is larger than the average for the group
in summer but is about 5 percent smaller than
its own summer HBS.

Ground Feeders

In the summer the ground feeders included
three permanent resident species (Table 4) .
All were rather evenly distributed throughout
the structural types of vegetation (Table 4) .
The average HBS was 1.650.

In winter the three ground feeders were
somewhat less general in their distribution
(Table 4) . The Abert Towhee (Pipilo aberti)
was found most frequently in structural type I.
The average HBg was about 5 percent smaller than
in summer.

The Flycatcher Guild

In summer the Ash-throated Flycatcher
(Myiarchus cinerascens) was the numerically
dominant flycatching species in all structural
types and was fairly evenly distributed with
somewhat greater density in structural type II
(Table 4) . The Western Kingbird (Tyrannus
verticalis) was found most frequently in
type II. The Wied Crested Flycatcher

3/ Manuscript in preparation discussing
seasonal changes in habitat breadth and overlap
among birds along the lower Colorado River.

Small Insectivores

In the summer there were three insectivorous
species weighing less than 15 g (Table 4) .
All three (two were permanent residents, one
visitor) were widely distributed among the
structural types (Table 4) as reflected by
their mean HBS of 1.694. The Lucy Warbler
(Vermivora luciae) was numerically dominant
in types I and II. The Verdin (Auriparus
f laviceps) was numerically dominant in all the
other types. In winter the number of small
insectivores increased to five. Among them
the Yellow-rumped Warbler (Dendroica coronata) ,

Table 4. — Densities (N/40 ha) of various birds by vegetation structure for winter 1975-76 and
summer 1976.

Structure Type

Species Groups I II III IV V VI HBg J

Flycatchers
Winter

Say Phoebe 1.00 2.67 3.00 0.60 - 0.50 1.37 0.765

Summer

Ash-throated Flycatcher

7.75

11.20

982

Wied Crested Flycatcher

586

Western Kingbird

0.50

1.50

837

Medium-sized Insectivores
Winter

Cactus Wren

949

amine r

Northern Oriole

921

Summer Tanager

458

Blue Grosbeak

988

Cactus Wren

971

Yellow-billed Cuckoo

647

Yellow-breasted Chat

742

Ground Feeders
Winter

Abert Towhee

22.00

.67

831

Crissal Thrasher

1.00

.33

938

Gambel Quail

.67

860

White-crowned Sparrow

• 0

636

Sage Sparrow

535

ammer

Abert Towhee

14.50

.33

50.

943

Crissal Thrasher

.33

691

Gambel Quail

0.50

.00

10.

898

Small Insectivores
Winter

Verdin

859

Black-tailed Gnatcatcher

882

Brown Creeper

882

Ruby-crowned Kinglet

69.

17.

770

Yellow- rumped Warbler

62.

10.

647

Bewick Wren

714

amine r

Verdin

14.

13.

954

Lucy Warbler

21.

26.

10.

938

Black-tailed Gnatcatcher

943

Woodpeckers
Winter

Ladder-backed Woodpecker

910

Common Flicker

893

ammer

Gila Woodpecker

798

Ladder-backed Woodpecker

926

Common Flicker

639

Brown Creeper (Certhia f amiliaris) and Ruby-
crowned Kinglet (Regulus calendula) , all
visitors, were the most specialized (Table 4).
The Black-tailed Gnatcatcher (Polioptila
melanura) and Verdin, both permanent residents,
were the structural generalists. The mean HBS
of the group was 18 percent lower than in the
summer (Table 4) .

Woodpeckers

The Ladder-backed Woodpecker (Picoides
scalaris) is the most generalized woodpecker
both in winter and' summer (Table 4). The Gila
Woodpecker (Melanerpes uropygialis) is the
most specialized in winter. The Flicker is
intermediate for both seasons. The average
HBS for the group was the same in summer and
winter (Table 4). No visitors occurred in
either season.

STRUCTURAL SPECIALISTS

Some species prefer certain structural
characteristics within a particular community
type; a few seem to be more specialized as to
structure and less specialized with regard to
the type of dominant vegetation. For this
analysis we have used the number of each species
in the structural types in each dominant vegeta-
tive type and calculated simple correlations
between the bird numbers and particular struc-
tural characteristics. The species we attempted
to find were those preferring, for example,
dense vegetation at 3.0 m and not discriminating
between different kinds of dominant vegetation.
A structural specialist found only in type I
cottonwood-willow would not qualify as an overall
structural specialist but would be a specialist
in two dimensions — structure and dominant
vegetation.

SPECIES OVERLAP IN USE OF STRUCTURE

Because two species have similar HBg does
not necessarily mean that the overlap in habitat
breadth by structure (R ) is 100 percent. We
quantified overlap between pairs of species
using Horn's (1966) formula. The method is
discussed more fully elsewhere in these pro-
ceedings (Anderson, Engel-Wilson, Wells, and
Ohmart) . The overlap in the five groups of
ecologically similar species was greatest in
the summer (Table 5) . Mean overlap of all of
these species was significantly smaller in
winter than in summer.

The analysis of HBg and habitat overlap
by structure (R0 ) points out several signif-
icant ecologicalsconsiderations . One is that
winter and summer visitors tend to be more
specialized with respect to their use of vege-
tation structure as well as the dominant types
of vegetation (Anderson and Ohmart, Ms£/) than
permanent residents. Both visitors and perma-
nent residents tend to be more specialized in
the use of vegetative structure in the cool
times of the year — woodpeckers being the
exception. This suggests that winter require-
ments may be different from summer requirements.
Our findings corroborate Fretwell's (1972)
prediction that winter residents in a given
area will have larger populations and be more
specialized than the local populations of
permanent residents. In the Colorado River
Valley winter residents do tend to have the
largest populations (Anderson and Ohmart, MS_/)
but are rather habitat specific. This indicates
that if these rather restricted habitat types
are destroyed, a relatively large breeding
population from srome more northerly area could
be reduced or eliminated.

Some Simple Cases

The Abert Towhee showed a slight, but not
statistically significant, preference in summer
for dense vegetation. In winter, however, the
correlation is significant (Table 6) . The
towhee cannot be considered a structural
specialist because it is also fairly common
in relatively sparse vegetation, but it does
seem to show a preference for dense vegetation
and this correlation becomes stronger in winter.

The Summer Tanager shows a significant
(Table 6) correlation with vegetation taller
than 9 m. Although much of the vegetation above
this height is cottonwood or willow, when salt
cedar or other vegetation is available, the
tanager uses it.

The Western Kin
ence for tall vegetati
tanager which is restr
area, the kingbird occ
are only a few tall tr
numbers in areas where
The Ruby-crowned Kingl
reaches peak densities
(Table 6) . It cannot
specialist, however, b
other structural types
vegetation completely

ird also shows a prefer-
on (Table 6) . Unlike the
icted mainly to the denser
urs in areas where there
ees but reaches peak

tall trees are dense,
et, a winter visitor,

in the tallest vegetation
be considered a structural
ecause it also accepts

including rather sparse
lacking in tall trees.

The three woodpeckers (Ladder-backed,
Gila, and Common Flicker) are all significantly
correlated with foliage greater than 9 m
(Table 6). The Ladder-backed Woodpecker,
while showing a preference for areas with taller
frees, was also quite common in some structural
types totally lacking in tall trees. The other
two, however, can be considered structural
specialists .

Table 5. — Overlap in use of vegetation structure by various groups of species in summer and winter
along the lower Colorado River.

Overlap Matrices

Mean R

Summer Winter

Flycatchers

Summer

WK.

Ash-throated Flycatcher

U • yOl

n 7^7

Western Kingbird

n £9 1
U . DZ X

Wied Crested Flycatcher

n 7RD

Winter

Say Phoebe

0.000

Medium— sized Insectivores

Summer

b 1

TIP

VRP, !

I DL-u

Northern Oriole

n 7/1/1

n Q7/1
u . y / ^

u . yoo

U . 07 J

Summer Tanager

u . 0 / 0

n ah?

U . OUZ

n Q A A

U • 0O0

Blue Grosbeak.

U . 7 J J

U.Oj/

Ui ODj

Cactus Wren

n 7 a q

u . ouz

Yellow— billed Cuclcoo

n Qf>?

Yellow— breasted Chat

n an

Winter

Cactus Wren

n nnn

Small Insectivores

Summer

Black-tailed Gnatcatcher

u. y 11

u. y;m

Verdin

n qoo
u. oyz

Lucy Warbler

u . yzo

Winter

~Q C

Ruby-crowned Kinglet

C\ (~. QQ

u . boy

C\ QQ 1

U. /UU

n 7 on

u. /yu

Yellow— rumped Warbler

n 7 0 q
U . //o

n Q 7 A

U . / J.4

n oi q
u . z±y

Black— tailed Gnatcatcher

u . yz 0

u . y ji

U . Zoo

verdin

n i a A

Bewick Wren

Brown Creeper

n 7QQ

u . / y 0

Ground Feeders

Summer

Abert Towhee

0.938

0.885

Crissal Thrasher

0.971

Gambel Quail

0. 931

Winter

Abert Towhee

0.659

0.892

0.509

0.527

Crissal Thrasher

0.833

0.808

0.825

0.748

Gambel Quail

0.928

White-crowned Sparrow

Sage Sparrow

0.748

Mean of all R 's

0.862(.

115) 0.602(.312)

p<0.01

* Abreviation for the species

in the vertical column

e.g.

WK^Western

Kingbird;

V=Verdin ; etc .

The relationship between the Yellow-
breasted Chat and structure is more complex.
The chat, a summer visitor, seems to prefer
areas with vegetation at both 3 m and >9 m
(Table 6) . It occurs in very low numbers or
1 is absent from areas lacking at least moderate
development in these two layers and is clearly

associated with vegetation of a specific
structural configuration.

The Crissal Thrasher (Toxostoma dorsale) ,
a permanent resident, seems to prefer areas
which are dense at 3 m but which lack vegetation
above 9 m. The line of prediction (Table 6)

Table 6

— Species which show preferences for vegetation structure along the lower Colorado River.

Correlated Prob. of

Species

with relative

Season

Regression

Correl .

assoc . being

volume at

Equation

Coef .

due to chance

Abert Towhee

Total Volume

Y=15.65x+8.51

0.375

1. 64<0. 100

Abert Towhee

Total Volume

Y=9.86x-0.62

0.580

2.38<0.050

Yellow-breasted Chat

3m+29m

Y=25. 54x-1.92

0. 752

4.63<0.001

Summer Tanager

>9m

Y=21.94x+0.48

0.731

4.58<0.000

Crissal Thrasher

3m-<9m

Y=24.75x+2.11

0.726

3.33<0.000

Ruby-crowned Kinglet

^9m

Y=17.30x+7.09

0.910

5.22<0.001

White-winged Dove

3m+4 . 5m+9m

Y=3.43x-10.36

0.782

3.58<0.002

Mourning Dove

3m+4 . 5m+9m

Y=316.30x+50.05

0.547

2.51<0.020

Yellow-rumped Warbler

0.1-0. 6m

W 1974-75

Y=266.74x-41.83

0.599

2.32<0.050

Yellow-rumped Warbler

>9m

W 1975-76

Y=93. 09x+3. 09

0.913

3.76<0.002

was obtained by subtracting the relative volume
greater than 9 m from that at 3 m. Since
thrashers occur commonly but in somewhat re-
duced numbers in vegetation with other struc-
tural configurations, they were not considered
structural specialists.

Some Complex Examples

The White-winged Dove (Zenaida asiatica)
and the Mourning Dove (Zenaida macroura) showed
very complex relationships with vegetative
structure. Both species reached greatest
nesting densities in areas with relatively
dense vegetation at 3.0, 4.5, and 6.0 m and
a lack of vegetation above 9 m (Table 6) .
The highly significant regression line was
obtained by subtracting the relative volume
above 9 m from the sum of the relative volumes
at the other three layers. The White-winged
Dove was more of a specialist in this regard
as nesting densities are low in other struc-
tural types. Mourning Doves on the other hand
reached moderate densities in other types, too.
The White-winged Dove seemed to suffer less
nest predation under these conditions of
vegetation structure (Butler 1977).

A Special Case

In winter 1975-76 Yellow-rumped Warbler
densities were found to be significantly
correlated with the volume of vegetation 0 to
0.6 m (Table 6); the following winter there
was a significant positive correlation with
volume above 9 m but not with volume at 0 to
0.6 m (r = 0.2, p<0.05). While this may appear
to defy explanation, it is apparently related
to climate. In the winter of 1975-76 the
Yellow-rumped Warbler population was about the
same in type I but much reduced in sparser areas
lacking tall vegetation. It seems possible that
the areas with tall trees are in fact preferred;

but when the limited amount of this structural
type is filled, the excess goes to the sparser
areas. A constellation of factors including
climate and food supply are probably important
in determining the number of Yellow-rumped
Warblers which move into and winter in the
Colorado River Valley.

The Yellow-rumped Warbler was considered
a structural specialist in winter. Obviously,
erroneous conclusions could be drawn if only
one year's data or data from only one community
type had been used in analyzing Yellow-rumped
Warbler wintering habitat requirements.

Significantly, of the nine species which
show structural preferences, six are visitors
and three are permanent residents in spite of
the fact that the number of species of permanent
residents and visitors are present in about
equal numbers. One of the permanent residents
showed a preference only in winter. All of the
species showing greatest structural preference
are visitors.

CONCLUSIONS

From data presented here, correlations
between bird population parameters and vegetation
structural characteristics vary seasonally in
the lower Colorado River Valley. Although the
relationships to structure were considered on
a rather coarse-grained level in this report,
the same trends are apparent at finer levels
of distinction as well as for other vegetative
characteristics (Anderson and Ohmart, unpubl.
data) . Habitat breadth is narrowest in winter
and broadest in summer; permanent residents
occupy the habitat more evenly than visitors.
RD of the various species are greater in
summer than winter. Narrower HBS and reduced
Ros in winter suggest that winter is potentially
the time of greatest stress. Permanent residents
tend to be less specialized in structural

preference than winter visitors. These facts
have management implications. First, since
winter requirements may be different but of
equal or greater importance than summer
(breeding) requirements, they should receive
at least equal attention. The requirements of
wintering visitors should receive particular
attention for they tend to be specialists with
large populations. If the portion of the
habitat in which they specialize is destroyed
or damaged, its loss could mean total loss of
a breeding population. Finally, the require-
ments of summer residents also need special
attention as they too tend to be specialized —
although probably not to the same degree as
winter visitors.

ACKNOWLEDGEMENTS

We wish to thank the many field biologists
who have helped in collecting data. We are
grateful to Jack Gildar for computerizing the
data. The efforts of the secretarial staff

;i in typing early drafts and Penny Dunlop and
Katherine Hildebrandt in typing the final

; manuscript are greatly appreciated. Linda
Cheney kindly prepared the illustrations.
We thank Jane Durham, Jake Rice, James Bays,

»: and Jeannie Anderson for critically reading
early drafts of the manuscript. The research

I was funded through grant number 14-06-300-2415

| from the U.S. Bureau of Reclamation.

LITERATURE CITED

Anderson, B. W. and R. J. Daugherty.

1974. Characteristics and reproductive
biology of grosbeaks (Pheucticus) in
the hybrid zone in South Dakota. Wilson
Bull. 86:1-11.
Anderson, S. H. and H. H. Shugart.

1974. Habitat selection of breeding birds
in an east Tennessee deciduous forest.
Ecology 55:828-837.
'• Balda, R. P.

1969. Foliage use by birds of the oak-
juniper woodland and ponderosa pine
forest in southeastern Arizona. Condor
71:399-412.

1975. Vegetation structure and breeding
bird diversity. Proc. of the Symp. on
Mgmt. of Forest and Range Habitats for
Nongame Birds, pp. 59-80.
I Butler, W.

1977. A White-winged Dove nesting study in
three riparian communities on the lower
Colorado River. M.S. thesis. Ariz.
State Univ. , Tempe.

Cody, M. L.

1968. On the methods of resource division
in grassland bird communities. Am. Nat.
102:107-147.
Conner, R. N. and C. S. Adkisson.

1977. Principle component analysis of

woodpecker nesting habitat. Wilson Bull.
89:122-129.
Fretwell, S. D.

1972. Populations in a seasonal environment.
Princeton Univ. Press, Princeton, N.J.

Harmeson, J. P.

1974. Breeding ecology of the Dickcissel.
Auk 91:348-359.

Holm, C. H.

1973. Breeding sex ratios, territoriality,
and reproductive success in the Red-winged
Blackbird (Agelaius phoeniceus) . Ecology
54:356-365.

Horn, H. S.

1966. Measurement of "overlap" in comparative
ecological studies. Am. Nat. 100:419-424.
James, F. C.

1971. Ordination of habitat relationships
among breeding birds. Wilson Bull.
82:215-236.
Karr, J. R.

1968. Habitat and avian diversity on strip-
mined land in east-central Illinois.
Condor 70:348-357.

Karr, J. R. and R. R. Roth.

1971. Vegetation structure and avian diver-
sity in several New World areas. Am.. Nat.
105:423-435.
Linsdale, J. M.

1938. Environmental responses of vertebrates
in the Great Basin. Amer. Midi. Nat.
19:1-206.

MacArthur, R. D. and J. W. MacArthur.

1961. On bird species diversity. Ecology
42:594-598.
Martin, S. G.

1971. Polygyny in the bobolink: Habitat
quality and the adaptive complex. Ph.D.
disser. , Oregon State Univ., Corvallis.
Raitt, R. J., and S. L. Pimm.

1976. Dynamics of bird communities in the
Chihuahuan Desert, New Mexico. Condor
78:427-446.

Shugart, H. H. , S. H. Anderson, and R. H. Strand.

1975. Dominant patterns in bird populations
of the eastern deciduous forest biome.
Proceedings of the Symp. on Mgmt. of Forest
and Range Habitats. pp. 90-95.

Sokal, R. R. and F. J. Rohlf.

1969. Biometry. W. H. Freeman and Co.,
San Francisco.

Verner, J. and G. H. Engelsen.

1970. Territories, multiple nest building,
and polygyny in the Long-billed Marsh Wren.
Auk 87:557-567.

Whitmore, R. C, Jr. Willson, M. F.

1975a. Habitat partitioning in a community 1974. Avian community organization and

of passerine birds. Ph.D. disser. , habitat structure. Ecology 55:1017-1029,
Brigham Young Univ., Provo, Utah.

1975b. Habitat ordination of the passerine
birds of the Virgin River Valley, south-
western Utah. Wilson Bull. 87:65-74.
Whittaker, R. H. , S. A. Levin, and R. B. Root.

1973. Niche, habitat, and ecotope. Am. Nat.
107:321-338.

1966. Breeding ecology of the Yellow-headed
Blackbird. Ecol. Monog. 36:51-77.

Zimmerman, J. L.

1966. Polygyny in the Dickcissel. Auk
83:534-546.

A Riparian Case History:
The Colorado River !q^o#

t \ 12/

Robert D. Ohmart, Wayne 0. Deason, and Constance Burke—

Abstract. — Historically to present cottonwood
communities have declined in abundance along the lower
Colorado River to the condition that the future of this
natural resource is precarious. Avian species showing
strong specialization to cottonwood communities may be
extirpated should the cottonwood community be lost from
the river. Only through the concern and action by responsible
agencies can we assure the persistance of this natural
resource .

An overview of the ecological changes that
have occurred on the lower Colorado River can
be obtained by selecting an important plant
community and examining its condition through
time. To adequately describe the lower
Colorado River riparian ecosystem and discuss
the ecological changes from our first written
records (1539) to the present is not possible
within the space constraints of these pro-
ceedings. Therefore, we have elected to
analyze an important plant association in this
ecosystem and examine its distribution from the
early 1600 's to present. Although we refer to
ithis association as the cottonwood (Populus
f remontii) community, it frequently occurs
mixed with such species as willow (Salix
gooddingii) and/or screwbean mesquite (Prosopis
pubescens) and infrequently with such species
as arrowweed (Tessar ia ser icea) , honey mesquite
(Prosopis julif lora) and recently the introduced
salt cedar (Tamarix chinensis) .

Some avian species that inhabit this
community along the lower river appear to be
very specific in their habitat requirements
(And erson and Ohmart 1975). Consequently, as
the unit area of this plant community changed,
in all probability, so did the abundance of

1/ Paper presented at the symposium on
Importance, Preservation and Management of
Riparian Habitat, Tucson, Arizona, 9 July 1977.

2/ Respectively, Associate Professor of
Zoology, Arizona State University, Dept.
Zoology and Center for Environmental Studies,
Tempe, Arizona; Biologist, Bureau of
Reclamation, Boulder City, Nevada; and
Graduate Student, Biogeography , Oregon State
University, Corvalis, Oregon.

these habitat-specific species. By examining
this biotic community from past to present,
we gain an appreciation of the areal changes
that have occurred and can make better
predictions as to the future of this community
type. This analysis should be helpful in
providing impetus for management decisions
relative to the cottonwood community along the
river .

It is tempting to speculate that early man
lived in harmony with his environment and that
the earliest descriptions of the plant communities
along the lower Colorado River reflect the
natural environment unaltered by man. Because
of space constraints, we will assume this is
true, but it must be kept in mind that man
could have had a strong influence on his
environment through burning and other habitat
alterations, especially at the local level.

If we assume that the Indians did not
drastically modify the environment, we can
examine alterations, classed as natural or
unnatural, brought about by the Spaniards and
later the Anglo-Americans. It has been argued
that man is a natural part of his environment
and that alterations brought about by his
activities are as natural as changes produced
by other animal activities (Malin 1956) . This
is academic when it is considered that natural
resources are finite, and regardless of the
causes of degradation and loss, if we value
these resources, then we must preserve and
manage them for their continued existence.

INDIANS AND SPANIARDS

size desired." (Dunne 1955: 31).

Long before European man viewed the waters
of the lower Colorado River, a variety of
Indian cultures evolved and became established
along its banks. However, they left no written
record, so Spanish documents provide our earliest
information about the region. The Spaniards
claimed the region of the lower Colorado for
over 250 years. Interested in converting the
Indians to Christianity, discovering a land
passage to the South Sea (Gulf of California)
and acquiring mineral wealth, their missionary-
military expeditions explored the Pimeria Alta.
These early explorers left diaries often
containing descriptive information on the areas
they traveled.

In February of 1699, Father Eusebio
Francisco Kino, accompanied by Juan Mathes
Manje, an excellent diarist, viewed the junction
of the Gila and Colorado rivers from a distance
of 15-13 miles. Manje wrote: "...we plainly
saw at a distance of six or seven leagues
[1 league =2.5 miles] the banks and junction
of the very great Rio Colorado with this one
[the Gila] , grown with dense groves with new
leaves and although we saw it at a long distance
the groves appeared to us to be more than a
league wide,..." (Bolton MS 203:20).

Kino, a student of both the Faith and
geography, spent twenty years in the area
known as the Pimeria Alta, delineated on the
east by the San Pedro River, the north by the
Gila Valley and on the west by the Rio Colorado
and the Gulf of California. Two of his major
goals were to convert the Indians to Christianity
and to find a land passage to California to
prove it was not an island. In 1700, Kino
was again near the junction of the Colorado
and Gila rivers. He ascended a hill in hopes
of viewing the Gulf of California: "...but
looking and sighting toward the south, the
west, and the southwest, ...we saw more than
thirty leagues of level country, without any
sea, and the junction of the Rio Colorado with
this Rio Grande. . . , and their many groves and
plains." (Bolton 1919: 1, 249). In the flood
plain in 1701, Kino wrote that the Indians:
"...showed themselves most affectionate toward
us, ...especially in opening for us some good,
and straight and short roads through the
thickets of the abundant and very dense woods,
which were on these most fertile banks."
(Bolton 1919: 1, 316) .

Cottonwoods are first mentioned specific-
ally by Father Sedelmayr in 1744 during his
endeavors along the Colorado at the junction
of the Bill Williams River: "The Indians burn
the trunks of the alders [willows] and
cottonwoods and when they fall they burn the
tops of them until they have a pole of the

Thirty years later, in 1774, Fathers Diaz
and Garces accompanied Captain Anza to the
Colorado River. After they crossed the river
near Yuma, Arizona, Diaz wrote: "On its banks
there are many cottonwoods, most of them small."
(Bolton 1930: 2, 264). The following day
(9 February), Diaz wrote: "...we camped on the
bank of this river below its junction with the
Gila... Its banks and adjacent lands are very
thickly grown with cottonwoods, which would
serve for any kind of building." (Bolton 1930:
2, 267-268). On that same day, Captain Anza
measured the Colorado River and commented on
the plant communities: "...one can see clearly
the junction of the rivers and the immense
grove of cottonwoods, willows useful for
thatching, and other trees, both upstream and
down." (Bolton 1930: 2, 169).

Diaz and Anza journeyed westward to the
coast, leaving Garces to continue his missionary
work. In his travels north along the river in
the vicinity of Ehrenberg, on 28 February 1774,
he wrote: "...I arrived at eleven o'clock in the
forenoon at the beaches and the groves of the
Colorado, halting at a very long and narrow
lagoon..." (Bolton 1930: 2, 379). As he
traveled north above the Bill Williams River,
he viewed the Chemehuevi Mountains and described
the area now under the upper end of Lake Havasu:
"In the vicinity of the sierra I saw much water,
groves and large beaches, which must be those
of the Colorado River." (Bolton 1930: 2, 385).

TRANSIENT PIONEERS

The crude Spanish Missions built on the
lower river were destroyed in 1781, and most
of the inhabitants, including Father Garces,
were killed. Spanish activity in the area
waned and between the late 1700' s and 1850,
only a few passing explorers, trappers and
pioneers visited the lower Colorado River.
One diary, that of Jose Joaquin Arrillaga,
contained worthwhile vegetative descriptions
where he approached the Colorado near its
mouth and then traveled north to Yuma. On
19 September 1796 he wrote: "At half past
three in the morning I took up my march to the
east,... and at sunrise I was already in sight
of the cottonwoods of the Rio Colorado." He
crossed the Colorado River near Yuma and
continued east: "After one leaves the banks of
the river there is not a useful tree to be
found, except the mezquite grove where I set
out, and this serves for nothing but firewood,"
(Arrillaga 1796 MS) .

A number of trappers were known to have
illegally entered the now Mexican territory
(Weber 1971) which included the region of the
Colorado River. They were tight-lipped and

only one, James Ohio Pattie, in 1827 left
records of the Colorado River environment:
"The river, below its junction with the Helay,
is from 2 to 300 yards wide, with high banks,
that have dilapidated by falling in. Its
course is west, and its timber chiefly cotton-
wood, which in the bottoms is lofty and thick
set." (Pattie 1831: 129) .

AMERICAN EXPLORERS

Colorado River, twelve miles below its
junction with the Gila, at a place
called "The Algodones," and soon after,
we halted upon its bank. It was much
swollen, and rushed by with great
velocity, washing away the banks and
carrying with it numberless snags and
trees. The road ran along the river's
bank, which, as well as the bottom-land,
was filled with a dense forest of
willows, cotton-woods, and mezquit.

The best and most complete records of
cottonwood distribution, abundance and size
occur from the late 1840' s on when soldiers
and scientists began working in the area of
the lower Colorado River either because of war
or to conduct various surveys. The United
States boundaries were expanded to the Pacific
Ocean in the 1840 's and in 1846, the United
States went to war with Mexico for land acqui-
sition, the Colorado River being part of this
region .

Lt. William H. Emory (1848: 99-100),
Topographical Engineer accompanying the "Army
of the West" in 1846, recorded his observa-
tions in the area of the Colorado-Gila junction:
"The banks are low, not more than four feet
high, and judging from indications, sometimes,
though not frequently, overflowed.... The
growth in the river bottom is cotton wood,
willow of different kinds,..."

A member of Emory's party, A. R. Johnson,
also kept a journal and observed that:
"The Colorado disappears from here in a vast
bottom; the last we can see of its cotton-
woods is in the southwest." (Emory 1848: 609).
The following day, after a 10 mile march, they
again reached the river and Johnson stated:
"...the river here is about ten miles wide,
and much of the land could bear cultivation;
it is all now overgrown with the most impene-
trable thickets of willows, mesquite, Fremontia,
etc." (Emory 1848: 609). The term "Fremontia"
was frequently used to denote cottonwoods
because the species was named in honor of the
naturalist J. C. Fremont.

One of the many commissioners of the
trouble-plagued boundary survey, John R.
Bartlett, published his "Personal Narrative of
Explorations" along the Colorado River in 1854.
His following description of the Algodones area
in 1852 is duplicated by a photograph (plate la)
some 50 years later and then again in 1976
(plate lb) . In his description of the area
(Bartlett 1854: 149-151) he wrote:

June 9th, 1852 - Our journey was
through a bottom filled with mezquit
and cotton-woods; . . .Our eyes were
greeted with a sight of the great

Lt. Amiel Weeks Whipple (1856: 3(1), 109) in
his exploration of a railroad route along the
35th parallel described the Chemehuevi Valley
(79 years after Garces): "On both banks are
strips of bottom lands, from half a mile to
a mile wide. The soil is alluvial, and seems
to contain less sand and more loam than is
found in the valley of the Rio del Norte. But
here, as there, are occasionally spots white
with efflorescent salts. A coarse grass grows
luxuriantly upon the bottoms. Bordering the
river are cotton-woods, willows, and mezquites,
or tornillas." Dr. J. M. Bigelow, surgeon and
botanist in Whipple's party, described the same
area: "From the mouth of Bill Williams' fork
to the point above where we crossed the Rio
Colorado, is about sixty miles.... Along the
valley of this river, alamo [cottonwood],
mezquite, and willow form the principal, and
almost entire, kinds of trees." (Bigelow 1856:
4, 13).

The interest in river navigability was so
strong in 1857 that a government expedition was
organized. Lt. Joseph C. Ives, Corps of
Topographical Engineers, was directed to deter-
mine how far and to what extent the Colorado
River was available for steamer traffic. At
Camps 50 and 53, Ives described Cottonwood
Valley and Round Island. The latter was
commonly called Cottonwood Island because of
the heavy forest of large cottonwood trees.
Of Cottonwood Valley, Ives (1861: 78) reported:
"Groves of cottonwood trees, of a larger growth
than any seen before, indicate that there is
some alluvial land, but the valley does not
appear to be of great extent.... The Cottonwood
valley was found to be only five or six miles
in length and completely hemmed in by wild-
looking mountains. The belt of bottom land is
narrow, and dotted with graceful clusters of
stately cottonwood in full and brilliant leaf."

AMERICAN EXPANSION

Ives' steamer exploration up the river,
and the demonstration that the river was navi-
gable, generated an abundance of steamer travel
and allowed shipment of goods to the mining
industry. In 1862, placer gold was discovered
midway between Fort Yuma and Mohave, and the

Plate la (1894). Looking southeast into Mexico at Mexican-American Boundary Monument 207.
Maturing cottonwood community with trees 20 to 30 feet tall can be seen in the background
following the cessation of the fuel wood industry for steam boats. A community of arrowwe
occurs between the marker and the cottonwoods.

Plate lb (1976). Looking southeast into Algodones, Mexico.

town of LaPaz was born (Renner 1974). Fuel
for steamers was readily available in the form
of Cottonwood, willow and mesquite, the latter
being less desirable because of its slow-burning
properties. Large quantities of fuel were
needed frequently and fuel stations were estab-
lished at 25-mile intervals. The Indians,
realizing that a profit could be made, cut and
corded the wood, selling it at $2.50 a cord
(Leavitt 1943) .

The continuing search for gold revealed
deposits near present-day Oatman and in El
Dorado Canyon (Dunning and Peplow 1959;
Casebier 1970) . The expanding mining activities
increased the military presence and businessmen
soon realized the potential for development.

G. W. Gilmore, a member of the Colorado
Steam Navigation Company, submitted a report
on the availability of fuel along the river
after having taken a trip on the steamer
Esmeralda in 1866 in which he stated: "...bends
of the river in the bottom lands, which, as
below Fort Yuma, are covered with vegetation
and timber; the trees of the varieties already
named are suitable for fuel, and are of very
rapid growth. It is found that upon new lands
formed by the cuttings of the river cottonwood,
willow, and mesquite trees will be produced in
three years large enough to cut for fuel. . .
Trees are quite abundant for most of the
distance, and plenty of fuel to be had."

His statements about Cottonwood Island
were as follows: "...about 10 miles long by
an average of about three miles wide, is a fine,
level island, fertile and covered with grass,
and having considerable timber.... An immense
quantity of this wood was upon the island,
estimated at several thousand cords" (Browne
1869: 462-464).

The need for exploration was almost over
and although further expeditions would be sent,
they would be of a very different nature. In
1875 and 1876, Lt. Eric Bergland, under the
direction of George Wheeler, leader of the
United States Surveys West of the 100th
Meridian, examined the Colorado River as a
potential irrigation source. Describing the
vegetation of the Colorado River, he stated:
"A most pleasant sight... Cottonwood Island,
with its majestic cottonwood trees and rich
vegetation, afforded a pleasant relief to the
eye... . Along the river there is a rich growth
of trees, principally cottonwood, and here the
fuel is obtained for the river steamers."
(Bergland 1876: 330-333).

In 1877, Lt. A. G. Tassin authored a
document entitled "Report on the Forestry,
Elevation, Rainfall, and Drainage of the
Colorado Valley together with an Apercue of

Its Principal Inhabitants the Mahhaos [Mohave]
Indians" compiled while he was stationed at
Camp Mohave. Never published, the report
remains in handwritten manuscript form, often
undecipherable. In his discussion of the
vegetation, Tassin (1877: 5-6) noted: "Finally
along the margin of ' lagunas ' the most substan-
tial of the Colorado timber the willow and
cottonwood. The largest of these in the entire
course of the river, are in what is called
'Cottonwood Island' between Camp Mohave and
the Grand Canyon where they have attained a
size which may be styled majestic... . The
cottonwood, mesquite and willow are the
principal, if not the only fuel of the country,
the first having a diameter varying from two
to twenty inches." In a latter section of his
report, Tassin (1877: 30) wrote: "Cottonwood
island the sole bottom-land between the Grand
Canyon and Camp Mohave, is, as denoted by its
appellation, celebrated for its splendid
cottonwood trees which here attain their full
size. Its area varies between from four to
six miles in length and from one to three-
fourths of a mile in width.... In a few years,
however, its beautiful trees will have disap-
peared, a large demand being made on them

yearly for fuel for the use of the

Mormons . "

The G eneral Land Office, now known as the
Bureau of Land Management, initiated the origi-
nal township surveys or cadastral mapping along
the river in 1855. Not all the land was survey-
ed during the same time period. Figure 1 shows
a reconstruction of the general vegetative types
below Blythe, California in 1879 derived by
interpreting floral descriptions contained in
original field notebooks and then transferring
these to the original field plats. The field
notes contain exact measurements from section
corners to points where the vegetation or topo-
graphy changed. At each change notes were taken
on soil, vegetation and general character of the
land. Once a section (1 square mile) was
chained, the surveyors took random walks (giving
specific localities) through the section and
again took notes at various places on vegetation,
etc. Insight into the maturity of the community
was also indicated when tree diameter values at
breast height were noted and when trees were
used as section corners. Although these data
are semi-quantitative and highly time consuming
to obtain, they yield the earliest aerial view
of plant community extent, abundance and place-
ment along the lower river. They further support
previous and subsequent written descriptions
in the historical record.

By about 1890, the use of and need for
steamboats had declined, as had the fuel supply.
Some steamer traffic ran north of Yuma, but to
insure adequate amounts of fuel, they had to
travel into the delta area (Sykes 1937: 37).
The decreased need for steamers was in part due

T.10S.

Davis Dam
Bullhead City

^°ha»e Co

bounty

Imperial Dam
Laguna Dam

VEGETATION STRUCTURAL TYPE

Mature

Figure 1. — Reconstruction of native plant community placement and species composition from
original surveyor notes and plats along the lower Colorado River in 1879. Area surveyed
by Benson.

to the completion of the railroad to Yuma (1877)
and Needles, California (1883).

By the late 1800's, the importance of
mining and other uses of the river had slowly
declined and agriculture along the river was
precarious because of the annual floods and
constant shifting of the channel. In 1892
Imperial Valley was rediscovered by the Arizona
and Sonora Land and Development Company.
George Chaffey was instrumental in bringing
water into the basin, and by 1904 the California
Development Company claimed seven hundred miles
of irrigation ditches and seventy-five thousand
acres were under cultivation.

The winter of 1905 was one not to be
forgotten. Unlike other years, the river
began to rise in February and "the condition
continued. . .until February of 1907" (Sykes
1937: 57). Despite attempts to control the
flooding waters, by August 1905 the entire
river was flowing into the intake of the
Imperial Valley canal. Thus, the Salton Sink
became the Salton Sea (Cory 1915; Tout 1931;
Sykes 1937; Hundley 1973).

The flood of 1905 brought heavy public
pressure for river management in the form of
flood control and water storage. The
Reclamation Act had been passed in 1902, the
settlers were having continual problems with
the development company, there were difficulties
with the Mexican government, and the disastrous
floods of 1905 and 1907 increased pressures
to evict the promoters and have the Reclamation
Service assume responsibility for the river.
From about this period on, the Reclamation
Service, now U. S. Bureau of Reclamation,
played the most Important role in developing
the river as a utility, although lesser
agencies such as the Imperial Irrigation
District also played important parts.

One of the major roles in development
was the installation of dams for flood control;
the first, Laguna Dam, became operational in
1909. A flood in the fall of that same year
"...was instrumental in completing the filling
of the basin above Laguna Dam with detrital
material within six months after the completion
of the dam itself." (Sykes 1937:152). Floods
continued and levees were raised until the big
flood of 1922 which convinced people in
Washington that larger dams were needed. In
1935 Hoover was operational. In 1943 Davis
Dam, 1938 Parker Dam and Imperial Dam were
completed. Lesser dams for water diversion
also were constructed.

Concomitant with and following dam
construction, engineers began to examine water
movement rates, channel siltation and bank
erosion. Many of these problems can best be

solved by channelization, riprapping of banks
and removal of sedimentation through dragline
or dredge. All of these methods were employed
to straighten and open channels to expedite
flows. New and old eroding banks were fortified
with riprap to increase bank stability.

Once the dangers of floods were reduced
and flows were controlled by Hoover Dam, this
allowed for agricultural expansion throughout
the floodplain. Settlers claimed new lands
and removed large and continuous tracts of
natural vegetation for agricultural purposes.
These activities, along with city and rural
development, continue to claim numerous acreages
once vegetated with natural communities. Much
of this land, especially near the river, once
supported cottonwood communities.

In the 1950 's through the 1960's, two
plans were formulated by the Bureau of Reclama-
tion to remove riparian vegetation for the
purpose of water salvage. Only one of these
was ever implemented and this was in the Yuma
flood plain area (Curtis W. Bowser, USBR, pers.
com.). Van Hylckama (1970 and in press) has
pointed out that water losses from a plant
varies considerably during the year and past
measurements have not taken this into consid-
eration, possibly resulting in large errors.

In the late 1960 's and early 1970' s, a small
flame of environmental concern was beginning
to burn throughout the nation. In 1969 the
National Environmental Policy Act was passed
and federal agencies began to review policies
and action decisions more closely. During
this period the Regional Director of the
Bureau of Reclamation, Mr. Edward A. Lundberg,
established an environmental office headed by
Mr. F. Phillip Sharpe. Efforts by these
individuals in 1972 resulted in an extensive
assessment of the little known flora and
fauna of the lower river and a comprehensive
and long-termed study was begun which was
entitled "Vegetation Management for the Enhance-
ment of Wildlife." These studies are still in
progress and the legacy of environmental concern
generated in the early 1970 's is strongly
supported by the current Regional Director,
Mr. Manuel Lopez, Jr.

DISCUSSION

How extensive was the cottonwood community
in historical times? As we swim through a sea
of qualitative data, there is little quantitative
information available to help answer this
question. We know from historical records that
cottonwoods were primarily restricted to the
river's edge where seedlings became established
in newly deposited soils. The following is an
attempt to give some quantification to the

extent of cottonwood communities in historical
times. If we assume that cottonwoods were
absent in areas where the river cuts through
canyons and thus remove 75 miles of the 275
miles of river between Davis Dam and the
Mexican boundary, this leaves approximately
200 miles of potentially suitable habitat for
cottonwoods. If the mean width of the cotton-
wood community along the river was 100 feet on
each side of the river, we can compute a
minimum area of 5,000 acres of cottonwood
habitat. This figure is conservative, but it
yields a value which will be instructive in
later discussion.

The Indians exerted some influence on the
ecology of the cottonwood community, especially
by using fire to fall and size timbers. Cotton-
woods cannot tolerate much heat and do not
resprout from the roots following a hot fire.
But to expedite discussion, in this report we
will assume the Indians' influence on the
cottonwood community along the river was
minimal and capricious.

The influence of the Spaniards on the
cottonwood community appears more minimal than
that of the Indians except for the introduction
of livestock in the early 1700's (Forbes 1965).
Spanish activity was concentrated primarily
around the Yuma area. The spread ^r.d extent
of use of domestic livestock by the Indians
is not well known, but Forbes (1965:287)
reported from hearsay evidence in 1842
that the Quechan and Mohave Indians "...own
large numbers of horses and cattle..." Not
all the Indians owned or had access to live-
stock or Browne (1864) would not have observed
them starving and eating rodents and reptiles.
The primary damage of livestock to cottonwood
communities would have been to seedling or
sapling stands which would have been important
foraging areas for domestic livestock. More
mature communities were sometimes cut if live-
stock forage was scarce; Pattie (1905 : 188), for
example, stated, "Our horses also fared well,
for we cut plenty of cotton-wood trees, the
bark of which serves them for food nearly as
well as corn."

The first and most widespread reduction
of cottonwood communities appears to have taken
place during the period of steamboat use on the
lower river (1855 to 1890) . Cottonwoods and
willows, fast burning woods, were located
nearest to the river and were one of the primary
fuels for powering the vessels. The extent of
reduction of the cottonwood community is
supported by Tassin (1877:30) who predicted
the denuding of cottonwoods from Cottonwood
Island, which came to pass. Another indication
of the reduction of maturing cottonwood communi-
ties is exemplified by the necessity for steamers
planning long trips up river from Yuma to go

into the delta for wood to insure an adequate
fuel supply (Sykes 1937) . As fuel demands
abated in the late 1880' s, cottonwood communities
began returning; and the photograph (plate la)
taken in April 1894 by Mearns shows redeveloping
cottonwood communities at the Mexican-United
States boundary.

The floods which occurred during the
years of 1905 and 1907 were of a greater
magnitude and longer duration than any described
in historical accounts. The destruction and
removal of natural communities must have been
far greater and more widespread than the 1852
flood which washed away banks and carried in
its waters "numberless snags and trees"
(Bartlett 1854:50). Presumably, the newly
deposited and moist soils again would have
provided the basic habitat for cottonwood seed
germination and the repetitious reforestation
process. Accounts by Grinnell (1914) in 1910
indicate that the cottonwood communities were
returning and the 1945 photograph (plate 2a)
provides pictorial testimony. Aerial photo-
graphs taken every three to five years, begin-
ning in 1942, show that the majority of the
trees were gone by 1967 and plate 2b, taken in
1976, revealed only four or five isolated
trees in the bottom right. Dense stands of
salt cedar presently cover areas previously
supporting cottonwoods.

Salt cedar was introduced in the New
World in the early 1800' s both as a soil
stabilizer and as an ornamental (Horton these
proceedings) . Its entry to the lower Colorado
River must have been sometime after 1910 when
Grinnell (1914) made extensive museum collections
of plants and animals. He makes no mention of
it in his publication or field notes, and had
it been at all common, he would have found and
collected it. The species appears to have
become established between 1910 and 1920 and
began to spread rapidly in the 1930' s and 1940' s.
By the 1940 's it dominated large areas along
the Gila (Marks 1950; Haase 1972; Turner 1974)
and Colorado (Robinson 1965) rivers.

Horton (these proceedings) discusses the
biology of salt cedar but a brief discussion is
necessary to gain insight into the events that
transpired between 1945 and 1976. The species
produces seeds over a long period of time and
seeds are both wind and water disseminated.
They germinate vigorously in newly deposited
alluvial soils. The species is deciduous and
when in relatively dense stands (periphery of
adjacent trees touching), the annual litter
accumulation after 10 to 12 years produces a
highly flammable condition. The above ground
portions are killed following a fire, but
suckers from the root stock reappear in one to
two weeks, the burn cycle repeats itself every
10 to 20 years (Anderson and Ohmart MS).

Plate 2a (1945) . Aerial oblique looking west into California about 25 miles north of Blythe.
The river is flowing in the foreground to the right and then to the left. In bottom right,
cottonwoods can be seen along the old braided stream channels on the cut bank. Understory is
primarily sparse arrowweed with taller willows or salt cedar near the cottonwoods. The
peninsula supports many cottonwoods, willows and arrowweed.

Plate 2b (1976) . The camera station is higher and closer than in the previous photograph.
Dense salt cedar and arrowweed have invaded the areas previously occupied by cottonwoods
Isolated cottonwoods persist in the lower right. On site remnant blackened tree stumps
are persistent testimonials of past fires. Mesquites have become established along the
peninsula on the higher and better drained soils.

Cottonwood communities were in the process
of returning following the floods; but when
salt cedar began invading the lower river, it
must have started mixing with the maturing
cottonwoods. The proximal location of the
cottonwood communities to the river and on
soils only inches above the water table provided
the type of habitat that salt cedar does best
on and aggressively spreads over. Some
individual cottonwoods were probably removed
for firewood by man and for food by beavers.
Today remaining cottonwoods occur as isolated
trees or as irregular rows with little under-
story, mixed with sparse willow communities,
in pure stands or temporarily mixed with salt
cedar. The latter being temporary as
exemplified by the area around Hunters Hole in
the Limitrophe Division which burned two years
ago and killed all the mature cottonwoods.
Many of the remaining cottonwoods along the
lower river are mixed with sparse willows and
these willows may have served as a buffer
which prevented or slowed the invasion of salt
cedar by shading out seedlings. Anderson and
Ohmart (MS), in studying rodent succession in
fire altered salt cedar communities along the
lower river, have only found two salt cedar
communities of 50 acres or more that have
survived fire for more than 20 years.

Many cottonwood communities have been
lost to expanding agriculture channelization
projects, inundation of lakes behind dams and
possibly the placement of dredge spoil materials.
Agriculture only poses a minor threat to
remaining cottonwood communities since there
are only so few left, and they occur primarily
on lands between the levees which are not
farmed. River management activities tempered
by environmental concern, require an Environ-
mental Impact Statement and mitigation for
project losses.

The demise of cottonwoods on the lower
Colorado River has been related to implemen-
tation of dams, and the data indicates that
dams expedited the natural loss by stopping
annual overflow. This periodic flooding and
water movement through the communities covered
or washed away litter accumulations. Litter
covered with sediment during overflow rapidly
decomposed to release nutrients and add humus
to the soil. River management stopped these
natural overflows and allowed litter accumu-
lation which in turn has resulted in the
increased frequency of communities being burned.
This has led to the loss of many cottonwood
communities from fire.

Cessation of annual overflows and natural
channel movements also curtailed the formation
of the basic cottonwood seedling habitat, bare
sandy soils with high water tables, which
appears essential for cottonwood seed germin-

ation. The rapid spread of salt cedar and slow
demise of cottonwoods began about the time
major dams were implemented, the mid 1930 's.
It is somewhat of a moot point whether major
dams tipped the ecological balance to favor
dominance of salt cedar over cottonwoods or if
cottonwoods could have retained their dominance
over the invading salt cedar on the lower
Colorado River. Currently the success of cotton-
wood regeneration has not been stopped, but it
has been lowered to the point where it is
negligible. Campbell and Dick-Peddie (1964)
reported that cottonwoods could maintain their
dominance over salt cedar in natural conditions
on the upper Rio Grande in New Mexico, but it
is doubtful this would be valid along the lower
Colorado River. Even without dams it appears
highly unlikely that cottonwood communities
could have maintained their dominance along
the lower Colorado River over the aggressive
and fire adapted salt cedar.

This conclusion is supported by examining
the loss and persistence of cottonwoods in
natural communities along other southwestern
streams. A reach of the Gila River in Arizona
between Kearny and Florence is still inter-
mittently flooded but contains few lone cotton-
woods and no gallery forest. Conversely, the
Verde River in Arizona above and below the
dams possesses good cottonwood gallery forest
and salt cedar appears to be having more
difficulty invading this riparian system than
it has had on the Salt or Gila rivers. Another
factor appears to be important — total dissolved
solids (TDS) . Further support of the importance
of TDS is indicated by the return of native
vegetation along the Salt River below the
Flushing Meadows sewage treatment plant in west
Phoenix, Arizona. In this area, salty native
ground water is being displaced by secondary
sewage water and following the flood conditions
in the 1960's which scoured away much of the
salt cedar in that area, native communities are
rapidly returning. It is highly improbable that
these native communities would have returned in
competition with salt cedar if salt cedar removal
by flood waters was the only cause. Many areas
have been cleared of salt cedar only to have it
promptly return. Observations along the Rio
Grande in New Mexico and Texas further support
the importance of low TDS and cottonwood
dominance over salt cedar. Along the upper
portion, around Albuquerque, New Mexico, cotton-
woods appear to be thriving and maintaining
their dominance. But between Las Cruces, New
Mexico, and El Paso, Texas, the frequency of
extensive gallery forests declines and individual
trees show heavy plant parasite infestations;
to the extent that they are dying. Further down
river in Texas between Presidio and Fort Quitman
there are no gallery forests remaining; only
lone cottonwoods remain isolated along ditches
or canals from the tall and dense salt cedar

forests. All along the river from Albuquerque
to El Paso extensive agricultural and industrial
effluent enters the river to slowly work its
way down stream. Salts from these and natural
sources can be seen covering many acres between
Fort Quitman and Presidio following the subsi-
dence of sluggish and intermittent winter floods.
Dams have stopped the once rapid moving floods
which once flushed and leached salts into the
gulf, leaving rejuvenated soils.

Other factors, both man caused and natural,
may or may not be involved in each case and
should be examined in depth before reaching a
final conclusion as to the reasons for cotton-
wood disappearances. We know that domestic
livestock concentrate in riparian communities
and heavily utilize young cottonwoods, but we
know nothing about possible climatic changes
and their effects postulated by Hastings and
Turner (1965)- What effects have these changes
had on cottonwood communities, if their thesis
is correct? Much remains to be learned about
the ecology of riparian communities and
unfortunately there is little information
available on the natural history of most of
these plant species.

Cottonwood communities have declined from
high abundance (5,000 acres plus) along the
lower Colorado River in the 1600 's to scattered
groves containing a few mature individuals
today. Anderson and Ohmart (1976) have
estimated that only 2,800 acres of cottonwood-
willow community remain along the lower river.
If one was to consider pure cottonwood communi-
ties, it would be less than 500 acres.

In conjunction with the loss of the cotton-
wood resource, we must have experienced popula-
tion reductions in bird species which show a
strong preference for cottonwood habitats.
Summer Tanagers (Piranga rubra) , Yellow-billed
Cuckoos (Coccyzus americanus) , Wied Crested
Flycatchers (Myiarchus tyrannulus) , Brown
Creepers (Certhia f amiliaris) , and many small
insectivorous birds (mostly warblers) breed
Oi. winter in these habitats. Numbers of some
of these species are very low (Anderson and
Ohmart 1975) and for all practical purposes
some species would be extirpated from the lower
river if cottonwood communities were eliminated.

Can anything be done and is anything being
done to prevent the further loss of this
resource? The U.S. Fish and Wildlife Service
has recently bought the remaining cottonwood
gallery forest that was not previously part of
the Havasu Wildlife Refuge on the Bill Williams
River and incorporated it into the refuge.
Although adjacent to the Colorado, along the
lower end of the Bill Williams River, there
is a young gallery forest of about 700-800
acres. If this area is properly managed and

prevented from burning, it should survive.
Willow Valley Estates, a private housing
development in the Mohave Valley was designed
with open space areas and has planted natural
vegetation (especially native cottonwoods) .
It is a small area, but some of the habitat
specific bird species are found in this commu-
nity. Recently, the Bureau of Reclamation has
begun experimenting with the redevelopment of
cottonwood-willow communities for operational
enhancement and mitigational measures.
Currently 25 acres of dredge spoil in the Cibola
Division are being revegetated with cuttings
or seedlings of native cottonwoods, willows
and honey mesquite, the results look promising.
A smaller area below Parker Dam also is being
revegetated with cottonwoods and the young
trees are doing well.

A look at the past allows us to examine
changes and postulate causes. Hopefully we
can then turn to the future, with the knowledge
of the past, and formulate management plans so
we can ultimately move with dispatch to manage
and expand the availability of a valuable
resource that is rapidly disappearing. To
insure the preservation and perpetuation of
this resource, all responsible agencies must
make special efforts to preserve what cottonwood
communities remain on their public trust lands
and even attempt to reestablish new communities
through transplants of native stock. It is
expensive and requires a lot of manpower and
attention but if this biotic community is to
be preserved for future generations the effort
must be undertaken soon.

ACKNOWLEDGEMENTS

We wish to thank Linda Cheney and Linda
Cross for their help in preparing the figure,
John Saunders for his help in transfering
surveyors' notes to plats, Eugene E. Hertzog
for duplicating the 1976 photographs and
Penny Dunlop and Katherine Hildebrandt for
their help in typing revised copies of the
manuscript. B. W. Anderson and W. A. Dick-
Peddie kindly reviewed the manuscript and made
helpful suggestions. Final thanks go to Jane
Durham whose patience preparing the literature
cited section and scrutinizing help in later
copies were invaluable. We deeply thank these
people .

LITERATURE CITED

Anderson, B. W. , and R. D. Ohmart.

1977. Densities and diversities of nocturnal
rodents in disturbed areas in the lower
Colorado River valley. MS in prep.

Anderson, B. W. , and R. D. Ohmart.

1976. An inventory of densities and diversi-
ties of birds and mammals in the lower
Colorado River valley — 1975. Submitted

to Bureau of Reclamation.
Anderson, B. W. , and R. D. Ohmart.

1977. An inventory of densities and diversi-
ties of birds and mammals in the lower
Colorado River valley — 1976. Submitted

to Bureau of Reclamation.

Arrillaga, J. J.

1796. MS. Diary relating to reconnoitering
expedition through the frontier region of
Baja California into Alta California as
far west as San Diego, and along the lower
Colorado River area. June 14 - Nov. 21,
1796. Translated by Nellie Vande Grift
Sanchez .

Bartlett, J. R.

1854. Personal narrative of explorations
and incidents in Texas, New Mexico,
California, Sonora, and Chihuahua,
connected with the United States and
Mexican Boundary Commission, during the
years 1850, '51, '52, and '53, with an
introduction by Odie B. Faulk. 2 vols.
D. Appleton and Co., New York.

Bergland, E.

1876. Preliminary report upon the operations
of Party No. 3, California Section, season
of 1875-1876, with a view to determine the
feasibility of diverting the Colorado River
for the purposes of irrigation. Annual
Report of Geographic Surveys West of the
100th Meridian for 1876, 1876, Appendix B
(pp. 109-125) also in Annual Report of
Chief of Engineers for 1876, Part 3:
329-345. Gov. Doc. W. 5. Serial Set
#1745.

Bigelow, J. M.

1856. General description of the botanical
character of the country. Ln Reports of
exploration and surveys, to ascertain the
most practicable and economical route for
a railroad from the Mississippi River to
the Pacific Ocean. 1853-4. 33d Congress,
2d Session, Senate Ex. Doc. No. 78, Vol. IV,
No. 1:1-26.

Bolton, H. E.

MS. Diaries and correspondence of the
expeditions of Juan Mateo Mange and
Eusebia Kino to the Indian countries,
1697-1706. Transcription and Translation,
Bolton Papers //203, Bancroft Library,
Univ. Calif., Berkeley, Calif.
1919. (Editor and Translator) Kino's
historical memoir of the Pimeria Alta.
Arthur Clark Co., Cleveland. Reprinted,
1948, Univ. Calif., Berkeley.
1930. Anza ' s California expeditions.
5 vols. Univ. Calif. Press, Berkeley.

Browne, J. R.

1864. Report of the commissioner of Indian
affairs, Arizona superintendency , No. 55.

In Report of the Secretary of the Interior,
Government Printing Office, Washington,
D.C., pp. 305-308.
1869. Resources of the Pacific slope. A
statistical and descriptive summary of the
mines and minerals, climate, topography,
agriculture, commerce, manufactures, and
miscellaneous productions, of the states
and territories west of the Rocky Mountains.
With a sketch of the settlement and
exploration of Lower California. D. Appleton
and Co., New York, 678 pp.
Campbell, C. J., and W. A. Dick-Peddie.

1964. Comparison of phreatophyte communities
on the Rio Grande in New Mexico. Ecology,
45:492-502.

Casebier, D.

1970. Camp El Dorado, Arizona Territory;

soldiers, steamboats and miners on the

upper Colorado River. Arizona Historical

Foundation, Tempe, Ariz., 103 pp.

(Arizona Monographs, No. 2).
Cory, H. T.

1915. The Imperial Valley and the Salton

Sink with introductory monograph by

W. P. Blake. John J. Newbegin, San

Francisco .
Dunne, P. M.

1955. Jacobo Sedelmayr, missionary, frontiers-
man, explorer in Arizona and Sonora. Four
original manuscript narratives (1744-1751).
Translated and annotated by P. M. Dunne.
Arizona Pioneers Historical Soc. MCMLV
(1955), 82 pp.

Dunning, C. H., and E. H. Peplow, Jr.
1959. Rock to riches; the story of

American mining — past, present and future —
as reflected in the colorful history of
mining in Arizona, the nation's greatest
bonanza. Southwest Publ. Co., Phoenix,
Ariz . , 406 pp .

Emory, W. H.

1848. Notes of a military reconnoissance ,
from Fort Leavenworth, in Missouri, to
San Diego, in California, including part
of the Arkansas, Del Norte, and Gila
rivers. Thirtieth Congress — First Session.
Ex. Doc. No. 41. Wendell and Van
Benthuysen, Printers, Washington, pp. 15-126.

Forbes, J.

1965. Warriors on the Colorado: The Yumas
of the Quechan Nation and their neighbors.
Univ. Oklahoma Press, Norman, 378 pp.

Grinnell, J.

1914. An account of the mammals and birds
of the lower Colorado Valley with especial
reference to the distributional problems
presented. Univ. Calif. Publ. Zool. ,
12(4) :51-294.

Haase, E. F.

1972. Survey of floodplain vegetation along
the lower Gila River in Southwestern
Arizona. J. Ariz. Acad. Sci., 7(2):66-81.

Hastings, J. R. , and R. M. Turner.

1965. The changing mile, an ecological
study of vegetation change with time in
the lower mile of an arid and semiarid
region. Univ. Arizona Press, Tucson,
Ariz . , 317 pp .

Hundley, N.

1973. The politics of reclamation:
California, the federal government, and
the origins of the Boulder Canyon Act —
a second look. California Historical
Soc. Quarterly, 2 (4) : 292-326 .

Ives, J. C.

1861. Report upon the Colorado River of
the west. 36th Congress, 1st Session.
Senate Ex. Doc. , Gov. Printing Office,
Washington, Part I, General Report: 13-131.

Leavitt, F. H.

1943. Steam navigation on the Colorado
River. California Historical Soc.
Quarterly, 22:1-25; 151-174.

Malin, J. C.

1956. The grassland of North America.

James C. Malin, Lawrence, Kansas, 486 pp.

Marks, J. B.

1950. Vegetation and soil relations in the
lower Colorado Desert. Ecology, 31(2):
176-193.

Pattie, J. 0.

1831. The personal narrative of James 0.
Pattie of Kentucky, edited by Timothy
Flint. John H. Wood, Cincinnati, xiii +
1-230 + Appendices, pp. 232-269. (The
Personal Narrative of James 0. Pattie,
the 1831 edition, unabridged, introduction
by William H. Goetzmann, Philadelphia and
New York, J. B. Lippincott Co., 1962.)

Renner, P.

1974. La Paz: Gateway to territorial
Arizona. M.A. thesis, Ariz. State Univ.,
Tempe, Ariz., 188 pp.

Robinson, T. W.

1965. Introduction, spread, and areal
extent of saltcedar (Tamarix) in western
states. U.S. Geol. Surv. , Prof. Pap.,
491-A, 12 pp.

Sykes, G.

1937. The Colorado Delta. Amer. Geographical
Soc, Spec. Publ. No. 19. Published jointly
by Carnegie Inst. Wash., and the Amer.
Geographical Soc. of New York, 193 pp.
Tassin, A. G.

1877. Unpublished manuscript. Report on
the forestry, elevation, rainfall, and
drainage of the Colorado Valley together
with an apercu of its principal inhabitants
the Mahhaos [Mojave] Indians. Called for
per Circular of October 5th, 1877, from
Engineer Office, Military Division of the
Pacific, for the Information of the United
States Department of Agriculture.

Tout, 0. B.

1931. The first thirty years 1901-1931:
being an account of the principal events
in the history of Imperial Valley, Southern
California, USA. Arts and Crafts Press,
San Diego, 429 pp.

Turner, R. M.

1974. Quantitative and historical evidence
of vegetation changes along the upper
Gila River, Arizona. U.S. Geol. Surv.,
Prof. Pap. , 655-H, 20 pp.

Van Hylckama, T.E.A.

1970. Water use by salt cedar. Water
Resour. Res., 6 ( 3) : 728-7 35 .

In press . Water use by salt cedar in the
lower Gila River Valley, Arizona. U.S.
Geol. Surv., Prof. Pap., in press.
Weber, D. J.

1971. The Taos trappers - The fur trade in
the far southwest, 1540-1846. Univ.
Oklahoma Press, Norman, 263 pp.

Whipple, A. W.

1856. Itinerary. J_n Reports of explorations
and surveys, to ascertain the most
practicable and economical route for a
railroad from the Mississippi River to the
Pacific Ocean. 1853-4. 33d Congress,
2d Session. House of Representatives,
Ex. Doc. No. 91, Vol. Ill, Pt. 1, 1854,
viii + 1-136.

Session II

Discussion Leader: Robert Jantzen
Director

Arizona Game and Fish Department
Phoenix, Arizona

Wildlife Conflicts
in Riparian Management: Grazing1

Charles R. Ames^

Abstract. — Grazing has a negative impact on riparian
zones o These zones constitute a small but critically im-
portant part of the range resource.

The riparian types in southern Arizona have increas-
ed from what they were 100 years ago. The increase has
occurred through stream eutrophication and is most notice-
able where the streams pass through the grassland type.

Protection of the riparian type where grazing is an
established use can only be effectively achieved through
fencing.

Wildlife managers frequently express their
concerns about the impact of grazing in the re-
parian zones and justifiably so. We have all
seen examples of riparian types where there is
virtually no reproduction or mixed age classes
of the trees or shrubs. The type is dominated
by mature and overmature trees. Trees growing
in riparian sites are usually relatively short-
lived. It is entirely possible for a riparian
zone to completely disappear within a span of
a man's lifetime where grazing use is prevalent.

Cattle exhibit a strong preference for the
riparian zones for a number of reasons. Cattle
prefer the quality and variety of forage avail-
able. Riparian forage is higher in palatability
because it has more moisture in it whether it
be shrubs, forbs, or grass. Moisture content,
probably more than any other factor, influences
palatability. A preferred species of forage
growing on a dry hillside will not be nearly as
palatable as the same species growing in a ri-
parian zone.

Availability of water in most riparian areas
provides a strong influence for livestock to
frequent the area.

If the surrounding country is rough and
rocky, livestock tend to concentrate along the

Paper presented at the Importance, Pres-
ervation and Management of the Riparian Habitat
Symposium, Tucson, Arizona, July 9, 1977.

Charles R. Ames is the Range and Wildlife
Staff Officer of the Coronado National Forest,
Tucson, Arizona.

riparian areas just to give their feet a rest.
In hot climates, livestock seek the shade avail-
able along the riparian areas. In cold climates,
they seek shelter from the cold winds.

If livestock are left to their own preference
the riparian zones get continued yearlong use
with no respite from grazing. These critical
zones represent a small but important percentage
of the total range area. This is where the non-
game birds and animals congregate unless it is
totally devastated.

On the Coronado National Forest, approximi-
mately 20% of the grazing allotments have signifi-
cant riparian zones. Southern Arizona perhaps is
unique in that we probably have more riparian
zones today than there were 100 years ago. The
increase has been due to overgrazing during the
1890 's and early 1900' s. .

Figure 1. — Monument No. 98 on Mexican border. Lo-
cated on west bank of San Pedro River. Photo
taken 1892. Drainage virtually devoid of any
riparian growth . San Pedro River was peren-
nial stream with fish, frogs and turtles.

Figure 2. — 1969 photo showing dense growth of
mesquite. Entire San Pedro River has dense
growth of mesquite and other riparian growth.

Figure 4. — 1969 photo shows heavy riparian growth
along water course.

During this period, there was a continual
buildup of cattle to a peak number of 173,000
head by 1900 in the area now encompassing Pima
and Santa Cruz Counties. Needless to say, the
country was devastated. During the rainy sea-
sons, the runoff resulted in serious flooding
causing gullying and heavy soil loss.

Prior to this period, the San Pedro and
Santa Cruz Rivers were perennial streams inhabit-
ed by fish, frogs, and turtles.

The resulting accumulation of silt in the
stream with the soil nutrients provided the seed-
bed for the riparian growth now so prevalent a-
long these streams. This process is called eu-
trophication.

Figure 3. — Monument 111 where Santa Cruz River
leaves U. S. No riparian growth showing a-
long stream bank. Photo taken 1892.

A recent example of this is along the Santa
Cruz River where the Nogales Highway was washed
out in the 1967 flood. This occurred about 20
miles north of Nogales. The resulting silt bed
formed in the bend of the river produced a dense
stand of cottonwood trees now 30 to 40 feet high.

The riparian types in the mountain areas
of the Coronado have probably remained fairly
stable with little change through the years.
The increase of this type has occurred by ex-
tension through the grassland type by the stream
eutrophication process.

In general, it would seem we can conclude
that riparian types undergo change. They recede
in some areas and increase in others.

It is a well established fact that we are
in trouble trying to retain riparian zones in a
reproductive condition. The question arises -
what can we do about it?

First of all, I believe we need to do an
intensive classification job on our riparian
types. These would logically fall in two cate-
gories: threatened and unthreatened. Of those

in the threatened category, select the key areas.
These would be the zones determined to be essen-
tial to keep. They may be critical habitat for
rare and endangered species. Therefore, re-
production of the tree and shrub species must
be ensured.

Figure 5. — photo of Monument 118 where Santa
Cruz River comes back into U.S. Note limited
scattered growth along water course.

Figure 6. — 1969 photo shows a veritable jungle
of heavy riparian growth. Original masonry
monument washed out in flood in 1914 and re-
placed with steel monument in 1917.

It has been our experience on the Coronado
Forest, that there is no known system of live-
stock management that will give adequate pro-
tection to a riparian zone. Even short term
use or seasonal use is inadequate. Because
these areas are usually extremely narrow and
linear in character, grazing for only a few days
can seriously impair its reproductive capability.
It's like having the milk cow get in the garden
for one night.

The only way we have been able to ensure
adequate protection of our riparian types is by
fencing them out from livestock use. This, of
course, is coordinated with the livestock manage-
ment plan to provide for watering places and
logical pasture divisions.

We have initiated a riparian fencing pro-
gram of our key drainages with excellent results.
One in Parker Canyon has been fenced for nearly
2 years. The response here has really been en-
couraging. We have another currently in progress
of being fenced which we hope will be equally
productive.

Wildlife Conflicts
in Riparian Management: Water1

Charles E. Kennedy'
INTRODUCTION (Fisher 1970) demonstra

This paper is a summary of observations
of the need for a better understanding of the
interactions of stream-riparian-vegetation-
energy-nutrients-water production-aquatic
life and terrestrial life. Most of the
riparian ecosystem interactions have had very
little attention in Arizona and New Mexico.

WHAT IS WATER

In its pure form water is a colorless,
clear liquid compound of hydrogen and oxygen.
Water in the riparian zone is never just H20.
It is a building block for photosynthesis by
riparian and aquatic vegetation. It carries
assorted dissolved salts (many of which are
nutrients). Water carries dissolved organic
matter, fine and coarse particulate organic
matter, and supports numerous aquatic life
forms, vertebrate and invertebrate, large and
small (fish plankton, bacteria, etc.) Water,
through the riparian vegetation, supports a
wide assortment of interesting and valuable
terrestrial wildlife species. Water is an
energy source in itself as it forms natural,
meandering channels and transports particles,
large and small.

ENERGY-RIPARIAN ECOSYSTEM

A number of studies have shown that fish
production is much lower where grazing occurs
in the riparian zone. For example, in the
Rock Creek Floodplain Investigation (Marcuson
1970) there were 63 pounds per acre of brown
trout in the heavily grazed area as compared
to 213 pounds per acre in the ungrazed area.

Bob Phillips (USFS) and others demonstra-
ted the presence of 31 steelhead in a 100-foot
heavily grazed section and 75 steelhead
present in a nearby lightly grazed section
(personal communication).

iPaper presented at the Symposium on
Importance, Preservation and Management of the
Riparian Habitat, Tucson, Arizona, July 9, 1977.

■'Fisheries and Non-Game Biologist, U.S.
Forest Service, Albuquerque, New Mexico.

ted that 99% of the
annual energy budget for Bear Brook comes from
the surrounding forested watershed or from
upstream areas. Even in large streams, such
as the Missouri River, fifty-four percent of
the organic matter ingested by fish is of
terrestrial origin (Berner 1951) .

(Cummins 1974) diagrammed the fate of
heterotrophic stream organic materials
(dissolved and particulate) and showed a
conceptual model of stream ecosystem structure
and function.

(Ensign 1957) found that in Mt. Vernon
Creek, southern Wisconsin, where cattle were
free to graze the streambanks, terrestrial
insects made up only 4% of the annual food of
brown trout. In Black Earth Creek (a few
kilometers from Mt. Vernon Creek) where stream-
banks were protected from grazing, terrestrial
insects comprised 15% of the annual diet of
brown trout.

Thus, we find in the literature that
streams are often energy dependent upon the
riparian vegetation and the watershed. (Likens
and Bormann 1974) have demonstrated the nutrient
linkages between streams and watersheds. They
state clearly that the key to wise management
of aquatic ecosystems is wise management of
the watershed.

We can extrapolate these works and assume
that many streams in Arizona and New Mexico
will also be dependent upon the riparian zone
and their associated watersheds for their
primary energy sources. But in this area, we
have streams which can begin at elevations up
to 11,000 feet on Mt. Baldy on the Apache
National Forest (where they are comparable to
streams in Northern United States or Canada )
descend to intermediate elevation where they
support warm water species comparable to
southern and Midwestern United States streams.
Others, purely desert streams, are unique in
the United States. Just as Arizona and New
Mexico are rich in the number of wildlife species
produced in the wide diversity of habitats,
Arizona and New Mexico streams are also rich
in diversity running the garnet from high altitude,
cold, clear, mountain streams, through warm,
algae rich mid-elevation reaches, finally to
low elevation pure desert reaches. For

instance, we have grayling, an arctic fish, in
a lake above the Mogollon River; while only
50 miles away there are channel catfish, a
warm water species, in the Verde River.

Energy interdependence will follow a
similar gradation. The high streams are most
likely to be dependent on outside sources of
energy for the aquatic organism food base.
The mid -elevation streams may have somewhat
more ability to capture energy in the stream
through algae, diatoms, and rooted vegetation.
The low desert streams with riparian vegetation
and with tributaries supporting riparian
vegetation may fix substantial energies in
the aquatic environment, but will also receive
substantial inflows of plant detritus during
storm flows. (Burns 1977)

We need to develop a stream classification
system which incorporates these energy sources
as a significant criteria, and we need to
study stream energy budgets on typical reaches
of several stream types, i.e. cold water,
intermediate, and warm water to document the
stream-energy system sources and gradations
of dependence upon terrestrial sources.

No doubt we will find streams which are
largely dependent upon the riparian vegetation
for a substantial portion of their organic-
energy and partially dependent upon the
watershed for dissolved organic matter.

As I said earlier, fish weigh less and
are less abundant in grazed portions of streams.
Putting this fact with the dependence upon
energy from the riparian zone, we can under-
stand that plant material eaten by cattle
in the streamside strip will not be available
for food for aquatic organisms in the stream.
Fish will have less food. I used the term
"streamside strip" here because on many miles
of our streams in the southwest free choice
grazing by cattle has brought about complete
type conversions in those immediate areas
alongside streams.

After many years (50 to 100 or more) of
grazing in this "most palatable area" the
old riparian trees have died, seedlings are
eaten and killed until only the most "grazing
resistent" unpalatable grasses and/or trees
remain. This type conversion at higher
altitude has eliminated alders and willows,
leaving only associated grasses. In the
middle elevations the sycamore cottonwood and
others are often entirely missing to be
replaced by bermuda grass-desert willow-seep
willow and at some elevations, tamarisk. Thus,
grazing is a significant force in altering
streamside composition - just as it is through-
out the watersheds.

Actual streamside composition varies from
those areas where all of the natural species
are gone with no seed sources remaining, to
other streams that have a few decadent
widely scattered specimens with most species
present. Fencing alone will start the stream
toward recovery, but plantings of seedlings
will be needed on many.

In figure 1 we see only a few remnants
of willow and narrowleaf cottonwood. The
stream is appropriately called the Rio de las
Vacas and is on the Cuba District of the Santa
Fe National Forest, at elevations from 7000
to 9000 feet. The loss of shade for the stream,
the loss of bird habitat and the premption by
cattle of often the only source of green feed
is obvious. The loss of energy to the stream
is not so obvious. In fact, all too many times
little thought has been directed towards
learning how energy used by the stream flows
through the ecosystem.

Figure 1

STREAM MORPHOLOGY

Another, more subtle, impact on the
fishery occurs when riparian trees are elimi-
nated by continual grazing. The stream is less
confined to its banks and will have a more
constant sediment load, especially from un-
vegetated stream banks. Overgrazing in associated
watersheds creates higher peak storm flows.
Overgrazing combined with hydraulic force of
these peak storm flows plus the grazing by
cattle on young seedlings keeps many streams
in a young, undeveloped and raw condition.

Region 3 of the Forest Service (Arizona
and New Mexico) has in National Forest streams
approximately 4000 fish habitat improvement
structures to make more pools in the miles and
miles of flat, shallow streams.

An alternative to these structures and
their maintenance is to fence cattle out of
the narrow riparian zone so that the streams
can progress through successional stages
toward more stable conditions. As vegetation
and trees become established in the immediate
water edge area, the stream will, over time,
become more narrow and deeper provided the
associated watershed is properly grazed.
Grazing levels must provide for suitable
vegetative cover to insure soil protection
and retard rapid runoff. The number of pools
and their suitability for fish habitat will
improve. Figures 2 and 3 show an area along
a one mile reach of the Rio de las Vacas that
has cattle fenced out. Stream profiles,
photos, etc., are being established to
document changes in stream morphology and
riparian composition. Water temperatures
in June 1977 reached 70°F. in this area.
Narrowleaf cottonwood (Populus angustif olia) ,
Arizona alder (Alnus oblongif olia) and willow
(several species) comprise the bulk of the
remaining riparian tree species. There are
only about 50 individual specimens of narrow-
leaf cottonwood remaining in eight miles of
the stream.

Figure 2

Figure 3 shows the remnants of an old
trash catcher type stream improvement structure,
entering the water at the arrow. Stones, silt,
etc., caught by the fence posts and wire have
somewhat constricted the stream making a slight-
ly deeper spot just to the left of the fence.
How much better for the fishery, the bird life,
the esthetics, and the cattle if the dead trees
had survived and reproduced until the roots
provided cover, formed a pool and dropped
leaves and insects into the stream.

Figure 3

(White and Brynildson 1967) have documented
successional stages with drawings which clearly
demonstrate the process (see figure 4) . Time
in these changes will no doubt be faster in
Arizona and New Mexico at low elevations with
long growing seasons and perhaps slower on
the Rio de las Vacas at 8500 feet with a short
growing season.

A great deal of research has gone into
ways to produce more waiter on National Forests
in Arizona. Much has been written about the
evapotranspiration of water by riparian species,
native and introduced. There have been no
concentrated, integrated efforts to determine
which mixture of riparian species might best
serve the needs of all resources, the fishery,
the bird and wildlife resource, esthetic needs
and water production.

As manipulations are applied to watersheds
(chapparal and timber) to produce more water,
it will become more important to manage the
riparian zone (which in one aspect becomes a
water "pipeline") to insure all the intrinsic
values while producing the maximum amounts of
high quality water for downstream users. It
is certain that a vigorous stand of well
established riparian trees will produce the
amenities we are interested in.

There may be ways to improve tree composi-
tion to favor energy flows for the fishery,
reduce evapotranspiration for water production,
and provide habitat for the bird life and
other animal needs for green forage and cover.
Perhaps leaves from Arizona walnut transpire
less water and are better food for aquatic
insects. Maybe the leaves have a higher
calorie count - a better mix of nutrients.

IIMIIIIIIIIII IIIIIHIIIII

Some stages in natural development of a fertile lowland Wiscon-
sin trout stream from overgrazed (A) to very productive (D-E-F)
to overforested (G&H) when protected from grazing. A hypothe-
tical 14-foot wide cross-section plus adjacent bank shown.

The complete sequence from stage A to stage E-F has been ob-
served on Black Earth and Mt. Vernon Creeks near Madison.

Later succession — stages G and H with many intermediates —
is to be seen on other streams. Details of this succession vary
from stream to stream, especially after stage E-F, but the pas-
sage from predominantly herbaceous to predominantly woody
vegetation generally has the same detrimental effects. Good
management for trout — and other wildlife — would be control
of vegetation to maintain stages D-E-F.

KEY

watercress

present water level

water level

soft sediments deposited
.original soft sediments

gravel "

0 5 10

of stage A
since stage A

MIDSUMMER CONDITIONS UNDER
HEAVY GRAZING BY LIVESTOCK:
Bank vegetation and watercress grazed and
trampled. Banks eroding, and stream bed
mostly covered by shifting silts. Submergent
plants grow poorly. Whole surface of water
and stream bed exposed to sun. Greatest
depth in cross-section only 9 inches (22 cm).
These conditions offer trout no shelter, no
place to spawn, little food, and frequently
unfavorable temperatures.

MIDSUMMER CONDITION AFTER 2 TO 4
YEARS OF PROTECTION AGAINST GRAZING:
Bank vegetation forming a turf. Abundant
watercress at edges of stream constricts
channel, thus deepening and speeding water.
Soft sediments scoured from much of stream
bed and trapped in cress beds. Submergent
plants thriving. Only about half the former
stream width exposed to sun. Greatest depth
about 20 inches (50 cm). Trout have ample
shelter beneath watercress, beside rock,
and among submergent plants. Firm stream
bed and many plants provide substrate for
many animals that trout eat. Newly
exposed gravel is a place to spawn.

LATE IN THE NEXT WINTER:
Watercress has withered and drifted away.
The silts it held slump into the channel,
smothering many of the trout eggs buried in
gravel and preventing fry from emerging
into stream. Food is scarce. Broad surface of
water exposed to cold. Shelter for trout
almost as poor as at stage A and will not
redevelop until May or June.

MIDSUMMER CONDITION IN ABOUT 3RD
TO 5TH YEAR AFTER GRAZING HALTED:
Further scouring of fine sediments from
stream bed. Silt bars at stream edges being
tied down by reed canary grass with its
tough system of roots and runners.
Watercress flourishing, and submergents at
peak of development. Only 4 feet of stream
width exposed to sky. and this shaded much
of day by high grasses. Greatest depth in
cross-section about 2 feet (60 cm). For trout,
shelter, food, and spawning gravels
are ample.

Figure 4

MIDSUMMER A FEW YEARS LATER:
Silt bars further stabilized by turl. Channel
narrowed by 40% to 50% since stage A.
Only 2 feet of stream width exposed;
therefore submergents less abundant. Also
less volume of watercress due to shade of
taller plants. Woody vegetation starting to
dominate.

LATE WINTER DURING STAGES D AND E:
Turf still holds bank materials firmly.
Overhanging fringes of matted grass provide
shelter for trout. Gravels remain clean
enough to allow normal hatching and
emergence of fry.

MIDSUMMER 10 TO 20 YEARS LATER:
Alders or other high bushes predominate
(saplings of ash, elrr or maple at left).
Turf completely shaded out. Water level high
due to clogging by debris. For trout, food
may be scarce, shelter is excellent beneath
banks, among roots and fallen branches.
But:

Innermost rows of alders will soon tip into
channel, further clogging flow and
destroying overhanging bank. The largely
vegetational processes of bank-building will
not be repeated as long as shade persists.

MANY YEARS LATER:

Mature forest . . . Dense shade. Few plants
on forest floor. Banks have eroded, channel
has spread and silts again cover stream bed.
Channel less than 1 foot deep. Little shelter
for trout. Even trees undermined by current
and toppled across the stream may provide
poor hiding cover. Conditions almost as
bad as in stage A.

Figure 4

This example reminds us that there are hundreds
of plants which regularly grow in the riparian.
We know very little about their intrinsic values
and how they interact in a normal , managed (not
overgrazed) riparian ecosystem. Certainly a
shaded stream with a nearly closed canopy over
a narrow, deep stream will produce cool, clear,
water and less sediment will reach the reser-
voirs, extending their lifetime. The fate of
many species such as the bald eagle may
ultimately depend upon the subtle energy flows
needed to produce the fish which the eagles
are dependent upon. The fate of several fish
like the endangered squawfish and others are
also dependent upon a properly functioning
riparian ecosystem.

This managed "riparian pipeline ecosystem"
will hopefully produce ample quality waters
for other downstream uses. The evapotranspir-
ation in the pipeline is not wasted, society
needs the products produced.

J. Stokley Ligon wrote 50 years ago,
"Cold water fish and fishing streams are as
seriously affected by overgrazed watersheds
as is game. Not only do the extremes of low
and high water, caused by floods and erosion,
affect the normal flow and temperature of
waters, but the destruction of willows, alders,
weeds and grasses eliminates both food and
shelter for cold water fish. No experienced
angler fishes in sun-exposed streams where
the water spreads shallow in unprotected flood-
ravished watercourses; he seeks the cool
shadows where the alders, willows or conifers
overhang the banks, where the stream is narrow
and banks with matted roots are secure along
New Mexico's cold water streams today. Abuse
by overgrazing of watersheds and watercourses,
as no other cause, has deteriorated New Mexico's
fishing. "

The creation or perpetuation of the little
winding stream jungles everywhere are a
national as well as a state need. The space
they occupy, whether on the farm, deep in
the creek bottom, canyon course or on overflow
lands, has no appreciable value frcm the stand-
point of agriculture or stock raising, but as
little jungles they have an intrinsic value.
As boys how many of us got our greatest thrills
and enjoyment from these little jungles - the
jungles we resorted to at every opportunity to
follow our dog after; rabbits, squirrels or
coons, or to hunt quail, fish, or to set our
traps for furbearers? The intensity of the
job and satisfaction thus derived demands that
this little institution, the wasteland jungle,
be perpetuated for the American boy and man.
These little spaces, properly protected, are
the only means of conserving the small game in
reclaimed canyons and valleys as commercialism
agressively overrides every weakling of Nature

that does not have the sympathetic support of
organized forces to oppose it."

CONCLUSIONS

1. The fishery resource is often energy
dependent upon the riparian vegetation and the
watershed .

2. Uncontrolled grazing brings about
complete type conversions in the riparian zone
and prevents streams from progressing to more
stable conditions.

3. Trees and other vegetation in the
riparian zone control sediments, provide stream
stability and tend to narrow and deepen channel
morphology, which benefits the fishery resource.

4. Research is vitally needed to document
and study the interactive and intrinsic value
of the many plant species in the riparian eco-
system.

5. The fishery, wildlife, esthetic res-
ources, and water quality and quantity are
dependent upon these interactions and our efforts
to integrate the needs of the various resources.
Free choice, uncontrolled grazing is incompatible
with these resources.

LITERATURE CITED

Anderson, T. W. 1976. Evapotranspiration losses
from flood-plain areas in central Arizona.
USGS, Open-File Report 76-864.

Babcock, H. M. 1968. The phreatophyte problem
in Arizona. 12th Annual Arizona Watershed
Symposium Proceedings, September 18, 1968.

Bruns, Dale Anthony Robert. 1977. Distribution
and abundance of benthic invertebrates in a
Sonoran desert stream. Arizona State Univ.

Campbell, C. J. and Win Green. 1968. Perpetual
succession of stream-channel vegetation in a
semiarid region. J. Az . Acac. Sci. 5:86-98.

Carothers, Steven W. and R. Roy Johnson. 1975.
Water management practices and their effects
on nongame birds in range habitats. For.
Serv. Gen. Tech. Rep. W0-1.

Chapman, Donald W. and Robert L. Demory. 1963
Seasonal changes in the food ingested by
aquatic insect larvae and nymphs in two
Oregon streams. Ecology, Vol. 44
No. 1

Cummins, K. . W., J. J. Klug, R. G. Wetzel, R. C.

Petersen, K. F. Suberkropp, B. A. Manny,

J. C. Wuycheck, and F. 0. Howard. 1972.

Organic enrichment with leaf leachate in

experimental lotic ecosystems. Bioscience

Vol. 22 No. 12.
Cummins, Kenneth W. 1974. Structure and

function of stream ecosystems. Bioscience

Vol. 24 No. 11.

Ensign, H. R. 1957, Foods eaten by brown

trout in two southern Wisconsin trout streams,
Mt. Vernon and Black Earth Creeks, Dane
County, Wisconsin. Wisconsin Dept. Nat.
Resources, Southern Area Invest. Mem. 181.

Fisher, Stuart G. and Stephen R. Carpenter.
1974. Ecosystem and macrophyte primary
production of the fort river, Massachusetts.
Hydrobiologia, Vol. 47, 2,"pag. 175-187
1976

Fisher, Stuart G. 1972. Stream ecosystem:
organic energy budget. Bioscience Vol. 22
No. 1

Fisher, Stuart G. and Gene E. Likens. 1973.
Energy flow in Bear Brook, New Hampshire:
an integrative approach to stream ecosystem
metabolism. Ecological Monographs, Vol 43
No. 4, pp. 421-439.

Fisher, Stuart G. and W. L. Minckley. 1977.
Chemical characteristics of a desert stream
in flash flood. Department of Zoology,
Arizona State Univ., Tempe, Ariz. 85281.

Fisher, Stuart G. 1971. Annual energy budget
of a small forest stream ecosystem: Bear
Brook. University Microfilms, Ann Arbor,
Mich.

Hibbert, Alden R. and Paul A. Ingebo. 1971.
Chaparral treatment effects on streamflow.
15th Annual Arizona Watershed Symposium
Proceedings

Horton, Jerome S. 1976. Management of moist-
site vegetation for water: past history,
present status, and future needs. U.S.
Forest Service, Region 5, San Francisco,
California.

Howarth, Robert W. and Stuart G. Fisher. 1976
Carbon, nitrogen, and phosphorus dynamics
during leaf decay in nutrient-enriched stream
microecosystems. Freshwater Biology (1976)
221-228.

Hubbard, John P. The riparian vegetation of
New Mexico. New Mexico Dept. of Game & Fish,
Santa Fe, NM,

Hunt, Robert L. 1975. Food relations and
behavior of salmonid fishes. Use of
terrestrial invertebrates as food by
salmonids (chap. 6.1). Spr inger-Verlag
New York Inc .

Johnson, Phil. 1975. More water for Arizona?

Forestry Research, USDA Forest Service,

Ft. Collins, Colorado 80521
Leopold, A. Starker. 1975. Ecosystem

deterioration under multiple use. Univ.

of Calif., Berkeley.
Lewis, Douglas D. 1961. Effects of controlling

riparian vegetation. Proceedings of 5th

Annual Arizona Watershed Symposium.
Ligon, J. Stokley. 1927. Wildlife of New Mexico

it's conservation and management. State

Game Commission, Dept. of Game and Fish,

Santa Fe, NM
Likens, Gene E. and F. Herbert Bormann. 1974.

Linkages between terrestrial and aquatic

ecosystems. Bioscience Vol. 24 No. 8.
McDiffett, Wayne F. 1970. The transformation

of energy by a stream detritivore, pteronarcys

scotti (plecoptera) . Ecology, Vol. 51, No. 6.
McDowell, William H. and Stuart G. Fisher. 1976.

Autumnal processing of dissolved organic

matter in a small woodland stream ecosystem.

Ecology, Vol. 57, No. 3.
Marcuson, Pat. 1970. Rock creek floodplain

investigation, July 1, 1968 to June 30, 1969.

Job completion report, project F-20-R-13.

Montana Fish and Game Dept., Helena.
Minckley, W. L.1976. Aquatic Habitats & Fishes

of the Lower Colorado River. Final Report

Contract No. 14-06-3002-529. Bureau of

Reclamation, Boulder City, Nevada 89005.
Minshall, G. Wayne. 1966. Role of allochthonous

detritus in the trophic structure of a

woodland springbrook community. Ecology,

Vol. 48, No. 1.
Odum, Howard T. 1955. Primary production in

flowing waters. Department of Zoology,

Duke University, Durham, N.C.
U.S. Forest Service. 1974. The effect of

cattle grazing on fish habitat. Region 6,

Portland, Oregon.
U.S. Forest Service. N.D. National Forests

provide water for Arizona. Region 3,

Albuquerque, NM
White, Ray J. and Oscar M. Brynildson. 1967.

Guidelines for management of trout stream

habitat in Wisconsin. Tech. Bui. 39

Dept. of Nat. Res. Madison, Wisconsin

Management Alternatives
for the Riparian Habitat
in the Southwest1 L? *

Gary A. Davis 2/

Abstract — ; Exploitation, by man, has significantly al-
tered the riparian- habitat in the Soufhwest. For decades,
the primary or dominant use of riparian habitat has been water
management; other values were not considered. Management al-
ternatives and objectives are evaluated for environmental con-
sequences.

Diversity and numbers of plant and animal
species are continually changing through geo-
logic time. Disappearance of some plant and
animal species and the emergence of others re-
sults from evolutionary processes of natural
selection. Plant and animal species are con-
stantly adapting to changing environmental
pressures. Fossil records indicate that ex-
tinction is the inevitable fate of all spe-
, cies. Continual variation in the physical and
biological environment initiate extinction in
nature. When an individual species is unable
to adapt to changing environmental stresses,
it is replaced by others.

Prior to the appearance of Homo sapiens
on this planet, extinction occurred as a con-
sequence of natural phenomena. With the ad-
vent of humans, an additional stress was ex-
erted on the physical environment. Some data
imply that the rate of extinction increased as
a result of human stress (Martin, 1967). Human
stress on the environment has many forms — ag-
ricultural practices, timber harvesting, do-
mestic animal grazing, industry, hunting, pred-
ator control, and pollution. Often it is the
interaction of numerous types of stress which
results in the extinction of a species.

1/ Paper presented at the Symposium on
Importance, Preservation, and Management of
the Riparian Habitat, Tucson, Arizona, July
9, 1977.

2/ Wildlife Biologist, USDA, Forest
Service, Apache-Sitgreaves National Forests,
P.O.Box 640, Springerville, Arizona 85938

The primary causal factors of animal ex-
tinction include, but are not limited to: eco-
system alteration, introduction of exotic spe-
cies, predator and pest control, pollution,
poaching, and the capture of wild animals for
legitimate and illegal purposes. Ecosystem al-
teration is one of the more significant causes
of extinction. When wildlife niches are altered,
animals must move to other areas, adapt to a new
environment, or die. Even though some habitats
are not totally destroyed, there may not be
enough suitable area remaining to maintain a
viable population. Habitat destruction is re-
sponsible for approximately 30 percent (%) of
the presently endangered species (Uetz & John-
son) .

Riparian habitat in the Southwest is a
classic example of the effect man can exert
on a particular habitat. Records of early ex-
plorers (Emory, 1948) reveal that riparian com-
munities have been altered significantly from
the original type. Significant man-caused im-
pact on the riparian type began approximately
450 years ago, when European man first jour-
neyed into the Southwest from Mexico. Early
day grazing undoubtedly had an effect on ri-
parian areas. In the last 100 years, the rate
of alteration has increased significantly.
This is due largely to ever-increasing human
pressures, land clearing for agriculture, dam
construction, grazing, pumping of ground and
surface water for irrigation, and increased
recreational pressures. For decades the pri-
mary or dominant use of riparian habitat in
the Southwest has been water management; other
values were not considered. The dominant use
was to supply metropolitan areas with water.

Wildlife populations have adapted to survive
in these alterations of the riparian type.

The importance of the riparian type for
wildlife has been well documented, particu-
larly for avian species. MacArthur (1964) es-
tablished a correlation between bird species
diversity (BSD) and floral height diversity
(FHD) . He also reported that habitats with
permanent water had higher avian populations
than those without. Johnson & Carothers (1975)
recorded the highest population of non-colonial
nesting birds ever reported in North America
in a homogenous cottonwood stand along the
Verde River in Arizona. Of the 70 breeding
species investigated in the riparian type, 50%
were obligate nesting species, 20% indicated
a decided preference for the riparian type and
30% nested in either the riparian type or non-
riparian without a significant preference for
either type (Carothers & Johnson, 1975).

on riparian zones or utilize them proportion-
ately more than any other habitat type. In
short, the riparian type is the most impor-
tant habitat type in the Southwest for wild-
life.

A substantial volume of literature docu-
menting the importance of the riparian type
has been published but some key questions need
to be answered prior to initiating a realistic
attempt to manage this type.

1. What f loristically is a riparian
community?

2. Where is it located and what is
its ecological condition?

3. What are the ecological factors
limiting perpetuation of the
community?

Gavin & Sowls (1975) found 476 pairs of
nesting birds per 40 hectares in a mesquite
(Prosopis juliflora) bosque in Southern Ari-
zona. The adjacent habitat type was temperate
and desert grassland. Balda (1967) found 31
and 46 pairs per 40 hectares (ha. ) in the
mixed grass and yucca-grassland types in
Southern Arizona. Carothers (1974) found 332
pairs/40 hectares (ha.) in the mixed broadleaf
type in the Verde Valley in Arizona. Beidleman
(1960) and Hering (1957) reported 30 pairs per
40 hectares in the adjacent pinyon- juniper type.
Obviously, bird densities are significantly
higher in the riparian type than in adjacent
communities .

Riparian vegetation enhances aquatic hab-
itats through reduction of solar radiation,
reduced erosion, decreased sedimentation, and
energy input in the form of vegetational debris
and terrestrial insects. Most of the food for
important aquatic insects comes from land veg-
etation. Several studies show that these sources
contribute a 50-70% of the energy responsible
for producing fish in a stream. (Fisher &
Likens, 1973).

Riparian habitats have three basic pre-
requisites for wildlife: food, water, and
cover. The cover component has proportionately
more ecotones than any other type. Ecotonal
areas are a result of horizontal and vertical
stratification of deciduous and evergreen
trees, water and hydrophilic plants, and the
undulating configuration of the type. Verte-
brates that either live or reproduce in water
are confined to these zones. Riparian hab-
itats receive proportionately more use per
unit area than any other type. A large per-
centage of terrestrial species known to occur
in a given area are either directly dependent

Is the community maintaining it-
self through natural reproduction?
If not, what are the factors pre-
venting perpetuation?

What should our management ob-
jectives be for riparian habitat?
Water production, habitat for
wildlife, water quality, recreation,
fuelwood, aesthetics, fisheries,
grazing, and agriculture are all
potential uses of the riparian
type.

What is the species composition
and age class of a healthy ri-
parian community?

Verbose definitions of what constitutes
a riparian type abound in the literature, but,
simply stated, it is an aggregation of floral
species which depend on a flow of water on or
near the surface for subsistence. Riparian
habitat occurs in every life zone in Arizona
with the possible exception of the Hudsonian.
Species composition changes with elevation.
Often the climatological conditions prevalent
in drainageways allow the downward extension
of a higher elevation species such as ponder-
osa pine (Pinus ponderosa) fingering down a
canyon into the Upper Sonoran Life Zone. These
inclusions are ecologically important as they
provide an additional ecotone within the arid
Upper Sonoran Life Zone and should properly be
classified as a riparian community.

These riparian communities should be
mapped and classified as to type and condition
rating. Until we know where they are and what
their ecological condition is, we cannot man-
age them. This should be an integral compon-

ent of our planning process.

Prior to implementation of any type of
management, the most critical need is know-
ledge of the ecological requirements of indi-
vidual plant and animal species for self prop-
agation. It is not realistic to believe that
we can artifically maintain a vegetative type
through perpetuity. Classification of ripar-
ian communities and documentation of their con-
dition class will reveal whether or not these
communities are maintaining themselves. If
not, the next logical step is to carefully de-
termine what are the causal factors. Generally
speaking, failure of the riparian type to re-
generate itself in Arizona can be related to
several factors either operating independently
or in conjunction with one another.

1. Loss of water flow as a result of
diversions for irrigation, impoundments for
metropolitan usage, and lowering of water
tables by pumping for sundry uses.

2. Loss of significant portions of en-
tire communities as a result of devestating
floods. These periodic floods are significant
because they remove substantial numbers of
older mature trees which serve as seed sources.
Many of the riparian species are adapted to
periodic flooding and an occasional flood is
necessary for germination and survival of the
seedlings, but floods of a significant magni-
tude are detrimental.

3. In areas of high recreational use,
soil compaction, trampling, and inability of
the soil to retain moisture prevent seedling
establishment. Also, loss of ground vegetation
(herbaceous) dries out the site and prevents
regeneration of some species.

4. Phreatophyte control essentially
eliminates the vegetation, removes the seed
source, and changes the micro-site relation-
ships.

5. Overgrazing by domestic livestock,
in ray opinion, is probably the major factor
contributing to the failure of riparian com-
munities to propagate themselves. Continued
overuse of riparian bottoms eliminates essen-
tially all reproduction as soon as it becomes
established. Overstocking and the consequent
loss of vegetative cover on the adjacent water-
sheds is probably the main reason for the fre-
quency of high intensity floods resulting in
drastic changes in the density and composition
of riparian bottoms.

An evaluation of a riparian community
necessitates making a judgment of whether the
type is in good, fair, or poor condition. In

the Southwest we are talking about many dif-
ferent species aggregations within the ripar-
ian type. Significant research data is needed
to answer some of these questions:

a. Should a certain percentage of the
vegetation be comprised of a particular spe-
cies?

b. What should the age class distribution
be in a healthy stand?

c. What is an ideal canopy coverage in
percent?

d. What should the composition and den-
sity of herbaceous vegetation be?

e. Does a particular site have potential
to develop a riparian community under proper
management ?

Research has been initiated here in the
Southwest in an attempt to answer some of
these questions. During the interim we have
developed the following scorecard to use in
our evaluations.

RIPARIAN STAND ANALYSIS

This rating of the riparian habitat will be
based principally on its attraction to asso-
ciated wildlife and ecological stability of
the type.

The 100-point transect described for browse
and the aspen stand analysis will be used.
Certain modifications in the technique and
score card will make it adaptable for the ri-
parian type. A description of this technique
follows :

A. Mapping

Riparian types will be delineated on aerial
photographs .

B. Establishing Transects

1. Riparian types to be analyzed will
be sampled with paced condition
transects .

2. Transect locations will be carefully
selected to fall within representative
portions of the type.

3. Additional transects will be run in
the same stand whenever a change in
condition is recognized.

4. Within the stand to be sampled, select
a route and pace interval that will

provide a good cross section of the stand.
The starting point should be identified and
pin pricked on an aerial photo.

5. Pace along the chosen route, walking
as straight as practically possible. Along a
meandering stream course, cross back and forth
across the channel but do not take sample points
in the channel.

lowing size classes. All specimens greater
than 12" dbh will be recorded individually
by species. Basal area in sq. ft/ acre will
be computed by using standard basal area table

9. At each tenth sampling point, a 1/100-
acre pellet group plot will be run. Include
all countable groups for deer, elk, cattle or
horses.

6. At each sample point, record whatever
is found with a 3/4 inch loop immediately in
front of a mark on the boot toe. This may

be bare ground or erosion pavement, rock, lit-
ter, grass, or forb. Grasses and forbs will
be identified and tallied by individual species
when all or part of the live root crown falls
inside the loop. Record as litter if more than
one-half of the loop covers dead plant material
older than that resulting from current growth.
Record hits on rock only for rock in place.
Small, loose moving rock should be tallied as
erosion pavement.

7. At each sampling point, the examiner
will record, by species, the nearest woody
riparian plant to the boot toe that occurs
within a 180 degree arc in front of the sample
point ("hit"). If the species involved can
be described in timber terminology as a sprout
(less than hh feet tall), a sapling (4% feet
tall to 4.9 inches diameter breast height
d.b.h. ), a pole (5 inches to 8.9 inches dbh),
or mature (over 9 inches dbh) , it should be
tallied as such. If, however, the species in-
volved is mature (at 4 inches dbh) , the ob-
server should use his best judgment on where
the specimen of that species fits into the
above described sale (i.e., if a species is
mature at 4 inches dbh and one is "hit" that
is 3 inches dbh. , it should be tallied as a
pole, not as a sapling.) If a dead riparian
species is "hit", tally it and then record
the size class for the nearest live riparian
species. This will result in a transect sam-
ple of 100 live riparian species. If a riparian
species is a sprout, determine if the sprout
has been browsed or not. Dot tally this in-
formation on the appropriate column.

8. At each tenth sampling point, obtain
the basal area and crown density of all woody
species. Crown density will be taken with a
spherical densiometer. Count each corner which
intersects an opening in the canopy. Each cor-
ner represents approximately 6% of the total
canopy. Multiply the number of corners which
intersect openings by 6 and subtract this fig-
ure from 100 for crown density percentage.
Basal area will be computed in the following
manner: using a l/100th acre plot (11' 9"
radius) record the dbh of all woody species

at breast height and dot tally into the fol-

C. Composition

"A species" (must be 4 or more) making up
75% or more of the composition. = H

"A species" (must be 2 or more) making up
35% or more of the composition. = M

"A species" comprise less than 35% of the
composition or only one "A species"
represented. = L

Species Rating - A

Cottonwood Ash Mulberry

Sycamore Willow

Walnut Alder

Hackberry Elm

Grape Box Elder

Rhus Oak

D. Crown Density

Crown density, as utilized in this partic-
ular scorecard, serves as a criterion of rel-
ative dominance, of potential productivity,
of the influence of plants on precipitation
interception and soil temperature, and of the
value of vegetation to animals. It is appli-
cable to almost all ecosystems, owing to the
universal importance of light coming from
above .

Crown density will be taken with a spherical
densiometer. Count each corner on the grid
which intersects an opening in the canopy.
Each intersection represents approximately 6%
of the total canopy. Multiply the number of
intersections which occur in openings in the
canopy by 6 and subtract the result from 100
for crown density percentage.

Crown Density Rating Guide

80%-100% = High (H)
50%-80% = Medium (M)
0-50% = Low (L)

E. Basal Area

Basal area refers to a comparison of species
as to the aggregate cross-sectional area of
the individual plants taken at or near ground

level, per unit of land area. Basal area gives
a relative indication of dominance and biomass
(by species) for the riparian community.

Basal area will be computed utilizing a l/100th
acre plot (11' 9" radius) at each tenth sampling
point. Record the d.b.h. of all woody species
at breast height and dot tally into size classes.
All specimens greater than 12" d.b.h. will be
measured and recorded individually by species.
Conversion factors for all d.b.h. size classes
from 0"-12" are included on the scorecard. For
those species greater than the 12" d.b.h. use
the standard basal area tables included in the
handbook.

Basal Area Rating Guide

the Riparian Stand Structure Rating Guide and
indicate score on Form.

Riparian Stand Structure Rating Guide

All age classes represented with sprouts/seed-
lings and saplings of "A species" making up
30% or more of the stand. = H

At least 3 age classes represented with sprouts/
seedlings and/or saplings of "A species" making
up 10% or more of the stand. = M

Less than 3 age classes of "A species" repre-
sented with sprouts/seedlings and/or saplings
of "A species" making up less than 10%
of the stand. = L

60 sq. ft/acre or greater = High (H)
30 sq. ft/acre - 60 sq. ft/acre = Medium (M)
0-30 ft. sq/acre = Low (L)

F. Vigor

Vigor is determined by utilizing three (3)
criteria: (1) the percentage of "A species"
which are sprouts, (2) the percent of "A
species" sprouts which have been browsed, (3)
the number of "hits" on dead "A species."
Summarize data for each measurement, apply to
Riparian Vigor Rating Guide and indicate ap-
propriate vigor rating (L-M-H) on the riparian
scorecard. (See example below)

Riparian Vigor Rating Guide

Riparian type No more than No more than

has at least and 25% of the and 10 "hits" on

10% sprouts/ sprouts/seed- dead ripar-

seedlings of lings are ian species

"A species" browsed

Riparian type

No more than

has over 5% and 75% of and 30 "hits" on

sprouts/seed- sprouts/seed- dead ripar-

lings of "A lings are ian species

species" browsed

Riparian type

More than

"Hits" on

has less than or of sprouts/ or dead ri-

5% sprouts/ seedlings are parian

seedlings of browsed species

"A species" exceed 30

G. Stand Structure

The age class distribution of "A species" de-
termines the stand structure rating which will
be applied to a riparian stand. This rating
is based on the percentage of sprouts and sap-
lings in relation to poles and mature "A spe-
cies". Summarize this percentage and apply to

The key question that needs to be answered
is what should our management objectives be for
the riparian habitat? Should the management
objective be identical for all the riparian
type, or should they be tailored to fit differ-
ent species aggregations?

The riparian type has many potential uses
but our primary objective should be to maintain
the type in a healthy ecological condition, a
condition which enables natural perpetuation
of the community. It should be managed as the
most sensitive habitat in the Southwest. This
is particularly important because it is an area
of maximum potential conflict between resources
such as timber, wildlife, grazing, recreation,
and water production. Past management has
tended to overlook or disregard the intangible
or non-economic uses of the community. Public
land management agencies, partially as a con-
sequence of public pressures, have had diffi-
culty recognizing uses that are superficially
lacking in tangible economic benefits. The
dominant use of riparian type has been grazing
and water production with little thought given
to its value for wildlife and recreation or
preservation as a unique community.

In order to evaluate management alter-
natives, an investigation of potential ben-
efits versus ecological consequences is needed.
Multiple use management should not assume that
all uses should necessarily occur on the same
acre of ground. Typically, management objec-
tives are complicated by a variety of environ-
mental situations and conflicting demands on
resources .

If our management objective is to maximize
the net gain in usable water , we should treat
the upper watersheds and eliminate the riparian
vegetation along the stream channel. Heindle
(1965) estimated that we were harvesting ap-

proximately 5 million acre feet of surface
water annually in Arizona, New Mexico and
western Texas and predicted this amount could
be doubled by treating upper watersheds, erad-
icating all riparian vegetation, suppressing
evaporation from reservoirs, salvaging exces-
sive surface water, diversions, and capturing
uncontrolled streamflow.

Predictable amounts of water salvaged as
a result of the complete removal of riparian
vegetation have not been thoroughly documented.
Estimates vary with different studies: Culler
(1970) estimated an approximate savings of 0.8
acre/ft. per acre when dense tamarix (Tamarix
Pentandra) and mesquite were completely clear-
ed. Bowie and Kam (1968) est imated that com-
plete removal of 22 acres of cottonwood (Pop-
ulus f remontii) , willow (Salix. spp), and seep-
willow (Baccharis spp.) would salvage approx-
imately 1.7 acre ft. /acre or a savings of 6
percent of the inflow. Converting 15 acres of
riparian shrubs and trees to grass in Southern
California increased water yield 17 acre feet
(1.1 acre ft. /acre) in eight months (Rowe,
1963). Average water savings in certain hab-
itat types is approximately 1 to 2 acre ft./
acre (Horton & Campbell, 1974).

Control of riparian vegetation for water
production appears to be most feasible on flood
plains where the water table is between 8 to
20 feet in depth and on upper watersheds above
7,000 feet in moist coniferous sites. Removal
of riparian vegetation along perennial streams
is probably not economically feasible because
evaporation exceeds transpiration (Horton &
Campbell, 1974).

Several logical assumptions can be pos-
tulated from the aforementioned studies: (1)
removal of riparian vegetation increases sur-
face flow but to what degree depends on the
species, composition, and density; (2) in-
creases in surface flow are modest because of
the attendant increased surface evaporation;
(3) re-treatment of the site is necessary as
a result of reinvasion. (Campbell, 1970)

Evaluating the data brings to mind an
interesting hypothesis. If we assume that
water is a natural resource and the demand
for water in large metropolitan areas for
municipal and industrial uses will increase
significantly, the price of water will also
increase. If the demand is such that we need
to increase our water yields we can accomplish
this task and also improve the condition of
our riparian habitat if we concentrate our
efforts on the upper forested watersheds and
the floodplains below 3500 feet with dense
stands of mesquite or tamarix.

Dortignac (1965) reported maximum water
yields emanate from forested high-elevation
watersheds. He estimated that, in the Rio
Grande Basin in New Mexico, 32 percent of the
total water yield comes from the spruce-fir-
aspen forest above 8,000 feet, while 40 percent
is derived from the ponderosa pine forest.
Horton & Campbell (1974) suggested that phreato-
phyte control is most effective on floodplains
in lower elevations which support a dense stand
of phreatophytes .

Riparian habitat that occurs between 7000-
3500 feet in elevation has the highest ecolog-
ical diversity, the greatest value to wildlife,
and is the most abused by overgrazing. Increased
streamflow through this elevational zone as a
result of treatment in the upper watersheds
would, if accompanied by reductions in domestic
livestock, change some ephemeral streams to
perennial, enhance regeneration potential as a
result of increased moisture conditions, enhance
density and vigor, improve aquatic habitat, and
reduce stream temperatures as a result of more
shading. Riparian vegetation in this zone, in
most cases, is relatively sparse. Increasing
the streamflow would increase the density of
vegetation with an attendant increase in the
amount of water lost through evapotranspiration.
However, if this anticipated increase flows
into perennial streams with a dense stand of
riparian vegetation, no significant increase
in evapotranspiration is predicted. (Campbell,
1970)

What would be the consequences of maxi-
mizing water yields without mitigating for
other resources? The answer must be specu-
lative, but the following results can be vis-
ualized :

1) All riparian plants will be temporarily
suppressed .

2) Erosion and sedimentation will increase
significantly because stream banks will lack
vegetation for stabilization.

3) Transpiration losses will be negligible,
but evaporation from the soil will increase as

a result of higher soil temperatures and shal-
lower water tables.

4) Rate of siltation of downstream res-
ervoirs will increase.

5) Degradation of aquatic habitat will
occur as a result of :

a. increased water temperatures

b. loss of energy from vegetational
debris

■ ■

c. loss of niches for aquatic
insects

d. increased algae growth

6) Riparian habitat for wildlife will
be lost; many species would be completely
extirpated.

7) Aesthetic quality would be signifi-
cantly diminished.

8) Potential recreational opportunities
would be eliminated.

9) Potential for torrential type floods
will increase.

10) Forage and cover for domestic live-
stock would be reduced.

What management strategies and alterna-
tives are available if the stated objective is
to manage the riparian type for production of
domestic livestock? Obviously, the riparian
type consists of many different aggregations
of species, occurs within many habitat types,
and is subjected to numerous management situ-
ations. Management strategies must, because
of the diversity of the type, be referred to
in a general sense. There is no panacea which
is applicable to all situations.

Logically, prior to proposing a manage-
ment strategy we need to know: What are the
problems and what are the desired consequences?
The problem is that the riparian areas are in
poor condition, particularly when their poten-
tial productivity is considered. In order to
correct a problem, one needs to determine what
was/is the cause. Overgrazing by domestic
livestock, in my opinion, is the obvious an-
swer. The desired consequence is to create a
situation within the riparian type which will
support an optimum number of domestic live-
stock on a sustained basis. This implies main-
taining a suitable forage base through perpe-
tuity to support livestock numbers for future
generations .

The effect overgrazing has had on the
riparian type is twofold: 1) increased potential
for devastating floods due to elimination of
vegetative cover on adjacent watersheds; 2)
removal of herbaceous material and seedlings
and/or sprouts of woody riparian in the bottoms.
Consequently, the following situation exists:

1) failure of the type to reproduce
itself ;

2) poor representation of age classes;

3) low vigor;

4) lack of sufficient vegetative cover
to prevent erosion;

5) elimination due to channel-scouring
floods of older mature trees which
constitute critical seed sources;

6) elimination of moist microsites re-
quired for reproduction of such spe-
cies as sycamore (Platanus wrightii) ;

Proper stocking on adjacent watersheds
is needed to reduce both the volume and fre-
quency of flooding. If this cannot be accom-
plished, efforts to obtain reproduction in the
riparian type will not be as effective.

An expedient procedure to rejuvenate ri-
parian stands is to exclude livestock by fencing
until reproduction is out of reach. In steep
canyons this can be accomplished easily because
of restricted accessibility, but in other areas
many miles of fence would be required. Riparian
species are prolific growers. If conditions are
amenable to growth, cotton (Populus spp.), alder
(Alnus spp. ) and sycamore can grow 10 to 15 feet
in several years if protected from grazing.

Once re-establishment has occurred, graz-
ing under a rest-rotation management program
accompanied by proper utilization factors, salt-
ing and riding can be utilized to maintain the
optimum species composition for a sustained
yield of domestic livestock.

Anticipated environmental and social con-
sequences of managing the riparian habitat for
domestic livestock are:

1) a significant reduction in stocking
rates would temporarily have an adverse economic
effect on many livestock operators;

2) decreased flooding potential;

3) improvement of terrestrial and
aquatic habitats;

4) reduced erosion and sedimentation;

5) improvement in water quality;

6) reduction in water yield;

7) retention of long term site pro-
ductivity;

8) improved forage production for
domestic livestock;

9) enhanced recreational opportunities;

10) increased esthetic quality;

LITERATURE CITED

Management of the riparian habitat for
wildlife could best be accomplished by the
total exclusion of domestic livestock with the
exception of water gaps for watering purposes.
A theoretical exception whereby periodic graz-
ing would be beneficial would be a marsh area
occupied by nesting waterfowl. Dense vegetation
along the periphery should be eliminated period-
ically by grazing to retain a terrestrial her-
baceous food source. A logical question as re-
gards a recommendation to exclude livestock
would be: Can livestock be prudently utilized
to maintain a desirable understory composition?
Realistically, the time necessary to restore
the riparian habitat to a healthy condition is
decades. The potential use of livestock to
manipulate vegetation in the riparian habitat
may be worthy of consideration in 30 years.
Horizontal and vertical stratification, diver-
sity of floral species, and floral volume is
needed for optimization of wildlife habitat
— regardless of what is done, this will not
be realized for many years.

Environmental consequences of managing
riparian habitat for wildlife are essentially
the same as listed for managing for livestock
with the following exceptions:

1. Adverse economic effect would be
permanent, i.e., production of domestic live-
stock from the riparian type.

2. Forage production for livestock
would not improve because they would be ex-
cluded.

3. Reduction in water yield would
increase.

Management for recreation would utilize
the procedures mentioned for wildlife, but
access should be provided by trails, camp-
grounds, etc. Environmental consequences are
the same.

Riparian habitat in the Southwest is
rapidly dwindling. Land managers need to
initiate management to stop the rate of loss
and insure the perpetuation of the community.

Balda, R.P.

1967. Ecological relationships of breeding
birds of the Chiricahua Mountains, Arizona.
Unpublished Ph.D. thesis. Univ. of 111.

Beidleman, R.G.

1960. Breeding bird census pinyon pine -
Rocky Mountain juniper forest. Audubon
Field Notes 14: 495-496.

Bowie, James E. , and William Kam

1968. Use of water by riparian vegetation.
Cottonwood Wash, Arizona. U.S. Geol.
Survey. Water Supply Paper. 1858, 62 p.

Campbell, C.J.

1970. Ecological implications of riparian
vegetation management. Journal of Soil
and Water Conservation. 25: 49-52.

Carothers, S.W. , R. Roy Johnson and S.W.
Aitchison

1974. Population structure and social
organization of Southwestern riparian
birds. American Zoologist. 14: 97-108

Carothers, S.W. and R. Roy Johnson

1975. Water management practices and their
effects on non-game birds in range habi-
tats. Proc. of the Symposium on Manage-
ment of Forest and Range Habitats for Non-
Game Birds. U.S.D.A. Forest Service.
Gen. Tech. Rep. WO-1

Culler, Richard G.

1970. Water conservation by removal of
phreatophytes. Am. Geophys. Union Trans.
51: 684-689.

Dortignac, E.J.

1956. Watershed resources and problems of
the Upper Rio Grande Basin. Rocky Mtn.
For. & Range Exp. Sta. , Fort Collins,
Colo. 107 pp.

Emory, W.H.

1848. Notes of a military reconnoissance
(sic) from Fort Leavenworth, in Missouri,
to San Diego, in California, including
part of the Arkansas, Del Norte, and Gila
Rivers. 30th Congress, First Sess. , Dec,
Washington. Wendell and Van Benthuysen.

Fisher, S.G. , and G.E. Likens

1973. An integrative approach to stream
ecosystem metabolism. Ecol. Mono. 43(4)
Autumn 1973.

Gavin, T. A. and L.K. Sowls

1975. Avian fauna of a San Pedro Valley
mesqulte forest. Jour. Ariz. Acad, of
Sci. 10: 33-41

Heindl, L.A.

1965. Ground water in the Southwest - a
perspective. Ecology of Groundwater in
the Southwestern U.S. Az. State Univ.,
Tempe, Arizona. pp. 4-26

Hering, L.

1957. Breeding bird census, pinyon-j uniper
forest. Audubon Field Notes. 11: 448-449

Horton, J.S. and C.J. Campbell

1974. Management of phreatophyte and ri-
parian vegetation for maximum multiple
use values. U.S.D.A. Forest Service Res.
Paper. RM-117

MacArthur, R.H.

1964. Environmental factors affecting bird
species diversity. Amer. Naturalist
98: 387-397.

Martin, P.S.

1967. "Prehistoric Overkill," Pleistocene
Extinction: The Search for a Cause. P.S.
Martin and H.E. Wright, Eds. Princeton
Univ. Press, N.J. pp. 75-120.

Rowe, P.B.

1963. Streamflow increases after removing
woodland-riparian vegetation from a
southern California watershed. Jour, of
For. 61(5): 365-370.

Johnson, R.R. and S.W. Carothers

1975. The effects of stream channel mod-
ifications on birds in the Southwestern

United States. Symposium on Stream Chan- Uetz, George and D.L. Johnson

nel Modification Proceedings. Aug. 15-17, — Breaking the Web. National Geographic

Harrisburg, Virginia. Society.

Endangered Species vs.
Endangered Habitats: A Concept1

1,2-2 3
R. Roy Johnson , Lois T. Haight , and James M. Simpson

Abstract. - Although the great diversity within
riparian ecosystems was recognized earlier, their extreme
productivity was not discovered until this decade. The
highest densities of nesting birds for North America have
been reported from Southwest cottonwood riparian forests.
Complete loss of riverine habitat in the Southwest lowlands
could result in extirpation of 47 percent of the 166
species of birds which nest in this region.

INTRODUCTION

Since 1600 more than 120 bird and mammal
species have become extinct while more than
300 are now threatened (Fisher et al. 1969).
In addition, dozens of fishes, amphibians and
reptiles have become extinct or are endangered
to say nothing of invertebrate species. Habi-
tat disruption and destruction have been a
major cause of extinction. Only 24 percent of
the birds and 25 percent of the mammals became
extinct through natural causes. Of the 76
percent of the birds and 75 percent of the
mammals which died from human related causes,
well over half have been through indirect
means, such as introduction of exotic speices
and habitat disruption (Fisher et al. 1969 and
IUCN Red Data Books issued periodically) .

In an attempt to reduce the numbers of
species which will soon become extinct, se-
veral steps have been taken. A major step
involves the formation of recovery teams,
comprised of authorities on a given species,
such as the Bald Eagle. The activities of
these teams have apparently been beneficial
in slowing down rates of loss in wildlife
species. However, the efforts of recovery
teams cannot possibly prevent continued
extirpation if we continue to disrupt habitat
through activities such as overgrazing, urban-
ization, "modern, clean" agricultural prac-
tices, dam construction and channelization.
Continued research is needed to provide answers
to questions posed by management regarding
means through which critical wildlife habitat
may be preserved.

DISCUSSION

Extirpation

The extirpation of wild animal species
has been a cause for concern for decades.
People only mildly interested in conservation
can bring to mind the examples of the Passenger
Pigeon (Ectopistes migratorius - extinct 1914),
the Carolina Parakeet (Conuropsis carol inensis-
extinct 1914) , the Dodo (Raphus cucullatus-
extinct 1681) and the Great Auk (Pinguinus
impennis - extinct 1844) . Dates for extinction
are from Pettingill (1970) and Van Tyne and
Berger (1971) . An entire book has been written
about the Passenger Pigeon (Schoerger 1955) and
people are still trying to find out whether
or not the Ivory-billed Woodpecker (Campephilus
principalis) is now extinct. Several recent
books have been written appealing to citizens
of the world to help save these rapidly
diminishing species (Greenway 1958, Fisher et
al. 1969, Prince Phillip and Fisher 1970,
Simon and Geroudet 1970, Tylinek and Ullrich
1972, and Ziswi]er 1967). Information from
the International Union for Conservation of
Nature and Natural Resources (I. U.C.N.) Red
Data Books (issued periodically) presents a
dismal picture (Table 1) .

1_ Paper presented at the Symposium on
Importance, Preservation and Management of the
Riparian Habitat, Tucson, Arizona, July 9, 1977.

2^ National Park Service, Grand Canyon
National Park

3^ Associate, Museum of Northern Arizona,
Flagstaff

Table 1. — A history of species' extirpation.

(adapted from I. U.C.N. Red Data
Books)

Number of

1 2
Date extinctions Direct Indirect

1600s 21

1700s 36

1800s 84
1900-

1974 85

86% 14%

84% 16%

24% 76%

28% 72%

Direct = Hunting for food or commercial
causes .

Indirect = Habitat disruption, introduction
of exotics, etc.

Attempts to Prevent Extinctions

The concern over the increasing numbers of
species being exterminated in the United States
caused the U.S. Fish and Wildlife Service
to begin work on classification of "threatened"
wildlife in the early 1960's. The 1st edition
of the "Redbook" was issued in July 1966. We
use the words "threatened" and "endangered*1 in
an unofficial sense (see U.S. Fish and Wildlife
Service 1973 for official definition). Endan-
gered species are assigned in the United States
according to the Endangered Species Conserva-
tion Act of 1969 and listed periodically in the
U.S. Federal Register by the U.S. Fish and Wild-
life Service. It is not our intent to go into
great depth regarding endangered speices pro-
grams. The f orementioned I.U.C.N., U.S. Fish
and Wildlife Service and others (e.g. American
Committee on International Wildlife Protection,
National Audubon Society and World Wildlife
Fund) publish periodic information on endan-
gered wildlife (e.g. Arbib 1976). Other
governmental agencies besides the U.S. Fish
and Wildlife Service publish information re-
garding endangered wildlife (Arizona Game and
Fish Department 1977-, Behnke and Zarn 1976,
U.S. Forest Service 1975, U.S. National Park
Service 1974). Symposia are held periodically
focusing on general problems of endangered
wildlife (New Mexico Game and Fish Department
1972) or even devoted to a single species such
as the Peregrine Falcon (Falco peregrinus)
(Hickey 1969) or the Red-cockaded Woodpecker
(Dendrocopos borealis) (Thompson 1971) .
Periodically, reports are issued on endangered
species such as the Southern Bald Eagle (U.S.
Fish and Wildlife Service 1976). Recovery

teams to address the problem of impending ex-
tinction have been set up by the U.S. Fish and
Wildlife Service for many species of endan-
gered wildlife. For example, several avian
species are now being raised by methods of
direct intervention such as egg manipulation
(Zimmerman 1976) . In addition, several
agencies are now involved in establishing
endangered plant lists.

In addition to teams concerned with the
protection of terrestrial wildlife, such as
the Peregrine Falcon and Southern Bald
Eagle, other recovery teams have been organ-
ized to focus on one or more fish species.
Recently, however (Johnson-) the U.S. Fish
and Wildlife Service has designated teams
which focus on river systems instead of
individual species, e.g. the Colorado River
Fishes Recovery Team. This approach has
been advocated for years by many of us who
have seen the wholesale extermination of
species in certain areas as a result of
habitat destruction. Nowhere is this chain
of destruction more certain than in riverine
ecosystems. This has long been recognized
by ichthyologists such as Deacon and Minckley
(1974), Holden and Stalnaker (1975), Minckley
and Deacon (1968) and Sigler and Miller (1963) ,

Glen Canyon Dam: An Example

The construction of Glen Canyon Dam on
the Colorado River above Grand Canyon is an
outstanding example of habitat modification.
The effect on the aquatic ecosystem has been
devastating. The original heavy silt burden
which rendered the Colorado River "too thick
to drink and too thin to plow" is dropped in
Lake Powell before the water enters Grand
Canyon. The river waters are now clear. The
reddish color for which the Colorado was named
can be seen only after flooding from tributaries
which enter below the dam. This has created
an entirely new riverine ecosystem (Carothers
et al. in press, Dolan et al. 1974 and in
press, Johnson and Martin 1976, and Laursen
and Silverston 1976). The management impli-
cations are staggering. On one hand, a new
riparian ecosystem has developed, protected
from the scouring and siltation of pre-dam
floods. On the other hand this white water
river has been converted from a stream which
was warm in the summer and cold in the winter
to a relatively constant 9-10°C (48-50°F)
along most of its length. The only insect
family recorded using these cold waters are
Chironomid midges (Stevens 1976) while the

5 Arizona Game and Fish Department.
1977. Endangered and threatened species in
Arizona; 3 p. memo

_ Johnson, J. Paper presented at New
Mexico-Arizona section meeting, the Wildlife
Society, Farmington, N.M., Feb 5, 1977.

small crustacean, Gammarus lucustris, abounds.
The cold, clear water is conducive to the
rapid growth of exotic species such as rain-
bow trout, which commonly reach lengths of
more than 2 feet and weigh over 5 pounds
(personal observation) . While exotic fish
flourish, our native species are declining.
In the 277 miles of the Colorado River in
Grand Canyon National Park several species
listed in Fishes of Arizona (W. Minckley
1973 and pers. comm.) occur either in low
numbers, or cannot be found at all, e.g. the
Humpback Chub (Gila cypha) , Bonytail Chub (G.
elegans) , Colorado Squawfish (Ptychocheilus
lucius) and Razorback Sucker (Xyrauchen
texanus) (Johnson 1977, C. Minckley and Blinn
1976, Miller 1975^, and Suttkus et al. 1976).

Endangered Species and Related Acts

When the Endangered Species Act of 1973
(PL #93-205) was passed it was hoped by many
of us concerned about extirpation of wildlife
that this might prevent further wholesale
extinctions through degradation of habitat.
It seemed that the Endangered Species Act
combined with the National Environmental
Policy Act of 1969 (PL #91-190) should slow
down direct extermination as well as massive
destruction of the type that has converted
nearly all southwestern rivers to poor or
impossible habitat for most native species.
Just how effective these laws will be remains
to be seen. Legal decisions involving the
case of the Tennessee Valley's Tellico Dam on
the Little Tennessee River vs. the Snail
Darter (Percina tanasi) may have important
implications regarding the future interpreta-
tion of Section 7 of the Endangered Species
Act, including possible amendation by congress
(Holden 1977).

It seems inevitable that riverine eco-
systems will become the battleground for
those advocating the "progress of civilizing
processes," e.g. hydroelectric and irrigation
projects. Economic interests oppose those
who advocate saving a few rivers to protect
associated wildlife and recreational values
and perhaps, "just to let them run."

The two f orementioned acts coupled with
the Wild and Scenic Rivers Act of 1968 (PL
#90-542) would seem to be sufficient to reduce
further decimation of river ecosystems. How-
ever, it is a difficult, uphill battle. Pre-

b Miller, R.R. 1975. Report on fishes
of the Colorado River drainage between Lees
Ferry and Surprise Canyon, Arizona. Unpublished
Grand Canyon Natl. Park Res. Rpt. 6 p.

vention of the use of streams for waste disposal
is gradually becoming an accepted philosophy.
Conversely, industrial, domestic and irrigation
demands for water for a growing population con-
tinue to escalate.

Major Causes of Habitat Loss

The impact of dams on aquatic ecosystems
has long been understood by biologists even if
ignored by dam builders and water users. The
area above the dam is converted into a lake,
rapidly filling with sediment. The area below
the dam too commonly becomes a dry stream bed,
as is the situation with most of the Salt and
Gila River dams of the Lower Colorado River
drainage. Neither habitat is conducive to most
of the pre-dam riverine plants or wildlife.
Other rivers are greatly reduced in volume by
practices such as pumping of underground water
which dries up spring sources, or by modifica-
tion of runoff patterns through overgrazing.
The latter often results in the development of
vegetation types which demand more water than
the original vegetation. The area may be
denuded, resulting in flash floods followed by
quick drying up of streams rather than a slower,
steady runoff. The effects of such practices
on native fishes have been well documented
(Minckley and Deacon 1968). However, we have
only recently begun to understand the impacts
on riparian ecosystems.

Recent work by various investigators
(Boster and Davis 1972, Clary et al. 1974, and
Hibbert et al. 1974) advocates the conversion
of shrub types, commonly resulting from over-
grazing, to grassland. This conversion to
grassland usually results in increased water
yield which, in turn, often results in an in-
crease in acreage of riparian vegetation (per-
sonal observation, Sierra Ancha and Three Bar
watersheds) .

Some investigators propose large scale
"phreatophyte control" projects as well as the
conversion of shrub types to grassland (see
Ffolliott and Thorud 1974 for discussion) .
These "water salvage" projects are often advo-
cated even at the expense of both game and non-
game wildlife values. Earlier work commonly
featured "pure" scientists as well as "applied"
scientists, all concentrating on single purpose
management of watersheds and their runoff for
man, his farms and cattle (Barr 1956, Duisberg
1957, and Warnock and Gardner 1960). In recent
years there has been a gradual trend toward
multiple use of this critical resource, water
(Horton and Campbell 1974) . The Arizona Annual
Watershed Symposia reflect this change in philo-
sophy (Arizona Water Commission; annually)

placing increasing emphasis on wildlife values,
recreation and even aesthetics (Arizona Water
Commission 1972) .

Riparian Exploration,
Development and Research

It seems incredible that man would so
badly mistreat riverine ecosystems. We have
used them for exploratory routes, fur trapping,
temporary settlements and forts, agricultural
land and cities. Finally, we have dammed them
up, dried them up, and turned them into sewers
and garbage disposals.

Early explorers commonly were army officers,
geologists, engineers or "soldiers of fortune"
who left incomplete to poor records regarding
the riparian habitat. This is true throughout
the Southwest. Thus, early notes from rivers
such as the Gila (Emory 1858) and even the
mighty Colorado (Powell 1961) often mention
vegetation and wildlife only in general terms.
We do not even have good species' lists for the
pre-dam ecosystems, much less information on
population densities or other more sophisti-
cated data. Even as late as the 1950' s (Woodbury
et al. 1959) scientists gathered information
regarding the area to be inundated by Lake
Powell, above Glen Canyon Dam. However, the
more than 250 miles of river between Glen Can-
yon Dam site and the upper reaches of Lake
Mead, which were also to be heavily impacted
by the dam, were totally ignored.

Riverine environments, including their
riparian ecosystems, have been ignored by bio-
logists as well as geologists, explorers and
laymen for many reasons. Riparian ecosystems
have several characteristics which make them
interesting but involved, difficult systems
to study. Riparian habitat may be considered
an ecotone between the aquatic habitat of the
stream itself and the surrounding terrestrial
habitat. As such, the riparian ecosystem con-
tains elements of both the aquatic and terres-
trial ecosystems plus retaining unique charac-
teristics found in none of the other ecosystems
exemplifying the edge effect. The concept of the
edge effect is relatively new. Earlier trea-
tises did not even mention this phenomenon and it
was not until the mid-1900s that ecology texts,
e.g. Allee et al. (1949) contained a discussion
of the edge effect. Odum (1959) defines the
edge effect as "the tendency for increased
variety and density at community junctions."

Ornithologists and birders have long recog-
nized the importance of riparian habitats to
birds. We chose at random 20 inland Christmas
Bird Counts for 1974 (National Audubon Society
1975). Nineteen (95%) of the 20 contained

streamside and/or lake side vegetation. The
large number of species utilizing riparian wood-
land has been documented by numerous studies
(Carothers and Johnson 1975b). In California,
Miller (1951) emphasized the importance of
riparian avifaunas, stating "the number of
species of birds associated with riparian
woodland is larger than that of any other
formation." However, the extremely high
densities of riparian avian populations was
not recognized until this decade (Carothers et
al. 1974, Carothers and Johnson 1971 and
1975b, Gaines 1974, Johnson 1970, O'Brien et
al. 1976, and Table 2) .

The ecological analysis of riparian birds
is complicated at best. Studies are further
complicated by recent changes, some of which
are related to man's activities and others
which may be operating independently of man.
One cannot help postulating however, that
nearly all of the recorded recent changes are
due to man's activities. For example, there
are records for the arrival of several species
of birds which have moved into Arizona as
breeding species within historic times. This
includes the Mississippi Kite, Inca Dove,
Thick-billed Kingbird, Starling, House Sparrow,
Great-tailed Grackle and Bronzed Cowbird.
The Starling and House Sparrow are European
introductions. The Inca Dove, Great-tailed
Grackle and Bronzed Cowbird are closely
associated with man and his animals. Their
movements are discussed by Phillips et al.
(1964) and Phillips (1968). Other cases are
not as clear but may have profound effects on
the native avifauna. The subtleness with
which human activity may affect the natural
ecosystem can be shown through a discussion
of the Brown-headed Cowbird. Phillips (1968)
discusses at length the historic expansion of
range by Brown-headed Cowbirds. Of the 33
species of Southwestern lowland birds listed
by Friedmann (1929) as hosts to the Brown-
headed Cowbird, 22 (2/3) are obligate or
preferential riparian nesting species. The
role of these brood parasites in reducing
populations of riparian birds in the Sacramento
Valley, California, is discussed by Gaines
(1974). Thus, Brown-headed Cowbirds may be
suspected of causing problems in Arizona and
other southwestern areas similar to those
reported for California.

SUMMARY AND CONCLUSIONS

During our recent analysis of the de-
pendency of the breeding avifauna of the
Southwest lowlands on water related habitat
(Table 3), we discovered some sobering facts.
166 species of nesting birds were analyzed
from southern Arizona, southern New Mexico
and west Texas, south through the lower

Table 2. — A compar ision of breeding bird
Johnson 1975b) .

densities in selected habitats. (After Carothers and

Habitat Type
(Community)

Locality

Authority

Breeding Bird Density
Males or Estimated
Pairs/40 ha [or 100 acres]
nonriparian riparian

Boreal Forest^

Spruce-Alpine Fir

Temperate Forest
Spruce-Douglas Fir
Ponderosa Pine
Ponderosa Pine
Mature Deciduous
Virgin Spruce
Forest Bird Sanctuary

Relict Conifer Forest
Cypress post climax

Riparian Deciduous Forest
Mixed Broadleaf
Mixed Broadleaf
Cottonwood
Cottonwood

Flood-plain Deciduous

Temperate Woodland
Pinyon- Juniper
Pinyon- Juniper

Encinal

Subtropical Woodland

Mesquite Bosque (riparian)
Mesquite

Grassland

Temperate Grassland
Tropical Grassland

Desert Grassland
Yucca/ Grassland

Chihuahuan Desert Scrub
Creosotebush

Sonoran Desert Scrub
Paloverde/Sahuaro

Temperate Marshland
Cattail Marsh

Arizona

West Virginia

Germany

Arizona

Arizona
Arizona
Arizona
Arizona
Illinois

Arizona
Arizona
Arizona

Arizona
Arizona

Arizona
Tanganyika

Arizona

New Mexico

Arizona

Cultivated, Urban and Suburban Lands
Park (zoological garden) Germany
Bird Sanctuary (Whipsnade) England
Urban Arizona
Cottonwood Arizona

Carothers et al. (1973)

Balda (1967)
Balda (1967)
Haldeman et al. (1973)
Audubon F.N. (1948)
Audubon F.N. (1948)
Bruns (1955)

178

380
336
232
724 2
762 |
5600 -

Johnston and Carothers (1975) 93

Balda (1967)

Carothers et al. (1974)

Ohmart (no date) -

Fawver (1947)

Hering (1957) 33
Beidleman (1960) 30
Balda (1967) 224

Gavin and Sowls (1975)

Ohmart (no date) - 236

Balda (1967) 64

Winterbottom (1947) 4000

Balda (1967) 31
Raitt and Maze (1968) 8.5-17.7
Tomoff(1974 & pers.comm.) 105-150
Carothers and Johnson (1975b)

Steinbacher (1942)
Huxley (1936)
Emlen (1976)
Carothers and Johnson (1975a)

1170 \
5800 ^
1230 -

304
332
847
683
216 2

476 2

175-176

605.2

1_ Arizona vegetation types after Brown and Lowe (1974).

2^ Density given in number of adult birds per 40 hectares (100 acres) instead of males or
nesting pairs (after Welty 1962).

J3 Average density for April and May, ;the height of breeding activity in the mesquite bosque.

4_ Riparian cottonwood habitat disturbed by urbanization. Two years prior, when the habitat
was undisturbed, the density was 1058.8 pairs/100 acres.

_5 Ohmart, R.D. and N. Stamp. No date. Final report on the field studies of the nongame
birds and small mammals of the proposed Orme Dam site. Bur. of Reel. Proj . , Boulder City,
Ariz. 54 ms. p.

Table 3. — Nesting birds of the Southwest Lowlands (Modified from Haight and Johnson 1977) 1

1.
2.
3.
4.

1.
2.
3.

4.
5.
6.
7.

10.
11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

28.

29.

30.
31.

10.

11.

12.

13.

14.

WETLANDS (2%)

Clapper Rail Rallus longirostris
Black Rail Laterallus jamaicensis
American Avocet Recurvirostra americana
Snowy Plover Charadrius alexandrinus

WETLANDS AND OBLIGATE RIPARIAN (19%)

Least Grebe Podiceps dominicus
Pied-billed Grebe Podilymbus podiceps
Double-crested Cormorant Phalacrocorax
auritus

Olivaceous Cormorant Phalacrocorax olivaceus
Great Blue Heron Ardea herodias

Green Heron Butorides striatus
Great Egret Casmerodius albus
Snowy Egret Egreta thula
Black-crowned Night Heron Nycticorax
nycticorax

Least Bittern Ixobrychus exilis

Black-bellied Whistling-Duck

Dendrocygna autumnalis

Mallard Anas platyrhynchos

Mexican Duck Anas diazi

Gadwall Anas strepera

Blue-winged Teal Anas discors

Cinnamon Teal Anas cyanoptera

Redhead Aythya americana

Ruddy Duck Oxyura jamaicensis

Osprey Pandion haliaetus

Virginia Rail Rallus limicola

Sora Porzana Carolina

Common Gallinule Gallinula chloropus

American Coot Fulica americana

Black-necked Stilt Himantopus mexicanus

Killdeer Charadrius vocif erus

Long-billed Marsh Wren Cistothorus palustris

Common Yellowthroat Geothlypis trichas

Yellow-breasted Chat Icteria virens

Yellow-headed Blackbird

Xanthocephalus xanthocephalus

Red-winged Blackbird Agelaius phoeniceus

Song Sparrow Melospiza melodia

OBLIGATE RIPARIAN (26%)

Common Merganser Mergus merganser
Mississippi Kite Ictinia mississippiensis
Cooper's Hawk Accipiter cooperii
Zone-tailed Hawk Buteo albonotatus
Gray Hawk Buteo nitidus

Common Black Hawk Buteogallus anthracinus
Bald Eagle Haliaeetus leucocephalus
Spotted Sandpiper Actitis macularia
Red-billed Pigeon Columba f lavirostris
Yellow-billed Cuckoo Coccyzus americanus
Violet-crowned Hummingbird Amazilia verticalis
Buff-bellied Hummingbird Amazilia yucatanensis
Broad-billed Hummingbird Cynanthus latirostris
Green Kingfisher Chloroceryle americana

15. Red-shafted Flicker Colaptes auratus cafer

16. Rose-throated Becard Platypsaris aglaiae

17. Tropical Kingbird Tyrannus melancholicus

18. Thick-billed Kingbird Tyrannus crassirostris

19. Kiskadee Flycatcher Pitangus sulphuratus

20. Black Phoebe Sayornis nigricans

21. Willow Flycatcher Empidonax traillii

22. Western Wood Pewee Contopus sordidulus

23. Vermilion Flycatcher Pyrocephalus rubinus

24. Northern Beardless Flycatcher
Camp to stoma imberbe

25. Bank Swallow Riparia riparia

26. Cliff Swallow Petrochelidon pyrrhonota

27. Bridled Titmouse Parus wollweberi

28. White-breasted Nuthatch Sitta carolinensis

29. Bewick's Wren Thryomanes bewickii

30. American Robin Turdus migratorius

31. Bell's Vireo Vireo bellii

32. Yellow-green Vireo Vireo f lavoviridis

33. Tropical Parula Parula pitiayumi

34. Yellow Warbler Dendroica petechia

35. Hooded Oriole Icterus cucullatus

36. Northern Oriole Icterus galbula

37. Bronzed Cowbird Molothrus aeneus

38. Summer Tanager Piranga rubra

39. Blue Grosbeak Guiraca caerulea

40. Painted Bunting Passerina ciris

41. White-collared Seedeater Sporophila torqueola

42. Lesser Goldfinch Carduelis psaltria

43. Albert's Towhee Pipilo aberti

PREFERENTIAL RIPARIAN (26%)

1. Peregrine Fal con Falco peregrinus

2. American Kestrel Falco sparverius

3. Gambel's Quail Lophortyx gambelii

4. White-winged Dove Zenaida asiatica

5. Mourning Dove Zenaida macroura

6 . Common Ground Dove Columbina passerina

7. White-fronted Dove Leptotila verreauxi

8. Greater Roadrunner Geococcyx calif ornianus

9. Groove-billed Ani Crotophaga sulcirostris

10. Barn Owl Tyto alba

11. Common Screech Owl Otus asio

12. Ferruginous Pygmy Owl Glaucidium brasilianum

13. Lesser Nighthawk Chordeiles acutipennis

14. Black-chinned Hummingbird Archilochus alexandri

15. Anna's Hummingbird Calypte anna

16. Gila Woodpecker Melanerpes uropygialis

17. Golden-fronted Woodpecker Melanerpes aurif rons

18. Ladder-backed Woodpecker Picoides scalaris

19. Western Kingbird Tyrannus verticalis

20. Cassin's Kingbird Tyrannus vocif erans

21. Wied's Crested Flycatcher Myiarchus tyrannulus

22. Ash- throated Flycatcher Myiarchus cinerascens

23. Rough-winged Swallow Stelgidopteryx ruf icollis

24. Green Jay Cyanocorax yncas

25. Common Raven Corvus corax

26. Verdin Auriparus f laviceps

27. Northern Mockingbird Mimus polyglottos

28. Long-billed Thrasher Toxostoma longirostre

29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.

Curve-billed Thrasher Toxostoma curvlrostre
Crissal Thrasher Toxostoma dorsale
Black-tailed Gnatcatcher Polioptila melanura
Phainopepla Phainopepla nitens
Common Starling Sturnus vulgaris
Lucy's Warbler Vermivora luciae
Lichtenstein' s Oriole Icterus gularis
Brown-headed Cowbird Molothrus ater
Cardinal Cardinalis cardinalis
Pyrrhuloxia Cardinalis sinuata
Indigo Bunting Passerina cyanea
Lazuli Bunting Passerina amoena
House Finch Carpodacus mexicanus
Olive Sparrow Arremonops ruf ivirgatus
Rufous-winged Sparrow Aimophila carpalis

SUBURBAN AND AGRICULTURAL (4%)

Black Vulture Coragyps atratus

Rock Dove Columba livia

Inca Dove Scardaf ella inca

Barn Swallow Hirundo rustica

House Sparrow Passer domesticus

Great-tailed Grackle Quiscalus mexicanus

NON-RIPARIAN (23%)

Turkey Vulture Cathartes aura
Red-tailed Hawk Buteo jamaicensis
Swainson's Hawk Buteo swainsoni
Ferruginous Hawk Buteo regalis
Harris' Hawk Parabuteo unicinctus
Caracara Caracara cheriway
Prairie Falcon Falco mexicanus
Common Bobwhite Colinus virginianus

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

Scaled Quail Callipepla squamata
Great Horned Owl Bubo virginianus
Elf Owl Micrathene whitneyi
Burrowing Owl Athene cunicularia
Long-eared Owl Asio otus
Poor-will
Pauraque

Phalaenoptilus nuttallii
Nyctidromus albicollis
White-throated Swift Aeronautes saxatalis
Lucifer Hummingbird Calothorax lucif er
Costa's Hummingbird Calypte costae
Gilded Flicker Colaptes auratus chrysoides
Say's Phoebe Sayornis saya
Horned Lark Eremophila alpestris
Purple Martin Progne subis
White-necked Raven Corvus cryptoleucus
Cactus Wren Campy lorhynchus brunneicapillus
Canyon Wren Catherpes mexicanus
Rock Wren Salpinctes obsoletus
Bendire's Thrasher Toxostoma bendirei
LeConte's Thrasher Toxostoma lecontei
Loggerhead Shrike Lanius ludovicianus
Eastern Meadowlark Sturnella magna
Western Meadowlark Sturnella neglecta
Scott's Oriole Icterus parisorum
Varied Bunting
Brown Towhee

Passerina versicolor

Pipilo f uscus
Grasshopper Sparrow Ammodramus savannarum
Lark Sparrow Chondestes grammacus
Rufous-crowned Sparrow Aimophila ruf iceps
Cassin's Sparrow Aimophila cassinii
Black-throated Sparrow Amphispiza bilineata

166 Total

(Information from A.O.U. 1958, Bailey 1928, Bent-various dates, Hubbard 1970 and 1971, Johnson

et al. 1973 , Johnson et al. -manuscript , Monson and Phillips 1964, Monson-personal communications,

Oberholser 1974, Phillips et al . 1964, Rea 1977, Todd 1975 and undated, Wauer 1973, and Wolfe 1956)

^ Haight, L.T. and R.R. Johnson. Paper presented at annual meeting of the Arizona Academy

of Science, April 17, 1977.

Johnson, R.R., S.W. Carothers and D.B. Wertheimer, 1973. The importance of the Lower Gila
River, New Mexico, as a refuge for threatened wildlife. Unpubl. Rpt. to U.S. Fish and Wildl.
Serv., Albuquerque. 53 p.

Johnson, R.R., J.M. Simpson and J.R. Werner. Unpublished manuscript. Birds of the Salt
River Valley, Maricopa Co. , Arizona

Rio Grande Valley. Habitats up through desert
grasslands were considered, stopping at the
lower edge of woodland and forests. 127 (or
77%) of the 166 nesting species were in some
manner dependent on water related habitat.
Of this 77% dependent on water related habitat
well over half, 84 of the 166 species, are
completely dependent on water related habitat.
Only 39 species are non riparian nesting
birds. Thus, if water dependent habitats
were completely destroyed in the Southwest
(not including suburban and agricultural) we
could completely lose 47% of our lowland
nesting birds while only 23% of our lowland
nesting species would probably not be affected.
43 (26%) of the 166 species would be partially
affected. Granted, several of the species
which are preferential riparian at lower
elevations, such as the Western and Cassin's
Kingbirds, extensively use non riparian
habitat at higher elevations. Still, the
overall populations of these species would
diminish with the reduction or loss of riparian
habitat at lower elevations. In a dissertation
on "Historic Changes in the Avifauna of the
Gila Indian Reservation," near Phoenix, Rea
(1977) uncovered the following information.
Through the use of archaeological, ethnographic
and historic sources he found that 101 species
breed or have bred on the reservation with 5
more species that probably bred and 7 species
that could have bred, based on biogeographic
distributions. During the past 100 years, 22
breeding species were extirpated of which 18
were related to the former riverine ecosystem.
Six speices of non-nesting birds dependent on
the Gila River, now dry, are also gone. At
least 13 species have recently recolonized
the area as a result of reestablishment of a
depauparate form of the original riparian
habitat. This newly established habitat has
developed as a result of the use of the Salt
and Gila Rivers for disposal of effluent from
the Phoenix sewage treatment plants.

Others, e.g. Hubbard (1972) have pointed
out the lack of attention given to song birds
when designating threatened and endangered
species. However, to our knowledge, ours is
the first attempt to quantify the number of
species threatened or endangered by practices
which greatly modify or destroy riparian
habitat .

Some proponents of water salvage projects
have pointed out that many breeding species
of the Southwest lowlands are at the northern
limits of their range. This, of course, is
an attempt to justify phreatophyte control,
channelization, dam construction, grazing and
other practices which reduce riparian vegeta-
tion and consequently riparian wildlife. The

main populations are found in Mexico for a
large percentage of the birds that also occur
in the Southwest lowlands. Thus, it is
argued, even complete loss of riparian and
marshy habitat should cause no great problem
at the total population level for that species.
No argument could be further from the truth.
The destruction of riparian habitat in northern
Mexico is progressing at an alarming rate.
One need but drive a few hundred miles south
from the United States-Mexico border to observe
the frantic rate at which Mexicans are draining
their streams and clearing riparian forests
and woodlands in an attempt to feed a rapidly
expanding population. One reads with nostalgia
Sutton's book, "At a Bend in a Mexican River"
(1972). His accounts from travels in Mexico
only four decades ago tell of ferrying across
rivers such as the Rio Purificacion and of
the lush growth in the Valley of the Rio
Corona. The riparian groves along these
rivers are being cut at a rapid rate to make
room for houses and fields. Rivers throughout
Mexico as well as the United States are being
dammed to provide water for municipal and
industrial use and for large irrigation
proj ects .

Thus, the same basic stages of "develop-
ment of natural resources" which took place in
the United States during two centuries promise
to occur in Mexico in a matter of decades. When
adding the available improved technology to
Mexico's great wealth of natural resources,
synergism may result. This may effect an even
greater cummulative ecological disaster in a
much shorter period of time than we have expe-
rienced in riverine ecosystems in the United
States. Thus, when evaluating the ecological
health of riparian species we must approach
the problem from the standpoint of a systems
analyst. One may start with his or her area
of responsibility whether it be a few yards
of small stream or several hundred miles of a
large river. However, we must be cognizant
of the resources up and downstream from our
area. We must show concern for the entire
drainage system, even if primary responsibility
for its management rests elsewhere. The
managers of resource plots, cities, counties,
states, and countries need to recognize that
streams commonly flow thru lands in different
ownership and across political boundaries.

MANAGEMENT RECOMMENDATIONS

1. The riparian habitat is the most productive
and possibly the most sensitive of North
American habitats and should be managed
accordingly . Due to the complexity of riverine
ecosystems, scientists have only recently

developed techniques to document the impor-
tance of these ecosystems to wildlife.

2. In addition to the importance of riparian
habitat from an ecological standpoint, other
values include:

(a) Recreational uses including hunting,
fishing (Meehan et al. this symposium)
and bird watching.

(b) Reservoirs for preservation of gene
pools and to allow recolonization of
areas hit by disasters such as
forest fires, severe droughts and
storms .

Thus, recreational, wildlife, and aesthetic
values should be weighed against other values
and alternative uses. This is especially im-
portant in land use planning for a habitat
which has high pressures from alternative uses
such as water for industrial and domestic pur-
poses, irrigation, grazing and urbanization.

3. Use interdisciplinary teams, including
recreation specialists, economists , etc., to
develop improved means for determining wildlife
values. This is especially important in figu-
ring cost-benefit ratios for determining the
best use for an area. We hope there will
never be a need for putting a dollar figure on
everything in order to establish its "value."
(What is the value of 2 or 3 days vacationing
along a streamside?) However, economic values
have been placed, in part, on recreation such
as hunting, fishing, and "general rural recrea-
tion" (Davis 1967, and Martin et al. 1974).
Attempts to quantify these values should make
them more competitive with other uses, such as
those mentioned in No. 2 (above).

4 . Finally, encourage investigations to
clarify areas of knowledge which are currently
poorly, if at all, known. We have discussed
the complexity of riverine ecosystems and
further reasons for the late development of
this area of ecology.

Problems which need to be solved include:

(a) The minimum area and suitable config-
urations necessary to retain both
plant and wildlife values in dif-
ferent riparian habitats.

(b) The maximum distance which can sepa-
rate islands of a given habitat type
before the loss of wildlife species
or a great reduction in populations
occurs.

(c) Optimal as well as minimal require-
ments for enhancing wildlife values
for a given habitat type. These in-
clude ground cover, trees and shrubs
per hectare, foliage volume, plant
species present, and disturbance
types and frequencies.

We will close by quoting Carothers and
Johnson (1975a),

"Determining these factors may be
the most important problem facing
us today. All the 'threatened
species recovery teams' we can
possibly amass will not prevent
many species from becoming extinct
in their native habitat if we de-
grade their habitats past the point
of no return."

LITERATURE CITED

Allee, W.C., A.E. Emerson, 0. Park, T. Park
and K.P. Schmidt. 1949. Principles of
animal ecology. W.B. Saunders, Philadel-
phia, xii + 837 p.

American Ornithologist's Union. 1957. Check-
list of North American birds. Fifth ed.
A.O.U. Baltimore. xiii + 691 p.

Arbib, R. 1976. The bluelist for 1977. Amer.
Birds 30:1031-1039.

Arizona Game and Fish Department. 1977. Memo
concerning endangered and threatened species
in Arizona. Ariz. Game and Fish Dept.,
Phoenix. 3 p.

Arizona Water Commission. 1972. 16th Annual
Arizona Watershed Symposium Proc. Ariz.
Water Comm. Rpt. 43 p.

Bailey, F.M. 1928. Birds of New Mexico. N.
Mex. Dept. Game and Fish. Judd and
Detweiler, Inc., Wash., D.C. xxiv + 807 p.

Barr, G.W. 1956. Recovering rainfall, Part I,
Arizona watershed program. Coop. Proj .
Ariz. State Land Dept., Water Div. , Salt
River Valley Water User's Assn., Univ. of
Ariz. 33 p.

Behnke, R.J. and M. Zarn. 1976. Biology and
management of threatened and endangered
western trouts. USDA For. Serv. Tech. Rpt.
RM-28, Rocky Mtn. For. and Range Exp. Sta.,
Ft. Collins, Colo. 45 p.

Bent, A.C. Various dates. Life histories of
North American birds. Dover Publ., New
York. (Republication of U.S. Natl. Mus.
Bulls, issued from 1919-1968).

Boster, R.S. and L.R. Davis. 1972. Soil-loss
considerations in chaparral-to-grass con-
version, p. 243-250. In Proc. Watersheds
in Transition Symp., Am. Water Resour.
Assoc., Urbana, 111.

Carothers, S.W.,S.W. Aitchison and R.R. Johnson.
In press. Natural resources in Grand
Canyon National Park and river management
alternatives on the Colorado River. In
Proc. First Conference on Scientific Re-
search in the National Parks. USDI Natl.
Park Serv., Wash. D.C.

Carothers, S.W. and R.R. Johnson. 1971. A
summary of the Verde Valley breeding bird
survey. Ariz. Game and Fish Dept. Land and
Water Projects Investigations. Verde
River Studies. Project FW-16-10. 20 p.

Carothers, S.W., R.R. Johnson and S.W.

Aitchison. 1974. Population structure
and social organization of southwestern
riparian birds. Amer. Zool. 14:97-108.

Carothers, S.W. and R.R. Johnson. 1975a. Water
management practices and their effects on
nongame birds in range habitats, p. 210-
222. In Proc. of Symp. on Management of
Forest and Range Habitats for Nongame
Birds. USDA For. Serv. Gen. Tech. Rpt.
No. 1. Wash., D.C.

Carothers, S.W. and R.R. Johnson. 1975b. The
effects of stream channel modification on
birds in the southwestern United States,
p. 60-76. In Proc. of Symp. on Stream
Channel Modification. USDI Fish and Wildl.
Serv., Off. Biol. Serv., Wash., D.C.

Clary, W.P. et al. 1974. Effects of pinyon-
juniper removal on natural resource pro-
ducts and uses in Arizona. USDA For. Serv.
Res. Paper RM-128. Rocky Mtn. For. and
Range Exp. Sta. , Flagstaff, Ariz. 28 p.

Davis, W.C. 1967. Values of hunting and

fishing in Arizona in 1965. Coll. of Busi-
ness and Public Adm. , Ariz. State Univ.,
Tempe. vii + 91 p.

Deacon, J.E. and W.L. Minckley. 1974. Desert
fishes, p. 385-488. In Desert Biology.
Academic Press, New York.

Dolan, R. , B. Hayden, A. Howard and R.R.

Johnson. In press. Man's impact on Colo-
rado River fluvial deposits within Grand
Canyon. In Proc. First Conference on
Scientific Research in the National Parks.
USDI Natl. Park Serv., Wash., D.C.

Dolan, R. , A. Howard and A. Gallenson. 1974.
Man's impact on the Colorado River in the
Grand Canyon. Amer. Sci. 62:392-401.

Duisberg, P.C., Ed. 1957. Problems of the
upper Rio Grande: an arid zone river.
Report of AAAS in coop, with UNESCO, U.S.
Comm. for Arid Resour. Improvement and
Development. Wash., D.C. 69 p.

Emory, W.T. 1858. Notes of military recon-
naissance from Fort Leavenworth in Missouri
to San Diego in California. Wendell and
Van Benthvysen, Wash., D.C. 614 p.

Ffolliott, P.F. and D.B. Thorud. 1974. Vege-
tation management for increased water yield
in Arizona. Agr. Exp. Sta., Univ. of Ariz.,
Tucson. iii + 38 p.

Fisher, J., N. Simon and J. Vincent. 1969.
Wildlife in danger. Viking press, New
York. 368 p.

Friedman, H. 1929. The cowbirds. Thomas
Publ., Baltimore, Md. xv + 421 p.

Gaines, D. 1974. A new look at the nesting
riparian avifauna of the Sacramento Valley,
California. Western Birds. 5:61-80.

Greenway, J.C. 1958. Extinct and vanishing
birds of the world. Amer. Comm. for Int.
Wildl. Protection, Spec. Pub. No. 13.
New York. 518 p.

Hibbert, A.R., E.A. Davis and D.G. Scholl.
1974. Chaparral conversion potential in
Arizona Part I: Water yield response and
effects on other resources. USDA For. Serv.
Res. Paper RM-126. Rocky Mtn. For. and
Range Exp. Sta., Ft. Collins, Colo. 36 p.

Hickey, J.J., Ed. 1969. Peregrine Falcon pop-
ulations, their biology and decline. Univ.
of Wis. Press, Madison, xxii + 596 p.

Holden, C. 1977. Review of law triggered by
Tellico impasse. Sci. 196:1426-1428.

Holden, P.B. and C.B. Stalnaker . 1975. Distri-
but ion and abundance of mainstream fishes
of the middle and upper Colorado River
basins, 1967-1973. Trans, of the Amer.
Fisheries Soc . 104 (2) : 217-231 .

Horton, J.S. and C.J. Campbell. 1974.

Management of phreatophyte and riparian
vegetation for maximum multiple use values.
USDA For. Serv. Res. Paper RM-117. Rocky
Mtn. For. and Range Exp. Sta., Ft. Collins,
Colo. 23 p.

Hubbard, J. P. 1970. Checklist of the birds
of New Mexico. New Mexico Ornith. Soc.
Publ. No. 3. McLeod Printing Co.,
Albuquerque. 108 p.

Hubbard, J. P. 1971. The summer birds of the
Gila Valley, New Mexico. Nemouria, Occas.
Papers of the Delaware Mus. Nat. His. 2:1-35.

Hubbard, J. P. 1972. Southwestern songbirds,
p. 79-96. In Proc. of Symp. on Rare and
Endangered Wildlife of the Southwestern
United States. New Mexico Game and Fish
Dept. , Santa Fe. 167 p.

Hubbard, J. P. 1977. A biological inventory
of the Lower Gila River Valley, New Mexico.
New Mexico Dept. Game and Fish, 56 p.

Johnson, R.R. 1970. Tree removal along

southwestern rivers and effects on associated
organisms. Amer. Phil. Soc, Yearb. 1970.
p. 321-322.

Johnson, R.R. 1977. Synthesis and management
implications of the Colorado River research
program. Colo. R. Res. Tech. Rpt. No. 17.
USDI Natl. Park Serv., Grand Canyon Natl.
Park. 50 ms. p.

Johnson, R.R. and S.P. Martin. 1976. The

Colorado River research project: a National
Park Service multidisciplinary , interdisci-
plinary research project, p. 29-41. In
Proc. of the 3rd Resources Management Conf.
USDI Natl. Park Serv., Wash., D.C.

Laursen, E.M. and E. Silverston. 1976. Hydro-
logy and sedimentology of the Colorado
River in Grand Canyon. Colo. R. Res.
Tech. Rpt. No. 13, USDI Natl. Park Serv.,
Grand Canyon Natl. Park. 27 p.

Lowe, C.H., Ed. 1964. The vertebrates of
Arizona. Univ. of Ariz. Press, Tucson,
270 p.

Martin, W.E., R.L. Gum and A.H. Smith. 1974.
The demand for and value of hunting, fishing,
and general rural outdoor recreation in
Arizona. Agr. Exp. Sta., Ariz. State Univ.,
Tempe. ix + 56 p.

Miller, A.H. 1951. An analysis of the distri-
bution of the birds of California. Univ. of
Cal. Publ. Zool. 50:531-643.

Minckley, W.L. 1973. Fishes of Arizona. Ariz.
Game and Fish Dept., Phoenix, xvi + 293 p.

Minckley, CO. and D.W. Blinn. 1976. Summer
distribution and reproductive status of fish
of the Colorado River in Grand Canyon
National Park and vicinity during 1975-76.
Colo. R. Res. Tech. Rpt. No. 14, USDI Natl.
Park Serv., Grand Canyon Natl. Park. 17 p.

Minckley, W.L. and J.E. Deacon. 1968. South-
western fishes and the enigma of "endangered
species." Sci. 159:1424-1432.

Monson, G. and A.R. Phillips. 1964. Species
of birds in Arizona. In The Vertebrates
of Arizona. Univ. of Ariz. Press, Tucson.
270 p.

National Audubon Society. 1975. The seventy-
fifth Christmas bird count. Amer. Birds
29(2): 151-638.

New Mexico Game and Fish Dept. 1972. Proc.
of the Symp. on rare and endangered wild-
life of the southwestern United States.
New Mexico Game and Fish Dept., Santa Fe.
167 p.

Oberholser, H.C. 1974. Bird life of Texas,
E.B. Kincaid, Jr., ed., Vol. I and II.
Univ. of Texas Press, Austin. 1069 p.

O'Brien, J. 1976. Observations on furbearers
within the riparian habitat of the upper
Sacramento River. Calif. Dept. Fish and
Game Memo Rpt. iii + 12 p.

Odum, E. 1959. Fundamentals of ecology. W.B.
Saunders Co., Philadelphia. xvii + 546 p.

Pettingill, G.S. 1970. Ornithology in

laboratory and field. Burgess Publ. Co.,
Minneapolis. xvii + 524 p.

Phillips, A.R. 1968. The instability of the
distribution of land birds in the Southwest,
p. 129-162. In Collected Papers in Honor
of Lyndon Lane Hargrave, Papers of the
Arch. Soc. of N. Mex. : 1. Mus. of New
Mexico. Press, Santa Fe. v + 170 p.

Phillips, A., J. Marshall and G. Monson.
1964. The birds of Arizona. Univ. of
Ariz. Press, Tucson, xviii + 212 p.

Powell, J.W. 1961. The exploration of the
Colorado River and its canyons. Dover
Publ., Inc., New York, xiv + 397 p.
(Unabridged republication of work publ.
in 1895 by Flood and Vincent, Canyons of
the Colorado) .

Prince Phillip and J. Fisher. 1970. Wildlife
crisis. Hamish Hamilton Ltd., Cowles Book
Co., Inc., and World Wildlife Fund., Chicago
and New York. 256 p.

Rea, A.M. 1977. Historic changes in the avi-
fauna of the Gila River Indian Reservation,
Central Arizona. Doctoral Dissertation,
Univ. of Ariz., Tucson. xiii + 346 p.

Schorger, A.W. 1955. The passenger pigeon:
its natural history and extinction. Univ.
of Wisconsin Press, Madison. xiii + 424 p.

Sigler, W.F. and R.R. Miller. 1963. Fishes of
Utah. Utah Dept. of Fish and Game, Salt
Lake City. 203 p.

Simon, N. and P. Geroudet. 1970. Last survi-
vors: the natural history of animals in
danager of extinction. World Publ. Co.,
New York. 275 p.

Stevens, L.E. 1976. An insect inventory of
Grand Canyon, p. 123-128. J_n An ecologi-
cal survey of the riparian zone of the
Colorado River between Lees Ferry and Grand
Wash Cliffs. Colo. R. Res. Tech. Rpt. No.
10, USDI Natl. Park Serv. , Grand Canyon
Natl. Park. 251 p.

Suttkus, R.D., G.H. Clemmer, C. Jones and
C.R. Shoop. 1976. Survey of fishes,
mammals and herpetofauna of the Colorado
River in Grand Canyon. Colo. R. Res. Tech.
Rpt. No. 5. USDI Natl. Park Serv., Grand
Canyon Natl. Park. 48 p.

Sutton, G.M. 1972. At a bend in a Mexican
river. Paul Eriksson, Inc., New York,
xvii + 124 p.

Thompson, R.L., Ed. 1971. The ecology and
management of the Red-cockaded Woodpecker.
USDI Bureau of Sport Fisheries and Wildlife,
Tall Timbers Res. Sta. , Tallahassee, Fla.
ii + 188 p.

Todd, R.L. No date. Field checklist of Arizona
birds. Ariz. Game and Fish Dept. Proj .
W-53-R-26. 4 p.

Todd, R.L. 1975. The species of birds and
mammals occurring on the watersheds com-
prising the Colorado River drainage of
Arizona. Ariz. Game and Fish Dept. Spec.
Rpt. W-53-R-26. 53 p.

Tylinek, I. and W. Ullrich. 1972. Endangered
species. Hart Publ. Co., Inc., New York.
284 p.

U.S. Fish and Wildlife Service. 1968. Rare
and endangered fish and wildlife of the
United States. USDI Fish and Wildl. Serv.,
Bur. of Sport Fisheries and Wildl., Wash.,
D.C. xi + 171 p.

U.S. Fish and Wildlife Service. 1973.

Threatened wildlife of the United States.
USDI Fish and Wildl. Serv., Bur. of Sport
Fisheries and Wildl., Office of End.
Species and Int. Activities, Wash., D.C.
xiii + 289 p.

U.S. Fish and Wildlife Service. 1976. The
Southern Bald Eagle in Arizona: a status
report. End. Species Rpt. No. I. USDI
Fish and Wildl. Serv., Div. of Wildl.
Serv., Phoenix, Az . 33 p.

U.S. Forest Service. 1975. Endangered and

unique fish and wildlife of the southwestern
national forests. USDA For. Serv. Wildl.
Habitat Manag. Staff Group, Albuquerque,
N. Mex. viii + 203 p.

U.S. National Park Service. 1974. Threatened
wildlife in the Western Region of the
National Park Service. USDI Natl. Park
Serv. West. Reg. Office, San Francisco,
vi + 68 p.

Van Tyne, J. and A.J. Berger. 1971.

Fundamentals of ornithology. Dover Publ.,
Inc., New York, xi + 624 p.

Warnock, B.H. and J.L. Gardner, Ed. 1960.
Water yield in relation to environment in
the southwestern United States. Iri Proc .
of Symp. SWARM Section AAAS , Sul Ross St.
Coll, Alpine, Texas. 74 p.

Wauer, R.W. 1973. Birds of Big Bend National
Park and vicinity. Big Bend Nat. Hist.
Assn. xv + 223 p.

Welty, J.C. 1975. The life of birds. Second
ed. W.B. Saunders, Philadelphia,
xv 4- 623 p.

Wolfe, L..R. 1956. Checklist of the birds of
Texas. Intelligencer Print. Co., Lancaster,
Pa. 89 p.

Woodbury, A.M. et al . 1959. Ecological studies
of the flora and fauna in Glen Canyon. Univ
of Utah Anthro. Papers (Glen Canyon Series
No. 7), Univ. of Utah Press, Salt Lake City.
226 p.

Zimmerman, D.R. 1976. Endangered bird species:
habitat manipulation methods. Sci.
192:876-878,

Ziswiler, V. 1967. Extinct and vanishing
animals. Springer- Ver lag Press, New
York. ix f 133 p.

Riparian Research Needs1 j)

David R. Patton

Abstract. — Approximately 22 studies on riparian habitat
are in progress in the western United States. Six categories
of studies are needed to provide managers with data for making
decisions about the riparian ecosystem. The concept of "vali-
dation sites" can be used in a team approach to solve plant
and animal problems in the riparian zone.

INTRODUCTION

Riparian zones, characterized by mesic
vegetation and more or less permanent surface
water, contrast with adjacent semiarid or dry
subhumid environments (fig. 1). Early settlers
gravitated to these limited river and stream
areas. As a result, ranches, farms, and towns
all have taken their toll of riparian zones.
Human population growth has increased this pres-
sure in the years following settlement. Intro-
duction of livestock, use of streamflow for
irrigation, and construction of dams and roads
have compartmentalized the riparian zone with
varied but mostly harmful effects to native
fauna. The importance of riparian vegetation
to wildlife habitat has become apparent only in
this decade. Even now its overall importance
is not widely recognized.

A search of some 24,000 research resumes
in the U.S. Department of Agriculture's Current
Research Information System (CRIS) revealed 10
related to riparian habitat. CRIS information
and other sources indicate 22 active studies
(10 plant, 5 animal, and 7 combined) in the
western United States. These figures show that
riparian vegetation is not receiving research
at a level comparable to its importance as wild-
life habitat. A concentrated effort directed at
specific problems with realistic goals is needed.

A problem analysis for Forest Service wild-
life habitat research in the Southwest that

identified riparian research needs (Patton 1976)
can serve as a guide for developing priorities
for other areas. Six general categories of
studies are necessary to provide natural re-
source managers with sound data for making
decisions (fig. 2).

Paper presented at the Symposium on Impor-
tance, Preservation, and Management of Riparian
Habitats, Tucson, Ariz., July 9, 1977.

Principal Wildlife Biologist, USDA Forest
Service, Rocky Mountain Forest and Range Experi-
ment Station, at the Station's Research Work
Unit at Arizona State University, Tempe. Central
headquarters is maintained at Fort Collins in
cooperation with Colorado State University.

Figure 1. — Riparian habitat along East Verde
River, Tonto National Forest, Arizona.

Patton, David R. 1976. Habitat criteria
development for southwestern wildlife. A prob-
len analysis. USDA For. Serv. , Rocky Mt. For.
and Range Exp. Stn. , Tempe, Arizona. Unpubl.
rep .

I nventorie s
&

Maps

Influence ol
Man & Nature

Figure 2. — Riparian Research Needs

These categories only provide an outline
for research. Each researcher will have to
determine his own priorities depending on the
need for data at the local, state, or regional
level. Most of the problems, however, can be
placed into one of these groups. The research
categories are not mutually exclusive, and can
be dealt with using a team or multidisciplinary
approach .

INVENTORIES AND MAPS

A map and inventory of the riparian resource
are very much needed. Aldo Leopold always empha-
sized that these tools were basic to all manage-
ment decisions. Preparation of maps and inven-
tories may not be a researcher's job, but he can
provide some of the necessary information. In
the Southwest, the vegetation map by Brown et al.
(1977a) is a major step in the inventory and
mapping process. In addition, Brown et al.
(1977b) prepared a stream map for Arizona that
identifies areas where riparian vegetation
should or could exist.

VEGETATION CLASSIFICATION

No comprehensive classification of riparian
vegetation suitable either for research or
management has been prepared for the Southwest.
The hierarchical system proposed by Brown and
Lowe (1974) as part of the Arizona Resource
Inventory System (ARIS) is an important step
in characterizing vegetation on a regional
basis to the community and association level.
Studies are needed to determine plant species
composition and abundance for every identifiable
successional stage for riparian vegetation,
from low-elevation desert to high-elevation

spruce-fir forests. An important part of the
work should be to develop techniques for eval-
uating the riparian habitat's "state of health."
Emphasis should be placed on studies of indicator
plants .

PLANT-ANIMAL ASSOCIATIONS

Once successional stages have been deter-
mined, there is a need to identify animals that
depend on a given stage or stages for their life
requirements. Such studies do not have to be
complex and can provide excellent field training
for graduate students interested in the habitat
approach to wildlife management.

Riparian habitats are oases in arid envi-
ronments and support a great density and diver-
sity of bird species. More is known about bird
life in riparian vegetation than any other
vertebrate group, but a check of the literature
quickly shows a lack of detailed information on
any animal. Some birds, such as the black
hawk (Buteogallus anthracinus) and zone-tailed
hawk (Buteo albonotatus) , that inhabit or are
semiripar ian-dependent , are on state threatened
lists and probably will become endangered unless
research provides managers with information on
their habitat requirements.

The needs of mammals that inhabit or are
closely associated with riparian vegetation
have not been well identified. Bats, squirrels
and skunks often are seen in or near riparian
vegetation, but their use of the type for food
and cover is known only in general terms. Small
rodents are probably the least understood and
documented group of animals in the riparian
habitat. These animals could be an important
link in the food chain of threatened hawks.

Researchers have neglected amphibians and
reptiles in favor of more economically important
animals. Herps are not as esthetically pleasing
to recreationists because of a lack of under-
standing of their ecological value. Yet, data
on the two herp groups are necessary-to under-
stand the complete riparian ecosystem.

Although fish are not restricted to streams
with riparian vegetation, the presence or absence
of streamside cover affects temperature which
determines species composition. There probably
is more detailed information available on game
fish than on any single group influenced by
riparian vegetation. Less data are available
on threatened or endangered species. In addi-
tion many remote streams have not been surveyed
or were insufficiently surveyed.

Invertebrates must be included in plant-
animal relationship research. These animals are

an important source of energy in the food chain
of riparian vertebrates. For this reason,
there needs to be an understanding of the
invertebrate populations associated with the
hundreds of terrestrial and aquatic mi crohabita ts
in riparian vegetation.

INFLUENCE OF MAN AND NATURE

Grazing, pollution, recreation, flooding,
and water reclamation projects all can influence
plant and animal species composition and abun-
dance. Managers need to know if desirable trees
and shrubs can survive under any grazing system,
or whether protection will be required during
critical seasons to restore the balance between
production and utilization.

Most natural systems can withstand some
pollution, but overloads soon become toxic to
both plants and animals. At that point it may
be too late for recovery. Studies are needed
to determine recovery rates associated with
different amounts of pollutants and their chem-
ical toxicity of the plant-animal complex.
Recreational activities can cause pollution
from human wastes and physical damage. Managers
need information on carrying capacity to regu-
late access when overuse becomes a problem.

The riparian habitat must have water. In
most of the West, demands for water lead to
conflicts between human uses and needs and other
uses and needs. Floods, caused by either man
or nature, can be either beneficial or detri-
mental. Some riparian plants need occasional
flooding to perpetuate the species. However,
removal of large amounts of vegetation in water-
shed treatments may create conditions conducive
to flooding that can destroy desirable plants.
Minimum flows to maintain the riparian habitat
as a viable biological system must be determined
soon, or managers will not have the information
they need to mitigate habitat losses from dam
and reservoir projects.

SILVICS OF TREE SPECIES

Life history data on the important riparian
deciduous tree species — sycamore, cottonwood,
willow, ash, and walnut — are almost nonexistent.
The lack of data on these species resulted from
a research emphasis on species with direct
economic value. Studies are needed to document
site requirements for germination and sprouting
of seedlings or suckers, effects of insects and
fire, and techniques for artificial regeneration.

LIFE HISTORY OF VERTEBRATE SPECIES

Life history information provides biological
data necessary for understanding each species'
role in the ecosystem and the effects of man's
activities. Because of the large number of
vertebrates that live in or are influenced by
the riparian habitat, the task of documenting
all their relationships with the plant complex
will be a difficult one; Species on the state
and federal threatened and endangered lists
should receive top priority. In many cases
it may be possible to group species for study
by common requirements or life forms.

USEFUL CONCEPTS

Two concepts may help to plan and initiate
research in the riparian habitat. The first is
establishing validation sites. A validation
site is an area that represents a given vege-
tation condition or successional stage. It is
permanently documented by maps and aerial photo-
graphs and used for long-term studies. In some
areas, validation sites may need permanent
exclosures for protection from grazing. By
using validation sites, scientists of many
different disciplines can work together on
separate studies to solve common problems.

Roy Johnson discussed the second concept,
"endangered habitat", earlier. Several years
ago I found the same terminology useful for
describing riparian vegetation in a wildlife
problem analysis. These select words directed
attention to a habitat containing endangered
and other species that depend on a vegetation
type that itself is in a precarious position.
The concept of threatened and endangered
habitat, when properly used, may increase the
chances of getting funds for research for a
variety of species, living in that habitat,
that otherwise would not receive high priority.

LITERATURE CITED

Brown, David E. , Charles H. Lowe, and Charles
P. Pase, 1977a„ Biotic communities of
the Southwest (map). USDA For. Serv. Gen.
Tech. Rep. RM-41. Rocky Mt. For. and Range
Exp. Stn., Fort Collins, Colo.

Brown, David E., N.B. Carmony, and R.M. Turner.
1977b. Drainage map of Arizona showing
perennial streams and some important wet-
lands. Ariz. Game and Fish. Dep., Federal
Aid Project W-53-R.

Brown, David E., and Charles H. Lowe. 1974.

The Arizona system for natural and poten-
tial vegetation — illustrated summary through
the fifth digit for the North American
Southwest. J. Ariz. Acad. Sci. Vol. 9,
Supply 3o 56 p.

Riparian Habitat Symposium
Closing Remarks1

M. J. Hassell'

I don't know about you folks, but I
personally have found this symposium to be
very interesting and enlightening.

My only regret is that Bill Hurst, the
man I replaced as Regional Forester, was not
here to participate. When he first came to
the Southwestern Region, I was a staff assistant
in the Range and Wildlife Division. The first
time I remer.ber hearing the word "riparian,"
Bill was remarking worriedly that there was
something wrong with our sycamore and cotton-
wood stands. No trees, shrubs and forbs in
the younger age classes were represented, and
this was a worry to this keen-eyed forester.

Now I know the reason. Age class "poverty"
has been the subject of many papers. But,
when Bill Hurst was expressing concern; few,
if any, had even recognized the problem. He
deserves much credit for the fact we are here
today.

Bill Morris, you made a good point which
I agree with. Therefore, another reason this
symposium has been so meaningful to me is
personal and perhaps even selfish. Our Region
of the Forest Service has been attempting to
put out a policy statement on the riparian
type. As competition for resources intensifies,
we desperately need basic information that
identifies the trade-offs of resource conflicts.
I don't believe it is acceptable to jump up
and down totally in the dark - we need more
information to make better choices. The
speakers today have supplied much of the needed
information, and I, for one, am extremely
grateful. It is my hope that you have been
able to glean some helpful information from
the discussions.

Steve Carothers set the stage for our
concerns and our reasons for being at the
symposium. Of course, the basic reason is
that changes are taking place in riparian

habitat, and these changes and their direction
are significant.

The riparian type is a key type to many
kinds of uses and species. Both Kel Fox and
John Hubbard covered this well in their
discussions on domestic livestock, wildlife
and fish; humans, water source and recreation.

Earle Layser and Charlie Pase highlighted
the fact that this valuable type is very limited
and the need we have to classify it. The
framework provided through classification would
permit scientists and managers better communi-
cation about what the problems and possibilities
are.

Dave Brown highpointed the need for
inventory in the riparian type, and we were
able to see some of the difficulty of getting
this basic data.

The challenge is to understand the
importance of the type for all the various
uses, to exchange what is known, to research
what is not known, and to finally reach
workable adjustments and compromises.

We realize it has been difficult to
digest all the information you have heard
today. But, in case you missed a point or
two, Director Herrick of the Rocky Mountain
Station has volunteered to print these
transactions. They should be available for
all registrants at a later date.

On behalf of all the cosponsoring groups,
we would like to thank you for making this
symposium a success. If riparian habitat and
its associated fauna receive the attention
it has long deserved, we will all be the richer
for this experience.

■"■Paper presented at the Symposium on
Importance, Preservation and Management of the
Riparian Habitat, Tucson, Arizona, July 9, 1977.

2Regional Forester, U.S. Forest Service,
Region 3, Albuquerque, New Mexico.

Contributed Papers

Classification of Riparian Vegetation1

I 2 ,3

William A. Dick-Peddie and John P. Hubbard

Abstract — Historically, little attention has been given
to vegetation associated with water courses. The reasons for
this neglect are reviewed. Today there is considerable interest
in riparian vegetation and a classification system would be of
value. A classification system is proposed for riparian vegeta-
tion of New Mexico.

INTRODUCTION

Vegetation growing along rivers, streams,
arroyos, and drainages in general has seldom
been separately classified as a unit.

Reasons For No
Classification System

Apparently there have been a number of
conditions contributing to this omission. One
condition is that historically, vegetation
occurring on open flats, rolling hills, and
mountain slopes has been of far greater economic
importance than the vegetation associated with
water courses. As a consequence, these areas of
grassland and forest have tended to monopolize
the attention of researchers.

Another condition responsible for the lack
of separate classifications is that the area
occupied by drainage associated vegetation often
constitutes a small fraction of the total area
being considered. The scale frequently used for
mapping (continent, region, state, etc.) makes
the recognition of this vegetation impractical
or impossible. This is particularly true in
mountainous regions where due to steep slopes,
the drainages are relatively narrow and the
vegetation associated with these systems consti-
tutes only a thin band bordering either side
of the reach. In addition the reach may be at
the bottom of a gorge, canyon, or arroyo gully.
In the plains and rolling country of the midwest
and east, vegetation which is associated with
wide old meandering drainage systems (flood-

plains) may make up a considerable portion of
the total vegetation. When it does , these
areas have sometimes been classified and mapped.
Kuchler's (1964) southern Floodplain Forest
(113) and Northern Floodplains Forest (98) are
examples of this situation.

A third condition which has undoubtedly
provided resistance to classification is the
apparent lack of sufficiently discreet boundaries
to the species aggregations as one proceeds up
or down a drainage. Complicating this strong
continuum condition is the fact that most species
which grow out in the open are also able to grow
along drainages. These species may be associated
with typical drainage species in aggregation
which are highly varied in both diversities and
densities making creation of vegetative units
exceedingly difficult.

Need for a Classification System

During the last fifteen or twenty years
some conditions which prevented or restricted
the development of classification systems for
drainage associated vegetation have changed or
disappeared and new priorities have even created
a demand for a classification system.

There has been a considerable increase in
the economic importance of wildlife which has
resulted in an increase of research activity
associated with the vegetation which supports
wildlife. Much of this wildlife habitat is along
drainage systems. A vegetation classification
system would be of value in the management of
these habitats.

Contributed paper, Symposium on the
Importance, Preservation and Management of the
Riparian Habitat, July 9, 1977, Tucson, Arizona.

2 ;
Prof. Dep't of Biology

New Mexico State University

Las Cruces , New Mexico

Supervisor, Endangered Species Program
New Mexico Dep't Game & Fish
Santa Fe, New Mexico

The public has become interested in safe-
guarding examples of its natural heritage and the
resultant competition between preservation and
economic development of sites , necessitates biotic
inventories. Biotic inventories are facilitated
by a workable classification system.

Increased demand for water has focused an
inordinate amount of attention on drainage systems
in western United States and particularly in
the southwest. This attention has led to various
studies of water salvage. A natural consequence

of this intensive activity is an interest in the
role of drainage associated vegetation and water
use. For an extensive literature review of this
activity see Horton (1973). A realistic vegeta-
tion classification would be valuable for this
continuing research effort.

The following is quoted from Lowe and
Brown (1973) because it incorporates some of what
we have said and indicates why we should be
concerned about this type of vegetation.

"The riparian communities are not shown on
the color map. In total they comprise a
limited geographic area that is entirely
disproportionate to their landscape impor-
tance and recreational value and their
immense biological interest."

"It should not go unnoticed that in Arizona
these riparian woodlands and streams
forests have been rapidly dwindling just as
the water table has been rapidly lowering,
and our broadleaf trees are now the native
phreatophytes of the water users. With
present plans for increasing dams, flood
control, water salvage, water forests are
now truly endangered. Unless those con-
cerned with projects such as the above do
not quickly reasses their commitments, it
will not be long before our riparian
forests are destroyed forever in Arizona.
They cannot be replaced."

Because of the changed and new conditions
just explained, it is our opinion that a class-
ification system for drainage associated
vegetation is timely and highly desirable. We
are also of the opinion that such a system is
feasible .

TERMINOLOGY

Some terms used when discussing and des-
cribing this type of vegetation need clarifica-
tion because their use has not always been con-
sistent. In our view the following definitions
are the most consistent and useful for these
terms .

Riparian

Riparian - associated with water courses.
Riparian may refer to vegetation associated
with large rivers or with small, even inter-
mittent drainages such as arroyos.

Phre atophytes

Phreatophytes - plants whose roots are
growing in the water table or its capillary
fring during a major portion of the growing
season. Riparian vegetation may or may not
include plants which are growing as preato-

phytes. Horton and Campbell (1974) have used
riparian for situations along water courses
where conditions are not suitable for phreato-
phytes. This use of riparian and phreatophytic
as conditions which are mutually exclusive is a
restrictive use and it can be confusing.

Bosque

Bosque - a stand of riparian vegetation
including plants which are growing as phreato-
phytes. Consequently, the term bosque is
usually limited to stands along major rivers or
floodplains. Bosque is not as common a term as
are riparian and phreatophyte . Bosque has a
historical use in the southwest and has a
different meaning here than in Mexico where it
usually means merely forest, woods, or grove.

Bosques may vary from gallery forest-like
stands of cottonwood (Populus fremontii) with
their associated shrubs to impenetrable thickets
which may include combinations or pure stands of
such plants as screwbean (Prosopis pubesoens)
mesquite (Prosopis glandulosa) , seepwillow
(Baodharis glutinosa) t saltcedar (Tamarix spp.)}
arrowweed (Pluohea sericea) 3 and various species
of willow (Salix spp.).

Obligate Riparian Species

Obligate Riparian Species - species
restricted to riparian or riparian-like situa-
tions. Riparian-like situations are those
typlified by areas receiving large amounts of
run-off water from surfaces with zero infiltra-
tion such as from the boulders of a talus slope
or when the local topography creates an area of
catchment which slows run-off sufficiently to
allow greater infiltration than that of
surrounding areas on the slope. We are aware
that the term "obligate" implies a restrictive
occurrence which has not been definitely
ascertained for most of the species so designated
However, the term serves to segregate the species
which may occur in riparian situations but are
more commonly found elsewhere.

Facultative Riparian Species

Facultative Riparian Species - species
other than "obligates" found in a riparian
situation. Some riparian habitats are virtually
restricted to obligate species because of poor
aeration and/or high slainity. This is often the
case along large river systems where there may
be high water tables and phreatophytic conditions
However, in montane situations where there is
good drainage or where running water is inter-
mittent or sporadic (arroyos) , any species
including obligate species can be expected.
These facultative species can usually be found
out in the open at higher elevations. This is a
result of the increased water availability in

riparian situations. This increased water
availability is due to the additional surface
or ground water and/or the reduced evapotrans-
piration rates (canyon effect) associated with
riparian situations.

RIPARIAN CLASSIFICATION

Recent Research
Related to Classification

During the past two decades there has been
considerable interest in riparian vegetation.
Particularly in the west (Horton 1973) . This
interest has slowly been heading toward attempts
at classification.

In the southwest, the biologist who has
perhaps led the way toward recognizing the
importance and character of riparian biotas is
Charles H. Lowe, Jr. For example, Lowe (1961)
defines riparian associations - or communities -
and discusses various aspects of them in his
treatment of the Sub-Mogollion region of New
Mexico, Arizona, Chihuahua, and Sonora. He
defines the associations as those occurring
"in or adjacent to drainage ways and/or their
floodplains and which are further characterized
by different species and/or life-forms than
those of the immediately surrounding non-
riparian climax." He emphasizes that "it is
incorrect to regard this biotic formation as
merely a temporary, unstable serai community,"
as "it is an evolutionary entity with an
enduring stability equivalent to that of the
landscape drainageways which form its physical
habitat."

In recent years the pronouncements of Lowe
have underlain an increasing concern and interest
among biologists and others in riparian associa-
tions and their biotas. For example, several
studies (e.g. Campbell and Dick-Peddie, 1964;
Campbell and Green, 1968; Freeman and Dick-
Peddie, 1970; Campbell, 19 74) have focused
directly on riparian plant communities, while
others have related such communities to their
use by such elements as birds (e.g. Hubbard, 1972;
Carothers and Johnson, 1973; Schmitt, 1976).
Even earlier papers were published in
ichthyology (e.g. Miller, 1961) touching on the
importance of riparian habitats to native
fishes, but in general these views have been
little appreciated by biologists until recently.

The awakening among biologists as to the
importance of riparian biotas has also sparked
increasing study, as indicated above. Initial
studies have been valuable but limited to date,
and a great deal remains to be learned. One
thing that is apparent however, is the fact that
a great deal of complexity exists in riparian
biotas. Not the least item among this complex-
ity is the matter of sorting riparian vegetation

into a workable and valid system of classifica-
tion. Attempts have already been made to
achieve this classification, but to date none of
these is entirely satisfactory.

Inclusion of Riparian Elements
In Vegetation Classification Systems

Lowe and David E. Brown have collaborated
in developing a vegetation classification system
for the southwest, to include riparian types.
In their latest endeavor, for example, Brown and
Lowe (1974) recognized twelve different riparian
communities. Following are two examples:

Forest Formation

Boreal Forest

Sub-Alpine Conifer Forest

Bristlecone-Limber Pine Communities
Spruce-Alpine Fir Communities

Temperate Forest

Montane Conifer Forest

Douglas-fir - White Fir Communities
Pine Communities

Relict Conifer Forest

Cypress postclimax Communities

Riparian Diciduous Forest

Mixed Broadleaf Communities
Cottonwood-Willow Communities

Woodland Formation

Boreal Woodland

Sub-Alpine Riparian Woodland
Willow Communities

Temperate Woodland

Rocky Mountain Conifer Woodland
Pinyon-Juniper Communities

Madrean Evergreen Woodland

Mexican Oak-Pine Communities
Encinal (Oak) Communities

California Evergreen Woodland
Oak-Pine Communities
Encinal (Oak) Communities

Subtropical Woodland

Riparian Deciduous Woodland
Mesquite Bosque Communities
Tamarisk disclimax Communities

In a preliminary classification of New
Mexican vegetation, Moir (1975) included

^-Moir, W.H. 1975. Vegetation Classifi-
cation System for use in New Mexico. Developed
for the New Mexico Natural Areas Problem, UNBUB.

riparian elements as parts of his major
categories as follows :

Coniferous Forest Association
Blue Spruce Series

Blue Spruce/Grass Streamside Association
Aspen Community
Willow-Alder Community

Deciduous (Riparian) Woodland Formation
Fremont Cottonwood Bosque

Arizona Sycamore Series

Arizona Sycamore /Fremont Cottonwood

Association
Arizona Sycamore /Arizona White Oak

Association
Arizona Walnut Series
Bigtooth Maple Series
Other Series

Forest Steppe (Mountain Grassland) Formation
Wet Meadow Series

Mesic Forb Community
Sedge-Grass Community
Willow-Sedge Community

Pase and Layser (1977)^ have further
refined ghe Brown and Lowe system in a tentative
classification. Following are two examples:

Temperate Riparian Deciduous Forest Biome
Mixed Broadleaf Series

Mixed Broadleaf Associations
Acer negundo Associations
Alnus oblongi folia Associations
Platanus wrightii Associations
Fraxinus velutina Associations
Juglans major Associations

Cottonwood-willow Series

Populus fremontii - mixed broadleaf

Associations
Populus fremontii Associations
Salix bonplandiana Associations
Populus fremontii - Salix goodingii

Associations

Subtropical Riparian Deciduous Woodland Biome
Mesquite bosque series

Prosopis juli flora Associations
Prosopis juli flora - mixed narrowleaf
(Tamarix, Chilopsis, Celtis)
Associations

Pase, CP. and Earl F. Layser. 1977.
Classification of Riparian Habitats. Paper
presented at Riparian Habitat Symposium,
Tucson, Arizona.

A Riparian
Classification System

The previous examples of the inclusion
of riparian categories in classification are
all worthy and valid. However, in light of its
importance and the fact that it represents an
unbroken continuum somewhat independent (partic-
ularly in the case of obligate species) of the
surrounding vegetation, it would appear that
riparian vegetation might well be treated as an
independent vegetative unit. By concentrating
upon obligate species , the seemingly bewildering
array of combinations, is greatly reduced.
Research on correlations of obligate species with
such variables as surface and subsurface hydrol-
ogy; canyon or valley cross sections and dimen-
sions; and reach elevations and exposures should
render more predictable the presently unpredic-
table occurrences of many obligate species. Such
occurrence predictability would enhance the
validity of a classification system.

Selection of
Obligate Riparian Species

Some of the obligate riparian species
found in one area have counterparts in their
genera as obligates in other areas. For example
some dominant floodplain genera in eastern
United States are Acer (maple), Fraxinus (ash),
Platanus (sycamore), Populus (cottonwood) ,
Juglans (walnut), and Salix (willow). These
same genera include major obligate riparian
species in the southwest. Other species are
obligate riparian species in the southwest but
their counterparts in the same genera are merely
facultatives elsewhere. Some examples of these
genera are Betula (birch), Celtis (hackberry) ,
and Cornus (dogwood) .

Even though the obligate nature of the
species so designated has not been scientifically
established it is not difficult to arrive at a
list of these species for any given area. For
example, following is a list of woody and grass
or grass-like obligate riparian species in New
Mexico. This list is undoubtedly incomplete
and some specialists may take exception to a
few of the included items. The list is the
result of lists compiled by the authors and a
list compiled by Dr. Hal McKay while working as
a consultant for the New Mexico Heritage Program.
Concensus on a list of obligate forbs is equally
easy to compile but due to the large number of
forb species a list of forbs is not included in
this paper.

Major Obligate Riparian Plants
Found in New Mexico

*Exotic (introduced)

Trees

Acer negundo
Alnus oblangifolia
Betula fontinalis
Celtvs reticulata
Fraxinus spp.
Jug tans major
Juglans micro carpa
Morus microphylla
Picea pungens
Platanus wrightii
Populus acuminata
Pcpulus angustifolia
Populus fremontii
Populus sargentii
Prunus virginiana
Salix gooddingii

Sapindus saponaria

Shrub -Trees

Acer grandidentatum
Alnus tenui folia
Amelanchier spp.
Amorpha fruticosa
Cerais occidentalis
Chilopsis linearis
Crataegus spp.

*Elaegnus angustifolia
Prosopis glandulosa
P tele a angustifolia

*Tamarix spp.

Rhus microphylla
Salix spp.

Sarcobatus vermiculatus
Shepherdia argentea

Grasses & Grass-like

Little-leaf sumac
Willow
Greasewood
Buffalo-berry

Box elder

*Alopercurus spp.

Fox-tail

New Mexican alder

*Arundo donax

Giant -reed

Birch

Bulbostylis spp.

Desert hackberry

Carex spp.

Sedge

Velvet ash

*Catabrosa aquatica

Brook grass

Walnut , Nogal

cyperus spp.

Flat -sedge

Little walnut

Distichlis striata

Salt grass

Mulberry

Eleocharis spp.

Spike-rush

Blue spruce

Equisetum spp.

Horsetail

Sycamore

Fimbristylis spp.

Glyceria spp.

Manna grass

Narrow-leaf cottonwood

*Hordeum hystrix

Barley

Fremont cottonwood

*Hordeum jubatum

Barley

Plains cottonwood

Juncus spp.

Rush

Common choke cherry

*Leersia oryzoides

Cut grass

Southwestern black

Luzula spp.

Wood-rush

willow

*Phragmites communis

Reed

Soapberry

Polypogon spp.

Rabbitf oot

Scirpus spp.

Bulrush

Typha lati folia

Cat -tail

Bigtooth maple
Thin-leaf alder
Service-berry
False indigo
Redbud

Desert willow
Hawthorn
Russian olive
Mesquite
Hop -tree
Salt cedar

Shrubs

Acacia greggii
Allenrolfea occiden-
talis
Apocynum spp.
Baccharis emoryi
Baccharis glutinosa
Baccharis sarothroides
Brickella calif ornica
Brickella laciniata
Chrysothamnus nauseosus
Var. graveolens
Var. bigelovii
Chrys othamnus

pu Iche I lus
Cornus stolonifera
Fallugia paradoxa
Forestiera neomexicana
Hymenoclea monogyra
Lonicera involucrata
Lycium torreyi
Philadelphus micro-

phy llus
Pluchea sericia
Rhamnus betulaefolia

Cat claw
Iodine bush

Dogbane
Baccharis
Seep willow
Desert broom
Brickel bush
Brickel bush

Rabbit-brush
Rabbit-brush

Rabbit -brush
Red-osier dogwood
Apache -plume
New Mexico Olive
Burro weed
Inkverry
Wolfberry
Mock orange

Arrow weed
Birchleaf buckthorn

Potential Classification of
New Mexican Riparian Vegetation

We present a potential classification
of riparian vegetation in New Mexico. This is
done to illustrate the validity of a system
based upon obligate riparian species and the
major topographic features which dictate their
presence. The classification does not rely upon
categories such as Boreal, Temperate, Subtro-
pical, Forest, Woodland, and Scrubland which
were developed primarily for non-riparian
vegetation.

380 RIPARIAN FORMATION

381 Alpine Sub -Formation
381.1 Forb Series

381.11 Rush Association

381.12 Spike Rush Association

381.13 Sedge Association

Montane Sub-Formation
Willow-Alder Series
Willow Association
Alder Association
Willow-Alder Association

Blue Spruce Series

Blue Spruce Association

Mixed Dedicuous Series

Willow -Dogwood Association
Alder-Willow Association
Boxelder-Ash-Walnut Association
Sycamore Association

Hackberry Association

Arroyo-Floodplain Sub-Formation
Arroyo Scrub Series

Greasewood Association
Rabbitbrush Association
Desert Willow-Brickelbush

Associ ation
Burroweed-Four-Winged Saltbush
Association

Floodplain (Bosque) Series
Cottonwood Association
Cottonwood-Willow Association
Mesquite Association
Arrowweed -Seep -willow

Associat ion
Saltcedar Association
Mixed Bosque Association

It would not be difficult to incorporate
this riparian classification into existing
systems. We have included digits for the
first series as an illustration of how this
classification could utilize the Brown and
Lowe system. We have used the major unit
terminology of Formation, sub-Formation Series,
and Association to illustrate the applicability
of this proposed classification to the system
being developed by the Forest Service.

The proposed classification is inten-
tionally labelled "potential." Some units may
not be valid or have utility when applied in
the field. Possibly, some units should be
raised or lowered in the hierarchy. There will
undoubtedly be additions and possibly a need
for deletions. Continued research and use of
the classification should serve to remove any
such errors.

LITERATURE CITED

Brown, D.E. and C.H. Lowe. 1974. The Arizona
system for natural and potential vegetation
— illustrated summary through the fifth
digit for the North American Southwest.
J. Arizona Acad. Sci. 9:1-7.

Campbell, C.J. 1970. Ecological implications
of riparian vegetation management. J. Soil
Water Conserv. 25(2): 45-52, illus.

Campbell, C.J. and W.A. Dick-Peddie. 1964.
Comparison of phreatophyte communities on
the Rio Grande in New Mexico. Ecology
45:492-502.

Carothers, S.W. and R.R. Johnson. 1973.

Population structure and social organiza-
tion of southwestern riparian birds.
Unpublished ms. 22 pp., 4 tables.
Freeman, C.E. and W.A. Dick-Peddie. 1970.

Woody riparian vegetative in the Black and
Sacramento mountain ranges, southern New
Mexico. Southwestern Nat. 15:145-164.

Horton, Jerome S. 1973. Evapotranspiration
and Water Research as Related to Riparian
and Phreatophyte Management. Forest
Service-United States Department of
Agriculture .

Horton, Jerome S. and Campbell, C.J. 1974.
Management of Phreatophyte and Riparian
Vegetation for Maximum Multiple use values.
USDA Forest Service Research Paper RM-117,
Fort Collins, Colorado.

Hubbard, J. P. 1977. A biological inventory

of the lower Gila River Valley, New Mexico.
U.S. Fish and Wildlife Service, Albuq.,
50+ pp.

Hubbard, J. P. and B.J. Hayward. 1973. A
biological survey of the San Francisco
Valley (Greenlee County, Arizona to Catron
County, New Mexico), with emphasis on
habitats and vertebrates. Unpublished ms .
(U.S. Forest Service) 23 pp.

Kuchler, A.W. 1964. Potential Natural Vege-
tation of the Conterminous United States.
Map and manual American Geographical Soc,
New York.

Lowe, C.H. 1961. Biotic communities in the

Sub-Mogollon Region of the inland Southwest.
J. Arizona Acad. Sci. 2:40-49.

Lowe, C.H. and D.E. Brown. 1973. The natural
vegetation of Arizona. ARIS Coop. Pub. #2
53 p., Arizona Resources Information System.

Schmitt, C.G. 1976. Summer birds of the San
Juan Valley, New Mexico. N. Mex. Orn.
Soc. Publ. 4.

Fishes Inhabiting the Rio Grande,

Texas and Mexico, Between
El Paso and the Pecos Confluence1

12 I 3

by Clark Hubbs , Robert Rush Miller ,

12 12
Robert J. Edwards , Kenneth W. Thompson ,

2 V- ' 2 4

Edie Marsh , [Gary P. Garrett , Gary L. Powell,

V 5^ I 5

D. J. Morris , and Robert W. Zerr

Abstract — The fishes of the middle part of the Rio Grande can
be divided into three faunal assemblages: The saline Rio Grande fauna
(made up of widely distributed and salt tolerant species) upstream
from the Conchos confluence; the Rio Conchos-Rio Grande fauna (mostly
south Texas and Mexican species) in the Rio Grande between the Conchos
and Pecos; the tributary creek fauna (Chihuahuan species plus some
derivatives) that depend on tributary creeks for all or part of their
life history stages. Endangered species are found in the last assem-
blage but two presumed endangered species (Notropis simus and Scaphi-
rhynchus platyrynchus) seem to have been eliminated already.

INTRODUCTION

The fishes of the Rio Grande (Belcher,
1975: fig. 3) have been intermittently studied
for the past 130 years. Reasonably extensive
reports exist for Colorado (Beckman, 1952), New
Mexico (Koster, 1957), and the Rio Grande down-
stream from its confluence with the Pecos River
(Trevino-Robinson, 1959) . No comparable summar-
ization exists for the intervening segment, al-
though Miller (1977) treated the Mexican part
of the middle Rio Grande basin. Proposals to
channelize about 300 kilometers of the river
and to designate another 200 kilometers as a
wild river underscored the absence of a summar-
ization of the fish fauna. The bulk of this
paper is a report on fishes collected on two
recent visits to the Rio Grande in the two pro-
ject areas. We also include a summarization of

Contributed paper, Symposium on the

Importance, Preservation and Management of the

Riparian Habitat, July 9, 1977, Tucson, Arizona.
2

Department of Zoology, The University of
Texas, Austin, Texas 78712

^Museum of Zoology, The University of Mich-
igan, Ann Arbor, Michigan 48109.

^Texas Water Development Board, Austin,
Texas, 78701

^Texas Parks and Wildlife Department,
Austin, Texas 78701

a large number of collections made from the Rio
Grande in Big Bend National Park between 1954
and 1976.

The Rio Grande "enters" Texas as a small
stream most or all of which is diverted to irri-
gate fields south and east of El Paso. Commonly,
the stream is dry over much of the distance be-
tween El Paso and Ft. Hancock. Southeast of
this town the valley narrows and the ground water
surfaces to form a salty stream. The river re-
mains small for the next 300 km until it "receives"
the Rio Conchos. Small volumes of water are
added by small salt laden springs (such as Indian
Hot Springs) and fresh tributary creeks (such as
Capote Creek) . These increases are commonly ex-
ceeded by losses from evaporation or irrigation
diversions. Drastic increases in flow periodi-
cally follow intense desert rains. These torrents
soon subside and the Rio Grande again becomes
a small, sometimes intermittent stream. This
pattern is of long duration as Emory (1859) re-
ported periodic dry stream beds and occasional
severe flooding and Thomas (1963) reported high
salinities in the Rio Grande in 1936. This
reach of the river has been extensively impacted
by human activities. Much of the flow (and most
of the low saline water) is diverted at or
north of El Paso. The northwestern 150 kilometers
have been leveed and channelized. A 16 kilometer
segment around the Conchos confluence has also

been leveed and channelized. The intervening
300 kilometers has been proposed for channel
"rectification" and is extensively leveed
already .

Much of the flow of the Rio Grande east
of the Conchos is dependent upon that "tributary".
Historically, the contribution of the Conchos
has been considerably greater than that of the
Rio Grande above the confluence and that dif-
ference has been magnified by the Rio Grande
diversions upstream. Present flow rates depend
chiefly on releases from Luis L. Leon Reservoir;
at the time of our visit on 18 March, 1977, the
Conchos flow was nearly 2 orders of magnitude
greater than that of the Rio Grande. Strangely,
the man-made conjunction has the Conchos enter-
ing at a right angle; in effect a forced right
angle turn of the huge stream where it enters
the small stream. Between the end of its recti-
fied channel below Presidio and the upper part
of Amistad Reservoir (just upstream from the
Pecos confluence) , the Rio Grande has not been
substantially impacted by human activities.
The major items are the stream measurement weirs
just below Alamito Creek and just above Amistad
Reservoir, river fords at Stillwater Crossing
and Boquillas, and a bridge near Stillwater
Crossing. Other impacts are indirect such as
minor irrigation diversions, overgrazing, exotic
plants and fishes, pesticides washed from nearby
fields, leaching from mine tailings, etc. Much
of this distance is little .disturbed and one can
see the diverse geology and magnificent land
formations .

COLLECTION SITES

Most of the newly reported locality records
are based on two collecting trips, 14-18 March
and 3-7 April, 1977. Collections were concen-
trated in the channelization and wild river
segments, respectively. Previously, only one
sample had been obtained from each of those
reaches. The 1977 and previous (1954) locations
are plotted on figures 1 and 2. The bulk of the
1954 (and subsequent) Rio Grande collections
were from the Big Bend National Park and have
been reported in Hubbs (1958) , Hubbs and Wauer
(1973) and Hubbs and Williams (in press).

RESULTS

The 15 collections from the Rio Grande
west of the influence of the irrigation water
from the Rio Conchos that enters the Rio Grande
between Stations 15 and 16 contain 11 fish
species (Table 1) . The redundant nature of
these samples is reflected by the presence
of 7 fishes (Dorosoma cepedianum, Cypr inus
carpio , Notropis lutrensis , Carpiodes carpio ,
Ictalurus punctat'us , Gambusia af f inis , and
Lepomis cyanellus) in 9 or more collections.

Figure 1. — Location of collection stations in
the Rio Grande from and adjacent to the pro-
posed channelization.

Son Francisco

Canyon.^ 28

Terrell Co.

Sanderson

Conyon Dryden
Crossing

Val Verde Co.

Brewster Co

Reagan
Canyon

Moravi I las <
Canyon

10km

Stillwell I
Crossing

Figure 2. — Location of collection stations in the Lower Canyons of the Rio Grande.

Their widespread abundance suggests that they
would be expected anywhere in this stream seg-
ment. The second listed species (Cyprinus carpio)
is an exotic but the others are all widely dis-
tributed native fishes. Two of the other four
species (Lepomis megalotis and Morone chrysops)
were collected at widely separated sites. The
former was found at very brushy sites. It is
likely that this species can be obtained wher-
ever those conditions prevail. The latter (un-
doubtedly, derived from fishes stocked near Del
Rio) is an open-water top carnivore. This fish
would be expected to be sparsely distributed
because of dependence upon a complex food chain
and consequently high primary productivity per
fish. In a similar way, a large fish like
Ictalurus f urcatus would be expected to be rare
in a small stream like the Rio Grande in this
reach. The last species Pimephales vigilax, lias
not previously been taken east of Val Verde Co.
As this fish is commonly used as a bait minnow,
it is likely that the samples obtained are
descendants from escaped bait.

The six collections from the vicinity of
Presidio (16B on Table 1) contain 20 species;

11 were not taken upstream but 2 from there were
absent. We expect that increased effort would
have produced an Ictalurus punctatus , but that
Pimephales vigilax is not present. Nine of the
additional 11 species, Astyanax mexicanus
(as A. f asciatus) , Hybopsis aestivalis , Notropis
chihuahua, Notropis braytoni, Notropis j emezanus ,
Pimephales promelas , Campostoma ornatum, Pylo-
dictis olivaris , and Lepomis macrochirus , were
reported from the Big Bend region by Hubbs
(1958). One exception, Cyprinodon eximius
has subsequently been reported from Terlingua
creek by Miller (1977). The other, Menidia
beryllina , is undoubtedly derived from descendants
of bait-released stocks now abundant in Amistad
Reservoir. We expect that Menidia (a euryhaline
species) will soon spread and become abundant
in the saline Rio Grande waters upstream from
the Conchos confluence.

The distinct difference between the Rio
Grande fishes on either side of the Conchos
confluence is reflected by similar differences
between the fishes inhabiting the tributary
creeks, Capote and Alamito (stations 13 and
19, respectively) .

Table 1. — Numbers of fishes collected from the Rio Grande from and adjacent to the proposed channelization between
Presidio and El Paso

Species

Dorosoma cepedianum
As tyanax mexicanus
Cypi inus carpio
Hybops is aestivalis
Motropis chihuahua
No tr op is bray ton!
Notropis lutrensis
Notrop is jemezanus
Fimephales vigilax
rimephqles promelas
Campos toma orna turn
Carpiodes carpio
Ictalurus punctatus
1c talurus f urea tus
Py Iodic ti s ollvaris
Cyprlnodon eximius
Gamhusia af finis
Ken id i a bery llina
Ho rone chrys ops
Lepomis cyanellus
Lepoinls megalotis
Lepomis macrochirus

% Introduced
Fish/s»ine hour

10 1 3

1 1

1 6

17 19

108

133

4 3

7 I

1 59

1 8

1 A

125 6

1
1

8
7

1 7

3 3

1 3

1 30

242

4 9

335 138

354 166

262

]

122

4 10

105

1.68

1.7

1.3

1.5

0.4

1.5

0.9

l.l

1.8

1.4

1.7 0.3

2.1

1.8

2.0

2.2

1.6

0.5 1.6

2.3

26 5

1 2

101

1.19

193

156

136

166

351

316 248

151

161

445

500

381 527

133

Table 2 — List of fishes collected from the Lower Canyons of the Rio Grande.

Stat ions

Species

Lepisosteus osseus

DoroGoma cepedianum

1.1

Cycleptus elongatus

108

Carpiodes carpio

IctioDus bubalus

Astyanax mexicanus

Cvprinus carpio

Rhynichfhys cataractae

188

182

209

184

141

191

168

850

167

Hybopsis aestivalis

Pimephales promelas

Notropis chihuahua

Notropis jemezanus

Notropis lutrensis

112

Notropis braytoni

Ictalurus punctatus

Ictalurus furcatus

Pylodictus olivaris

Fundulus kansae

Gambusia afflnis

Menidia beryllina

Micropterus salmoides

Lepomis cyaneilus

Lepomis macrochlrus

Chemical and physical conditions in the two
creeks are reasonably similar - Capote
is slightly smaller and has been reported dry
near the mouth. We attribute the fish faunal
differences to the impact of seasonal migra-
tions into the Rio Grande as reported for Ter-
lingua Creek by Hubbs and Wauer (1973) . Those
salty Rio Grande waters at the mouth of Capote
Creek may exclude the typical Rio Grande tribu-
tary creek fauna from any upstream tributary.
Regardless of the cause, these fishes were not
found in Capote Creek. We looked carefully for
fishes in the waters of Indian Hot Springs to
determine if an endemic fauna were there. These
warm, salt-laden springs were fishless. We did
note Gambusia af f inis was concentrated in the
warm outflow waters emptying into the colder
Rio Grande waters during our March visit.

The faunistic difference between the
fishes of the two segments is of long duration.
The 1954 samples (A and B) are well representa-
tive of the faunal units found in 1977 samples.

The 13 collections from the Lower Rio
Grande Canyons contained 23 species (Table 2) .
Thirteen (Dorosoma cepedianum, Carpiodes carpio ,
Astyanax mexicanus, Hybopsis aestivalis , Pime-
phales promelas , Notropis chihuahua, Notropis
jemezanus , Notropis lutrensis, Notropis bray-
toni, Ictalurus f urcatus , Pylodicitis olivaris ,
Gambusia af finis, and Lepomis macrochirus)
were found in the collections near Presidio
and reported from the Rio Grande in Big Bend
National Park (Hubbs, 1958). The absence of
Cycleptus elongatus and Rhinichthys catarac-
tae from the upstream stations is likely to
be a seasonal artifact because both have been
reported from the Rio Conchos. Both species
are also absent in the August 1954 collection
(C). All of the Cyclep tus collected down-
stream were young of the year. Similarly,
the bulk of the Rhinichthys were young. It
is unlikely that adult Cycleptus would be
collected with the seines used in such high
water (and none were) . Samples taken near the
mouth of Tornillo Creek in April commonly have
many young Cycleptus but no adults are in col-
lections from that spot. Similarly, Rhinichthys
are likely to be most abundant just after the
breeding season. Our station 20 (Rio Grande
just east of the mouth of Alamito Creek) in-
cluded one fish tentatively identified as a
Rhinichthys that escaped prior to being pre-
served . Rh inichthys abundance in the area is
supported by its presence in a collection from
the Rio Grande just upstream from Mariscal Canyon
in Big Bend National Park. Specimens have also
been taken from the Conchos system in Chihuahua.
The absence of Lepisosteus osseus, Ictiobus
bubalus and Micropterus salmoides in the col-
lections near Presidio is likely to be a samp-
ling artifact. The high-water flows made it
very difficult to sample deep-water environments

commonly occupied by these (especially Ictiobus
and Ilicropterus) and our downstream samples were
sufficiently infrequent that chance occurrence in
the Presidio samples is likely. Two species
(Menidia beryllina and Lepomis cyanellus) were
collected near Presidio but not reported from Big
Bend National Park by Hubbs (1958) . It is un-
likely that the former existed in the region
before 1960 as Tilton and White (1964) showed
that this fish was then just being distributed
across Texas. Hubbs and Echelle (1972), docu-
mented a similar and recent spread of this
fish in the Pecos Basin. Lepomis cyanellus is
now known from Big Bend National Park (Hubbs
and Williams, in press), supporting Hubbs'
(1958) prediction that it existed within the
park. Similar to Menidia audens , Fundulus
kansae has recently been introduced into the
region. Its introduction and subsequent spread
was reported by Hubbs and Wauer (1973) . The
fish from the Rio Grande at the mouth of Mara-
villas Cr. surely reflects an additional
spread. We herein also report the presence of
Fundulus kansae in McKinney Spring in Big Bend
National Park. It is likely that the speci-
mens of Micropterus salmoides reflect a modest
population of indigenous fishes that can serve
as a recreational resource. We expect that
largemouth bass occur throughout the Rio Grande
east of the Conchos confluence (and also in
much of the Conchos system) .

The 23 species extensively overlap those
reported from the Big Bend by Hubbs (1958) who
recorded 7 additional fishes. Four of them,
[Dionda episcopa, Gambusia gaigei, Lepomis
(=Chaenobryttus) gulosus , and Lepomis micro-
lophus] were recorded as inhabiting small clear
tributaries and would be rare or absent in the
river proper. Moxostoma conges turn was subse-
quently reported from Tornillo Cr. by Hubbs and
Wauer (1973). Three fishes, Anguilla rostrata,
Hybognathus placitus, and Aplodinotus grunniens)
were reported from Big Bend by Hubbs (1958) ,
but not obtained in the 1977 samples. The first
is catadromous and upstream migrants would be
unlikely to pass Falcon Dam (much less Amistad);
samples have not been obtained since Falcon was
filled. The other two would be expected to occur
in the area. Aplodinotus could easily have been
overlooked but the absence of Hybognathus is
inexplicable .

DISCUSSION

The fishes inhabiting the Rio Grande in
west Texas can be placed in three faunal assem-
blages: Saline Rio Grande fauna, Rio Conchos-
Rio Grande fauna, Tributary Creek fauna.

The Saline Rio Grande Faunal assemblage
is dominated by four wide spread species,
Dorosoma cepedianum, Cyprinus carpio, Notropis

lutrensis , and Lepomis cyanellus . The limited
diversity (Shannon H' values are generally
well below 2) seems to be due to harsh condi-
tions - salinity and periodic interrupted stream
flows. The latter may be most critical as the
fish present are ones expected in pools in
west Texas streams. Our repeated efforts in
riffles were generally unproductive. This assem-
blage has been impacted by human activities.
Certainly the three exotics (Cyprinus , Morone,
Pimephales) must have some impact. It is
likely that Cyprinus has depressed Carpiodes
abundance but the impact of Morone and Pime-
phales is difficult to assess, and the absence
of prior studies makes any conclusions con-
jectural.

The Rio Conchos - Rio Grande faunal assem-
blage is made up of those species living in the
Rio Grande and not dependent upon tributary creeks
for a part of their life history. The abundance
of these fishes is not correlated with the pre-
sence of tributary flows. Typical fishes of this
assemblage are Notropis j emezanus , N_. lutrensis ,
_N. braytoni , Rhinichthys cataractae , Hybopsis
aestivalis , Ictalurus punctatus , Ictalurus
furcatus, Pylodictis olivaris, Dorosoma cepe-
dianum, Cycleptus elongatus , and Carpiodes
carpio. Seven other fishes (Lepisosteus osseus,
Ictiobus bubalus , Pimephales promelas , Men-
idia beryllina, Micropterus salmoides , Aplodin-
otus grunniens , and Hybognathus nuchalis) are
reasonably abundant in the Rio Conchos-Rio Grande
faunal assemblage.

Only one (Menidia beryllina) of those 18
species is introduced. Its impact is not yet
fully assessed as its entry into the region is
so recent that its abundance may be in a growth
phase. It is not likely that this quiet water
euryhaline form will become excessively abundant
in the fresh-flowing waters of the Rio Grande.
Rhinichthys cataractae is not only a prominent
member of this faunal assemblage, it also seems
to be absent or very scarce in adjacent areas.
This population is isolated from other stocks by
the saline and frequently dry Rio Grande up-
stream from Presidio. It is likely that it re-
presents a race adapted to deep canyons with re-
latively warm water. Essentially, a collection
from Texas with numerous Rhinichthys and/or
Cycleptus is likely to be from the Rio Grande
between Presidio and Amistad Reservoir. The
Rio Conchos - Rio Grande faunal assemblage will
often be supplemented by representatives from
the tributary creek faunal assemblage.

Two fishes (Scaphirhynchus platyrynchus
and Notropis simus) may once have inhabited
the Rio Conchos - Rio Grande faunal assemblage.
Scaphirhynchus was reported from the Rio Grande
near Albuquerque by Cope and Yarrow (1875) .
We have obtained hearsay reports of a sturgeon
from near Dryden Crossing (and also from Mexican

tributaries in Coahuila) that support the former
occurrence of shovelnose sturgeon in the river.
Notropis simus has been recorded from the Rio
Grande in New Mexico and downstream from Del
Rio but the collections preceded or were at a
similar time interval as the first collections
from our study area. We doubt that Notropis
simus now lives in the Lower Canyons of the Rio
Grande and suggest that work to ascertain if it
still exists concentrate on the lower Rio Conchos.
We have no suggestions as to the conditions
that may have led to the extinction or substantial
decline of these two fishes that once were part
of this faunal assemblage. Both species are com-
monly found on listings of endangered species and
N. simus may be extinct in U.S. waters. Its ab-
sence in Trevino-Robinson' s collections is par-
ticularly alarming as most Texas records are
from that stream segment. The New Mexico re-
cords are from the Rio Grande in areas that now
have reduced flow or are dry.

The tributary creek faunal assemblage i~
made up of a group of fishes that spends all or
a substantial fraction of their time in the small
tributaries. Three species (Notropis lutrensis ,
Pimephales promelas, Notropis braytoni) may
occur in the creeks or Rio Grande. Except for
the first, they are seldom found far from the
creek mouth. Three (Moxo stoma congestum
Carpiodes carpio, Cycleptus elongatus) are
creek inhabitants only as young and the adults
may be found with equal abundance elsewhere
in the Rio Grande. Eleven species (Cyprinodon
eximius , Campostoma ornatum, Notropis chihuahua,
Fundulus kansae , Astyanax mexicanus , Gambusia
af finis and the sunfishes, Lepomis gulosus ,
cyanellus , microlophus , macrochirus , and mega-
lotis) are most commonly collected in creeks
but have been found in the Rio Grande. The
first six are listed by relative frequency of
creek vs. river abundance. Hubbs and Wauer
(1973) had reported that this assemblage moved
out of the creeks seasonally but our 1977 sam-
ples of the first two are the first demonstration
of fish that must have moved into the river.
Samples of the five sunfishes are sufficiently
infrequent that definite patterns are difficult
to ascertain. Two species (Gambusia gaigei,
Dionda episcopa) are limited to the tributary
waters. The former is on all lists of endan-
gered fishes; its status has been discussed
recently by Hubbs and Williams (in press) .
The fishes in the tributary creek assemblage
often present special problems. Three of them
(Cyprinodon eximius , Campostoma ornatum,
Notropis chihuahua) are commonly found on
endangered species listings as their U.S. dis-
tribution is restricted to the creek mouths.
These areas should be watched with care to
reduce the possibility of extermination of
this fragile assemblage. The spread of the
introduced Fundulus kansae is of primary con-
cern (Hubbs and Wauer, 1973) . Future intro-

ductions of bait minnows should be avoided.

ACKNOWLEDGMENTS

We have benefitted from the encourage-
ment by and discussion with James E. Johnson
in planning our program. David S. Marsh, Linda
Garrett, Danny Swepston, Floyd Potter, and
Dwane Kippes participated in collecting the
fishes. Ross Foster, John Vandertulip, and
Pollard Rogers were hospitable hosts, pro-
vided helpful information, and permitted tres-
pass access to or from collection sites.
William Provine, Floyd Potter, and W. S.
Swanson helpfully participated in the plan-
ning phases of our studies. Texas Parks and
Wildlife and Departamento de Pesca de Mexico
gave permission to collect fishes from the
Rio Grande.

LITERATURE CITED

Beckman, W. C. 1952. Guide to the fishes of
Colorado. Univ. Colo. Mus. Leaflet 11.

Belcher, R. C. 1975. The geomorphic evolution

of the Rio Grande. Baylor Geol. Stud., Bull.
29.

Cope, E.D., and H.C. Yarrow. 1875. Report

upon the collections of fishes made in por-
tions of Nevada, Utah, California, Colorado,
New Mexico and Arizona, during the years 1871,
1872, 1873, and 1874. Rept. Geog. and Geol.
Expl. and Surv. W 100th Merid. (Wheeler
Survey) 5:637-703.

Emory, William C. 1857. Report of the United
States and Mexican Boundary Survey. Vol. I.

Hubbs, Clark. 1958. List of fishes known or

expected to belong to the fauna of Big Bend
National Park. Mimeograph report to Big
Bend National History Association.

, and Antony A. Echelle. 1973. En-
dangered non-game fishes in the Upper Rio
Grande Basin. In: Endangered Vertebrates
in the Southwest. William C. Huey (ed.)
New Mexico Game and Fish : 147-167 .

, and Roland Wauer . 1973. Seasonal

changes in the fish fauna of Tornillo Creek,
Brewster County, Texas. The Southwestern
Nat. 4:375-370.

, and John G. Williams. in press. A

review of circumstances affecting the abun-
dance of Gambusia gaigei, an endangered
fish endemic to Big Bend National Park.
In: Proceedings of the First Conference
on Scientific Research in the National
Parks. Robert Linn and George Sprugle,
(eds.) U.S. Government Printing Office.

Koster, W.J. 1957. Guide to the Fishes of

New Mexico. University of New Mexico Press.

Miller, R.R. 1977. Composition and deriva-
tion of the native fish fauna of the Chi-
huahuan desert region. In: Transactions,
Symposium on the Biological Resources of
the Chihuahuan Desert Region, U.S. and
Mexico. R.H. Wauer and D.H. Riskind (eds.).
National Park Service.

Thomas, H.E. and others. 1963. Effects of

drought in the Rio Grande Basin. U.S. Geol.
Surv. Prof. Pap. 372-D=Dl-59.

Tilton, J.E., and R.L. White. 1964. Meni-

dia from several central Texas impoundments.
Texas J. Sci. 16:120.

Trevino-Robinson, D. 1959. The ichthyofauna

of the lower Rio Grande, Texas and Mexico.
Copeia 1959:253-256.

An Overview of
Riparian Forests in California:
Their Ecology and Conservation1

Anne Sands and Greg Howe

This paper is comprised of abstracts from presentations
made at the Symposium on Riparian Forests in California:
Their Ecology and Conservation held in Davis, California on
May 14, 1977. Sponsors of the symposium were the Institute
of Ecology (University of California at Davis) and the Davis
Audubon Society, Inc. The purpose of this symposium was to
encourage a strong alliance of individuals and agencies which
will work together to establish protection for the endangered
riparian ecosystems of California. Complete texts of these
papers will appear in proceedings of the Davis symposium.

A SHORT REVIEW OF THE STATUS OF
RIPARIAN FORESTS IN CALIFORNIA

Felix Smith, Field Supervisor
Division of Ecological Services
United States Fish and Wildlife Service
Sacramento, California

Riparian vegetation along streambanks,
where there is usually fertile soil and an
ample water supply, is a most striking feature
of California's landscape. These forests appear
as a green belt along permanent and intermittent
water courses, sloughs, flood plains, overflow
channels and oxbows, drainage ditches and lakes.

One can quickly see that the riparian
community with its soil, water, and vegetation
is a complex ecosystem. Cheatham and Haller
(California Fish and Game, 1965), in their
"Annotated List of California Habitat Types"
have identified four major riparian habitats
with 11 subhabitat types. Of the 29 habitat
types listed in the "Inventory of Wildlife

■'•Contributed paper, Symposium on the
Importance, Preservation and Management of the
Riparian Habitat, July 9, 1977, Tucson, Arizona.

^Anne Sands, Institute of Ecology,
University of California, Davis, California
95616; Greg Howe, Department of Wildlife and
Fisheries Biology, University of California,
Davis, California 95616. Both editors are
also representing the Davis Audubon Society,
Incorporated, P.O. Box 886, Davis, California
95616.

Resources, California Fish and Wildlife Plan"
(Vol. Ill), riparian habitat provides living
conditions for a greater variety of wildlife
than any other habitat type found in California.
It was estimated in 1963 that riparian vegetation
covered about 347,000 acres — less than one-half
of one percent of the total land area of the
State.

Factors affecting or adversely impacting
riparian vegetation include upstream reservoir
construction, levee and channelization projects,
and water conservation. The reservoir, levee
and channelization activities, along with clearing
for agriculture, are common activities that have
occurred throughout the State. Removal of vege-
tation is a common practice in the Colorado
River area. Let's look at a riparian area from
a local viewpoint. In An Island Called Califor-
nia, Elna Bakker (1971) states that no natural
landscape in California has been so altered by
man as its bottom lands. It was in the Central
Valley that riparian forests were most extensive
and were called gallery forests. Coupled with
the extensive grasslands and rivers, large and
small, a unique setting was created. It is now
one of the richest agricultural areas in the
world, blessed with good climate, rich soil, and
until the last couple of years, ample water
supplies .

The Sacramento River from Red Bluff to its
mouth in the Delta is a meandering alluvial
stream. The Sacramento Valley extends about 150
miles north-south and spreads about 45 miles
east-west at its widest point, averaging about
30 miles wide. The area of the Sacramento Valley,

so defined, is about 5,000 square miles; the
area of the entire Sacramento River drainage
basin is 26,150 square miles.

The Sacramento Valley is bounded by the
Coast Ranges on the west, the Klamath Mountains
on the north, and the southern Cascade Range and
northern Sierra Nevadas on the east. The
southern margin is extremely low terrain, cut
by numerous branching channels of the Delta.
This whole low- lying and level area is formed
by the combined Delta of the Sacramento and San
Joaquin Rivers. Lands of the Sacramento Valley,
excluding the Sutter Buttes, are essentially
flat, almost featureless, and were formed by the
long-continued accumulation of sediments in a
great structural trough lying between the Coast
Ranges and the Cascades-Sierra Nevada. Large
and small streams break up the landscape. Each
had its green belt of riparian vegetation that
stretched from the base of the foothills to the
big river and adjacent wetlands.

Vegetation will grow on any portion of a
streambed and its banks if the soil or other
substrate is exposed long enough during the
growing season. The fertile loam soils of the
Sacramento River riparian land coupled with
favorable ground water conditions and a long
growing season provide near optimum conditions
for the establishment of the extensive riparian
forests.

The riparian woodlands occurred on the
natural levees formed by the Sacramento, Lower
Feather, American, and other aggrading streams.
These levees rose from 5 to 20 feet above the
streambed, and ranged in width from 1 to 10
miles. Based on historical accounts, it has
been estimated that there were about 775,000
acres of riparian woodlands in 1848-1850.
Diaries and field notes written in the early
1800' s describe the extent of the forests. They
also describe the lush jungles of oak, sycamore,
ash, willow, walnut, alder, poplar, and wild
grape which comprised almost impenetrable walls
of vegetation on both sides of all the major
valley rivers and their tributaries. Notes were
made of giant sycamore 75 to 100 feet tall and
of oaks 27 feet in circumference. By the late
1800' s, however, vast tracts of riparian forests
had already been cut by settlers for fuel,
fences, and building materials. In addition,
many thousands of acres were cleared to free the
fertile alluvial soil for agricultural use. By
1952, only about 20,000 acres of riparian
forests remained. Today's estimate of 12,000
acres is probably generous.

Prior to 1960, few people showed any
concern for the demise of California's Riparian
Forest communities. In addition, very little
botanical data had been collected. During the
early 60' s, the first major work at removing the

riparian forest remnants in an effort to protect
levees occurred in the Delta. The removal of this
riparian vegetation from along the lower Sacra-
mento River was viewed with great concern by the
public. Statements, both written and oral, voiced
strong opposition to the methods of levee main-
tenance and stated that better methods should be
employed so as not to destroy the esthetic beauty
and wildlife habitat of the Delta waterways. Most
of the same concerns exist today. However, today
dedicated and enthusiastic botanists, ornitholo-
gists, mammalogists , entomologists and other field
scientists are compiling species lists, recording
observations, and beginning to publicize their
findings. People now realize that public aware-
ness must be coupled with political pressure.

The previously expressed concerns demonstrate
a clear need for a higher order of planning and
evaluation before additional irreparable alter-
ations to this river system occur. Although no
governmental body, agency, interest, or person
would deliberately set out to destroy the Sacra-
mento and other California Rivers, adjacent lands
and natural resources, all too often there has
been insufficient concern about the singular or
cumulative effects of work accomplished by one
agency or interest on the resources under the
jurisdiction or responsibility of another, or
how such work affects the entire riverine eco-
system and the public interest.

The realization of a Sacramento River
environmental/open space corridor is a valid and
long-term planning objective. Implementation is
the difficult part. However, it can be done. It
will require the formulation of a multigovern-
mental agency and concerned citizen group to see
that modifications and developments are accom-
plished without further deterioration of the
existing resources and that efforts are under-
taken to enhance these same resources in the
public interest while at the same time protecting
the integrity of the levees and communities of
the Sacramento Valley. Can one imagine a Sacra-
mento River Parkway from the Redding area to
Collinsville patterned after the American River
Parkway? What a valuable recreational resource
it would be to the public and especially for
future generations.

LITERATURE CITED

Bakker, Elna S. 1971. An Island Called Cali-
fornia. University of California Press.

California Department of Fish and Game. 1965
California Fish and Game Plan.

GEOLOGICAL HISTORY OF THE RIPARIAN FORESTS
OF CALIFORNIA

Robert Robichaux
Department of Botany
University of California, Davis

The plant communities that we see today are
the products of evolutionary processes acting
over long periods of geological time. As indi-
vidual species have evolved and migrated in
response to changing environments, the corres-
ponding plant communities have changed in
composition and distribution. The evidence
relating to the rates and direction of this
change comes from an analysis of numerous fossil
deposits laid down at various times and in
various regions in the past (Axelrod, 1967b).
This analysis involves the systematic descrip-
tion of the component species in each fossil
flora and the reconstruction of the ancient
topographical, climatic, and vegetational
settings at the site of deposition. Regional
comparison of various floras then allows us to
piece together the history of individual lin-
eages and the corresponding plant communities.

The history of California's vegetation as
a whole has recently been reviewed by Axelrod
(1977). Three general principles emerge from
his discussions. First, the modern plant com-
munities of California are composed of taxa of
diverse geographical sources. The two princi-
pal floristic elements are a "Madro-Tertiary"
or southern, and an "Arcto-Tertiary" or northern
element. The former includes species in such
genera as Arbutus, Arctostaphylos , Ceanothus ,
Cercocarpus, Cupressus , Quercus (some species),
and Umbellularia, while the latter includes
species in such genera as Acer, Alnus , Casta-
nopsis, Fraxinus , Picea , Quercus (some species) ,
and Sequoia. Second, the modern communities
are relatively impoverished representatives of
richer, more generalized ancestral communities
that include taxa related to species now found
only in the southwestern United States and
northern Mexico, the eastern United States, or
eastern Asia. These "exotic" taxa were gradu-
ally eliminated from this region during the
later Tertiary in response to a general trend
to a cooler and drier climate, to a shift in
the seasonal distribution of precipitation, and
to progressively decreasing equability (Axelrod,
1968). Third, some of the species that are
associated in these modern communities have
apparently been associated, as ancestral forms
in fossil communities, throughout most of Cali-
fornia's later Tertiary and Quanternary history,
covering a time span of at least 20 million
years .

Distributions of the Species

One of the most important factors facili-
tating a species entry into the fossil record is
a proximity to a site of sedimentation. As many
fossil deposits accumulate along stream and lake
borders, riparian taxa are generally well-repre-
sented in the record. Many of the dominant
species in the modern riparian community of the
Sacramento River have counterparts in the fossil
record of the western United States. The present
and past distributions of eight of these species
are particularly informative in terms of under-
standing the floristic sources of the modern
forest. These include Acer negundo , Alnus rhombi-
folia, Fraxinus latif olia, Platanus racemosa,
Populus f remontii , Quercus lobata, Salix lasiandra,
and Salix lasiolepis .

The California sycamore, Platanus racemosa,
ranges in distribution from the upper reaches of
the Sacramento River southward into Baja Cali-
fornia (Griffin & Critchfield, 1972; Little,
1976). In the Central Valley, this species is
locally abundant along the Sacramento and San
Joaquin Rivers, ascending their main tributaries
to low elevations in the Sierran foothills. It
is notably absent from the North Coast Ranges and
the western side of the Sacramento Valley (Jepson,
1910). It is distributed throughout the South
Coast Ranges, where it is "one of the most widely
distributed aboreous species" (Jepson, 1910), and
occurs in the Transverse and Peninsular Ranges
of southern California (Griffin & Critchfield,
1972). The California sycamore is generally
confined to sites with an abundant water supply,
as along perennial streams, around springs, and
in moist gulches (Sudworth, 1908). Two distinct
late Tertiary species have been referred to the
modern P_. racemosa. To the north, Platanus dis-
secta is a characteristic species in the Miocene
floras of the Columbia Plateau and northern Great
Basin (Chaney & Axelrod, 1959). It survived into
the Pliocene in this region as evidenced by the
Dalles flora of Oregon (Chaney, 1944a) and the
Upper Ellensburg flora of Washington (Smiley,
1963). To the south, Platanus paucidentata is
a characteristic species in both the Miocene and
Pliocene floras of southern California (Axelrod,
1939, 1940, 1950c, d) . The distributions of
these two species overlapped in central Nevada
in the Miocene (Axelrod, 1956) and in central
California in the Miocene and Pliocene (see
Axelrod, 1944a, b; Renny, 1972). The question
arises as to which of these species is more
closely allied to the modern _P. racemosa. Judging
from the available record, it appears that P_.
paucidentata shows more definite relationship to
the modern species (Axelrod, 1939, 1956, 1967),
while P_. dissecta may be more nearly related to
the modern _P. orientalis of the Middle East or
P_. occidentalis of the eastern U.S. (Axelrod,

100

1956; Renney, 1972). The two fossil species
probably diverged from a common ancestor during
the early or middle Tertiary. Leaves of the
modern J?, racemosa appear in abundance in the
Pleistocene Soboba flora of southern California
(Axelrod, 1966).

Discussion

The evidence from these fossil floras sug-
gests that lowland riparian forests comparable
to that along the modern Sacramento River have
had a long and virtually continuous history in
the western United States during the last 20
million years. These widespread ancestral com-
munities showed regional variation as a conse-
quence of major climatic differences from north
to south. In southern regions, the riparian
communities originally included several species
with relatives in the modern forest (P. race-
mosa, P_. f remontii, and j3. lasiolepis) plus
numerous taxa now restricted to the summer-wet
region of the southwestern U.S. and Northern
Mexico. In contrast, the original riparian
communities of northern regions included several
other species with relatives in the modern
forest (A. negundo, A. rhombif olia, J_. latifo-
lia, 0. lobata, and j5. lasiandra) plus many
others now confined to the summer-wet regions
of the eastern U.S. and eastern Asia. It is in
the intermediate areas that we first see the
intermingling of these northern and southern
riparian taxa that is apparent in the modern
community. This is first evident in the in-
terior (Middlegate) , where the northward mi-
gration of southern taxa with spreading aridity
was apparently aided by the Sierra Nevadan rain-
shadow. This mixed type of community subsequen-
tly appeared on the western slopes of the Sierra
Nevada (Remington Hill) and disappeared from
western Nevada. It became well-established
over lowland west-central California by the
middle Pliocene (Mulholland) and persisted in
this region with some modifications down to the
present. In all of these regions, we see the
gradual loss of the exotic taxa in the communi-
ties as climate became progressively cooler,
drier, and less equable, and as summer rainfall
was reduced.

LITERATURE CITED

Axelrod, D.I. 1939. A miocene flora from the
western border of the Mohave Desert. Car-
negie Inst. Wash. Pub. 516:1-129.

• 1940. The Mint Canyon flora of

southern California: a preliminary state-
ment. Amer. Jour. Sci. 238:577-585.

• 1944a. The Black Hawk Ranch

flora. Carnegie Inst. Wash. Pub. 553:91-101.

• 1944b. The Mulholland flora.

Carnegie Inst. Wash. Pub. 553:103-145.

. 1950. The Piru Gorge flora of

southern California. Carnegie Inst. Wash.
Pub. 590:159-124.

. 1956. Mio-Pliocene floras from

west-central Nevada. Univ. Calif. Publ.
Geol. Sci. 33:1-316.

. 1966. The Pleistocene Soboba

flora of southern California. Univ. Calif.
Publ. Geol. Sci. 60:1-109.

. 1967a. Evolution of the Cali-

fornian closed-cone pine forest. In: Pro-
ceedings of the Symposium on the Biology of
the California Islands. (Ed. by R.N. Phil-
brick), pp. 93-150. Santa Barbara Botanical
Garden, Santa Barbara.

. 1967b. Geologic history of the

California insular flora. In: Proceedings
of the Symposium on the Biology of the Cali-
fornia Islands (Ed. by R.N. Philbrick) ,
pp. 267-316. Santa Barbara Botanical Garden,
Santa Barbara.

. 1968. Developments, trends, and

outlooks in paleontology. Late Tertiary
plants (Oligocene-Pliocene) . J. Paleont.
42:1358.

. 1977. Outline history of Cali-
fornia vegetation. In: Terrestrial Vege-
tation of California (Ed. by M.G. Barbour &
J. Major), pp. 139-193. Wiley-Inter science,
New York.

Chaney, R.W. 1944. The Dalles flora. Carnegie

Inst. Wash. Pub. 553:285-321.
Chaney, R.W. , and D.I. Axelrod, 1959. Miocene

floras of the Columbia Plateau. II. Syste-
matic considerations. Carnegie Inst. Wash.

Pub. 617:135-237.
Griffin, J.R. , and W.B. Critchfield. 1972. The

Distribution of Forest Trees in California.

U.S. Dept. of Agriculture, Berkeley.
Jepson, W.L. 1910. The Silva of California.

The University Press, Berkeley.
Little, E.L. 1976. Atlas of United States Trees.

Vol. 3. Minor Western Hardwoods. U.S. Dept.

of Agriculture, Washington, D.C.
Renney, K.M. 1972. The Miocene Temblor flora of

west-central California. M.Sc. Thesis,

University of California, Davis.
Smiley, C.J. 1963. The Ellensburg flora of

Washington. Univ. Calif. Publ. Geol. Sci.

35:157-275.

Sudworth, G.B. 1908. Forest Trees of the Pacific
Slope. U.S. Dept. of Agriculture, Washington,
D.C.

RIPARIAN FORESTS OF THE
SACRAMENTO VALLEY, CALIFORNIA

Kenneth Thompson
Department of Geography
University of California, Davis

Although edaphic and biotic influences pre-
cluded trees from most of the Sacramento Valley

101

in its pristine condition, the riparian lands
(Mainly natural levees) supported a flourishing
tree growth — valley oak, sycamore, cottonwood,
willow, and other species. A number of factors
contributed to their presence — principally sub-
irrigation, fertile alluvial loam soils, and
relative freedom from surface waterlogging and
fire. These riparian forests varied consider-
ably in width, from a narrow strip to several
miles. They also varied greatly in the spacing
of the trees, from irregular open to fairly
crowded stands, but were generally of sufficient
extent and closeness to justify the term "for-
est".

Pristine Condition of the Riparian Lands

Among the first outsiders to visit the
Sacramento Valley were fur trappers of the
Hudson's Bay Company in the period prior to
1814. The Spaniard Luis Antonio Arguello inves-
tigated the valley in 1817 and again in 1821,
and Jedediah Smith, in 1825, may have been the
first American to reach the Sacramento River.
However, it was not until the 1840' s that
significant outside influence was felt in the
northern end of the Central Valley. This
seclusion, however, could not survive the
meteoric developments of the Americanization of
California. After 1849 came a huge influx of
population, lured by gold but often quickly to
adopt other pursuits. These immigrants, mostly
with rural backgrounds, could not overlook the
agricultural promise of the Sacramento Valley.
Heightening the attractions of the Sacramento
Valley for agricultural settlement was its
virtually vacant condition. Its relatively
sparse and peaceful aboriginal population,
having been greatly reduced in numbers by an
epidemic in the early 1830' s was unable to
offer more than token resistance to the Ameri-
can invaders (Cook, 1955) .

After recognizing the promise of the Sacra-
mento Valley, the invading Americans quickly
set about its realization. To do this called
for new patterns of occupance and land use; and
in the initiation of these the environment was
substantially modified. The agencies of change
were sufficiently drastic to transform the
physical, biotic, and cultural landscape. One
of the very first transformations concerned the
natural levees and riparian lands, which were
thickly forested in their pristine condition.

Because of the brief period between initial
investigation and development, little informa-
tion was accumulated on the aboriginal condition
of the Sacramento Valley. One of the earliest
observers to report on the riparian forests was
John Work (in Mahoney, 1945) in the course of a
fur-trapping expedition from his headquarters
at Fort Vancouver. Writing in 1832, he de-

scribed the riparian forests of the Sacramento
Valley, below Red Bluff as follows:

All the way along the river here
there is a belt of woods principally
oak which is surrounded by a plain
with tufts of wood here and there which
extend to the foot of the mountain,
where the hills are again wooded.

Another early visitor to the Sacramento
Valley, Captain Sir Edward Belcher, R.N., noted
the profusion of oak, ash, plane, laurel, sumach
(sic), hiccory (sic), walnut, roses, wild grapes,
arbutus, and other small shrubs in the vicinity
of the river (Belcher, 1843). He described its
lower course as follows:

Having entered the Sacramento,
we soon found that it increased in
width as we advanced, and at our noon
station of the second day was about
one-third of a mile wide. The marshy
land now gave way to firm ground,
preserving its level in a most remark-
able manner, succeeded by banks well
wooded with oak, planes, ash, willow,
chestnut (sic), walnut, poplar, and
brushwood. Wild grapes in great abun-
dance overhung the lower trees, clus-
tering to the river, at times completely
overpowering the trees on which they
climbed, and producing beautiful
varieties of tint. . . . Our course lay
between banks. . . . These were, for
the most part, belted with willow, ash,
oak, or plane (Platanus racemosa) ,
which latter, of immense size, overhung
the stream, without apparently a suf-
ficient hold in the soil to support
them, so much had the force of the
stream denuded their roots.

Within, and at the very verge of
the banks, oaks of immense size were
plentiful. These appeared to form a
band on each side, about three hundred
yards in depth, and within (on the
immense park-like extent, which we
generally explored when landing for
positions) they were seen to be dis-
posed in clumps, which served to
relieve the eye, wandering over what
might otherwise be described as one
level plain or sea of grass. Several
of these oaks were examined, and some
of the small felled. The two most
remarkable measured respectively
twenty-seven feet and nineteen feet in
circumference, at three feet above
ground. The latter rose perpendicularly
at a (computed) height of sixty feet
before expanding its branches, and was
truly a noble sight.

102

Most of the historical reports give no in-
dication of the actual depth of the woodland.
Where Belcher examined the lower Sacramento
banks, probably the delta section, in 1837 he
noted a belt of large oaks (including one with
a trunk 27 feet in circumference at 3 feet
above the ground) "about three hundred yards in
depth" (Belcher 1843). John Work (Mahoney,
1945) in 1832, probably referring to French
Camp Creek, a Sierra stream that flows to the
delta, wrote: "the plain is overflowed and we
had to encamp at the skirt of the woods about
two miles from the river." Derby's report of
1849 (Farquhar, 1932) noted a two-mile-wide
belt of woods on both sides of the lower Feather
River. The map accompanying this report shows
forest bordering all the major and minor streams
in the lower Sacramento River system. Thus,
riparian forest seems to have bordered the
entire mapped portion of the river system from
the vicinity of Clarksburg in the south to
Glenn in the north. These riparian forests are
shown as being fairly uniform in width, about
four to five miles. Derby's map also shows
riparian forests along the tributary streams
almost equal in width to those of the main
stream, and flanking the tributaries to the
edge of the valley. On the Derby map Cache and
Putah creeks have forests about three miles
wide, the American and Feather rivers about
four miles wide (which checks with a section of
his report), and Butte Creek and Yuba and Bear
rivers each have levee forests about two miles
wide. A note of caution should be inserted
here. Derby, although a topographical engineer,
performed only a reconnaissance type of survey
of the valley. This being so, together with
the undoubted fact that the tree symbols are
intended to be approximate rather than precise,
his map should not be invested with undeserved
(and unintended) accuracy. However, even with
these limitations the Derby map does suggest
riparian forest of substantial width and conti-
nuity, and in 1849 these were, of course, still
virtually in their pristine condition.

It is highly improbable that the forest
belt was of uniform width along both banks of
the streams. Indeed, historical accounts
clearly indicate the irregular occurrence of the
trees. Belcher (1837) refers to the trees as
being "disposed in clumps." Derby also speaks
of "clusters of beautiful trees - oaks sycamore
and ash" on the banks of the Yuba River to
differentiate the forests there from those of
the Sacramento and Feather rivers, which were
"thickly wooded." Elsewhere he speaks of
riparian forests along the Feather River "dotted"
for two or three miles back from the river.

The Railroad Reports of a few years later
(1855) speak of the riparian forest as being a
"varying breadth, from a mile or more. . . to a
meager border. Even more generally, but clearly

indicating the variation of width in the riparian
forests, the Railroad Reports refer to the
riparian forests as "of greater or less width."
Moreover, the riparian forests varied not only
in width but also in tree size and density, "the
number and size of trees being apparently propor-
tioned to the size of the stream and the quantity
of moisture derived from it."

The preceding discussion shows that in their
pristine condition the streams of the lower
Sacramento River system were flanked by forests.
The historical evidence suggests that these
riparian forests had varied characteristics.
They included trees of all sizes, from brush to
very large valley oaks or sycamores, 75 to 100
feet high, growing closely spaced or scattered
irregularly in groves. On the banks of the
lower Sacramento, where the natural levees are
widest, the riparian forests achieved their
greatest width, four to five miles. On the
lesser streams and in the delta, with smaller
levees, the forests formed a narrower belt,
generally about two miles wide but less in the
delta. Dominant species in the riparian forest
were valley oak (Quercus lobata) , interior live
oak (Quercus wislizenii) , California sycamore
(Platanus racemosa) , Oregon Ash (Fraxinus ore-
gana) , Cottonwood (Populus f remontii) , alder
(Alnus rhombif olia) , and several willows, (Salix
gooddingii, S_. exigua, j>. Hindoiana, S^. Lasiandra,
and jS. Laevigata) .

Present Condition of the Riparian Forests

Although the Sacramento Valley riparian
forests were an early casualty of the white man,
their destruction, far-reaching as it was, was
not complete. Today, parts of both banks of the
Sacramento and its tributaries are bordered by
many shrunken remnants of the once extensive
riparian woodland. The numerous traces that
remain corroborate the historical evidence
examined by the author. The same tree species
mentioned in the historical records - mainly
valley oaks, cottonwoods, willows, sycamores,
and ash - still grow on the river banks, natural
levees, and channel ridges. Typically, cotton-
woods and willows predominate on the immediate
stream banks, whereas valley oaks are spread
irregularly over the natural levees farther away
from the river.

Instead of a strip measurable in miles, the
forested zones along the Sacramento Valley
streams are now often only yards deep, and dis-
continuous at that. Generally, the remaining
fragments (not necessarily virgin stands, of
course) form a belt less than 100 yards wide and
are largely confined to bank slopes of streams
and sloughs, abandoned meanders, and on the river
side of artificial levees.

103

Examination of the Sacramento River levees
reveals hundreds of larger relict stands of
riparian forest. Some cover only a few acres;
others several hundred. Most prominent are
fully mature specimens of valley oaks in the
"weeping" stage of development described by
Jepson (1893) as indicating an age between 125
and 300 years. Such trees occur mostly on
natural levee or channel ridge sites and are
frequently around older settlements, presumably
preserved for shade and ornament. Even small
house lots may contain two or more oaks that
predate the Anglo-American settlement period,
presumably relicts of a more extensive stand.
Some tracts of uncleared land near the Sacra-
mento River (including two in Yolo County be-
tween Knights Landing and Elkhorn Ferry) are
still so thickly studded with trees, including
many valley oaks in the "weeping" stage, that
they form the definite, if open, forest
described by early visitors to the region.

Near Woodson Bridge, Tehama County, another
expanse of apparently virgin riparian forest can
be seen. It is still subject to almost annual
overflow and is composed mainly of mature valley
oaks, forming an open woodland that extends
discontinuously for about a mile from the river's
edge. Some splendid mature specimens of valley
oak remaining from the Cache Creek riparian
forest can be seen in the older residential
sections of Woodland in Yolo County, which is
named for the fine oak forest in which the
settlement was established in 1855. Again, in
and around Davis, also in Yolo County, there
are many large relict oaks of the Putah Creek
Forests .

In view of the general lack of trees in the
Sacramento Valley, the riparian forests must
have served as a source of fuel, construction,
and other types of wood for a wide area. There
was doubtless little incentive to conserve the
riparian forests, since few of the tree species
have much value as lumber. Typically the ripar-
ian forest species are fit only for low eco-
nomic uses. For example, the numerous members
of the genus Salix (willow) generally yield
soft, light, and brittle wood of poor form for
saw timber. Rather similar is the cottonwood,
which is soft, brittle, not durable, and espe-
cially liable to cracking. The largest, and
probably most numerous, riparian tree, the
valley oak, is "very brittle, firm, often cross-
grained and difficult to split or work. On
account of its poor timber form the trees are
rarely if ever cut for anything but fuel, for
which, however, they are much used" (Sudworth,
1908).

The clearing of the riparian forest for
fuel and construction also served another end:
it made available for agricultural use some of
the most fertile and easily managed land in the

valley. In its pristine, or nearly pristine
condition, much of the valley was more or less
unusable for agriculture because of waterlogging
and inundations. The original limitations of
many valley areas have been partially overcome
in recent decades with improved drainage, irri-
gation, and other technical advances. However,
initially these limitations were such as to
discourage permanent settlement and agriculture
on much of the valley floor with the exception
of the natural levee lands. There both settle-
ment and cultivation were concentrated; utili-
zation of the remainder of the valley was un-
certain and irregular, with much attention paid
to livestock raising. The general superiority
of the levee lands still holds. The most profi-
table form of land use in the valley, orchards,
shows a very marked concentration on levee soils,
a final confirmation of their inherent suitability
for tree growth.

Perhaps because the riparian forests were
largely effaced during the first two or three
decades of Anglo-American occupance, their exis-
tence is largely overlooked by modern students
of the Sacramento Valley. But this neglected
element in the landscape is by no means of
negligible importance. The riparian trees
served to reinforce the river banks and provide
greater stability to the stream channels. They
also acted as windbreaks, reducing evaporation,
transpiration, and wind damage. In addition, the
riparian forests provided a haven for the wild-
life of the valley, furnishing cover and food
sources for land and arboreal animals. Even
more important was the fact that acorns, mainly
from Quercus lobata, were a staple foodstuff of
the Indian population. Furthermore, the forests
furnished an important source of wood in an area
otherwise poorly supplied.

The mere existence of the riparian forests,
however, inevitably spelled their doom. The
conditions, characteristic of natural levee
sites, that permitted their development — compara-
tive freedom from flood and waterlogging, high
soil fertility, and favorable soil moisture -
eventually led to their destruction, for the
existence of the forest was incompatible with
the modes of land use initiated by the Anglo-
Americans. Today, only a few traces of the
formerly extensive riparian forests remain, and
the Sacramento Valley exhibits a striking lack
of trees.

LITERATURE CITED

Belcher, R.N., "Narrative of a Voyage Round the
World Performed in Her Majesty's Ship Sulphur
During the Years 1836-1842". Vol. I (London:
Henry Colburn, 1843), p. 130.

Cook, S.F. "The Epidemic of 1830-1833 in Cali-
fornia and Oregon". University of California

104

Publications in American Archaeology and

Ethnology, Vol. 43, No. 3 (1955).
Farquhar, Francis P. (ed.), "The Topographical

Reports of Lieutenant George H. Derby,"

California Historical Society Quarterly, Vol.

II (1932), p. 115.
Jepson, Willis L. "The Riparian Botany of the

Lower Sacramento" (Erythea Vol. I 1893, p.

242).

Mahoney, Alice B. (ed.) Fur Brigade to the
Bonaventura, John Work's California Expedition
1832-33 for the Hudson's Bay Company (San
Francisco: California Historical Society,
1945), p. 18.

Sudworth, George B. Forest Trees of the Pacific
Slope (Washington, D.C.: Department of
Agriculture, 1908), pp 212-278.

THE FLUVIAL SYSTEM:
SELECTED OBSERVATIONS

Edward A. Keller
Environmental Studies and
Department of Geological Sciences
University of California, Santa Barbara
Santa Barbara, California 93106

Human use and interest in the riverine
environment extends back to earliest recorded
history. We have used the river system as an
avenue for transportation and communication, a
water supply, a waste disposal site, and a
source of power. Massive dams and channel works
to dissipate the disastrous effects of floods
and droughts have been constructed, and even
though we can sometimes control a river we
still know little about the processes which
form and maintain the natural fluvial system.
Only recently have we realized that rivers are
natural resources that must be conserved and
properly managed if we are to continue a
meaningful existence.

The natural stream channel generally has
sufficient discharge to emerge from its banks
and flood areas adjacent to its banks on the
average of once every year or two. It is this
natural process of overbank flow which slowly
but relentlessly builds floodplain features
such as natural levees along the stream channels.
The overbank flows also supply water to adja-
cent lowlands on the floodplain which serve as
a storage site for excess runoff, much of which
may enter the groundwater system. A main philo-
sophical concession that must be recognized by
more communities which compete with the river-
ine environment is that overbank flow (flooding)
is a natural process rather than a natural
hazard and that, if we are to maintain the
integrity of the riverine system, we must
consider the channel and floodplain as a comple-
mentary system.

Human use and interest in the fluvial en-
vironment has historically included significant
drainage modification. This modification —
whether termed channelization, channel works, or
channel improvement — generally is controversial
because of its potential adverse effects on the
biological communities in the riverine environ-
ment. The loss of fish and wildlife habitat
due to channel modification generally leads to
simplification with less variation in the bio-
logical communities of the fluvial environment.

The reduced variability of the biological
community in response to channel modification
is directly attributed to the loss of variability
in the physical environment. That is, stream
channel modification tends to reduce the diversity
of flow conditions, the diversity of bed-material
distribution, and the diversity of bed forms.
If environmental deterioration caused by stream
channel modification is to be minimized then
new design criteria must be developed such that
the stream's natural tendency to converge and
diverge flow and sort the bed material is main-
tained. That is, we must apply environmental
determinism or "designing with nature" to our
channel works if we are to maintain a quality
fluvial environment.

The natural fluvial environment is an open
system in which the channel-f loodplain form and
processes evolve in harmony. Significant changes
in the fluvial system often occur when a geo-
morphic or hydraulic threshold is exceeded.
These changes are partly responsible for main-
taining the quasi- or dynamic equilibrium state
of the stream system. Human use and interest in
the fluvial environment has led to human inter-
ference with the fluvial system. This inter-
ference generally reduces the channel, flood-
plain and hydraulic variability and thus the
biologic variability which depends on the physical
environment.

The behavior of natural streams is not
completely understood. Particularly important
is the need to know more about relationships
between erosion, deposition, and sediment con-
centration, as well as the effect of organic
debris on stream channel morphology. In addi-
tion, if we are going to understand more about
relationships between the biology of stream
channels and the geomorphology , then we must
begin to study complex interactions between the
two. That is, we must learn more about processes
which produce channel morphology necessary for
biological productivity and thresholds that
control the maintenance and development of the
physical and biological environment.

105

RIPARIAN VEGETATION AND FLORA
OF THE SACRAMENTO VALLEY

Susan G. Conard
Ronald L. MacDonald
Robert F. Holland
Department of Botany
University of California, Davis

Research on Sacramento Valley riparian
vegetation has primarily concerned land-use
patterns (McGill, 1975; Brumley, 1976) or dis-
tribution and ecology of birds and mammals in
riparian habitat (Stone, 1976: Michny, Boos and
Wernette, 1975; Brumley, 1976). These studies
frequently include partial floristic lists or
brief vegetation descriptions. Michny et_ al.
(1975) provide quantitative vegetation data at
each nine study sites. Most of these stands
were apparently less than 15 m. in width and
several were highly disturbed.

The objectives of this study were 1) to
obtain preliminary floristic and vegetation data
on several major riparian vegetation types, and
2) to use this data to a) delineate important
vegetation units, b) describe structure of
mature stands of riparian forest, and c)
describe major serai and topographic relation-
ships within the riparian vegetation.

The major riparian vegetation types were
1) Valley oak woodland, 2) Riparian forest
dominated by cottonwood, 3) Gravel bar thickets,
4) Open floodplain communities, 5) Hydric com-
munities .

Valley Oak Woodland

The valley oak phase of the riparian forest
is typical of high terrace deposits and cut
banks along the outside of meanders. These
forests are dominated almost exclusively by
Valley oak (Quercus lobata) . Common associates
include Sycamore (Platanus racemosa) , willows
(Salix spp.), Box elder (Acer negundo) , Oregon
ash (Fraxinus latif olia) and Black walnut
( Juglans hindsii) . Canopy height is 15-20 m.
and tree cover ranges from 30-60%. A typical
valley oak woodland sampled at the Cosumnes
site had a density of 124.5 trees/ha and basal
area of 18.35 m /ha. The relative density of
C£. lobata in this stand is .73 and its relative
basal area is .81, indicating strong dominance
by lobata at this site.

Valley oak woodlands are characteristically
heterogeneous with areas of high density,
smaller trees interspersed with more open areas
of larger trees. Openings contain typical
grassland species of genera such as Avena,

Lolium, Hordeum and Elymus. Where tree cover is
higher, the understory is characterized by poison
hemlock (Conium maculatum) , poison oak (Rhus
diversiloba) , ripgut brome (Bromus diandrus) ,
soap plant (Chlorogalum pomeridianum) , several
species of Carex and Erigeron sp.

Riparian Forest

Cottonwood (Populus f remontii) dominates the
riparian forest of lower terrace deposits and
stabilized gravel bars along the Sacramento
River. Common associates are similar to those
in the valley oak woodland including willows
(Salix lasiolepis, S_. goddingii, S^. laevigata,
j>. lasiandra) , Fraxinus latif olia, Acer negundo,
Juglans hindsii, and, on higher ground, Quercus
lobata and Platanus racemosa. Canopy height is
approximately 30 m. in a mature riparian forest,
with a tree cover of 20-30%. Tree density in
these forests is about 250 stems/ha — double that
of the valley oak woodland sampled. Basal area
is about 50 m^/ha. The relative basal area of
Populus fremontii is .75, reflecting its high
dominance in the vegetation. The low relative
density (.33-. 44) of cottonwood in these stands
reflects the large number of small subcanopy
(10-12 m) trees (particularly Acer negundo,
Fraxinus latif olia, and Salix spp) . Understory
species are mostly shrubs (Sambucus mexicana,
Cephalanthus occidentalis, Rubus spp, Rosa Cali-
fornica) . Lianas such as Rhus diversiloba and
Vitis calif ornica are a dominant feature, fre-
quently providing 30-50% ground cover and fes-
tooning trees to heights of 20-30 m. Herbaceous
vetetation is < 1% cover except in openings where
species such as Artemisia douglasiana , Urtica
dioica, and various shade tolerant grasses may
occur.

Gravel Bar Thickets

Well-stabilized gravel bar deposits are
dominated by sand bar willow (Salix hindsiana)
which forms dense thickets 3-5 m. tall of up to
95% cover. Common associates include saplings
of Alnus rhombif olia, Acer negundo, Fraxinus
latif olia , and Populus fremontii, and shrubs of
mule fat (Baccharis viminea) . Scattered her-
baceous species are also present but cover is
generally low due to the dense canopy.

Open Floodplain Communities

Sand and gravel bars which are flooded
annually support a sparse vegetation cover (5-
25%) dominated by small (1 m) shrubby and her-
baceous perennials and annuals. The frequent
disturbance normal to this habitat has favored
invasion by many introduced species such as
Bromus diandrus, B.- tectorum, Salsola kali,
Raphanus and Brassica spp, Tunica prolif era,

106

Polypogon monspeliensis , and Verbascum thapsus.
Native species of floodplains include the small
shrubs Chrysopis oregona, Trichostema lanatum,
and Bidens laevis.

Hydric Communities

In old oxbows and low areas a series of
hydric communities occurs. Open water supports
emergent and free-floating mat vegetation con-
taining plants such as Polygonum hydropiperoides ,
_P. coccineum, Ludwegia peploides , Azolla f ilicu-
loides , Potamogeton crispus , Elodea spp, and
Myriophyllum spicatum ssp exalbescens. Shallow
water and low mud flats are dominated by Scirpus
acutus (50-100% cover) 2-3 m. tall. On higher
areas, where Scirpus acutus is less dominant,
the species diversity of the fresh water marsh
increases considerably. Hummocks in higher
areas of the marsh support shrub thickets of
Cornus stolonif era, Cephalanthus occidentalis ,
Rubus vitif olius with occasional Alnus rhomb i-
f olia and Fraxinus latif olia. It is also in
this zone that the rare Hibiscus calif ornica
i may be found. The Cornus and Cephalanthus
hummocks are in turn invaded by understory
(Alnus rhombif olia , Salix spp, Fraxinus lati-
f olia, Rubus vitif olia, Rosa calif ornica)
species typical of the riparian forest, as well
as Populus f remontii. This turns higher hum-
mocks into Alnus dominated thickets and even-
tually Populus forests.

The riparian zone is a dynamic habitat:
the vegetation of a given site reflects the
history of flooding, aggradation, and degrada-
tion by the river. These habitats are subject
to varying frequencies of flooding and of lat-
eral erosion by the meandering river. The
major riparian plant communities can be aligned
along topographic gradients. The low, recent,
gravel bar deposits are flooded frequently.
Plant cover is low and is dominated by intro-
duced annuals and low perennials. As gravel
bars become more removed from the river and
begin to stabilize, they are colonized by
thickets of tall shrub and tree saplings
generally dominated by Salix hinds iana. Ripar-
ian forest will become established (on lower
terrace deposits) as flood frequency decreases.
These junglelike gallery forests are dominated
by Populus f remontii and characterized by heavy
cover of lianas. Higher ground in these forests
supports Quercus lobata and Platanus racemosa.
The older, higher terrace deposits support
stands of valley oak woodland dominated by (£.
lobata. These woodlands gradually thin out and
grade into valley grassland vegetation with in-
creasing distance from the river.

Oxbows and overflow basins are character-
ized by a series of hydric communities. Fresh
water marsh in low, wet areas is dominated by

Scirpus acutus. On higher ground, this is
succeeded by shrubs such as Cornus Stolonif era
and Cephalanthus occidentalis. These shrub-
dominated habitats appear transitional to typi-
cal Populus f remontii dominated riparian forests
on higher ground.

LITERATURE CITED

Brumley, Terry D. 1976. Upper Butte Basin
Study 1974-1975. State of California
Resources Agency, Wildlife Management Branch.
Admin. Report No. 76-1. 30 pp. + Appendix.

McGill, Robert R. , Jr. 1975. Land use changes
in the Sacramento River riparian zone, Redding
to Colusa. State of Cal. Resources Agency,
Department of Water Resources. April, 1975.
23 pp.

Michny, Frank J., David Boos, and Frank Wernette
1975. Riparian habitats and avian densities
along the Sacramento River. Cal. Resources
Agency, Dept. of Fish and Game. Admin. Rpt.
No. 75-1. March, 1975. 42 pp.

Stone, Thomas B. 1976. Birds in riparian

habitat of the upper Sacramento River. State
of Cal. Resources Agency, Dept. of Fish and
Game. Memorandum Report. Nov. 1976. 22 pp.
+ Appendix.

THE VALLEY RIPARIAN FORESTS OF CALIFORNIA:
THEIR IMPORTANCE TO BIRD POPULATIONS

David A. Gaines
Institute of Ecology
University of California, Davis

Those who have heard the spring chorus of
songbirds, watched herons feed their young in
tree-top nests, glimpsed swarms of warblers in
the early autumn greenery and tried to count
wintering flocks of sparrows know first-hand the
wealth and diversity of California's valley
riparian forest avifauna. Today, with the last
extensive remnants of these forests in jeopardy,
it behooves us to weigh the importance of
riparian habitat to birds and other wildlife
(Gaines 1976).

Diversity

California's riparian forests support a
high diversity of breeding birds (Miller 1951) .
Excluding Ring-necked Pheasant and Western
Meadowlark (included because some census plots
edge on grassland) 67 species are known to
nest in the forests of the Sacramento Valley.
Species richness (number of species) equals or
exceeds that in any habitat for which census
data is available (Gaines 1974b) . Using the
Shannon-Weaver species diversity index, the
average species diversity for the cottonwood-

107

willow census plots (3.17) is considerably
higher than that for the oak-cottonwood plots
(2.51). Species richness, however, is only
slightly higher (27 to 24). Thus the high di-
versity values in cottonwood-willow reflect a
large number of species with relatively even
densities. This high diversity seems to depend,
not on edge effect or plant species diversity,
but on foliage volume and foliage height profile.
One of the most interesting census results is
the lack of correlation of diversity with the
extent that riparian forest habitat edges on
openings or other types of vegetation. Most
species are more or less evenly dispersed with-
in the forest with little or no tendency to
concentrate near the edge (DeSante 1972) . Thus
the theory that diversity is enhanced by the
mixture of species from adjacent habitats may
not apply to riparian forests.

Beginning with MacArthur and MacArthur
(1961) a series of studies has linked bird
species diversity in forest communities with
foliage height diversity, foliage volume, and
other habitat characteristics. This complex,
fascinating subject has recently been summarized
by Balda (1975) . In addition to foliage, such
factors as food resources, nest sites, nesting
material, song posts, proximity to water, extent
of habitat, geological history, and human dis-
turbance need to be considered. Understanding
these factors is important to assuring a diverse
avifauna in sanctuaries, state parks, and other
lands set aside as riparian forest preserves.

The percentage of breeding individuals
which are migratory differs strikingly between
the cottonwood-willow and oak-cottonwood census
plots. In the former a large influx of birds
which winter in subtropical areas, such as
Western Wood Pewee, Yellow Warbler, and Northern
(Bullock's) Oriole, account for 36% of the
nesting bird density. In the valley oak forest,
in contrast, only 4% of the nesting birds are
migratory. Moister conditions in the cotton-
wood-willow forests may promote lusher plant
growth, higher invertebrate populations and,
therefore, more available food for flycatchers,
warblers, and other migratory, insectivorous
birds.

Based on Miller's (1951) analysis of the
California avifauna, 43% of the species and 38%
of the individuals breeding in cottonwood-willow
habitat have a "primary affinity" to riparian
forest (Table 1). In other words, in compari-
son to 21 other California vegetation types,
these forests probably support the highest con-
centrations of these species. In cismontane
California Red-shouldered Hawk, Yellow-billed
Cuckoo, Willow Flycatcher, Bell's Vireo, Yellow
Warbler, Yellow-breasted Chat, and Blue Gros-
beak breed in no other forest habitat.

The breeding avifauna of California's
riparian forests has intriguing affinities to
that of the similarly winter-deciduous hardwood
forests of eastern North America (Miller 1951).
Many typically "Eastern" or "Mid-eastern"
species, such as Red-shouldered Hawk, Yellow-
Billed Cuckoo, Downy Woodpecker, Bell's Vireo,
Warbling Vireo, Yellow Warbler, Yellow-breasted
Chat, Blue Grosbeak, American Goldfinch, and
Song Sparrow, have been able to colonize the
arid West primarily because humid, broad-leaved
riparian forests offered congenial haunts.
Interestingly, all of these birds have evolved
western subspecies (American Ornithologist's
Union 1957). Three of these races, the Red-
shouldered Hawk Buteo lineatus elegans, the
Bell's Vireo Vireo belli pusillus, and the Blue
Grosbeak Guiraca caerulea salicaria, breed only
in the valleys of California.

Breeding Densities

The average density of nesting birds on
the cottonwood-willow census plots (2088/km^)
is strikingly higher than that on the oak-cotton-
wood plots (1279/km2) . This difference is due
primarily to migratory species. If we only
consider residents, the density in cottonwood-
willow (1336/km2) is only slightly higher than
that in oak (1227/km2). Breeding bird densities
in cottonwood-willow forests equal or exceed
those in any California vegetation type for
which census data is available (Gaines 1974).
The dense, stratified cottonwood-willow forest
vegetation may facilitate these high breeding
bird densities. With increased trunk, branch,
and foliage foraging space, bird territories
may occupy less ground surface area.

The large number of migratory birds implies
a seasonal abundance of insect food during the
warmer months (DeSante 1972). A recent study,
however, suggests that bird densities do not
depend on habitat productivity (Willson 1974) .
In this regard it would be interesting to try
to correlate bird densities in riparian forest
habitats with plant productivity and inverte-
brate populations.

Wintering Densities

The average density of wintering birds on
the valley oak plots (2439/km2) is strikingly
higher than that on the cottonwood-willow plots
(997/km2) . It is interesting to compare these
figures with breeding bird densities. The data
suggests that oak forests support 90% more
wintering than nesting birds, and cottonwood-
willow forests almost the reverse. This same
trend, although less pronounced, is reflected
by the data on species richness and diversity.

108

These seasonal changes are due primarily
to migrants. The large number of breeding birds
which leave cottonwood-willow forests before the
autumn leaf-fall deplete wintering bird densi-
ties. In the oak forests, in contrast, a large
influx of migratory wintering species augments
the largely resident breeding population. Most
(69%) of these migrant birds subsist on seeds
and/or fruits. More open conditions in the oak
forests may promote the growth of herbaceous,
seed-producing forbs and grasses. Berry-pro-
ducing plants are probably more abundant . The
available census data suggests that average
bird density in oak riparian forest exceeds
that in coastal mixed forests, coastal coni-
ferous forests and chaparral (Stewart 1972).
Wintering bird diversity is also high; 60
species are known to winter in the riparian
forests of the Sacramento Valley.

Migration

Large numbers of passerine birds forage
and shelter in riparian forest habitat during
their migratory journeys. Most are foliage-
gleaning or sallying insectivorous species
which winter in subtropical Mexico and Central
America. During the spring migration, these
birds pass northwards on a broad front through
the forests and woodlands of lowland California.
The hills are green, the deciduous foothill
oaks have just leafed out, and insect life is
everywhere abundant. By late summer, however,
the long dry period has seared the hills to
golden brown. At this season riparian forests
provide the only lush, insect-rich forest habi-
tat in lowland, cismontane California. The
importance of these forests to southward (fall)
migrants cannot be underestimated.

An Endangered Habitat

Nothing better illustrates the destruction
of riparian forest habitat than the decline in
Californian populations of the Yellow-billed
Cuckoo. This sinuous bird is closely restrict-
ed to broad expanses of cottonwood-willow
forest. In the early part of this century the
clearing of these forests was recognized as a
threat to the cuckoo's survival (Jay 1911). At
that time, they were still "fairly common"
(Grinnell 1915). Only three decades later, how-
ever, Grinnell and Miller (1944) concluded that
"because of removal widely of essential habitat
conditions, this bird is now wanting in exten-
sive areas where once found." Recent studies
have confirmed this gloomy picture. Only in
the relatively large remnants of forest that
hug the Sacramento River between Colusa and
Red Bluff are a few pairs still known to nest
within cismontane California (Gaines 1974b) .

Over most of this area, once extensive
riparian forest habitat has been sacrificed to
civilization. The Santa Ana River in the San
Bernardino Valley of Southern California is an
excellent example. Here the Yellow-billed
Cuckoo was first discovered nesting in Cali-
fornia by Stephens in 1882 (Bendire 1895).
During the 1920' s Hanna (1937) found 24 nests
in the "miles of cottonwood and willow" watered
by the river. "In contrast with those good
old days," he writes, "we now have very little
water in Warm Creek and seldom any surface
water in the Santa Ana River, the large thicket
have been replaced by farms and pastures, the
trees cut down, and the evergrowing population
has crowded in on the old haunts of the cuckoos
to such an extent that if they come here now at
all they must be exceedingly rare."

In California, as throughout western North
America, the last remaining groves of valley
riparian forest are in jeopardy. Each year
more of these forests are bulldozed and cut
for pulpwood, or to make way for orchards,
gravel extraction, rip-rap bank protection and
urban development. Unless immediate measures
are taken, this endangered habitat will no
longer provide a home for the Yellow-billed
Cuckoo and the many other birds and animals
which dwell there.

As Eleanor Pugh (1965) recognized a decade
ago, the choice is ours. "As long as housing
tracts start landscaping from bare soil," she
writes, "rather than plan around existing
mature willows, cottonwoods, sycamores and oaks
with their entangled undergrowths so rich in
the shyer birds; as long as willow shrub ripar-
ian cover is scraped away and replaced with
ubly concrete channeling, breeding success will
be low for many species . . . small wonder that
Willow Flycatchers, Swainson's Thrushes, Yellow
throats, Yellow Warblers and Yellow-breasted
Chats, though quite adaptive and once numerous,
are becoming a rare sight to behold or even
hear above the roar of traffic on the nearby
freeway. "

109

Table 1. The breeding riparian forest avifauna
of the Sacramento Valley, California.

Table 1, continued.

Species

CO 4J

U C

cd *rH _

(X m

Guild

2 ^ Foraging Nesting

Double-crested res?
Cormorant

(Phalacrocorax
auritus)

Great Blue Heron res
(Ardea herodias )

Green Heron mig
(Butorides
virescens)

Great Egret res?
(Casmerodius albus)

Wood Duck res
(Aix sponsa)

Common Merganser res?
(Mergus merganser)

Turkey Vluture
(Cathartes aura)

White-tailed Kite
(Elanus leucurus)

Cooper's Hawk
(Accipiter cooper i)

Red- tailed Hawk
(Buteo jamaicensis)

Red- shouldered Hawk
(Buteo lineatus)

Swain son's Hawk mig
(Buteo swainsoni)

Bald Eagle res
(Haliaeetus
leucocephalus)

Osprey mig
(Pandion haliaetus)

tree

tree
hole*

mig 8 ground tree
carrion stump

res 1 ground tree
mammal

res 1 foliage tree
bird

res 5 ground tree
mammal

res 1 ground tree
mammal

tree

"^res = resident; mig = migratory.

^scale 1-8; 1 = primary affinity; 8 =
species breeds in greater density in 7 other
habitats (Miller 1951).

*does not excavate tree hole nesting cavity.
**does excavate tree hole nesting cavity

Species

c >>

•rH *H
CO *H

a. 14-1

•H 14-1

Guild

2 5 Foraging Nesting

American Kestrel res
(Falco sparverius)

California Quail res
(Lophortyx calif ornicus)

Ring-necked Pheasant res
(Phasianus colchicus)

Mourning Dove mig
(Zenaida macroura)

Yellow-billed Cuckoo mig
(Coccyzus americanus)

Screech Owl res
(Otus asio)

Great Horned Owl res
(Bubo virginianus)

Long-eared Owl res?
(Asio otus)

Anna's Hummingbird res?
(Calypte anna)

Black-chinned mig
Hummingbird
(Archilochus alexandri)

Common Flicker res
(Colaptes auratus)

Acorn Woodpecker res
(Melanerpes f ormicivorus)

Downy Woodpecker res
(Picoides pubescens)

Nuttall's Woodpecker res
(Picoides nuttalli)

Western Kingbird mig
(Tyr annus verticalis)

Ash-throated mig
Flycatcher
(Myiarchus cinerascens)

Black Phoebe res
(Sayornis nigricans)

Willow Flycatcher mig
(Empidonax thraili)

4 ground
insect

ground
seed

3 ground
seed

1 foliage
insect

2 ground
insect?

4 ground
mammal

1 ground
mammal

foliage
nectar

1 foliage
nectar

1 bark
insect

- air
insect

tree
hole*

ground

tree

tree
hole*

tree

1 ground tree

insect hole**

foliage tree

seed hole**

tree
hole**

2 bark tree
insect hole**

tree

5 air tree
insect hole*

2 air
insect

1 air tree
insect

110

Table 1, continued.

Species

c >>

(fl 4-1

•H -H

4-1

M C

(0 *H

CU <4-l

C/3

•H M— l

Guild

^ <3 Foraging Nesting

Western Wood Pewee mig
(Contopus sordidulus)

Tree Swallow mig
(Iridoprocne bicolor)

Purple Martin mig
(Progne subis)

Scrub Jay res
(Aphelocoma
coerulescens)

Yellow-billed Magpie res
(Pica nuttalli)

Plain Titmouse
(Parus inornatus)

res

Bushtit res
(Psaltriparus minimus)

White-breasted
Nuthatch
(Sitta carolinensis)

res

4 air
insect

1 air
insect

- air
insect

genera-
list om-
nivore

4 genera-
list om-
nivore

- bark
insect

4 foliage
insect

- bark
insect

tree

tree
hole*

tree

tree
hole*

tree

tree
hole*

Table 1, continued.

Species

tn n) 4-i
4-1 u C

a. u-i .

Guild

•H 14-1

a! <

Foraging Nesting

Bell's Vireo
(Vireo bellii)

Warbling Vireo
(Vireo gilvus)

mig

Yellow Warbler mig
(Dendroica petechia)

Common Yellowthroat mig
(Geothlypis trichas)

Yellow-breasted Chat mig
(Icteria virens)

House Sparrow res
(Passer domesticus)

Western Meadowlark res
(Sturnella neglecta)

Northern Oriole mig
(Icterus galbula)

Brown-headed Cowbird mig
(Molothrus ater )

foliage tree
insect

foliage shrub
insect

ground
seed

ground ground
insect

foliage tree
insect

ground -
seed

Wrentit res - foliage

(Chaemaea f asciata) insect

House Wren mig 2 foliage

(Troglodytes aedon) insect

Bewick' s Wren res 3 foliage

(Thryomanes bewickii) Insect

Mockingbird res - foliage

(Mimus polyglottos) insect

California Thrasher res - ground
(Toxostoma redivivum) insect

shrub

tree
hole*

tree

shrub

Black-headed Grosbeak mig ;
(Pheucticus melanocephalus)

Blue Grosbeak mig
(Guiraca caerulea)

Lazuli Bunting mig '.

(Passerina amoena)

House Finch res (

(Carpodacus mexicanus)

American Goldfinch res?
(Carduelis tristis)

foliage tree
insect

foliage shrub
insect

ground
seed

tree

foliage tree
seed

American Robin res?
(Turdus migratorius)

Swainson' s Thrush mig
(Catharus ustulata)

Blue-gray Gnatcatcher mig
(Polioptila caerulea)

European Starling res
(Sturnus vulgaris)

Hutton's Vireo
(Vireo huttoni)

res i

6 ground
insect

1 ground
insect

4 foliage
insect

- genera-
list om-

3 foliage
insect

tree

tree
hole*

tree

Lesser Goldfinch res?
(Carduelis psaltria)

Rufous-sided Towhee res
(Pipilo erythrophthalmus)

Brown Towhee
(Pipilo fuscus )

Lark Sparrow res?
(Chondestes grammacus)

ground
seed

ground

seed

insect

ground

seed

insect

ground

seed

insect

tree

ground

shrub

ground

111

Table 1, continued.

•H -H

u c

Guild

Species

*** Foraging Nesting

Song Sparrow
(Melospiza melodia)

ground

seed

insect

shrub

HABITATS OF NATIVE FISHES IN THE
SACRAMENTO RIVER BASIN

Donald Alley
David H. Dettman
Hiram W. Li
Peter B. Moyle
Wildlife and Fisheries Biology
University of California, Davis

LITERATURE CITED

American Ornithologist's Union 1957. Check-list
of North American Birds. Fifth Edition.

Balda, R.P. 1975. Vegetational structure and
breeding bird diversity. Proceedings Sympo-
sium on management forest and range habitats
for nongame birds. U.S.D.A. Forest Service
technical report WO-1: 59-80.

Bendire, C.E. 1895. Life histories of North
American birds. U.S. Nat. Mus. Spec. Bull. 3.

Desante, D. 1972. Breeding bird census.
Riparian willow woodland. Amer. Birds 26:
1002-1003.

Gaines, D. 1974. A new look at the nesting
riparian avifauna of the Sacramento Valley,
California. Western Birds 5:61-80.

. ed. 1976. Abstracts from the con-
ference on the riparian forests of the Sacra-
mento Valley. 25 pp. California Syllabus,
Oakland, Calif.

Grinnell, J. 1915. A distributional list of
the birds of California. Pac. Coast Avifauna
11.

Grinnell, J. and A.H. Miller. 1944. The dis-
tribution of the birds of California. Pac
Coast Avifauna 27.

Hanna, W.C. 1937. California Cuckoo in the San
Bernardino Valley, California. Condor 39:
57-59.

Jay, A. 1911. Nesting of the California
Cuckoo in Los Angeles County, California.
Condor 13:69-73.

MacArthur, R.H. and J.W. MacArthur. 1961. On
bird species diversity. Ecology 42:594-598.

Miller, A.H. 1951. An analysis of the distri-
bution of the birds of California. Univ. of
Calif. Publ. Zool. 50:531-643.

Pugh, E.A. 1965. Southern Pacific Coast region
report. Audobon Field Notes. 19:577.

Stewart, R.M. 1972. A summary of bird surveys
in California. Pt. Reyes Bird Observatory
Newsletter 21:3.

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

Fish habitat in the Sacramento River Basin
has been degraded severely through placer mining,
dredging, wetland reclamation, destruction of
stream side vegetation, livestock grazing, lumber
operations, irrigation and water diversion, dams,
stream channelization and bank stabilization,
dewatering, and domestic pollution. The general
effect has been severe. Several species are now
so rare as to be virtually extinct. Salmon and
steelhead runs are a fraction of previously
recorded levels. But the specific effect on
many species is unclear because historically the
study of the ecology of native species was un-
fashionable. Through survey and experimental
studies we have been trying to reconstruct habi-
tat requirements and preferences of native fishes
in order to estimate the impact of human activi-
ties.

The task of understanding the stream fish
communities has been difficult because the
streams of California have been badly disturbed.
The destruction of the riparian forests has
been only one part of this perturbation, although
one of the most visible. One of the first major
disturbances was placer mining which destroyed
salmonid spawning grounds, increased siltation,
removed or covered up riparian vegetation, and
drastically changed stream morphology. As agri-
culture became more and more important to Cali-
fornia's economy the deterioration of aquatic
habitats continued (and continues) at an ever-
increasing rate. Then, as irrigation and flood
control became necessary, channelization of
streams started to become as common as did
irrigation diversions and the construction of
bypasses for flood waters. Channelization
consists of vegetation removal, straightening
channels (thus removing meanders) , dredging the
stream bed and stabilizing the banks with loose
material (riprapping) . This type of habitat
alteration has been well documented in terms of
its effect (Whitney and Baily 1959, Peters and
Alfond 1964, Funk and Ruhr 1971, Barton et al.
1972, Moyle 1976a). Essentially, the environment
has been simplified: cover by stream side vege-
tation is removed, pools are eliminated, and
undercut banks are destroyed. The substrate is
made more uniform as snags and fallen logs are
removed. As expected, species richness and
standing crops diminish as a result. Irrigation
diversions and flood bypasses often divert migra-

112

tory young of anadromous fishes from the main
streams. The degree of impact of these diver-
sions is not presently known; however, there is
some concern that substantial mortality of
young may contribute to declining chinook salmon
and steelhead runs. Dewatering streams for
irrigation also reduces flows, which triggers a
series of changes: water temperatures increase,
current is reduced, silt deposition increases,
dissolved oxygen decreases, the stream becomes
more shallow, and finally production decreases.

Bad forestry practices can lead to severe
problems very similar to overgrazing of live-
stock. Small streams are often used as chutes
to transport downed trees, badly damaging banks
and substrate. If slash is dumped into creeks,
this will cause dams to form which will impede
spawning migrations, decrease flow, and increase
siltation. Major problems in logging areas have
also been caused by poorly designed roads which
often follow stream courses. Such roads can
accelerate soil erosion tremendously which
dramatically increases the silt burden of the
stream (Platts and Megahan 1975, Megahan and
Kidd 1972, Arnold and Lundeen 1968). Fine
sediment can smother embryos, alevins, and fry.
Fish migrations may also be impeded when roads
cross the stream and improperly designed
conduits are constructed.

Numerous water diversion projects completed
in California during the past 60 years have
drastically altered natural hydrologic factors
and increased water temperatures. A 90%
reduction in flow caused average width, depth,
and velocity to decrease by 22%, 44%, and 75%
(Curtis 1959) . A similar reduction in flow can
result in a 75% decrease in riffle area, a 55%
increase in shallow runs, and 96% decrease in
deep, fast runs (Kraft 1972). This type of
disruption of the natural hydrologic regime can
explain recent imbalances in native fish popu-
lations and is more probable than interspecific
competition.

The destruction of riparian forests in the
Central Valley has been an important factor
contributing to the changes in the fish commu-
nities, mostly because of the effect on water
temperature. However, there is much we do not
understand about their relationship to fish
populations, particularly in regard to the use
of flooded vegetation by young fish and the
role of logs and other debris in increasing
habitat diversity.

LITERATURE CITED

Arnold, J.F. and L. Lundeen. 1968. South fork
of the Salmon River special survey — soils and
hydrology. USDA For. Serv. , Intermountain
Region. Mimeographed report.

Barton, J.R. , E.J. Peters, D.A. White and P.V.
Winger. 1972. Bibliography on the physical
alteration of the aquatic habitat (Channeli-
zation) and stream improvement. Brigham Young
Univ. Publ., Provo, Utah. 30 pp.

Curtis, B. 1959. Changes in a river's physical
characteristics under substantial reduction in
flow due to hydroelectric diversion. Calif.
Fish and Game 45:181-188.

Funk, J.S. and C.E. Ruhr. 1971. Stream channel-
ization in the midwest. In: E. Schneberger
and J.L. Funk, eds. Stream channelization: a
symposium. N. Cent. Div. Amer. Fish. Soc. Spec.
Publ. 2. p. 5-11.

Kraft, M.E. 1972. Effects of controlled flow
reduction on a trout stream. J. Fish. Res.
Board Can. 29:1405-1411.

Megahan, W.F. and W.J. Kidd. 1972 Effects of
logging and logging roads on erosion and
sediment deposition from steep terrain. J.
For. 80:136-141.

Moyle, P.B. 1976. Some effects of channelization
on the fishes and invertebrates of Rush Creek,
Modoc County, California. Calif. Fish, Game
62(3):179-186.

Peters, J.C. and W. Alvord. 1964. Man-made

channel alterations in thirteen Montana streams
and rivers. Trans. 29th North Amer. Wildl.
and Nat. Resour. Conf. pp. 93-102.

Platts, W.S. and W.F. Megahan. 1975. Time
trends in riverbed sediment composition in
salmon and steelhead spawning areas: South
Fork Salmon River, Idaho. Trans. 40th North
Amer. Wildl. and Nat. Resour. Conf. pp. 229-
239.

Whitney, A.N. and J.E. Bailey. 1959. Detri-
mental effects of highway construction on a
Montana trout stream. Trans. Amer. Fish.
Soc. 88(l):72-23.

ENVIRONMENTAL APPLICATIONS IN CORPS OF
ENGINEERS WORK WITH REFERENCE TO
RIPARIAN VEGETATION MANAGEMENT

Fred Kindel
Chief of the Environmental Planning Section
Sacramento District
U.S. Army Corps of Engineers

Two aspects of the Corps activities in
which environmental applications are important
are protecting the Sacramento Valley levee
system with rock bank protection and projects
for which plans have been developed to protect
riparian trees and vegetation.

A system of about 1,000 miles of levees has
been constructed to provide flood protection to
about one million acres and about 800,000 persons
living in the flood plain of the Sacramento River
(Environmental Statement, 1972). The levee
system is threatened by continuing erosion, and
normal maintenance and even emergency measures

113

are not adequate to cope with the danger to the
levees (Sacramento River Flood Control Project,
1960). In 1960 at the request of the State of
California, Congress authorized the Sacramento
River Bank Protection project to protect the
levees (Sacramento River Flood Control Project,
1960).

Although there are some variations in the
work, the usual circumstance is that erosion
has progressed into or near the levee which is
in danger of failure. To provide protection,
a section of levee is prepared by sloping to a
1 on 2 or a 1 on 3 slope and placing the rock
bank protection. All trees and vegetation in
the area to be rocked must be removed to slope
the bank to retain the rock. In the past,
trees and vegetation were also removed from
some areas adjacent to the actual worksite to
facilitate equipment operation while the rock
is being placed.

The following design changes were made in
recent years to reduce the environmental impact
of bank protection work (Environmental State-
ment, 1972; Bank Protection General Design,
1974):

Where feasible, contractors have been re-
quired to avoid disturbing any significant
vegetation outside the limits of where the rock
is placed. Besides careful equipment operation
from the top of the levee, work is sometimes
accomplished from a barge on the river which
avoids unnecessary disturbance of vegetation to
the maximum. However, barges can only navigate
the deeper reaches of the river south of Colusa.

Trees have been surveyed and evaluated at
the edge of the bank protection areas and all
individual trees which would not interfere with
construction and could be saved are marked.

At some erosion sites there is still some
berm area remaining between the river and the
levee. By placing rock only to the top of the
berm, three things are accomplished: erosion
is arrested and the levee is protected; the
berm is protected, permitting vegetation growth;
and there is much less rock required for con-
struction. Protecting the berm means that trees
and other vegetation on the berm will not have
to be removed. Placing rock only to the top of
the berm means there is much less visible rock
when the river is at low flow. This appears to
be the most desirable of the protection methods
for environmental application. More of this
type of work could be done if additional funds
were available (with only limited funds work is
restricted to the critical erosion sites and
the other protection methods are utilized) .

At some locations, the circumstances of
the erosion and other factors led to a different

design than adding rock for protection of the
levee. Where more economical, the existing
levees may be set back or relocated further from
the river bank. Rock protection is placed on
the riverside of the new berm, and vegetation
may be planted on the berm. An example of this
type of design is at a location near Monument
Bend located on the right bank about one mile
upstream from the Interstate 880 bridge crossing.

As each unit of bank protection is completed
and turned over to the State for operation and
maintenance, a supplement is provided to the
standard operation and maintenance manual which
covers specifically the operation and maintenance
needs of that unit. Where measures are instituted
for added vegetation in our construction work,
it is required that this vegetation must be
properly maintained.

On berm areas where there are significant
trees and vegetation, the Corps has stipulated
that the protected trees should remain when such
sites are provided with bank protection. The
State Reclamation Board has adopted a program of
acquiring a stronger easement than solely for
flood control purposes; this provides the land-
owner a higher price and requires him to leave
the native riparian vegetation in place. This
is an important companion feature to the berm
protection design change (Bank Protection General
Design, 1974).

Over the past several years, a number of
experimental measures have been tested. The
experimental program has had two primary pur-
poses: to test the effectiveness of alternative
bank protection methods and materials, and to
determine costs of such alternative methods.
The testing has been to determine engineering
and economic characteristics on the effective-
ness of the alternative methods as well as their
environmental contribution. One important
factor is whether alternative or supplemental
methods are more costly to operate and maintain.
Where possible, alternatives should be found that
do not add significant maintenance expense.

A pilot levee maintenance study was conduc-
ted by the State of California and reported on
in 1967 (Pilot Levee Maintenance Study, 1967) .
The study demonstrated that certain types of
ground cover were compatible on levees, that
some trees and shrubs may be allowed on some
levees, and that in most cases unrestricted
growth may be allowed on berms. The study
indicated that costs of maintenance of levees
would be increased with this vegetation.

The Corps has planted trees and shrubs at
several selected sites along the Sacramento
River (Environmental Statement, 1972) to demon-
strate that such vegetation can be successfully
grown, can be compatible with flood control

114

requirements, and can offer a significant im-
provement to aesthetics and other environmental
aspects of the river. The most outstanding ex-
ample of such a demonstration is near Monument
Bend just upstream from Interstate 880 bridge.
In 1967 we planted a variety of trees and shrubs
along about three miles of the riverbank where
the levee had been set back and the new berm
protected by rock. In 1970 after three years,
the vegetation has provided a significant im-
provement (Environmental Statement, 1972) and
this is still in evidence today. The State
Department of Water Resources conducted some
maintenance studies on this vegetation demon-
stration site and in 1973 reported on the sur-
vival rates of the various species in relation
to the effects of inundation by floodwaters and
accidental losses by fire. Cost of manpower for
levee maintenance with the planted vegetation
was increased by 64 percent over costs without
vegetation on similar adjacent levee areas
(Sacramento River Levee Revegetation Study, 1973).

The Sacramento River and Tributaries Bank
Protection and Erosion Control Investigation,
authorized by the House Public Works Committee,
was initiated in 1977. The purpose of this
study is: to determine the Federal interest in,
and responsibility for, providing bank protection
and erosion control; to study alternative means
and the feasibility of providing a comprehensive
program to stabilize the streams, protect the
levees and banks, preserve riparian vegetation,
wildlife habitat and aesthetic values, and pro-
vide outdoor recreation opportunities along the
river; and to select and recommend the best and
most balanced plan of improvement, provided that
such a plan is found feasible. Completion of
the study is scheduled for 1982.

CONCLUSION

The fate of Riparian Forests in California
depends upon public education and protective
legislation. The first "public hearing" of the
plight of these habitats was a conference in
Chico, California, on May 22, 1976, which was
sponsored by the Davis and Altacal Audubon
Societies. A second conference was held in Davis,
California, on May 14, 1977, and sponsored by
the Institute of Ecology at the University of
California and the Davis Audubon Society.
Public awareness of the demise of Riparian Eco-
systems must now be coupled with political pres-
sure. Governmental agencies which have juris-
diction over the fate of riverbanks must be made

aware of the significance and uniqueness of these
ecosystems. We must study these agencies'
surveys, participate in their hearings, join
their advisory committees and become well armed
with facts and determination. However, even
federal and state agencies have restrictions on
their spheres of influence. Almost 95% of the
yet unspoiled remnants of riparian hardwoods in
California are in private ownership. Each year
more of these areas are bulldozed for orchards,
cut for pulpwood and cleared for "stream bank
protection. "

Several approaches can be made to solve the
riparian protection problem. Land use plans
must be established at county and state levels
to encourage recreational and open space ease-
ments as well as wildlife sanctuaries. Zoning
laws should be altered to relieve land owners
from heavy taxes on riparian forest (many far-
mers are taxed on their forests as if they were
fruit orchards) . Forestry management acts
should be amended to protect riparian species.
Private landowners should be offered reasonable
alternatives to tree cutting, such as tax deduc-
table donations of land to non-profit, private
organizations like the American Land Trust,
Audubon, and the Nature Conservancy. Prime
riverine forest land should be purchased by
conservation groups if all other measures fail.
We must all publicize what we know about the
Riparian Forests and work together to bring
about the political changes necessary to preserve
these very special and vulnerable ecosystems.
Interested persons should contact the Riverlands
Council, P.O. Box 886, Davis, California 95616,
to receive fact sheets and legislative updates.

LITERATURE CITED

Bank Protection General Design, Design Memorandum
No. 2, Sacramento River Bank Protection Project,
1974.

Environmental Statement, Sacramento River Bank
Protection Project, Sacramento District, Corps
of Engineers, November 1972.

Pilot Levee Maintenance Study, Bulletin No. 167,
Department of Water Resources, State of Cali-
fornia, June 1967.

Sacramento River Flood Control Project, Califor-
nia, Senate Document No. 103, 86th Congress,
2nd Session, 26 May 1960.

Sacramento River Levee Revegetation Study,

Department of Water Resources, State of Cali-
fornia, July 1973 (Central District, D.W.R.).

115

Regeneration and Distribution of
Sycamore and Cottonwood Trees
Along Sonoita Creek,
Santa Cruz County, Arizona1 ^

Richard L. Glinski 2/

Abstract. — This study describes the effects of livestock
grazing and streambed erosion on the regeneration and distribution
of sycamore and cottonwood trees. Sycamores reproduced from
root and trunk sprouts and because of this their distribution
is not as likely to change significantly. Cottonwood reproduction
was nearly absent in areas grazed by cattle , and was confined to
the narrow erosion channel. If this regeneration pattern
continues, the future maximum width of the cottonwood forest
will decrease nearly 60%.

INTRODUCTION

Many Riparian Deciduous Forests in the
Southwest contain extensive groves of sycamore
(Platanus wrightii) and cottonwood (Populus
fremontii) trees. (Plant community terminology
follows Brown and Lowe 197^ • ) In many areas
either one or both of these species are the
sole tree component of the riverine habitat.
They increase habitat diversity and create
invaluable niches for a variety of wildlife,
particularly birds (Bottorff 197^, Carothers
et al. 197^i Johnson and Simpson 1971 > and
others) .

Along Sonoita Creek in southeastern Arizona
cottonwoods exclusively are used as nest trees
by rare birds like the Gray Hawk (Buteo nitidus) ,
Zone- tailed Hawk (B. albonotatus) and Black Hawk
(Buteogallus anthracinus ) , and sycamores are the
favorite nest tree of the Rose-throated Becard
(Platypsaris aglaiae) (pers. obser.).

This study is an assessment of the numbers ,
condition, regeneration and distribution of these
important trees along Sonoita Creek. It provides
some basis for comparing populations of these
species in areas of varied livestock grazing use.

Contributed paper, Symposium on the
Importance, Preservation and Management of the
Riparian Habitat, July 9, 1977, Tucson, Arizona.

£/ Staff Research Biologist, Department of
Zoology, Arizona S'tate University, Tempe,
Arizona 85281

This paper also describes the effects of
streambed erosion on the regeneration and future
distribution of these trees in riparian habitats
of the Southwest.

THE STUDY AREA

Sonoita Creek originates in the Plains
Grassland about 1 km northwest of Sonoita,
Santa Cruz County, Arizona, and flows in a
southeasterly direction through the Desert Grass-
land to its confluence with the Santa Cruz River
14 km north of the Mexico-United States border.
Its 51 km reach occurs along an elevational
gradient ranging from 1^80 to 1035 m. Its tribu-
taries drain habitats from Boreal Forests at
2865 m elevation through Temperate Woodlands
(Madrean Evergreen Woodlands) and Desert Grass-
lands at 1035 m elevation.

The low hills immediately bordering the
creek are covered in moderate density with oaks
(Quercus spp.), mesquite (Prosopis juliflora) ,
juniper ( Juniperus monosperma), cliff -rose
( Cowan ia mexicana) , mountain-mahogany ( Cerco-
carpus sp~T) and ocotillo (Fouquieria splendens) .
Groundcover, consisting mainly of gramma grasses
(Bouteloua spp.) and love grasses (Eragrostis
spp . ) , is moderately dense near Sonoita where
soils are better developed, and sparser along
the middle and lower reaches where steeper hill-
sides and rockier soils prevail.

The upper 19 km of the alluvial valley-
floor from the headwaters to the town of

116

Patagonia consist of eroded grass- and scrub-
covered plains, irrigated pastures and croplands,
and remnants of a Sacaton Grass-Scrub Community.
Here surface waterflow is mainly ephemeral.
Streambed erosion as reported by Bryan (1925) is
a common feature of this upper valley-floor. The
erosion, which begins about 2.9 km southwest of
Soniota, forms vertical banks 1 to 5 m high and
separated by a channel from 15 to 37 m wide. The
evidence of this erosion decreases progressively
downstream for 5 km, first through weedy grass-
lands then past irrigated pastureland beginning
near Adobe Canyon. For the next 5 km flood-
irrigated pastures border the dry creekbed.
Erosion produces cuts here less than 2 m high
and occurs only locally where the creekbed bor-
ders steep hillsides.

The widest reach of this upper alluvial
valley is nearly 1 km across and continues for
5.6 km downstream from the pastureland. The
flood plain here supports a Sacaton Grass-
Scrub Community. In early 1975 many large mesquite
trees with basal diameters up to 0.6 m and a few
walnut trees ( Juglans major) with basal diameters
of nearly 0.8 m were removed from here. This area
is gradually being cleared of natural vegetation
and transformed into irrigated fields. Streambed
erosion is absent in the upper 4.0 km of this
Sacaton Grass-Scrub plain, where the creekbed
branches out among dense sacaton clumps that
reach over 2 m in height, but begins about 1.6 km
upstream from the lowest reach of this community.
Through this eroded reach, vertical banks are 1
to 2 m high and 10 to 40 m apart. Irrigated
cropland borders remnants of the Sacaton Grass-
Scrub plain about 0.8 km above Patagonia.

Scattered individuals and small clumps of
large cottonwood and sycamore trees occur along
this upper valley-floor. The predominant tree of
this upper floodplain is mesquite, which occurs
mainly as a shrub or small tree up to 7 m tall.

About 0.8 km below Patagonia perennial sur-
face water begins and continues downstream for
about 19.3 km. Along this reach there is a near-
continuous belt of Riparian Deciduous Forest,
consisting of cottonwood, sycamore, willow
(Salix gooddingii) , ash (Fraxinus vel utina) and
walnut trees. This riparian forest varies in
width from 15 to 150 m and in density of trees.
It is bordered frequently by small bosques of
mesquite and hackberry (Celtis reticulata) .

Willow is the most abundant large riparian
tree at the head of this forest, and ash is the
commonest large tree along the middle and lower
reaches. Cottonwood, the third most prevalent
tree, is irregularly distributed throughout the
forest, and densities of mature tree and sapling
cottonwoods vary considerably. Sycamore and
walnut are the least common large trees, and occur
in small clumps or singly, mainly along the lower
reaches of the forest.

The forest floor is covered with annual and
perennial grasses and forbs that in some open
areas form dense thickets up to 2 m tall after

the onset of summer rains in July. In many
places livestock grazing and human recreation
eliminate summer groundcover by the following
spring.

The watercourse, which normally varies in
width from 2 to 5 m, is usually less than 10 cm
deep, and is often lined with seep-willow
(Baccharis glutinosa) . In areas that are not
grazed by livestock, water-cress (Rorippa
nasturtium-aquaticum) blankets the flowing
water most of the year.

Since much of the Riparian Deciduous Forest
is within the confines of a narrow rocky canyon
less than 0.5 km across, alluvial deposits are
limited and extensive creekbed erosion is
restricted usually to reaches near the entrances
of side canyons where alluvium has accumulated.
In many areas this erosion has resulted in
vertically cut banks up to 4.9 m high. In 1881
a railroad track was completed along the entire
length of Sonoita Creek, altering the water-
course along many reaches, including that of
the Riparian Deciduous Forest. The railroad
levee often constitutes the bank of the creek
and contains floodwaters.

The continuity of the Riparian Deciduous
Forest was disrupted in the early 1960's with
the construction of Lake Patagonia. This lake
covers 650 hectares beginning 11.6 km below
the head of the Riparian Deciduous Forest and
continuing downstream for 3«5 km. Below the dam,
surface flow and patches of Riparian Deciduous
Forest persist for about 2.9 km.

Thereafter, for the next 11.7 km to its
confluence with the Santa Cruz River, Sonoita
Creek flows intermittently with scatterings of
large cottonwood, ash and willow trees.

The creek proper dissects two general soil
types (Richardson in press). Ending about
10 km below Patagonia, the upper type is
usually more than 1.5 m deep in old alluvium
from igneous and sedimentary rocks and is
composed of fine to moderately coarse textured
soils with about 35 percent gravel and cobble
throughout. The lower type consists of rocky,
very cobbly and sandy loams less than 0.5 m
deep on weathered granitic, tuff -conglomerate,
or andesite-tuff bedrock. Here sediment yield
is low.

The most complete climatic data for this
area is from Nogales, Arizona, which lies just
south of the watershed 27 km southwest of
Patagonia and is 1158 m in elevation. Records
from 1893 through 1962 show that maximum daily
temperatures occurred during June and averaged
34.9 Cj minimum daily temperatures occurred
during January and averaged -1.4 C. Temperature
extremes for the area were -14.4 and 43.3 C.
Precipitation was biseasonal, and about 60
percent of the annual average of 396 mm fell
during July through September. The driest months
were May and April, respectively (Green and
Sellers 1964). The spring of 1974 was an
extremely dry year and some reaches of the

117

Riparian Deciduous Forest were without surface
flow.

METHODS

From April 1974 through April 1977 I surveyed
the trees of the Riparian Deciduous Forest on
foot and recorded numbers, height classes, con-
ditions, locations and evidence of regeneration
of sycamores and cottonwoods. Multi-trunked trees
were counted as one tree if the trunks were
joined above ground level. To estimate percentage
of sprouting, closely clumped sycamores were
counted as an individual tree since they probably
shared a common root system. I estimated the
numbers of densely clustered seedlings and sap-
lings. Tree heights were estimated and grouped
into three size classes: seedling ( < 2 m),
sapling (2-9 m), and trees ( > 9 m). I made brief
notes on the conditions of mature trees, especially
cottonwoods that had experienced excessive leaf-
drop or canopy die-out, or ones with fully leafed
canopies that had fallen over for no apparent
reason. The numbers and locations of other
riparian trees were only casually recorded.

At 100-m intervals I measured the width of
the creekbed between the tops of the banks, and
the height and approximate slope (to the nearest
30 angle) of the creek banks. To determine the
maximum distance that cottonwoods occurred from
the creekbed, within each 100-m interval I
measured the distance between the edge of the
creekbed and the cottonwood farthest removed
from the creekbed on both sides of the creek.
Lateral bounds of the creekbed often were easily
defined by vertically cut banks , but at times
were estimated for banks with shallow slopes.
Frequently the railroad levee obviously confined
the watercourse and was counted as a bank. Bank
height and slope were not recorded where the
hillside bordered the creekbed. Bank Slopes were
classified as shallow (0-30 ), intermediate
(30-60 ) or vertical (60-90°) to the plane
of the creekbed.

The Riparian Deciduous Forest was divided
into five segments that differed in livestock-
grazing use. The grazing practices along
Sonoita Creek have varied greatly within the past
100 years, and exact grazing use, past or present,
was not measured. However, it is possible to
place each of the five segments in different use
categories based upon livestock class or use
intensity. Segment 1 is 2.7 km long, lies between
1234 and 1204 m elevation, and consists of the
Patagonia-Sonoita Creek Sanctuary of the Nature
Conservancy, an area fenced to exclude all live-
stock grazing in 1966. Segment 2 is 2.1 km long,
lies between 1204 and 1189 m elevation, and has
been grazed by cattle and horses for at least
50 years. Segment 3 is 0.8 km long, lies between
1189 and 1183 m elevation, encompasses the RL
Ranch, and has been grazed by horses only since
1966. Segment 4 is 5»9 km long, lies between

1183 and 1143 m elevation at the head of Lake
Patagonia, and has been grazed by both horses
and cattle for at least 50 years. Segment 5
is 2.9 km long, lies below Lake Patagonia at
elevations between 1116 and 1097 m, and has
been grazed mainly by cattle for at least 50
years. The soil and vegetation in the upper
third of Segment 5 w&s highly modified in the
early 1960*s when the dam was built.

RESULTS

Sycamore regeneration,
distribution and condition

Sycamores reproduced mainly vegetatively
by sprouts from lateral roots and trunk bases.
Either root or trunk sprouts were found in 74%
of mature sycamores (clumped and individual
trees). I never encountered sycamore seedlings,
but in Segment 4 I located one sapling that
was more than 30 m from mature sycamores and
was not a root sprout (Table l).

Root sprouts occurred on 397° of the syca-
mores. On the average, root sprouts grew
a distance of 1,0 m from the trunk bases
(SD 0.5, r 3.0), were 3 m tall (SD 5, r 23),
and 5 cm in diameter (SD 10, r 46), Trunk
sprouts occurred on §6% of the sycamores and,
on the average, sprouted 0.1 m above ground
level (SD 0.1, r 0.6), were 3 m tall (SD 3,
r 13) and 4 cm in diameter (SD 6, r 28). Trees
with root sprouts averaged 8 sprouts per tree
(SD 12, r 31)1 and those with trunk sprouts
averaged 6 sprouts per tree (SD 10, r 30).
Both root and trunk sprouts occurred on 22%
of the sycamores . Root sprouts that were either
less than 0.1 m tall or growing among dense
debris and brush may have been overlooked,
while probably all trunk sprouts were counted.

I was unable to determine what stimulates
a sycamore to sprout. There was no significant
correlation between either the presence of
sprouts or the number of sprouts and the
percent of canopy die-out, soil texture at
tree base, or proximity of tree to surface
water. Possibly the more latent factor of sub-
surface rock formations affected the presence
of groundwater near individual trees and thus
their tendency to sprout. The dependence of
sprouting sycamores on abundant shallow water
remains to be examined.

The large variance in the size of
sprouts, some of which themselves were mature
trees nearly 30 cm in diameter, indicates that
sprouting has been a major means of regenera-
tion along Sonoita Creek for some time. Many
even-aged, large sycamores were growing in
circular clumps with their trunk bases touching
or nearly touching (fig. l), suggesting that
these clumped trees had been sprouts from a
common parent tree that has long since
decomposed.

118

Table 1. — Length, grazing
tree sycamores

use, and numbers of sprouts, seedling, sapling, and mature
and cottonwoods for each of five stream segments.

Sycamore

Sprouts

(Sprouts /km)

Seedlings ( <Z m)

(Seedlings /km)

Saplings (2-9 m)
(Saplings /km)

Trees ( > 9 m)

(Trees / km)

Cottonwood

Sprouts

(Sprouts /km)

Seedlings ( < 2 m)

(Seedlings /km)

Saplings (2-9 m)

(Saplings /km)

Trees ( > 9 m)

(Trees / km)

Segment length (km)

Grazing use

Stream Segment

Total

160

259

(o)

(1)

(0)

(27)

(34)

(o)

(0)

(o)

(0)

(o)

(3)

(1)

(6)

(5)

(o)

(0)

400

500

700

1500

100

3200

{ IH-O )

( <yna\
\ 230/

(One \

1078

586

1697

(399)

(5)

(732)

(3)

(1)

638

385

612

1767

(236)

(183)

(85)

(104)

(22)

2.7

2.1

0.8

5.9

2.9

none

cattle &

horses

cattle &

cattle

horses

Sycamores were absent in Segment 1 and
most abundant in Segment 4 (Table l) except
along the reach where surface flow was absent
during the spring drought in 1974. They were
distributed most frequently as scattered clumps
of up to 10 trees. The clumps occurred from 0
to approximately 140 m from the watercourse
(m 33 f SD 48). Since regeneration was mostly
from sprouting, almost all seedlings and sap-
lings occurred near mature trees.

Livestock had browsed on only 13% of root
and trunk sprouts. Most sprouting sycamores
were growing on steep banks or amidst brush and
litter, which probably prevented cattle from
browsing sprouts. Aside from the limited canopy
die-out that some trees had experienced, mature
sycamore trees appeared healthy. They did not
respond to the severe drought in 1974 as did the
cottonwoods; however, their occurrence along
specific reaches of Sonoita Greek may indicate

the presence of abundant surface or subsurface
water, thus they would not be expected to re-
flect initial drought conditions that might
affect other trees.

Cottonwood regeneration,
distribution and condition

All cottonwood regeneration along Sonoita
Creek was from seeds. In July of 1975 cotton-
wood seedlings about 10 cm high were ubiquitous
on moist, sandy alluvium along the watercourse
in all segments of the Riparian Deciduous
Forest. By the following spring almost all
seedlings were absent except those that had
germinated in Segments 1 and 3> where cattle
were excluded. In January of 1975 cottonwood
seedlings were most abundant in Segment 3
(875 seedlings/km) and least abundant in

119

Figure 1. — Clumped mature sycamore trees.

Segment 5 (l sapling/km). Segments 1 and 3
averaged 399 and 732 saplings/km, respectively,
and Segments 2,4 and 5 averaged 5.3 and 1 sap-
lings/km, respectively (Table l).

All seedlings occurred below the banks,
many immediately along the watercourse. Many
cottonwood seeds had germinated on damp soil
around isolated, ephemeral pools of rain water
that had collected away from the watercourse,
but the resulting seedlings wilted and died
when the pools became dry several weeks after
germination. A continuous supply of supplemental
water (treated sewage water) existed in Segment
1, and sustained 93% of "the established seedlings
outside the creekbed in that segment.

Mature cottonwoods were not evenly distri-
buted throughout the Riparian Deciduous Forest,
but were densest in Segment 1 (236 trees/km) and
sparsest in Segment 5 (22 trees/km) (see Table l).
The construction of Lake Patagonia Dam eliminated
many mature trees from Segment 5« Within each
segment cottonwoods were not homogeneously dis-
tributed, but occurred with greatest density where
alluvial deposits were plentiful. Individual, small
clumps or large groves of cottonwoods occurred a
mean maximum distance of 26 m from either edge
of the creekbed (SD 40, r 196). The mean maximum
width of the cottonwood forest , including the
creekbed width, was 83 m.

On numerous occasions I observed cattle
grazing seedlings on sand deposits along the
watercourse. I never observed horses grazing
cottonwoods , although many times I saw them nose
aside a cottonwood seedling to get at a nearby
sprig of grass or water-cress.

The extremely dry spring of 1974 obviously
affected many mature and sapling cottonwoods
along the entire reach of the Riparian Deciduous
Forest. Extensive leaf-drop and canopy die-out
occurred in May and June, and many areas of the
creekbed and forest understory at that time were
blanketed with small, partially developed cotton-
wood leaves that ranged in color from yellow to

yellowish green. Leaf -drop occurred in all seg-
ments but was most obvious in Segment 4, where
the surface flow failed in May and June along a
reach about 150 m long. Many trees regained
fully developed canopies after the onset of
summer rain in July, but others remained bare
and lost large canopy limbs.

Several large cottonwood trees were up-
rooted in the spring of 1974, having fallen
over with canopies in full leaf. The exposed
roots of these fallen trees were dry and had
been severed less than 1 m below ground level
and less than 2 m laterally from the trunk.
Uprooting occurred mainly in Segment 1 , where
five trees fell over. Three trees in Segment 2
and two trees in Segment 4 also were uprooted
in early 1974.

Characteristics of valley-floor erosion

Along Sonoita Creek erosion has resulted
in banks that are sloped in a shallow, inter-
mediate or vertical manner to the plane of the
creekbed. In many instances the shallow and
intermediate slopes probably are the result of
gradual trampling of once-vertical slopes by
livestock. Including the railroad levee,
which accounted for 12% of the sampled banks,
the average slope of the banks was 56 (SD 17).
Nine percent of the sampled banks were vertically
cut (fig. 2). The mean height of the banks was
2.3 m (SD 1.6, r 5.8). The width of the creekbed
averaged 31 m (SD 23, r 104). There was no trend
in these characteristics. Each characteristic
was independently and heterogeneously distributed
along the valley-floor, and probably reflected
the variable alluvium deposits and geometry of
the valley-floor.

DISCUSSION

Effects of grazing

Regeneration and distribution of sycamore
and cottonwood were affected differently by
livestock grazing. The full effects of grazing
on sycamore are uncertain, largely because
sycamores were absent in Segment 1 where live-
stock were excluded so the chance of seeds
germinating in this segment was reduced. In
other segments livestock may have grazed some
small sycamore seedlings that I had overlooked,
but sycamore seedlings certainly were not
common along Sonoita Creek. The germination
requirements of sycamore seeds are unknown.
Perhaps the soil chemistry of the eroded valley-
floor has changed significantly and now inhibits
either germination of seeds or growth of
seedlings.

Livestock did not heavily browse sycamore
sprouts for several reasons : Sprouts were
necessarily associated with mature trees and

120

Figure 2. — Vertical-bank erosion along Sonoita
Creek. Cut bank here is 1.5 m high.

thus were scattered and were neither as readily-
encountered by livestock, nor as easily browsed
and trampled as dense clusters of seedlings on
exposed alluvium. Small sprouts were frequently
well protected and hidden by leaf and branch
litter from the nearby mature trees and often
the trunks of the parent trees obstructed
browsing attempts. Several sprouting sycamores
were growing on steep eroded banks or among
dense brush, both of which hindered browsing.
Also, the stems and leaves of sprouts may be
less palatable to livestock than those of
seedlings.

Because livestock may graze occasional
sycamore seedlings, grazing may be preventing
a limited increase in the distribution of syca-
mores along the streambed. Livestock did not
appear to be inhibiting regeneration of sycamores
by browsing on sprouts, which constituted the
vast majority of sycamore regeneration along
Sonoita Creek. These observations indicate that
the future distribution of sycamores along
Sonoita Creek will be nearly identical to the
present distribution, except that where mature
trees fail to reproduce vegetatively by sprouting
the sycamore will disappear.

Grazing of small seedlings by cattle was
the most obvious factor preventing regeneration
of cottonwood. Seedlings averaging 9 cm in
height and 3^5 seedlings/m in density were
commonly grazed and trampled by cattle. Both the
proximity to the creek water, which was used by
cattle for drinking, and the unprotected open-
ness of the creekbed alluvium where seedlings
occurred made seedlings vulnerable to trampling
and grazing by cattle. Horses did not graze
seedling cottonwoods, but frequently trampled
young seedlings that were growing on open
alluvium. Horses did strip the bark from some
sapling cottonwoods that had been bent over by
flooding.

Thus, by grazing seedlings cattle have
severely reduced the establishment of

cottonwood in the Riparian Deciduous Forest
of Sonoita Creek. In Segments 2, k and 5, where
cattle have grazed for at least 50 years, the
combination of decreased establishment and
normal mortality of the mature trees will
eventually severely reduce in number or elimi-
nate the cottonwoods from this forest. Dry
years such as the one reported for 197^ could
hasten the mortality of mature trees and the
elimination of the cottonwood.

Effects of streambed erosion

Sonoita Creek flows through the Desert
Grassland, a biome which has undergone major
vegetational changes in the past century
( Hastings and Turner 1965 , and others ) . The
erosion that is associated with these vegetation
changes consists, in part, of extensive vertical
cutting or channeling of the streambeds of
major drainages and their tributaries by flood-
waters. Both overgrazing by livestock and
climatic shifts have been associated with the
start of this erosion. Bryan (1925) reported
that "Nearly all streams in southwestern United
States flow between vertical banks of alluvium
that vary in height from 10 to as much as one
hundred feet. Although subject to periodic
floods, these streams no longer overflow their
banks, nor build up their adjacent flood-plains.
Floods merely deepen and widen the channels
(arroyos) which continually grow headward into
the undissected valley floors of the headwater
valleys and tributaries."

The extent of erosion along Sonoita Creek
is limited by both the width and depth of
alluvial deposits in this narrow valley-floor,
although the heights of some vertically cut
banks have increased up to 0.3 m from 1973
to 1976.

Two significant consequences of this
channeling are the containment of floodwaters
within the confines of the relatively narrow
channel , and the scouring of the vegetation
that occurs within the erosion channel.
Precipitation that now falls within the Sonoita
Creek watershed spends relatively less time in
this drainage since the channel quickly trans-
ports the water along the valley-floor,
preventing water dispersion laterally from the
creekbed onto the adjacent floodplain. Such
rapid transport may also affect the water table
recharge rate.

It seems that once streambed cutting begins
it is further perpetuated and accelerated by
the concentrated floodwaters in the erosion
channel, which transport and remove vegetation
and debris that would, prior to the channel,
have remained in place and promoted dispersal
of less forceful floods, thus decreasing the
water velocity and inducing silt deposition.
The overgrazed hillsides that border Sonoita
Creek no doubt assist in increasing the flood-
water velocity since they support relatively

121

less vegetation to intercept precipitation.

I witnessed the effects of channeling on
1 July 197^ when a heavy thundershower occurred
over the watershed starting about mid-afternoon.
At approximately 1700 hours the water level in
Segment 1 had risen almost 0.8 m, and only in
unchanneled areas did it overflow laterally
onto the shallow 'banks. The current was swift,
removing or flattening vegetation on the shallow
banks and scouring most aquatic plants and
streamside stands of seedling cottonwoods and
willows. Only seedlings that were growing among
dense stands of seep-willow remained. The next
morning scouring of creekbed vegetation was
evident all the way to Lake Patagonia. Relatively
few stream terraces had been flooded since the
erosion channels or railroad levee had contained
the floodwaters. In the afternoon of 2 July the
water was still turbid and about 10 cm above
pre-flood level in Segment 1. A layer of sand
and finer silt had been deposited only on the
slightly elevated banks that had been flooded
and along the edges of the creekbed where the
water had receded.

Coupled with recent erosion are factors
of future precipitation trends. Seemingly,
higher rates of precipitation would cause more
intense flooding and involve a greater volume
of water in the channel. Possibly the present
extent of channel erosion could not contain such
volumes of water and lateral flooding could
result. However, if precipitation rate decreases
in the future, the chance of lateral flooding
would be further reduced by the channel.

The single sycamore sapling in Segment 4
that was not a root sprout had been pushed over
by flooding and was partially covered by flood
debris. The bark of this sapling was scraped
in many places, and it appeared unhealthy and
not likely to survive. Possibly sycamore seed-
lings are intolerant of severe flooding, and
mortality from floods during the seedling stage
is responsible for decreased regeneration from
seeds. Root and trunk sprouts probably suffer
less mortality from flooding since they are
more solidly rooted and protected by parent
trees.

Most cottonwood seedlings were removed by
floodwaters. The cottonwood seedlings and
saplings that occurred in the erosion channel
usually were growing among dense stands of
seep-willow, where competition for sunlight
and water may affect their survival. Those
occurring away from seep-willow were bent over
by floodwaters and seemed unlikely to survive,
and many such saplings had leaves browsed by
cattle or bark stripped by horses.

While the streambed within the confines
of the erosion channel transports increasingly
more floodwater, the elevated terrace adjacent
to the erosion channel becomes increasingly
more xeric due to reduced overbank flooding
outside the channel and increased depth of the

water table as erosion progresses. Since
cottonwood seedlings were absent from these
benches, it is likely that in the absence of
saturating floods they seldom contain the
reliable surface water necessary for seedlings
to extend tap roots to permanent water. This
problem is compounded by increased distance to
the ground water on benches.

The ability of mature trees to survive on
the benches is questionable for they too must
cope with relatively drier conditions. Most
mature cottonwoods which had fallen over during
the drought of 197^ were growing away from the
creekbed on slightly elevated benches.
Z immermann ( 1969 ) noted that along the San Pedro
River cottonwoods occurred where ground water
exceeded 300 feet in depth, and he suspected
that some tree species "may depend for growth
during at least part of the year only on
moisture in the alluvium." Where channeling
has increased the depth of the water table
and prevented lateral flooding onto benches,
the only moisture available to vegetation on
these benches may be from precipitation. For
cottonwoods and sycamores this may not be
sufficient to sustain growth, especially if
bench alluvium and topography permit rapid
runoff of precipitation. If mature sycamores
require saturated alluvium for sprouting, the
frequency of sprouting may decrease with
increased erosion.

SUMMARY AND CONCLUSION

Sonoita Creek provided comparison of
regeneration of sycamore and cottonwood trees
in areas of various livestock grazing uses.
It also afforded observations on the relation
between regeneration and streambed erosion,
which along Sonoita Creek is limited yet
effective in containing and quickly transporting
floodwater along the valley-floor, and ultimately
in transforming the broad cottonwood forest into
a relatively narrow strip of trees that grow
in the erosion channel.

Livestock grazing did not appear to prevent
regeneration of sycamores, which produced by
sprouting from roots and trunk bases. The
apparent absence of sycamore seedlings may be
related to the erosion and turbid flooding that
now periodically occurs in this drainage.
Because of vegetative reproduction, sycamore
distribution along Sonoita Creek is riot likely
in the near future to change appreciably from
its present distribution unless mortality of
sprouts occurs. An increase in soil aridity
associated with the erosion channel may induce
sprout and parent tree mortality.

Cottonwood, which reproduced from seed,
was nearly absent in stream segments grazed
by cattle, but abundant in areas grazed by
horses only. Because stream flow needed for

122

cottonwood regeneration is confined to the
eroded channel , all cottonwood regeneration is
confined to this narrow habitat, which averaged
31 m wide. The present mean maximum width of
the cottonwood forest including the pre-erosion
remnants is 83 m, Thus the future maximum width
of the cottonwood forest along Sonoita Creek
will decrease nearly 60 percent if the present
natural regeneration pattern continues.

ACKNOWLEDGEMENTS

Special appreciation is extended to
S. J. Shellhorn and R. M. Turner for stimulating
discussions of riparian ecology and for critical
review of this paper. Thanks to S. B. Terrell
for assisting in the field work, to W. Van Asdall
and R. D. Ohmart for reviewing the paper.

LITERATURE CITED

Bottorff, R. L.

197^. Cottonwood habitat for birds in
Colorado. Amer. Birds 28(6) :975-979 .
Brown, D. E. and C. H. Lowe.

197^. A digitized computer-compatable
classification for natural and potential
vegetation in the Southwest with
particular reference to Arizona. J. Ariz.
Acad. Sci. 9:3-11.

Bryan , K .

1925. Date of channel trenching (arroyo
cutting) in the arid Southwest.
Science 62:338-3^.

Carothers , S . W . , R . R . Johnson and
S. W. Aitchison.

197^. Population structure and social
organization of southwestern riparian
birds. Amer. Zool. 14:97-108.

Green, C. R. and W. D. Sellers, eds.

1964. Arizona Climate. Univ. of Ariz
Press, Tucson. 503 PP-

Hastings, J. R. and R. M. Turner.

1965. The Changing Mile. Univ. of Ariz
Press, Tucson. 317 PP.

Johnson, R. R. and J. M. Simpson.

1971. Important birds from Blue Point
Cottonwoods, Maricopa County, Arizona.
Condor 73: 379- 380.

Richardson, M. L.

In press. Soil survey of Santa Cruz and
parts of Cochise and Pima Counties,
Arizona. U.S. Dept. of Agriculture.

Zimmermann, R. C.

1969. Plant ecology of an arid basis, Tre
Alamos-Redington area, southeastern
Arizona. U.S. Geol. Survey Prof.
Paper 485- D. 51PP«

123

The Development and Perpetuation
of the Permanent Tamarisk Type in the
Phreatophyte Zone of the Southwest1

Jerome S. Horton

Abstract. — Several species of tamarisk were introduced
into the United States in the 19th century for ornamental use.
Saltcedar (Tamarix chinensis Lour.) became naturalized and by
the 1920 's was a dominant shrub along the Southwestern rivers.
Its aggressive characters suit it to be a permanent dominant
in much of the phreatophyte vegetation of this region. Success-
ful management of this vegetation for any resource must care-
fully consider its ecological characteristics.

INTRODUCTION

Tamarisk ^Tamarix spp.), first introduced
into the United States for ornamental uses in
the early 1800' s (Horton 1964), soon spread
throughout the nation. Most dramatic, however,
was its invasion onto the flood plains of the
Southwestern rivers, where it soon became a
major vegetation type. These stands attracted
little attention until it was realized they
were using large amounts of water (Horton 1973)
Their aggressive spread, associated with local
water shortages, resulted in many action pro-
grams to remove phreatophytes .

plain vegetation can be managed for perpetuation
of wildlife habitat and still reduce water losses
(Horton and Campbell 1974).

SPECIES CHARACTERISTICS

To become an aggressive part of any vege-
tation community, a species must establish itself
successfully under existing conditions or to
spread into new habitats created by man's
modifications. Of primary importance are seed
production and germination, followed by success-
ful seedling establishment.

Robinson (1965) reported that saltcedar,
as the aggressive tamarisk (Tamarix chinensis
Lour.) is often called, was occupying an esti-
mated 900,000 acres of flood-plain land by 1961.
Although this acreage has now been considerably
reduced by agricultural and industrial develop-
ments and various projects for control of the
species for water salvage, the remaining stands
are becoming increasingly important for wild-
life and other resource management. In many
cases, these values outweigh those of the water
that might be saved by eradication of the cover.
Most of these values are dicussed in accompanying
papers. If should be kept in mind that flood-

Contributed paper, Symposium on the
Importance, Preservation and Management of the
Riparian Habitat, July 9, 1977, Tucson, Arizona.

Principal Plant Ecologist (Retired), USDA
Forest Service, Rocky Mountain Forest and Range
Experiment Station, at the Station's Research
Work Unit at Arizona State University, Tempe.
Central headquarters is maintained at Fort Collins
in cooperation with Colorado State University.

Many phreatophyte species — such as saltcedar,
cottonwood (Populus spp.), willow (Salix spp.)
and seepwillow (Baccharis glutinosa Pers.) — are
spread primarily by abundant wind-borne seeds
which germinate quickly on water or moist soil.
Seeds of these species will usually lose
viability rapidly, and must germinate within
2 to 4 months (Horton et al. 1960). Though
the seeds will germinate rapidly, the new
seedlings require wet soils for several weeks.
These species thrive best in open sun, such as
along sandbars or areas disturbed by floodflows.
Of the species disseminated by wind-borne seed,
tamarisk is the most aggressive, and when con-
ditions are ideal, invasion will be rapid.

Seed germination of mesquite (Prosopis
julif lora (Swartz) DC.) and associates is not
dependent on such rigid soil-moisture conditions.
While germination may be started by floodflow,
especially in gravel washes, seeds are spread
more by animal activity, such as defecation by
cattle, coyotes, etc. Thus, mesquite has spread
into the grasslands and hillsides of southern
Arizona where summer rains are more frequent
(Schuster 1969) . In the drier areas of central
Arizona, however, the species is more common
in alluvial soils above the deeper groundwater
tables.

124

Root systems of phreatophyte species
vary greatly. Mesquite is usually deep rooted
and saltcedar can also be deep rooted. In
contrast, seepwillow is shallow rooted, growing
only where the groundwater is close to the sur-
face. Arrowweed (Pluchea sericea (Nutt.) Coville)
shrubs send out lateral roots just below the
J surface of the soils which sprout to form dense
clusters over relatively large areas (Gary 1963).
Some seedlings of this species have been noted,
but the dense thickets are probably caused by
lateral spread.

All of the aboveground portions of salt-
cedar will develop adventitious roots and form
new shrubs if kept wet in moist soil. Gary
and Horton (1965) found that 100 percent of

1 stem cuttings would sprout at all times of
the year if they are kept moist and warm. If

I the stem cuttings are allowed to dry, even

1 as short a period as one day, the sprouting
ability is quickly reduced. Root cuttings

|! did not show any signs of sprouting. Wilkinson
(1966), however, reported that a small percent-
age of completely buried root cuttings formed

!' stem sprouts in a mist-bed in the greenhouse.
How frequently similar conditions might occur

j; in the field is not known. Spreading by lateral

I roots from established saltcedar shrubs has

■ never been observed.

After burning or cutting, saltcedar shrubs
I redevelop rapidly; the sprouts from the root
crown will grow as much as 10 to 12 feet in a
year under favorable conditions. In a study
I of the effect of grazing upon resprouting
I tamarisk shrubs, cattle removed approximately

50 percent of the foliage produced. The shrubs
f still grew vigorously, however, and by the
j second year the stand became so dense that
j cattle would not enter the area (Gary 1960) .

Cattle and probably sheep will also browse
heavily on young seedlings as well as the more
mature plants if the stand is open.

Mature saltcedar shrubs are more drought
resistant than the native species. They are
also long lived and will mature into small
trees. In New Mexico, individual trees report-
edly 75 to 100 years of age have not yet shown
signs of deterioration due to age.

VEGETATION OF THE PHREATOPHYTE FLOOD PLAINS

The original vegetation of the flood-plain
areas was determined primarily by the water
supply available to the plant roots. Undoubt-
i edly the rivers flowed rather constantly and the
I water tables were high in much of the valley
area. The area close to the river was usually
dominated by a wide band of trees, principally

Fremont cottonwood (Populus f remontii Wats.),
with associated willows. On the higher ground
were large areas dominated by mesquite. Arrow-
weed was dominant in many areas. In the more
saline sites, there were large patches of
salt- tolerant grasses such as saltgrass
(Distichlis stricta (Torr.) Rydb.), surrounded
by saltbushes (Atriplex spp.) and other
salt-tolerant plants.

The early pioneers used the cottonwood,
mesquite, and other trees and larger shrubs
for fuel and for building their homes. In the
Arizona desert, the lands dominated by mesquite
were some of the best soils in the valley and
were soon cleared for farming. Along the Rio
Grande, the first and finest farmland was created
by removal of cottonwood.

These activities soon removed or at least
greatly reduced the natural wooded areas along
the rivers. Thus, the saltcedar found conditions
ideal for rapid invasion of the flood plains.
Shortly after the turn of the century saltcedar
began spreading aggressively. By the 1940' s,
extensive areas were dominated by saltcedar along
the Gila (Marks 1950, Haase 1972, Turner 1974),
Salt (Turner and Skibitzke 1952, Gary 1965),
and Rio Grande (Campbell and Dick-Peddie 1964)
as well as along the Pecos and Colorado
(Robinson 1965) . It is now also found along
many smaller streams, around springs and seeps,
by roadsides, and in many other areas of the
West wherever there is sufficient moisture to
germinate the seeds and establish the seedlings.

In recent years, much of the land dominated
by saltcedar has been converted to farms or
industrial use near the towns and cities, or
cleared for water salvage projects.

In spite of these major changes, there are
still large areas occupied by wildland vegetation,
although they are usually altered by man. Haase
(1972), in his study of the lower Gila River,
indicates that saltcedar occupies about 50
percent of the total bottom-land area. Under
present conditions he feels this dominance will
not be changed unless there is some marked
fluctuation in the water table or in other
environmental conditions. His analysis and
breakdown of the communities is very similar
to Marks (1950).

Somewhat similar communities were studied
along the Salt River above Granite Reef dam
east of Tempe (Gary 1965) . The saltcedar
communities were separate and distinct fromthe
arrowweed, and occupied sites with shallower
water tables and a silt loam soil, contrasted
to the sandy loam found under the arrowweed and
mature mesquite. There were a few cottonwood

125

trees, but not enough to be included in the
analysis.

Along the Rio Grande, Campbell and
Dick-Peddie (1964) found that saltcedar was
the major dominant in southern New Mexico, but
that cottonwood, Russian-olive (Eleagnus
angustif olia L.), and other species increased
upstream. These authors observed that cotton-
wood assumes dominance over saltcedar if the
cottonwood is left to develop into a full tree
without disturbance. In mature stands of
cottonwood, saltcedar grows only in natural
openings and along the outer edge of the cotton-
wood stand.

Along some flood-plain reaches, dropping
water tables have reduced the stand of salt-
cedar, because ground water is now apparently
out of reach of its roots. In the 1940's a
dense stand of saltcedar extended along the
Salt River from east of Mesa through Tempe
and Phoenix to its confluence with the Gila
River. Shrubs are now growing along this
river only as widely spaced desert-type plants
dependent on floodflows and rain for survival.
In dry periods, these saltcedar shrubs will
make almost no growth and tend to drop their
leaves. They leaf out quickly when water
becomes available, however.

Fires burning through such stands kill
a fairly large number of plants and create an
even more open stand. It is probable, in this
desert climate, that shrubs must be spaced 15
or 20 feet or more apart to have sufficient root
systems to withstand lengthy droughts. A heavy,
dense stand will survive only where the water
table is within 15 or 20 feet of the surface.

Thus, although saltcedar has aggressively
spread over a large portion of the western
flood plains, it has probably reached its
maximum spread or is being reduced in most of
the area. However, it will always threaten to
invade aggressively after any change in local
conditions. Its ecology must be understood
if management of flood-plain vegetation is to
be successful.

FUTURE OF TAMARISK STANDS

Future changes in the vegetation cover of
flood plains now dominated by saltcedar is a
concern of many land managers. The aggressive-
ness of saltcedar suggests that it will remain a
dominant in most areas if conditions remain as
at present and often may invade where conditions
change in other types.

Seeds of cottonwood, willow, and seep-
willow have characteristics similar to salt-
cedar. Thus, in theory they are highly
competitive,, However, saltcedar produces seed

over a much longer period and also can become
established after a summer recession flow when
seeds of the other species are not present.
In addition, tamarisk seedlings can tolerate
drying at an earlier stage and, while often
grazed, are less sought after than cottonwood
and willow. Also the mature shrubs are more
drought resistant which tends to eliminate
many of the competing native shrubs and trees.

With the characteristics of the various
species in mind, let us consider the different
types of vegetation along the rivers. A dense,
mature stand of tamarisk would not have any
bare soil underneath, and thus there would be
no opportunity for regeneration of any species.
Unless subjected to fire or flood, the stand
would not deteriorate. However, if cottonwood
was present in the initial seedling establishment
stage, there can be a gradual increase of domin-
ance of this species as the tree grows. This
relationship can often be observed along the
Rio Grande south of Albuquerque and, very rarely,
at lower elevations such as along the San Pedro
River, south of Winkelman, Arizona. Thus, mature
saltcedar stands should not be expected to yield
to any invading vegetation type unless the water
table drops or the existing stand is altered by
man, fire, or flood.

Lowering water tables may kill a large
portion of the shrubs. The degree of damage
would depend upon the rapidity of the drop and
the depth of the final water table. In some
cases, shrubs may die back but readjust to the
lower groundwater if it stabilizes at 20 feet
or so. The resulting stand after the root
systems are extended downwards may be nearly
as dense as had previously existed.

If the water table is at 5 feet or less,
saltcedar does not develop densely and the inter-
shrub spaces are usually dominated by saltgrass
or Bermudagrass (Cynodon dactylon (L.) Pers.)
Dropping water tables in such an area will allow
the saltcedar to grow dramatically and replace
the grass.

Fire burning through a saltcedar stand
will not kill the shrubs, as they tend to sprout
vigorously unless they are growing under stress.
Then as many as half of the shrubs may not sur-
vive .

Floods do, at times, remove large areas of
saltcedar. If this occurs during or just before
the seeds are flying, seedlings will likely be
established along the edges of the receeding
flows more aggressively than other species.
Sometimes, such as after flash summer floods,
drying is too rapid for seedling survival,,
Buried root crowns or above-ground portions of
branches and smaller stems will often sprout,
however, even if conditions are not favorable
for seedling establishment.

126

STUDIES IN TAMARIX TAXONOMY

The identification of saltcedar and its
proper relationship to Old World form has long
been confused. The taxonomy of Tamarix is
difficult primarily because of the lack of
distinct identifying floral characteristics,
and the great variation among plants in the
same community. Major confusion is caused
by the length of the blooming season (March
to October in desert climates) with changing
inflorescence types and floral characters as
the season progresses. Thus, accurate species
identification requires several collections
from a shrub to sample seasonal variations.

The early floras usually listed T. gallica
as the introduced species. This terminology
continued until McClintock (1951) , in a study
of horticultural tamarisks, stated that J_. gallica
was a rare shrub in the West, and the common
aggressive saltcedar was Tamarix pentandra Pallo

Baum (1966), after extensive study of the
genus Tamarix at the Hebrew Univeristy, Jerusalem
abandoned the name T_. pentandra because it did
not follow the standard rules of nomenclature.
He considered the widespread American tamarisk,
after examining material from various American
herbaria, as consisting of two species: Tamarix
ramosissima Lebed. and Tamarix chinensis Lour.
(Baum 1967). J_. gallica was reported as
occurring mostly on the Texas Gulf Coast.
Tamarix af ricana Poiret and several others were
reported as horticultural species.

After detailed studies of many shrubs of
diverse species and forms obtained from various
American and Old World localities and grown on
the Arizona State Univeristy Farm as well as
herbarium specimens collected elsewhere in the
United States, I feel that our aggressive
saltcedar, though extremely variable, should
be considered as one species and not two as
outlined by Baum. The oldest synonym applied
to the aggressive tamarisk group is Tamarix
chinensis Lour.; thus this name should now
be accepted for the species so commonly
naturalized in the West.

LITERATURE CITED

Baum, 1966. Monographic revision of the genus
Tamarix. Final Res. Rep. for the USDA
Proj. No. A10-FS-9. Dep. Bot. Hebrew
Univ., Jerusalem. 193 p. Processed.

Baum, Bernard R. 1967. Introduced and

naturalized tamarisks in the United States
and Canada (Tamaricaceae) . Baileya 15:
19-25.

Campbell, C.J. and W.A. Dick-Peddie. 1964.

Comparison of phreatophyte communities on
the Rio Grande in New Mexico. Ecology
45(3):492-502.

Gary, Howard L. 1960. Utilization of five-
stamen tamarisk by cattle. U.S. Dep.
Agric, For. Serv., Rocky Mt . For. and
Range Exp. Stn. , Res. Note 51, 4 p.

Gary, Howard L. 1963. Root distribution of
five-stamen tamarisk, seepwillow, and
arrowweed. For. Sci. 9:311-314.

Gary, Howard L. 1965. Some site relations in
three flood-plain communities in central
Arizona. J. Ariz. Acad. Sci. 3 (4) : 209-212 .

Gary, Howard L, . and Jerome S. Horton. 1965.
Some sprouting characteristics of five-
stamen tamarisk. U.S. For. Serv. Res. Note
RM-39, 7 p. Rocky Mt. For. and Range Exp.
Stn., Fort Collins, Colo.

Haase, Edward F. 1972. Survey of floodplain
vegetation along the lower Gila River in
southwestern Arizona. J. Ariz. Acad. Sci.
7(2) :66-81.

Horton, Jerome S. 1964. Notes on the intro-
duction of deciduous-tamarisk. U.S. For.
Serv. Res. Note RM-16, 7 p. Rocky Mt. For.
and Range Exp. Stn., Fort Collins, Colo.

Horton, Jerome S. 1973. Evapotranspiration
and watershed research as related to
riparian and phreatophyte management, an
abstract bibliography. U.S. Dep. Agric.
Misc. Pub. 1234, 192 p.

Horton, Jerome S., and C. J. Campbell. 1974.
Management of phreatophyte and riparian
vegetation for maximum multiple use values.
USDA For. Serv. Res. Pap. RM-117, 23 p.
Rocky Mt. For. and Range Exp. Stn., Fort
Collins, Colo.

Horton, Jerome S., F. C. Mounts, and J. M.

Kraft. 1960. Seed germination and seedling
establishment of phreatophyte species. USDA
For. Serv., Rocky Mt. For. and Range Exp.
Stn. , Stn. Pap. 48, 26 p.

Marks, John Brady. 1950. Vegetation and soil
relations in the lower Colorado desert.
Ecology 31(2) :176-193.

McClintock, Elizabeth. 1951. Studies of
California plants. 3. The tamarisks.
J. of Calif. Hort. Sci. 12:76-83.

Robinson, T. W. 1965. Introduction, spread,
and areal extent of saltcedar (Tamarix)
in western states. U.S. Geol. Surv. Prof.

Schuster, Joseph L. (Ed. ) 1969. Literature
on the mesquite (Prosopis L.) of North
America. An annotated bibliography „
Texas Tech. Univ., Lubbock, Tex. Spec.
Rep. 26, 84 p.

Turner, Raymond M. 1974. Quantitative and

historical evidence of vegetation changes
along the upper Gila River, Arizona. U.S.
Geol. Surv. Prof. Pap. 655-H, 20 p.

Turner, S.F. and H.E. Skibitzke. 1952. Use
of water by phreatophy tes in 2,000 foot
channel between Granite Reef and Gillespie
Dams, Maricopa County, Arizona. Am.
Geophys. Union Trans. 33(1): 66-72.

Wilkinson, Robert E. 1966. Adventitious shoots
on saltcedar roots. Bot. Gaz. 127 (2-3):
103-104.

127

Avian Use of
Saltcedar Communities in the
Lower Colorado River Valley1

2 3 4-

Bertin W. Anderson_/ , Alton Higgins_/, and Robert D. Ohmart_/

Abstract. — Bird densities and bird species diversities
(BSD) in saltcedar (Tamarix chinensis) stands of the lower
Colorado River Valley were determined on a seasonal basis from
May 1974 through February 1977. Comparisons were made between
six saltcedar structural types as well as on a community level
with seven other vegetation types. A method of determining the
relative value of the communities, as well as the saltcedar
structural types, based on density, density with 10 percent
doves, BSD, BSD with 10 percent doves, number of species,
structural diversity, and size of census area is described.
Results showed the saltcedar community supported fewer birds
than native communities, although tall, dense stands were
valuable for nesting doves and rarer bird species in riparian
communities along the lower Colorado River.

INTRODUCTION

Events of the past century have resulted
in tremendous changes in the flora and fauna
of the lower Colorado River Valley. The
Colorado River has been channelized and
controlled, and vast stretches of honey mesquite
(Prosopus j ulif lora) have been converted to
agricultural use — a practice which has continued
at an accelerated rate in the past few years.
These conditions have favored the Brown-headed
Cowbird (Molothrus ater) and have reduced or
extirpated the breeding populations of such
species as the Yellow Warbler (Dendroica
petechia) and Bell Vireo (Vireo bellii) .

This loss of habitat has also been
accompanied by a deterioration of the remaining

1/ Paper presented at the symposium on
Importance, Preservation and Management of
Riparian Habitat, Tucson, Arizona 9 July 1977.

2/ Faculty Research Associate, Arizona
State University, Dept. Zoology and Center for
Environmental Studies, Tempe, Arizona 85281.

3/ Field Biologist, Arizona State
University, Dept. Zoology and Center for
Environmental Studies, Tempe, Arizona 85281.

4/ Associate Professor of Zoology,
Arizona State University, Dept. Zoology and
Center for Environmental Studies, Tempe,
Arizona 85281

bottomland by the now well-entrenched exotic
saltcedar (Tamarix chinenses) . First recorded
in Arizona in the late 1800' s, saltcedar was
not an important species until after 1910
(Robinson 1965). Nevertheless, it is now
present in pure communities or mixed with
virtually all riparian community types, being
absent only from a few stands of honey mesquite .
Knowledge concerning those avian species which
utilize saltcedar is essential for those
agencies involved with river or riparian
vegetation management.

Areas containing saltcedar are regularly
swept by fire, as demonstrated by the fact that
21 of the 25 stands involved in our study have
burned during the last 15 years. The other four
stands of saltcedar developed after some other
form of severe disturbance. Many of these areas
obviously supported another community type in
the past. Saltcedar is a fire-adapted species
and shows a greater recovery rate than the
native riparian species. Willow (Salix
gooddingii) and arrowweed (Tesseria sericea)
respond quickly after fire while honey mesquite
shows slower growth. Species such as cotton-
wood (Populus f remontii) are killed during fire.
With the initiation of a burn cycle, the
dominance of an area by saltcedar becomes
successively more complete (see Horton, these
proceedings) .

128

METHODS

Structural Types

The saltcedar community (stands in which
saltcedar is virtually the only tree was
divided into six structural types, based on
distribution and density of foliage at varying
heights, as explained elsewhere in these pro-
ceedings (Anderson, Engel-Wilson, Wells, and
Ohmart) . Structural types IV and V (trees
not dense and seldom taller than 5 or 3 m,
respectively) represent typical stands found
in the lower Colorado River Valley. Data were
gathered in these areas from the summer of
1974 through February 1977. Beginning in
1976 data were gathered from about 18
transects averaging over 0.8 km in length,
using censusing techniques described by
Anderson, Engel-Wilson, Wells, and Ohmart
(these proceedings) . These included one
transect in structural type I (dense vegetation
at 10 to 20 m) , established in March 1976;
one transect in type II (dense vegetation at
5 to 10 m) , established in June 1975; two
transects in type III (trees dense, seldom
exceeding 6 m) , established in March 1976;
four transects in type IV, eight transects
in type V and two in type VI (sparse vegetation
representing regrowth after disturbance) ,
established in 1974.

Ranking Technique

We developed the ranking technique for
assessing the relative value of structural
types of saltcedar stands and of saltcedar
compared to other community types. A rank
value for bird density in the structural types
of saltcedar was determined using all doves
and 10 percent doves, by assigning the smallest
value (1) to the structural type with the
greatest density and the largest value (n) to
the one with the smallest density. This was
repeated using numbers of species and BSD with
all doves and with 10 percent doves. A mean
rank for these parameters was calculated for
five seasonal periods throughout the year.
The average of these seasonal values was the
rank assigned to a particular type.

The relative value of saltcedar compared
with other community types was achieved by
assigning the smallest score to the community
type with the greatest average density (or
number of species, or BSD's, all structural
types combined) and the largest score to that
community type with the smallest density (or
number of species, or BSD's) as described above
for saltcedar structural types.

The number of species may increase with
the diversity or size of area censused. We
attempted to compensate for this by ranking
the most heterogeneous community or structural
types with the greatest diversity or largest
census area last.

We assumed that each of the parameters
considered were of equal importance, a point
of potential contention.

RESULTS

Densities and Diversities

Types IV and V, 1974-76

Data for three consecutive summers (May,
June, July) from structural types IV and V
were fairly consistent. Large dove densities
in type IV in 1976 and in type V in 1975
(Table 1) resulted in relatively depressed
BSD's in those years. Type IV diversities
with 10 percent doves were higher than those
of type V in 1974 and 1975. Dove densities
for 1976 in type IV increased threefold from

1975, depressing the diversity value just
lower than that of type V.

Fall (October, November) data for type V
were similar in the first two years but showed
a rather dramatic increase in all parameters
in 1976 (Table 1) . Type IV showed greater
values in 1975 than in 1974 and increased
further in 1976. Few doves were present at
this time of year and this is reflected in
slight differences in BSD's with 10 percent
doves and BSD. We feel that increased
densities in the fall of 1975 and 1976 over
1974 can be traced in part to the much milder
conditions which existed during the late fall
and winter seasons, allowing increased and
prolonged use of the saltcedar community,
particularly by small wintering insectivores .

Diversity values appeared to be more
closely correlated with the structural
parameters than were densities or species
numbers. For example, in the fall of 1974 and

1976, densities in type V were greater than
those of type IV; the reverse was true in 1975.
Diversity values, however, were always greater
in type IV.

Types I - VI, 1976

Bird densities in the six saltcedar
structural types generally follow the same
annual pattern of low winter numbers, increasing
in the spring and peaking in the summer (Table
2) . Densities dropped in late summer and
continued dropping through the following winter.
Spring (March-April) densities were apparently

129

Table 1. — Summer and fall densities, diversities, and number of species in saltcedar.
types IV and V, lower Colorado River Valley 1974-1976.

Structural

Structural Year Density Density with BSD BSD with Number

Type 10% Doves 10% Doves Species

SUMMER
IV
IV

IV
V
V
V

1974
1975
1976
1974
1975
1976

120
126
241
129
182
131

64
77
98
91
120
86

2801
4377
9255
4135
4022
4411

2.7009
2.6237
2.6055
2.5871
2.5631
2.6760

19
19
18
21
22
20

FALL
IV
IV
IV
V
V
V

1974
1975
1976
1974
1975
1976

42
76

105
60
75

110

40
75

103
55
63

110

3878
4033
6336
2369
0881
5772

.3062
,3644
,5934
.1274
.0126
,5772

14
16
22
14
14
22

related to structure. Abert's Towhee (PiRilo
aberti) provides a good example of a species
whose density was strongly correlated with
structure in saltcedar, with 1, 5, 14, 19
and 27 birds per 40 ha in types V through I
respectively .

The preference of nesting doves for
dense vegetation at 3 to 6 m is strongly
reflected in the bird density value of type II
saltcedar in the summer. There were, in
fact, as many doves in this type as birds of
all species in most of the other structural
types .

Type II continued to show a large dove
population In late summer (August-September) ,
although types I and III had higher populations
of birds of other species. Type V had the
greatest diversity values and a relatively
large number of species, but by far the lowest
density .

The dove population was extremely low in
the fall. Diversities and numbers of species,
however, continued to show an inverse relation-
ship with structure as in late summer.

Densities during the winter (December-
January-February) season of 1976-77 were high
compared with the fall, and especially high
compared with the previous winter (Table 2) .
The majority of these birds, however, were
small insectivores such as the Ruby-crowned
Kinglet (Regulus calendula) , Orange-crowned

Warbler (Vermivora celata) , and Yellow-rumped
Warbler (Dendroica coronata) . As previously
mentioned, the relatively mild winter was at
least partly responsible for the densities of
these birds. The monthly totals for these
species in the saltcedar community as a whole
decreased throughout the winter, whereas the
total found in the cottonwood-willow community
was higher in January and February than it was
in December. This demonstrates that cottonwood
willow maintained a high value for these specie
throughout the winter — unlike saltcedar
(Table 3).

COMPARISON OF COMMUNITIES

Knowledge of the value of the different
saltcedar structural types is necessary, but
more important is the relative value of the
saltcedar community as compared with other
community types — many of which are either being
displaced by saltcedar or lost in other ways.
Communities to be compared include six riparian
communities as well as desert wash and citrus
orchard communities.

Community Densities

Bird densities in saltcedar were consis-
tantly greater than those in arrowweed only
(Tables 4 and 5) while numbers of species in
seasons other than winter were comparable with
other communities (Table 6) . Winter densities

130

Table 2. — Densities, diversities and number of species in six saltcedar structural types, lower
Colorado River Valley, December 1975 - February 1977.

Structural December March May August October December

Type January April June September November January

February July February

Total Density (N/40 ha)
I

II 42
III

IV 25

V 29

VI 293

146
111
101
39
54
89

290
503
316
241
131
226

213
363
296
187
89
280

165
268
129
105
110
171

107
275
119
50
L25
153

Density 10% Doves (N/40 ha)
I

II
III
IV
V
VI

20
27
132

136
91
81
28
50
83

193
238
156
98
86
157

183
177
239
155

75
104

165
267
129
103
110
95

107
272
115
49
125
103

BSD

I
II
III
IV

V
VI

2.0383

2.4850
2.5825
1.6514

2.1739
2.2129
1.7179
2.5366
2.4147
2.2435

2.5036
2.0411
1.8521
1.9266
2.4411
2.5269

1.8976
1.8211
2.3969
2.4985
2.7965
1.5361

1.9097
2.2582
2.3934
2.6336
2.5772
1.8744

7062
9667
2683
3853
0141
0174

BSD 10% Doves
I

II
III

. IV
V
VI

1.9487

2.5191
2.5284
2.5715

2.0930
2.0597
1.6070
2.6939
2.3437
2.1687

2.7312
2.5425
2.1643
2.6055
2.6760
2.6214

1.7129
2.1160
2.3143
2.3894
2.7972
2.7467

1.9097
2.2397
2.3934
2.5934
2.5772

7062
9367
2128
3272
9908

2.5127

2.2271

Number Species (N/40 ha)
I

II 8
III

IV 12

V 15

VI 21

12
13
6
15
13
17

25
20
19
18
20
24

8
13
26
23
24
25

12
18
20
22
22
21

11
19
18
14
19
23

of birds in saltcedar are greater than those
found in saltcedar-honey mesquite and arrow-
weed but included the greatest percentage and
nearly the greatest dove densities of all
communities. Densities decreased from summer
through the winter while densities with 10
percent doves remained fairly stable through
the fall. Although doves comprised fully
50 percent of the summer density in the salt-
cedar community, there were actually more
doves in all of the other community types,

excepting arrowweed and desert washes. There
was a distinct relationship between departure
of doves and rising BSD values from August
through November (Tables 7 and 8) . Bird
densities in honey mesquite rose sharply in
October-November, and bird densities in desert
wash and saltcedar-honey mesquite not only
increased from late summer to fall, but the
greatest number of species occurred at this
time .

131

Table 3. — Winter densities of small insectivores in cottonwood-willow and saltcedar communities,
lower Colorado River Valley, 1976.

Community Month

Ruby-crowned

Yellow-rumped Orange-crowned Total

Percent of

Kinglet

Warbler

Total Population

Cottonwood- Dec

323

258

664

Willow Jan

340

516

984

Feb

321

327

109

757

Saltcedar Dec

152

535

734

Jan

176

130

338

Feb

100

195

Table 4. — Total densities for eight

community types December 1975

- November 19/6,

lower

Colorado River Valley.

Community Dec,

Jan, Feb

Mar, Apr

May, June, July

Aug, Sept

Oct, Nov

Cottonwood-

Willow

148

172

336

262

210

Screwbean

Mesquite

109

318

307

183

Honey Mesquite

193

323

195

270

Saltcedar-

Honey Mesquite

111

295

184

177

Saltcedar

216

177

129

Desert Wash

115

176

118

185

Arrowweed

124

141

Orchard

158

678

540

135

Table 5. — Densities including 10% doves for eight

community types

December 19/5 -

November iy/o,

lower Colorado River

Valley.

Community Dec,

Jan, Feb

Mar , Apr

May, June, July

Aug, Sept

Oct, Nov

Cottonwood-

Willow

134

151

223

195

201

Screwbean

Mesquite

174

218

159

Honey Mesquite

161

166

169

148

265

Saltcedar-

Honey Mesquite

170

151

176

Saltcedar

119

126

120

Desert Wash

106

121

185

Arrowweed

101

135

Orchard

144

132

178

128

132

Table 6. — Number of species for eight community types found in the lower Colorado River Valley from
December 1975 through November 1976.

Community

Dec, Jan, Feb

Mar, Apr May, June, July

Aug, Sept

Oct, Nov

Cottonwood-

Willow
Screwbean

Mesquite
Honey Mesquite
Saltcedar-

Honey Mesquite
Saltcedar
Desert Wash
Arrowweed
Orchard

16
19

16
10
16
8
17

27
30

20
19
20
13
20

24
22

20
25
20
21
18

33
28

19
27
21
23
25

26
27

22
26
30
18
17

Table 7. — BSD for eight community types found in the lower Colorado River Valley from December 1975
through November 1976.

Community

Dec, Jan, Feb

Mar, Apr May, June, July

Aug, Sept

Oct, Nov

Cottonwood-

Willow

7401

1762

8494

1817

9502

Screwbean

Mesquite

6422

9067

4015

4451

8087

Honey Mesquite

1850

8608

1850

6826

6206

Saltcedar-

Honey Mesquite

5428

4575

327

5476

5095

Saltcedar

8071

8537

405

7038

8167

Desert Wash

5047

2293

364

5706

7706

Arrowweed

9652

4643

665

7037

5160

Orchard

8823

0460

693

3052

2837

Table 8. — BSD with 10% doves for eight community types found in the lower Colorado River Valley from
December 1975 through November 1976.

Community

Dec, Jan, Feb

Mar , Apr

May, June, July

Aug, Sept

Oct, Nov

Cottonwood-

Willow

2.6941

3.2125

2225

3.3940

2.9067

Screwbean

Mesquite

2.7721

3.1914

9040

2.4758

2.8263

Honey Mesquite

2.1561

2:8937

8276

2.7997

2.5844

Saltcedar-

Honey Mesquite

2.4869

2.4888

675

2.4660

2.4971

Saltcedar

2.6848

2.8443

883

2.8544

2.7827

Desert Wash

2.4718

2.1516

564

2.5943

2.7706

Arrowweed

1.9652

2.4229

6108

2.6562

2.5160

Orchard

1.7105

2.2931

897

2.3454

2.1917

133

USE BY VARIOUS GUILDS

The percentage of the total lower Colorado
River Valley population of sixteen of the more
common breeding species (representing six
guilds) which would occur in saltcedar, using
40 ha of each of the six riparian community
types, should approximate 16.6 percent (1/6
the population of a species) if there were no
selection for a particular vegetative type by
any of these species, i.e. if they were evenly
distributed in all community types. Two of
three small (<15 gm) insectivorous species
apparently exhibited no selection against salt-
cedar (Table 9) , occurring in densities at or
slightly above the expected. Woodpeckers
demonstrated much less flexibility in adapting
to saltcedar, possibly as a result of body
size in relation to tree limbs and trunks
suitable for making nest cavities. The Ladder-
backed Woodpecker (Picoides scalaris) , the
smallest species, was more common than the Gila
Woodpecker (Melanerpes uropygialis) ; the Common
Flicker (Colaptes auratus) , the largest species,

did not occur in saltcedar at all. Fifty to
86 percent of the population of three medium-
sized insectivores were found in structural
types I and II, but only the Summer Tanager
(Piranga rubra) used saltcedar to any signifi-
cant extent (Table 9). The density of Abert's
Towhee was slightly above that which would be
expected by chance while other ground feeders
(Gambel's Quail, Lophortyx gambelii and the
Crissal Thrasher, Toxostoma dorsale) were
slightly below expected values (Table 9) . All
four species of granivores occurred at greater
than expected levels although a constantly wet
condition was probably the greater attractant
for the Song Sparrow (Melospiza melodia) ,
considered here to be a granivore.

OVERALL VALUE OF SALT CEDAR TO BIRDS

Value of Structure

The structural types found in saltcedar
(all types except type VI) were ranked to

Table 9. — Differential use of community and structural types by foraging guilds of birds in the
lower Colorado River Valley, 1976.

Species Total/240 ha Total in % total in % population in each structure,

all saltcedar saltcedar all communities

communities 40 ha I II III IV V VI

Small Insectivores

Verdin

108.

,60

10,

,17

15.

20,

25,

15.2

18.

Lucy's Warbler

87.

,37

17.

,50

20.

23.

29.

15,

15.

11.1

Black-tailed Gnatcatcher

34.

,03

.83

17.

27.

23.

20.8

Woodpeckers

Ladder-backed Woodpecker 31.80 2.17 6.8 26 . 9 29 . 9 15 . 1 12 . 1 6.2 9.9

Gila Woodpecker 11.57 .17 1.4 42.0 31.3 11.1 5.9 3.7 5.9

Medium-sized Insectivores

Northern Oriole

46.

.67

12.

18.

31.

22.

14.

Summer Tanager

,83

.00

34.

73.

13.

11.

Yellow-breasted Chat

11.

.17

33.

37.

17.

10.

Cactus Wren

.67

11.

23.

20.

16,

20.

Ground Feeders

Abert's Towhee

102.

20.

12.

25.

29.

13.

Crissal Thrasher

21.

.83

13.

21.

25.

21.

11.

Gambel's Quail

93.

.25

13.

14.

23.

12.

18.

24.

21.

ranivores

Song Sparrow

.17

21.

69.

21.

Blue Grosbeak

40.

,78

,50

23.

15.

22.

19.

14.

12.

16.

House Finch

.17

53.

56,

15,

Brown-headed Cowbird

105.

18.

.50

17.

18.

27.

20.

14.

11,

Lycatchers

Ash-throated Flycatcher

65.

.52

.83

10.

24.

17.

12.

17.

Western Kingbird

,37

.00

12.

45.

12.

17.

15.

134

determine their relative value. Type II can
be seen to be, overall, the preferred structure
by birds in general, followed by types I, III,
V, and IV, the values of the last two being
very close (Table 10). The changes in the avian
community that occurred when saltcedar reached
a structure of type II or I were significant
not only in terms of increasing densities of
some birds but also in the addition of species.
For example, the White-winged Dove (Zenaida
asiatica) and the Mourning Dove (Zenaida
macroura) , Abert's Towhee , Lucy's Warbler
(Vermivora luciae) and the Black-chinned
Hummingbird (Archilochus alexandri) were much
more abundant in type II than in type III.
Type I attracted the Song Sparrow and relatively
high densities of Abert's Towhee as well as
the Summer Tanager and Yellow-breasted Chat
(Icteria virens) in the summer.

Relative Community Value

The communities, including the two non-
riparian communities, were analyzed to determine
their overall relative value to birds during
1976 using the "ranking" technique discussed
above. Ranked in this way cottonwood-willow
communities proved the most valuable, followed
by honey mesquite, screwbean mesquite, salt-
cedar-honey mesquite, desert wash, saltcedar,
orchard and arrowweed (Table 10). Since
orchards do not represent a naturally occurring
community, it can be seen that saltcedar is
only slightly more valuable than arrowweed
(Table 11).

DISCUSSION

It has been demonstrated that the salt-
cedar community does not compare favorably
with essentially native communities (except
arrowweed, which lacks trees). Nevertheless,

in the face of present environmental conditions
and continuing loss of native vegetation, a
concomitant increase in the proportion of the
riparian habitat dominated by saltcedar is
inevitable. Of particular interest was the
comparison between saltcedar and orchards.
The occurrence of these communities in the
lower Colorado Valley has been relatively
recent, and both present a uniform monoculture
regardless of structural types. The birds
have thus responded in a similar overall manner
to these exotic communities.

Although it would appear that few species
of birds are actually attracted to saltcedar
during the breeding season, the addition of
one or more of the native tree species, even
in small numbers, would no doubt greatly enhance
the overall attractiveness of an area. Addition
of cottonwood or willow trees would add nest
site potential, an important community compo-
nent, especially for the woodpecker and
flycatcher guilds. Screwbean or honey mesquite,
if infested with mistletoe, would attract
frugivores, a guild entirely missing from pure
saltcedar .

Managing areas of saltcedar for structural
types I and II appears to have significant
potential (Ohmart and Anderson, MSjV). Salt-
cedar type II and mature orchards support the
greatest densities of doves (Mourning Dove in
orchards, both species in saltcedar type II),
which are important game species in the lower
Colorado River Valley. Saltcedar type I
provides a habitat for avian species which are
normally restricted to cottonwood-willow
communities, such as the Summer Tanager, and
is another important reason land managers
should strongly consider managing saltcedar
communities. Fire prevents saltcedar from
reaching maturity and/or persisting as mature
communities for any length of time along the
lower Colorado River. Maintenance of mature

Table 10. — Relative value of saltcedar structural types to birds as determined by Ranking Technique ,
lower Colorado River Valley, March 1976-February 1977. Lower rank indices indicate greater
relative value.

Structural

Density

BSD

Number

Size of

Grand

Type

Density

10% Doves

BSD

10% Doves

Species

Census Area

Rank

2.6

2.2

3.8

4.0

1.0

2.90

1.2

1.6

3.6

2.6

1.2

2.30

III

2.6

3.6

3.4

3.2

2.8

3.03

4.6

1.8

1.4

2.8

4.0

3.20

4.2

4.0

1.0

1.8

5.0

3.13

5/ Manuscript in preparation discussing
management alternatives of saltcedar communities
for wildlife.

135

Table 11. — Relative value of eight community types in 1976 using Ranking Technique. Lower rank
indices indicate greater relative value.

Total Density

Density with

10% Doves

Number of Species

Honey Mesquite

. 0 Cottonwood-Willow 2.0

Cottonwood-Willow

.4 Honey Mesquite

2.4

Honey Mesquite

Orchard

.6 Screwbean Mesquite 3.8

Screwbean Mesquite

.0 Orchard

4.0

Desert Wash

Saltcedar-

Honey Mesquite

.4 Honey Mesquite 4.4

Saltcedar

Desert Wash

.4 Desert Wash

4.8

Saltcedar-

Honey Mesquite

Saltcedar

.4 Saltcedar

6.8

Orchard

Arrowweed

. 8 Arrowweed

7.8

Arrowweed

Bird Species

Diversity (BSD)

BSD with 10% Doves

Community Diversity

Grand Rank

Cottonwood-

Saltcedar-

Cottonwood-

Willow 1.0

Willow 1.2

Honey Mesquite

Willow

Screwbean

Honey

Mesquite 3.2

Mesquite 2.6

Arrowweed

Mesquite
Screwbean

Saltcedar 3.4

Saltcedar 2.7
Honey

Desert Wash

Mesquite
Saltcedar-

Arrowweed 4.2

Mesquite 4.0

Orchard

Honey Mesquite

Honey

Mesquite 4.8

Arrowweed 5.2

Mesquite
Screwbean

Desert Wash

Desert Wash 4.8

Desert Wash 5.4

Mesquite

Saltcedar

Saltcedar-

Honey Mesquite 4.8

Honey Mesquite 5.4

Saltcedar
Cottonwood

Orchard

Orchard 7.8

Willow

Arrowweed

saltcedar communities

for 20 or more years

LITERATURE CITED

would enhance the overall value of this plant

species for birds.

Robinson, T. W

1965. Introduction, spread, and areal extent
of salt cedar (Tamarix) in western states.
ACKNOWLEDGEMENTS U.S. Geol. Surv./Prof. Pap. 491-A, 12p.

We wish to thank the many field biologists
who have helped in collecting data. We are
grateful to Jack Gildar for computerizing the
data. The efforts of the secretarial staff in
typing early drafts and of Penny Dunlop and
Katherine Hildebrandt in typing the final
manuscript are greatly appreciated. We thank
Jane Durham, Jake Rice, James Bays, and Jeannie
Anderson for critically reading early drafts
of the manuscript. The" research was funded
through grant number 14-06-300-2415 from the
U. S. Bureau of Reclamation.

136

Influences of Riparian Vegetation
on Aquatic Ecosystems
with Particular Reference to
Salmonid Fishes
and Their Food Supply1 2

William R. Meehan, Frederick J. Swanson, and James R. Sedell

Abstract. — The riparian zone has important influences
on the total stream ecosystem including the habitat of
salmonids . Shade and organic detritus from the riparian
zone control the food base of the stream and large woody
debris influences channel morphology. Temporal and spatial
changes in the riparian zone, the indirect influences of
riparian vegetation on salmonids, and the effects of man's
activities are discussed.

INTRODUCTION

Streamside vegetation strongly influ-
ences the quality of habitat for anadromous
and resident coldwater fishes. Riparian veg-
etation provides shade, preventing adverse
water temperature fluctuations. The roots of
trees, shrubs, and herbaceous vegetation sta-
bilize streambanks providing cover in the form
of overhanging banks. Streamside vegetation
acts as a "filter" to prevent sediment and
debris from man's activities from entering
the stream. Riparian vegetation also directly
controls the food chain of the stream eco-
system by shading the stream and providing
organic detritus and insects for the stream
organisms .

Contributed paper, Symposium on the
Importance, Preservation and Management of the
Riparian Habitat, July 9, 1977, Tucson, Arizona.
2

The work reported in this paper was
supported in part by National Science Foun-
dation Grant No. 7602656 to the Coniferous
Forest Biome and Grant No. BMS75-07333 to
the River Continuum Project. This is con-
tribution No. 283 from the Coniferous Forest
Biome and No. 5 from the River Continuum
Project.
3

Research Fishery Biologist, USDA For-
est Service, Pacific Northwest Forest and
Range Experiment Station, Forestry Sciences
Laboratory, Corvallis, Oregon 97331; Research
Associate and Research Assistant Professor,
Oregon State University, Corvallis, 97331.

WHAT IS RIPARIAN VEGETATION?

Riparian vegetation is at the interface
between aquatic and terrestrial environments.
It has, therefore, been defined and examined
from a number of perspectives. Plant ecolo-
gists speak in terms of riparian species
and plant communities. The riparian zone
may also be defined geographically in terms
of topography, soils, and hydrology. We pre-
fer to take a functional approach; that, is,
to consider riparian vegetation as any extra-
aquatic vegetation that directly influences
the stream environment.

Consequently, in defining riparian vege-
tation we must consider the full scope of its
biological and physical influences on the
stream. Riparian vegetation regulates the
energy base of the aquatic ecosystem by shading
and supplying plant and animal detritus to the
stream. Shading affects both stream temper-
ature and light available to drive primary
production; therefore, the balance between
autotrophy and heterotrophy is determined by
multiple functions of riparian vegetation.

Although imperfect , the stream order
system (Leopold et al. 1964) is a useful way
to classify elements of a drainage system.
In small and intermediate-sized streams (up
to about fourth-order) in the Pacific North-
west, riparian vegetation exercises important
controls over physical conditions in the stream
environment. Rooting by herbaceous and woody
vegetation tends to stabilize streambanks,
retards erosion, and, in places, creates over-

137

hanging banks which serve as cover for fish.
.Above ground woody riparian vegetation is an
obstruction to highwater streamflow, sediment
and detritus movement, and is a source of
large organic debris. Large organic debris
in streams (1) controls the routing of sedi-
ment and water through the system, (2) defines
habitat opportunities by shaping pools, riffles,
and depositional sites and by offering cover,
and (3) serves as a substrate for biological
activity by microbial and invertebrate organ-
isms (Triska and Sedell 1976; Swanson et al.
1976; Sedell and Triska 1977; Anderson et al.
in press) .

The influences of riparian vegetation on
coniferous forest stream ecosystems in the
Pacific Northwest are summarized in figure 1.
In a functional approach to defining riparian
vegetation, all floodplain vegetation as well
as trees on hillslope areas which shade the
stream or directly contribute coarse or fine
detritus to it are considered part of the
riparian zone. In the Pacific Northwest,
vegetation in the zone of riparian influence
includes herbaceous ground cover, understory
shrubby vegetation (commonly deciduous), and
overstory trees on the flood plain (generally
deciduous) and on hillslopes (generally conif-
erous ) .

VARIATIONS OF THE RIPARIAN ZONE IN
TIME AND SPACE

The character and importance of riparian
vegetation varies in time and space. Temporal
variation involves patterns of vegetative
succession following disturbances. Major
processes of vegetation disturbance include
wildfire and clearcutting (important to up-
slope vegetation) and damage due to impact of
sediment and floating ice or organic debris
during flood flows. Spatial variation occurs
along the continuum of increasing stream size
from small headwater streams to large rivers.

Temporal Variations of Riparian Zones

The effectiveness of a riparian zone in
regulating input of light, dissolved nutrients,
and litterfall to the stream varies through
time following wildfire, clearcutting, or
other disturbances (fig. 2). In the first
decade or two following deforestation, stream-
side vegetation may increase in height growth
and biomass more rapidly than upslope commun-
ities. Shading of the stream by riparian
vegetation gradually diminishes the potential
for aquatic primary production until maximum
canopy closure. Deciduous shrubs and trees
within the riparian zone will contribute most

Boundaries of Riparian
Zone

•••••• Xv:-.-r tills:

Hillslppe.

SITE

above ground -
above channel

stream banks

floodplain

RIPARIAN VEGETATION
COMPONENT FUNCTION

canopy S stems

1. Shade- controls temperature 8
in stream primary production

2. Source of large and fine plant

detritus

3. Source of terrestrial insects

in channel large debris I. Control routing of water and

derived from sediment

riparian veg 2. Shape habitat- pools, riffles,

cover

3. Substrate for biological activity

roots

1. Increase bank stability

2. Create overhanging banks -cover

stems S low
lying canopy

Retard movement of sediment,
water and floated organic
debris in flood flows

Figure 1. — Extent of riparian zone and functions of riparian vegetation as they relate to
aquatic ecosystems.

138

Figure 2. — Changes in the riparian zone through
time.

of the litter inputs during early watershed
recovery. These deciduous inputs will more
readily decompose than coniferous litter which
dominates inputs late in watershed recovery
and in old-growth forests (Sedell et al. 1975;
Triska and Sedell 1976).

The temporal development of riparian zones
causes a shift in the energy base of the stream
from algae to deciduous leaves to a combination
of deciduous and coniferous leaves. The last
stage in riparian succession is a complex mosaic
of coniferous overstory, deciduous shrub layer,
and herbaceous ground cover. Streams flowing
through older, stratified forests receive the
greatest variation in quality of food for
detritus-processing organisms. Herbaceous
vegetation is high in nutrient content, low in
fiber, and utilizable by stream organisms as
soon as it enters the stream. Leaves from the
deciduous shrub layer are higher in fiber con-
tent and take 60 to 90 days after entering
the stream to be utilized fully by stream
microbes and insects. The conifer leaves take
180-200 days to be processed. Thus there is a
sequencing of utilization of inputs from these

three distinctive riparian strata. The re-
sults for the stream are rich and diverse
populations of aquatic insects which are
keyed into the timing and varied quality of
the detrital food base.

Spatial Variation of Riparian Zones

A stream should be viewed as a continuum
from headwaters to mouth (Vannote, personal
communication; Cummins 1975, 1977). The in-
fluence and role of riparian vegetation will
vary with stream order and position along the
continuum. Some broad characteristics of
streams and rivers are depicted diagrammati-
cally in figure 3.

Extensive networks of small first to third
order streams comprise about 85 percent of the
total length of running waters (Leopold et al.
1964) . These headwater streams are maximally
influenced by riparian vegetation (the ratio
of shoreline to stream bottom is highest) , both
through shading and as the source of organic
matter inputs. Even in grasslands, the dis-
tribution of trees and shrubs follows perennial
and, occasionally, intermittent watercourses
except where land use practices have resulted
in removal or suppression of riparian vegetation.

These low light, high gradient, constant
temperature headwater streams receive signi-
ficant amounts of coarse particulate matter
(CPOM > 1-mm diameter). Their most striking
biological features are the paucity of green
plant life or primary producers (algae and
vascular plants) and the abundance of inver-
tebrates that feed on CPOM (Cummins 1974,
1975) . Shredders reduce detritus particle
size by feeding on CPOM and producing feces
which enter the fine particulate organic
matter (FPCM < 1-mm diameter) pool.

Although the transition is gradual and
varies with geographical region, the shift
from heterotrophy to autotrophy usually occurs
in the range of third- to fourth-order streams
(fig. 3). Rivers in the range of fourth- to
sixth-order are generally wide and the canopy
of riparian vegetation does not close over
them. Direct inputs of CPOM from the riparian
zone are lower in larger rivers because of the
reduced ratio of length of bank to area of
river bottom.

The importance of floodplain vegetation
(mainly deciduous) increases relative to the
hillslope species (mainly coniferous) and in a
downstream direction. Generally this is so
because the floodplain width increases down-
stream and the canopy opening over larger
streams allows greater arboreal expression of
deciduous riparian vegetation. Development
of deciduous riparian trees is suppressed by
shade along small streams.

139

Figure 3. — A diagrammatic
representation of some of
the changes that occur in
running water systems
from headwaters to mouth.
The organisms pictured are
possible representatives
of the various functional
groups occurring in the
size ranges of streams and
rivers. Although a large
network of smaller tribu-
taries coalesce into
larger rivers, the system
is shown diagrammatically
as a single headwater
through all orders to the
river mouth (orders and
approximate ranges of
stream or river width are
shown at the left margin) .
The decreasing direct in-
fluence of the adjacent
terrestrial vegetation of
the watershed and in-
creasing importance of

EOATORS

ORS

12- (70& METERS)

inputs from upstream tributary systems is a basic feature of the conceptual scheme. The pro-
portional diagrams at the right show the changes in relative dominance of invertebrate func-
tional groups from headwaters to mouth. Important shredders include certain species of
stoneflies, caddisflies, and craneflies that feed on CPOM (coarse particulate organic matter).
Dominant collectors are netspinning caddisflies, blackflies, clams, and certain midge species
which filter FPOM (fine particulate organic matter) from the passing water. Also, certain
species of mayflies, midges, oligochaetes , and amphipods (may also function as shredders)
gather particles from the sediments. Grazers or scrapers include certain species of caddisflies,
mayflies, snails, and beetles. In addition to the fish shown at the left, the major predators
are helgramites, dragonflies, tanypod midges, and certain species of stoneflies. The midregion
of the river system is seen as the major zone of plant growth (algae, or periphyton, and rooted
vascular plants) where the ratio of gross primary production (P) to community respiration (R)
is greater than 1. Fish populations grade from invertebrate eaters in the headwaters to fish
and benthic invertebrate eaters in the midreaches to benthic invertebrate and plankton feeders
in the large rivers. (Modified from Cummins 1975).

140

FOOD BASE AND BIOLOGY OF FORESTED STREAMS

The food base for the biological communi-
ties of forest streams consists of leaves,
needles, cones, twigs, wood, and bark. The
large boles which help shape the small stream
are usually biologically processed in place.
The input of bole material to the stream is
not a regular annual occurrence. Leaves,
cones, twigs, lichens, and other components
of fine litter have a reasonably predictable
timing of input to and export from streams.
Of the organic material which falls or slides
into first-order streams every year, only
18-35 percent may be flushed downstream to
higher order streams. These streams are very
retentive, not mere conduits exporting materials
quickly to the sea. Sixty to 70 percent of
the annual organic inputs are retained long
enough to be biologically utilized by stream
organisms. Big wood debris dams serve as
effective retention devices for fine organic
material, allowing time for microbial coloni-
zation and insect consumption of this material.
Functionally the invertebrates of streams
flowing through forests have evolved to gouge,
shred, and scrape wood and leaves and to
gather the fine organic matter derived from
breakdown of coarser material (Cummins 1974;
Anderson et al.in press).

Woody debris and leaves, the
two major allochthonous components enter-
ing a stream from the riparian zone, operate
in different ways in relation to quantity,
quality, and turnover time of standing crop.
The leaves form a small pool of readily avail-
able organic material, while the wood forms a
large pool of less available organic matter.
The slowly processed wood also constitutes a
long term reserve of essential nutrients and
energy. The composition, metabolic structure,
and nutrient turnover time of the particulate
organic pool effectively provide both flexi-
bility and stability within the system.

The amount of debris processed in a de-
fined reach of stream depends on two factors:
(1) the nature of the debris (abundance arid
species of wood or leaves) and (2) the capacity
of the stream to retain finely divided debris
for the period of time required to complete
processing. Debris undergoing utilization by
stream biota may either be utilized fully
within a stream reach or be exported to a
downstream reach. Processing continues as
small debris moves along the drainage because
export from one reach constitutes downstream
input. Processing includes both material used
metabolically by bacteria and fungi and those
debris pieces physically abraded by mineral
sediment or by insect consumption. In all
cases, the debris- is broken into smaller
pieces which increases the surface- to-volume
ratio and makes a debris particle increasingly
susceptible to microbial attack.

Wood in streams is a substrate for bio-
logical activity and it creates other habitat
opportunities by regulating the movement of
water and sediment. To measure the importance
of large organic debris from the riparian zone
in streams, Swanson and Lienkaemper (unpublished
data) examined several streams and measured
percent of stream area in (1) wood, (2)wood-
created habitat, principally depositional pools,
and (3) nonwood habitat such as bedrock and
boulder cascades. In a 245-m section of Mack
Creek, a third-order stream flowing through
an old-growth Douglas-fir stand in the western
Cascade Range, Oregon, 11 percent of the stream
area is in wood, 16 percent in wood-created
habitat, and 73 percent in nonwood habitat.
Figure 4 shows an example of the distribution
and quantity of debris in a section of Mack
Creek. In a first-order tributary draining
10 ha, wood comprises 25 percent of the stream
area and another 21 percent is habitat created
by wood. Much of the biological activity by
detritus-processing and consumer organisms is
concentrated in the areas of wood and wood-
created habitat. Each habitat type has a
different faunal composition.

Wood Habitat Community

Wood habitat communities are distinctive.
The primary utilizers are beetles, midges, and
snails. In addition to the food supplied to
the major wood eaters, the surface area and
large number of protective niches on wood
afford considerable living space and conceal-
ment. Wood is used for oviposition, as a
nursery area for early instars, for resting,
molting, pupation, and emergence. Because of
its unique capillary properties, it affords an
ideal air-water interface where gradients of
temperature and moisture can be selected by
different taxa for various activities.

Wood-Created Habitat

The depositional areas behind large debris
are prime areas for processing leaf material
and the fine organic matter derived from wood.
These areas are richer than the wood habitat
community both in numbers and biomass of inver-
tebrates. Leaves and the shredders (primarily
caddis- and craneflies) are concentrated in
these areas. Many of the shredders feeding
here will use the wood habitat to molt, pupate,
and emerge.

The difference in invertebrate biomass on
leaves and wood is attributed primarily to
differences in food quality. Although both
are low in nitrogen compared with periphyton,
seeds, or fresh macrophytes, the wood is so
high in the refractory components lignin and
cellulose that it becomes available at a very
slow rate. The greater surface area and pene-
trability of leaves results in microbial con-

141

METERS

WATER FLOW
\l LOG: HT ABOVE LOW WATER, (M)
FLOATED ORGANIC DEBRIS
~~||| TRAPPED SEDIMENT
^-(^ MINIMUM TIM€ AT SITE, YR.
® LARGE ROCK

MA-13^ ACCUMULATION NUMBER

CZZZ1 POTENTIAL STREAM DEBRIS, ABOVE CHANNEL (M)
I 0 I RECENT DEBRIS (1976)
BOULDER ISLAND

CHANNEL BOUNDARY
SPRING CHANNEL

Figure 4. — Distribution of debris in a section of Mack Creek, western Oregon. Courtesy of George W.
Lienkaemper .

ditioning occurring within months, compared
with years for wood. Conditioning is a kr -
factor in the debris becoming available as
food for the invertebrates.

RELATIONSHIP OF RIPARIAN VEGETATION
TO SALMONIDS

Indirect Influences

In addition to the effect of riparian zone
material which directly becomes a part of the
stream system, streamside vegetation has many
important indirect influences on the habitat
of salmonids.

Direct Influences

The previous discussion has described how
riparian vegetation contributes to primary
stream productivity through input of organic
material and nutrients which are utilized by
various components of the stream biota. These
relationships directly affect the production
of fish by establishing the basic components
of the food chain which eventually lead to the
fish themselves. Likewise, necessary portions
of salmonid habitat are created by large pieces
of debris from the riparian zone. Logs and
debris jams create pools and protective cover.
This type of habitat also provides communities
of benthic organisms different from those
associated with the shallower and faster waters
of riffles and runs. This increase in diver-
sity of invertebrates provides a more useable
food base for the fishes, which depend to a
great extent upon them. A large part of the
diet of fish in the family Salmonidae (the
various Pacific salmon, trout, and char) is
aquatic insects and other invertebrate organisms,

Water Temperature

The principal source of heat which raises
water temperatures is direct solar radiation
(Brown 196°). Consequently, streamside vege-
tation is important in maintaining water temp-
eratures suitable for spawning, egg and fry
incubation, and rearing of anadromous and
resident salmonids. Several studies in the
last decade have demonstrated how streamside
vegetation directly controls water temperature
(Levno and Rothacher 1967, Brown and Krygier
1970, Meehan 1970, Burns 1972). The literature
is also rich with documentation of the effects
of streamside canopy removal on stream temp-
eratures (Kail and Lantz 1969, Meehan et al.
1969, Brown and Krygier 1970, Burns 1972,
Moring 1975) .

Stream temperature is directly proportional
to surface area and solar energy input, and in-
versely proportional to streamflow (Gibbons
and Salo 1973). Therefore, small forested
streams are the most susceptible to temperature

142

change. The insulating effect of riparian
vegetation is thus of primary importance in
maintaining acceptable stream temperatures in
the many small streams which cumulatively pro-
duce a significant portion of the salmon and
trout populations of the Western United States.

Sediment

Another major function of riparian vege-
tation is to act as a buffer or "filter" against
sediment and debris which would otherwise be
deposited in the stream. Surface runoff is a
primary vehicle for the transportation of sedi-
ment to streams from adjacent sources, either
natural or man-created. The herbaceous communi-
ties within the riparian zone are effective in
reducing the impacts of this runoff, and the
larger shrubs and trees prevent larger debris
from entering the stream channel. The value of
streamside vegetation for stream protection has
been quantified in economic terms by Everest
(1975).

Sediment which affects salmonids occurs
in two general forms. As suspended sediment,
it can be harmful if concentrations are high
and persistent (Cordone and Kelley 1961).
Under these conditions, silt may accumulate
on the gill filaments and actually inhibit the
ability of the gills to aerate the blood,
eventually causing death by anoxemia and carbon
dioxide retention.

Bedload sediment, however, probably limits
salmonid production more than suspended sedi-
ment. Excessive deposited sediment reduces
the flow of intragravel water, which in turn
limits the supply of oxygen available to incu-
bating eggs and alevins, and hinders the re-
moval of metabolic waste products (Sheridan
1962, Vaux 1962, Cooper 1965, McNeil 1966).
Bedload sediment may also act as a physical
barrier, preventing the emergence of newly
hatched fry up through the gravel (Koski 1966,
Hall and Lantz 1969).

Another effect of sediment is the alter-
ation of habitat used by aquatic insects
(Wagner 1959) which directly relates to the
growth and condition of the fish which utilize
them. Although biomass may not decrease, the
species composition may change such that the
new forms are not as readily available to the
fish.

Cover

tion also acts as escape cover and in some
instances as a deterrent against predation by
birds and mammals.

Insects

As discussed earlier, riparian vegetation
contributes to the food base of stream biolog-
ical communities in the form of wood and other
organic debris. In addition, streamside vege-
tation is important in directly providing in-
sects to the stream which then become part of
the available fish food. Terrestrial insects
which are associated with the various strata
of the riparian zone become "accidental" fish
food items. Many of the aquatic insects use
streamside vegetation during emergence and in
the adult stages of their life cycle.

EFFECTS OF LAND USE PRACTICES

Many of man's activities affect the
riparian zone to varying degrees. We must
consider logging and road construction to be
among the most severe disturbances. Until
recently it was common practice to clearcut
timber to the stream's edge. In addition to
removing the trees which provided shade to
the stream surface, the understory vegetation
and ground cover were usually cut down or
severely disturbed. In recent years, the
importance of the smaller streams has been
more fully recognized and buffer strips along
streams are often left.

The riparian zone is also affected by
livestock grazing. In addition to cropping
off much of the herbaceous vegetation along
streambanVs , livestock also use the smaller
shrubs and young trees as forage. As a result,
much of the ground cover and many of the plants
which provide shade to small streams are re-
moved. The soil along the streams is compacted
by trampling, and together with the removal of
the "filtering' plants a situation is created
which promotes the addition of fine sediment to
the streams. Wild ungulates also utilize the
riparian zone, but their presence is much less
noticeable than that of cattle and sheep. A
workshop was conducted in Reno in May 1977 to
bring together existing knowledge on the rela-
tionships between livestock and fisheries,
w-ildlife, and range resources. A large part
of the material which was discussed at this
workshop concerned the riparian zone, and will
soon be available.^

The extensive rooting of herbaceous ripar-
ian vegetation aids in streambank stabilization.

As a result, where streamside vegetation is

intact, the occurrence of undercut banks is USDA Forest Service, Pacific Southwest

higher. This is prime habitat for trout and Forest and Range Experiment Station, Berkeley,

young salmon. Overhanging streamside vegeta- California (in press).

143

SUMMARY

The riparian zone is a very important area
influencing the habitat of salmonids. Much of
the wood which forms the food base for stream
biota comes from the riparian zone. This same
wood, when it falls or slides into a stream,
has an important role in shaping the stream
and creating its habitat types. Streamside
vegetation provides shade to the stream surface,
thereby maintaining water temperatures accept-
able to salmonid fishes. The roots of woody
and herbaceous plants provide streambank stabil-
ity and help to create overhanging banks , an
important component of salmonid babitat.
Streamside vegetation provides habitat for the
later life history stages of aquatic insects
and for terrestrial insects which accidentally
become part of the food utilized by salmonids.

When the riparian zone is affected by man's
activities, the quality of fish habitat will
likewise be affected.

LITERATURE CITED

Anderson, N. H. , J. R. Sedell, L. M. Roberts,
and F. J. Triska. In press. The role of
aquatic invertebrates in processing of wood
debris in coniferous forest streams. Am.
Midland Naturalist.

Brown, George W. 1969. Predicting temperatures
of small streams. Water Resour. Res. 5(1):
68-75, illus.

Brown, George W. and James T. Krygier. 1970.
Effects of clear-cutting on stream temper-
ature. Water Resour. Res. 6 (A) : 1133-1139 ,
illus.

Burns, James W. 1972. Some effects of logging
and associated road construction on northern
California streams. Trans. Am. Fish. Soc.
101(1) :1-17, illus.

Cooper, A. C. 1965. The effect of transported
stream sediments on the survival of sockeye
and pink salmon eggs and alevins. Int. Pac.
Salmon Fisb. Comm. Bull. 18, 71 p., illus.

Cordone, Almo J. and Don W. Kelley. 1961. The
influences of inorganic sediment on the
aquatic life of streams. Calif. Fish &
Game 47 (2) : 189-228 .

Cummins, Kenneth W. 1974. Structure and func-
tion in stream ecosystems. Biosci. 24(11):
631-641.

Cummins, Kenneth W. 1975. The ecology of
running waters. Theory and practice. In:
Proc, Sandusky River Basin Symp., May 2-3,
1975, Tiffin, Ohio, p. 278-293.

Cummins, Kenneth W. 1977. From streams to
rivers. Am. Biol. Teacher 39:305-312.

Everest, Fred H. 1975. A method of estimating
the value of streamside reserve trees. USDA
For. Serv. Siskiyou Natl. For., Grants Pass,
Oregon, 12 p.

Gibbons, Dave R. and Ernest 0. Salo. 1973.
An annotated bibliography of the effects of
logging on fish of the Western United States
and Canada. USDA For. Serv. Gen. Tech. Rep.
PNW-10, 145 p. Pac. Northwest For. and
Range Exp. Stn. , Portland, Oregon.

Kali, James D. and Richard L. Lantz. 1969.
Effects of logging on the habitat of Coho
salmon and cutthroat trout in coastal streams.
In: T. G. Northcote (ed.), Symp. on salmon
and trout in streams, p. 355-375, illus.
Univ. B.C., Vancouver, 388 p.

Koski, K Victor. 1966. The survival of coho
salmon (Oncorhy nahus kisutchjf rom egg depo-
sition to emergence in three Oregon coastal
streams. MS. Thesis, Oregon State Univ.,
Corvallis, 84 p., illus.

Leopold Luna E., M. Gordon Wolman, and John
P. Miller. 1964. Fluvial processes in
geomorphology . W. H. Freeman, San Francis-
co, 522 p.

Levno, Al and Jack Rothacher. 1967. Increases
in maximum stream temperatures after logging
in old-growth Douglas-fir watersheds. USDA
For. Serv. Res. Note PNW-65, 12 p., illus.
Pac. Northwest For. and Range Exp. Stn.,
Portland, Oregon.

McNeil, William J. 1966. Effect of the spawn-
ing bed environment on reproduction of pink
and chum salmon. U.S. Fish and Wildlife
Serv., Fish. Bull. 65 (2) : 495-523 , illus.

Meehan, W. P., W. A. Farr, D. M. Bishop, and
J. H. Patric. 1969. Some effects of clear-
cutting on salmon habitat of two southeast
Alaska streams. USDA For. Serv. Res. Pap.
PNW-82, 45 p. illus., Pac. Northwest For.
and Range Exp. Stn., Portland, Oregon.

Meehan, William R. 1970. Some effects of

shade cover on stream temperature in south-
east Alaska. USDA For. Serv. Res. Note
PNW-113, 9 p., illus. Pac. Northwest For.
and Range Exp. Stn., Portland, Oregon.

Moring, John R. 1975. The Alsea Watershed
Study: Effects of logging on the aquatic
resources of three headwater streams of the
Alsea River, Oregon. Part II - Changes in
environmental conditions. Oregon Dep. Fish
and Wildlife, Fisb. R s. Rep. No. 9, 39 p.,
illus .

Sedell, James R. and Frank J. Triska. 1977.
Biological consequences of large organic
debris in Northwest streams. Logging Debris
in Streams Workshop, Oregon State Univ.,
Corvallis, 10 p. March 21-22, 1977.

Sedell, James R. , Frank J. Triska, and Nancy S.
Triska. 1975. The processing of conifer
and hardwood leaves in two coniferous forest
streams: I. Weight loss and associated in-
vertebrates. Verb. Internat. Verein. Limnol.
19:1617-1627.

Sheridan, William L. 1962. Waterflow through
a salmon spawning riffle in southeastern
Alaska. U.S. Fish and Wildl. Serv. Spec.
Sci. Rep. Fish. No. 407, 20 p., illus.

144

Swanson, Fredrick J., George W. Lienkaemper, and
James R. Sedell. 1976. History, physical
effects, and management implications of large
organic debris in western Oregon streams.
USDA For. Serv. Gen. Tech. Rep. PNW-56, 15 p.
illus. Pac. Northwest For. and Range Exp. Stn
Portland, Oregon.

Triska, F. J. and J. R. Sedell. 1976. Decompo-
sition of four species of leaf litter in re-
sponse to nitrate manipulation. Ecol. 57(4):
783-792.

Vaux, Walter G. 1962. Interchange of stream
and iLntragravel water in a salmon spawning
riffle. U.S. Fish and Wildl. Serv. Spec.
Sci. Rep. Fish. No. 405., 11 p., illus.

Wagner, Richard. 1959. Sand and gravel oper-
ations. In: Proc. Fifth Symp. , Pac. North
west, on siltation — it sources and effect
on the aquatic environ. Water Supply and
Water Pollut. Control Prog., Portland,
Oregon (mimeo) .

145

Ecological Study of
Southwestern Riparian Habitats:
Techniques and Data Applicability1

Bert in W. Anderson, Ronald W. Engel- Wilson,

^ 2

Douglas Wells and Robert D. Ohmart _/

Abstract. — Techniques used in a comparative ecological
study of bird and rodent populations along the lower
Colorado River are presented. Data were gathered to
examine not only faunal community relationships to various
plant community types but also to gain detailed knowledge
of an individual species' vegetational preference and
its niche within the riparian habitat. Examination of
parameters such as habitat breadth, habitat and niche
overlap and dispersal is instructive in the determination
of a species' niche and is of use for the resource manager
and the theoretical ecologist alike.

INTRODUCTION

Agencies responsible for the management
of natural resources are beginning more to
recognize and approach management from the
ecosystem level as opposed to monospecific
research. As a result of this, many
governmental agencies are currently developing
ecosystem approaches to management or are
providing financial support for studies of this
nature. A central problem in implementing
management based on the ecosystem approach is
often a direct lack of knowledge concerning
species' requirements and the relationships
and composition of faunistic communities.

This paper reports on the field techniques
employed and the applicability of data gathered
during a comparative ecological study of the
avian and small mammal populations in riparian
habitat along the lower Colorado River. The
approach is primarily synecological in that
the focus is on entire floral and faunal

Contributed paper, Symposium on the
Importance, Preservation and Management of the
Riparian Habitat, July 9, 1977, Tucson, Arizona.

2 Respectively, Faculty Research
Associate, Field Biologist, Field Biologist,
and Associate Professor of Zoology, Arizona
State University, Dept. Zoology and Center for
Environmental Studies, Tempe, Arizona 85281.

populations and their relationships within the
riparian ecosystem; the design of the study also
allows collection of in-depth autecological data.

The specific objectives of the study were
as follows: 1) to determine the relative value
of each of the various plant communities to
birds and small mammals found within the riparian
system; 2) to examine ecological relationships
of the avifaunal and mammalian components within
each plant community type; 3) to determine how
these relationships, and other factors, are
important in population regulation and habitat
selection. In addition to gathering in-depth
and repeatable data which would satisfy the
above requirements, field techniques had to be
chosen and developed which would be suitable
for studying a large area with a limited number
of personnel.

The study area embraced all of the riparian
vegetation in the lower Colorado River Valley
between the Mexican boundary north to Davis
Dam — a distance of approximately 425 km. The
width of the valley varies from 0.8 km to 16 km.
Much of the natural vegetation has been cleared
and developed for agriculture or other purposes.
The remaining riparian vegetation is scattered
throughout the valley in tracts of various sizes.

146

VEGETATION MEASUREMENTS

Establishment of study sites in relatively
homogeneous stands of riparian vegetation was
begun in June 1973. As personnel increased
(currently 13 persons) , more study sites were
added, and by April 1976, 84 sites had been
established. Study sites varied in length
from 0.8 to 1.6 km. Initially a line was
cleared 0.7 m wide and 0.8 to 1.6 km long.
The sampled area was considered to extend
laterally 126 m from the center line on each
side. Each line was numbered and designated
as orienting in a specific compass direction.
As a vegetation type map was developed for
the entire valley, every attempt was made to
sample each plant community type by line
establishment, proportionate to their total
area in the valley. The level of quantification
of the plant community types was such that if
a sample area of a given structure was lost
(i.e. cleared or burned), another could be
established without drastically affecting
comparisons between years.

Although great care was taken to locate
sample sites in relatively homogeneous
communities, vegetation analyses indicated
that considerable variation existed along many
of the sites. To quantify this variation along
each study site and to better understand bird-
habitat relationships, each lateral portion
of the study site from the center line was
subdivided into 150 m long intervals. Each
150 m interval was marked with surveyor's tape
and a wooden stake on which was painted the
distance from the start of the line. A line
1,500 m long, orienting north to south, now
was conceptualized as being composed of twenty
distinct 150 m units — ten on the west side and
ten on the east side of the line. Hereafter,
each distinct (150 m) unit will be referred
to as a plot.

Tree Counts

A direct tree count was conducted along
all transects. The count area extended 15 m
laterally to the line on each side. Within
each 30 m advancement along the transect, all
trees were counted and listed as to tree
species, size class of trees (<3 m or >3 m)
and presence of mistletoe (Phoradendron sp.).
Tree counts could either be summed to yield
number of trees per plot or number of trees
along the line in the study site.

using the board technique (MacArthur and
MacArthur 1961) . The number of sampling points
per study site was dependent on the number of
plots comprising each study site. The vegeta-
tion within each plot was sampled in three
predetermined locations. The sampling points
were at 15, 76, and 137 m along the line from
the beginning point. Each foliage volume
sampling point was displaced laterally 4.6 m
from the line. At each sampling point a
measuring tape was used to determine the
distance to the nearest foliage that would
cover approximately 50 percent of the foliage
board. This was determined at 0.15, 0.6, 1.5,
3.0, 4.6, 6.0 m and every 3.0 m height interval
thereafter until the tallest foliage present
was measured. The plant species to which the
distance measurement was taken at each height
interval was also recorded. Measured distances
were converted to m of foliage per mJ of space
using the formula

l°8e2
K=—'

where K is foliage density and D the measured
distance. The foliage density within each plot
for the various height intervals and for total
foliage density was determined by using the
average of the measurements from the three
sample points within each plot. The foliage
density for the study site as a whole was
determined by using the average density of all
plots within the site.

Mature plant communities in each study
site were measured for foliage density once
between May and July. Plant communities
undergoing succession were measured each year
at the beginning and at the end of the growing
season to quantify growth changes. These data
were extrapolated for all seasons under the
rationale that measurements involving leaves
in the summer correspond to potentially leaf-
bearing parts in the winter. Thus a community
with dense leaves in summer should have more
leaf-bearing parts in winter than a community
that possessed few leaves in the summer. This
assumption was validated by field measurements
in selected study sites.

Foliage Height Diversity

Foliage height diversity for each transect
was computed using information theory (Shannon
and Weaver 1949) where

Relative Foliage Volume H = -I p loep

n=l^i &ti

The relative foliage volume within each In this instance H equals foliage height

study site and within each plot was determined diversity, and p. equals the proportion of the

147

total foliage volume contributed by the volume
at height level i.

Litter Height Diversity

Litter height diversity (LHD) was measured
every 100 feet along the center line in each
study site where litter was present in the
trees. The measurement consisted of recording
the presence or absence of litterwithin a 1.5 m
radius at height intervals similar to
those used in foliage volume measurements.
Information theory was also used in calculating
this parameter; p^ was the proportion of the
total points with litter occurring in the ith
layer .

In some vegetative types it was theorized
that the amount and distribution of leaf litter
above the ground was partly responsible for
high rodent populations. Litter may provide
additional nest sites, foraging substrate, or
cover for some rodent species and even some
birds and allow higher population numbers.

Phenology

Phenological data were recorded monthly
for trees located in study sites. The objective
was to record information which might explain
population movements or trends within the
mammalian and avian communities. Types of
data gathered included duration and initiation
of flowering, initiation and amount of stem
growth, and fruit production.

Vegetative Communities

The various plant associations within the
riparian habitat were classified as being
components of six communities. These
communities are listed in Table 1 along with
the criteria used to distinguish each. The
dominant tree species (total numbers) in some
communities classified as cottonwood-willow
was neither cottonwood nor willow, yet one or
both of these species were responsible for
the presence of one or more additional canopy
layers to which the numerically dominant tree
species did not contribute. If mature cotton-
woods or willows were in densities of 2 or
more per ha, the avian data indicated they
exerted enough influence in the study site
to be classified as cottonwood-willow.

Structural Types

Foliage density in the combined height
intervals of 0.15 to 0.6 m, 1.5 to 3.0 m and

>4.6 m was used in computing the amount of
overlap in foliage density and structure
between all pairs of study sites using Horn's
(1966) formula where:

Z (x-j+y^) logCx^+iO -^x^logx^-Ey^logy^
Ro = (X+Y)log(X+Y)-XlogX-YlogY *

For study sites x and y, x-^ and y-^ represent
the proportion of the volume occurring at
height interval i, and X and Y represent the
total foliage volume. From a matrix of these
overlap values a dendrogram was constructed
(fig. 1) showing study sites with greatest
affinities between foliage density and structure.
The dendrogram was constructed following Cody
(1974) where

„ ._ aC A + aC B
aC, AB = ^ •

This simply states that the overlap of C with
A and B is equal to the average of the overlap
of C with A and C with B. The dendrogram was
interpreted as showing the existence of six
structural types of vegetation within the
riparian habitat. Each study site within a
structural type is more closely related in
structure to the other members of that type
than to any other site or group of sites.

Structural types I and II are the most
heterogeneous, but as a group, they are
separated by a substantial number of units
from any other structural type.

Structural type I was the most dense
overall and was characterized by the amount
of volume over 9 m (fig. 2) although there
were relatively well developed layers below
the 9 m level. Type II was characterized by
having less vegetation above 9 m but more
volume between 3.0 and 6.0 m than type I.
The other types were mainly characterized by
having less volume at higher layers and more
at 0.0 to 0.6 m (fig. 2) .

RODENT POPULATION DATA

Small mammal populations were sampled by
snap-trapping. We used two parallel rows of
trap stations. One row was placed along the
line of the sample site, and the second row
was placed at a lateral distance of 15 m.
Each row consisted of 15 stations, each station
being 15 m apart. Two museum specials and
one Victor rat trap were set at each station
yielding a total of 90 traps. All traps were
set for three consecutive nights using an
oatmeal-peanut butter bait which contained an
ant repellant (Anderson and Ohmart 1977) .
Traps were checked daily and the catch
recorded. All mammal densities are expressed

14 8

Table 1. — Vegetative communities and criteria used to classify study sites within a community.

Community

Criteria

I. Cottonwood (Populus fremontii)
Willow (Salix gooddingii)
II. Screwbean mesquite (Prosopis pubescens)
Salt cedar (Tamarix chinensis)
III. Honey mesquite (Prosopis velutina)
Salt cedar
IV. Salt cedar
V. Honey mesquite
VI. Arrowweed (Tessaria sericea)

Populus and/or Salix constituting at least 20%
of the total trees

P_. pubescens constituting at least 20% of trees

Approximately equal numbers of each

Constituting 95-100% of total trees
Constituting 95-100% of total trees
Constituting 95-100% of total vegetation in area

Vegetation Type

I.0H

0.9-

0.8-

0.7-

0.6

I „ll // in

IV // V // VI

Figure 1. — Dendrogram showing relationships between all transects based on overlap in foliage
density and structure.

14 9

.63

Volume (m2/m3)

Figure 2. — Foliage volume characteristics of
structural types.

as the number per 270 trap nights. No area
was trapped consecutively within a six-week
period .

Trapping of small mammals was organized
so that all structural types represented within
a community were sampled on an equal basis.
Even with 60,000 trap nights in 1976, the data
were not sufficiently adequate to examine some
populations for more than two seasons, April
through October and November through March.
Any analyses utilizing finer divisions of time
suffered from inadequate sample sizes for
several species.

Standard reproductive measurements were
recorded on all mammals to examine possible
differential breeding rates and timing between
communities. Cheek pouch contents were saved
from all heteromyids for later analysis.

In an effort to better understand prefer-
ences of small mammals for particular micro-
habitat types, the vegetation present at three
levels around each trap station was recorded.
Species caught at a particular trap station
could then be correlated with the vegetation
around that station.

the total density per 270 trap nights contributed
by species i.

AVIAN POPULATION PARAMETERS

Avian Densities

Estimates of avian densities in each study
site were calculated using a technique developed
by Emlen (1971) . Bird detections were recorded
as being within 0 to 15, 15 to 30, and 30 to
60 m lateral distance from the study site center
line. At the termination of each census of a
study site the distribution of detection points
for each species was used to determine the
density per sampled area. All densities were
converted to the number of individuals per
species per 40 ha. Each study site was censused
two to three times per month, and the mean was
used as a monthly population estimate for each
site. The number of censuses required per unit
of plant community is discussed elsewhere in
these proceedings (Anderson and OhmartjV). In
addition, mean monthly densities for each
structural type in each community type allows
analysis of seasonal changes in number of
individuals and species composition.

As part of the regular census, bird detec-
tions were also recorded in the specific plot
in which they occurred (fig. 3). This refine-
ment was not necessary to obtain density
estimates but served to document preferred
vegetation profiles of each species as discussed
later in the text.

Bird Species Diversity

Bird species diversity (BSD) was determined
using information theory where p. is the propor-
tion of the total bird density contributed by
the ith species. BSD's are calculated over a
monthly and seasonal basis for each study site
as well as for each structural type in each
community .

Foraging Behavior

Foraging behavior of birds was recorded in
all study sites to better understand the for-
aging niche of each species and to determine
how each species utilized specific parts of its
environment. By gathering these data in specific
study sites, it was possible to compare foraging
behavior to the structure, density, and vege-
tative species composition of the sites. Much
of the behavior was gathered while conducting

Rodent Species Diversity

Rodent species diversity for all community
and structural types was computed using infor-
mation theory where p^ is the proportion of

3/ Climatological and Physical Character-
istics Affecting Avian Population Estimates in
Southwest Riparian Communities using Transect
Counts .

150

READNorthTO South Colorado River BIRD DATA 06504 page_1 of L_

tran PvlO hatf 10 May '77t,mf 0600-0715 CLOUDS Clear wind Calm Tfmp 60 F reader_DGW

DIRECTION: R^st. DIRECTION: West

SPECIES

INT.

0-15m

15-30m

30-6C

5?20

0-15m

15-30m

30-6Q

BIRD
HT

TREE
HT

TREE
SP

POS

Verdin

Maii y~r\ "i n rr Do

Abert 1 s Tow
Cactus Wren
Verdin
Gambel's Qu
Lucy 1 s Warb

°150
iee

480

ail
ler

2
1

Figure 3. — Sample bird census form showing location of detection points within plots and at
lateral intervals.

regular censuses. Specific parameters recorded
for each species included:

Climatic conditions
Bird species and sex
Foraging method
Substrate

Height of substrate
Height of bird
Branch diameter
Location within tree
Shade or exposed

APPLICABILITY OF DATA

Data obtained in this study provide infor-
mation concerning the extent of wildlife use
in each of the various plant communities.
Statistical analysis of many plant communities
over a relatively large area permits examination
of some of the factors responsible for these
use values — be they specific vegetational
configurations or the presence or absence of
an ecologically close species. Through a
comparison of species compositions and densities
between community types and structural groupings,
insight is gained into particular aspects of
community ecology. A greater understanding of
these relationships is necessary for management
decisions. Specific examples are given below
of ways in which these data may be examined in
order to satisfy the outlined objectives. The
analyses do not exhaust the list of possibilities
but are primary ones found to be useful in this
particular study.

Wildlife Use

The various avian and rodent population
parameters and the vegetational parameters

with which they are associated are as follows:

Avian Population Parameters
Bird species diversity
Number of species

Bird density (total and by individual
species)

Rodent Population Parameters
Rodent species diversity
Rodent density (total and by individual
species)

Vegetation Parameters

Foliage volume (total and by height

classes)
Foliage height diversity
Tree counts (of a species by height

class)

Litter height diversity

Each comparison can be made on the basis
of different area and time units. The possible
units include:

Area Units

Individual transects
Communities
Structural types
Total riparian habitat

Time

Individual months
Seasons

Different years

The specific ones chosen are dependent upon the
specific needs and requirements of the study.
We have found analyses of data by groups of
months or seasons to be practical in determining
wildlife use values, but we also utilize data

151

from individual months in certain analyses
where more precise resolution is needed.

In examining wildlife use values one begins
to see not only which species utilize an area,
but also what factors in that environment have
an important influence or at least show strong
correlations with the various population
parameters. Figure 4 presents correlation
coefficients between several rodent populations
and vegetational parameters in a honey mesquite
community. It can be seen (fig. 4) that there
is generally an inverse relationship between
rodent density and rodent species diversity.
One can predict that in an area with dense
foliage at 1.5 to 3.0 m the rodent density
would be relatively high, whereas the rodent
species diversity would be depressed. This
example only serves to show a small part of
the total picture, but the implications in
this case are evident concerning possible
options to achieve the desired management goals .

Habitat Breadth

Habitat breadth values indicate the extent
to which a species' population is evenly
distributed throughout the habitat. It can be
calculated using a species' distribution among
the six categories of dominant vegetation and
six structural types separately or combined.
Habitat breadths are calculated on a monthly
and seasonal basis. Since we currently (May
1977) have about 100 study sites in riparian
vegetation and since each site is censused
about three times per month, monthly habitat
breadths are based on about 300 censuses. On
a seasonal basis this increases to 900 censuses
for winter (December, January, February) and
summer (May, June, July); and 600 censuses for
spring (March, April), late summer (August,
September), and fall (October, November).
This allows monitoring of monthly and seasonal
changes in habitat breadth of each species.

Habitat breadth for each species is
calculated using information theory where p.
represents the proportion of a species' total
population contributed by its density within
a community or structural type i. When expressed
as the percent of maximum (J) , habitat breadth
can be used to designate species as generalists
or specialists within communities or structural
types. Any species with a percent value of
maximum habitat breadth below an arbitrarily
set limit would be classified as a specialist.
It is possible for a species to be a specialist
with regard to communities and a generalist
with regard to structural types, and vice versa.
Designating species as specialists or generalists
is especially valuable to decision makers for
management of critical habitats (i.e. those
habitats containing several specialist species) .
It is these species which have the most exact-

ing habitat requirements and which require
special management efforts.

Niche Breadth

Niche breadth values are calculated in a
manner similar to habitat breadth values but
are based on the distribution of a species
within a particular community. The proportion
of the total community population of a species
contributed by its density per individual study
site in that community represents the p. values
in the information theory formula.

Calculation of niche breadth is an attempt
at a more sensitive analysis of distribution of
species within a portion of the habitat — the
community. In combination with other measures
of resource utilization, niche breadth is
instructive in understanding the vegetational
preferences of a species and how this changes
through time and in comparison to other closely
related species.

Preferred Vegetation Profiles

Using the data gathered on occurrence of
birds within plots, it is possible to compute
the preferred vegetation of each bird species.
All bird detections within a plot located
within a lateral distance of 15 m (0.2 ha) from
the center line of the study site are compared
to the vegetative characteristics of the plot
in which the detection occurred. The 15 m
lateral distance was selected because it was
thought to be a complete census coverage of
that area (i.e. all birds in that area were
detected) and because it corresponded closely
with the area of vegetation measurements, such
as tree counts and foliage volume estimates.
Given a suitable sample size of detections for
each bird species, it is possible to calculate
an average vegetation profile where a species
is most often found. Collectively there are
between 1,000 and 1,100 of these 0.2 ha areas.
This profile will vary seasonally and will
indicate the changing temporal requirements of
a species. This can be done for all individuals
of a species or during the breeding season for
adult males only. The data can be further sub-
divided by dominant vegetation and/or structural
types. These data are suitable for multivariate
techniques such as discriminate function and
principal component analyses.

Figure 5 (Anderson and Ohmart 1975) presents
the preferred vegetation profiles of the Verdin
(Auriparus f laviceps) , Lucy's Warbler (Vermivora
luciae) , and Black-tailed Gnatcatcher (Polioptila
melanura) . Each is a small (<10 gm) insectivore
with generally similar foraging behavior patterns .
However, they each show a significantly (p<0.05)
different preferred foliage profile. Lucy's
Warblers generally inhabit the areas with

152

-1.0

November - March

April -October

Volume at: .15-.6 1.5-30 4.5+ Total FHD35-.6 15-3.0 4.5+ Total FHD
meters Volume Volume

Vegetation Parameters

Figure 4. — Correlations between rodent
population parameters and vegetation
parameters in honey mesquite (Anderson and
Ohmart 1975) .

Lucy s Warbler

\ \

1 Verdir

Black- tailed Gnatcatcher

Height of Vegetation

Figure 5. — Preferred structural configuration
and foliage density of three bird species
April through August 1975 (Anderson and
Ohmart 1975) .

in each plot each month. The times of inter-
specific compatibility are indicated by the
times of greatest average number per plot;
dispersal is indicated when the average drops.
These data are calculated monthly and are
based on 1,000 to 1,100 plots.

denser vegetation at 1.5 to 3.0 m and the
Black-tailed Gnatcatchers inhabit the sparser
areas. These subtle divisions of the habitat
are sometimes obscured by an analysis at the
study site level due to the amount of hetero-
geneity within each site. By utilizing bird
detections and vegetational measurements made
within plots, some of the heterogeneity is
reduced and more definite relationships can
be observed.

Dispersal

We obtained a measure of the extent and
time of dispersal in birds by calculating the
average number of individuals of each species

Habitat and Horizontal Niche Overlap

Habitat and horizontal niche overlap are
similar in that they compare the level of
coexistence of two species within an area.
As in the calculation of habitat breadths,
habitat overlap can be calculated using two
approaches to the composition of the habitat-
i.e. community and structural types. Habitat
overlap values are computed using Horn's
formula (1966) for ecological overlap where

and y^ represent the density (per 40 ha for
birds; number per 270 trap nights for rodents)
in community or structural type i, and X and
Y represent the sum of the x^ and y^ values.

153

Horizontal niche overlap values computed
for birds only are based on the co-occurrence
of species within plots. It is based on the
ratio of the number of plots two species hold
in common to the geometric mean of the number
of plots they occupy separately as shown by
Cody (1974) where:

H,12 = PU/[(PU + P12) (P22+P12»

1/2

Here ~P\2 equals the number of plots the two
species hold in common and ¥\\ and P22 represent
the number of plots occupied by one species
but not the other.

Overlap values, either habitat or niche,
when organized into a matrix of species by
species overlaps, can be used to construct a
dendrogram (Cody 1974) . A dendrogram illus-
trates those groups of species which are most
similar to each other with respect to habitat
or horizontal niche overlap. Figure 6 presents
a simple dendrogram of horizontal niche over-
lap among the birds in salt cedar. The
Mourning Dove (Zenaida macroura) and the
White-winged Dove: (Zenaida asiatica) are
the only species with a high level of overlap.
Each of these species obtains most of its food
from agricultural areas (Anderson and Ohmart
1975) . It is possible that the low overlap
values exhibited among the other bird species
may be due to the fact that salt cedar is an
exotic plant to which the native avifauna has
only partially adapted. Examination of overlap
patterns within other communities as well as
examination of patterns shown by other measures
of resource utilization is necessary before
any conclusions can be drawn.

Balda (1975) mentioned different approaches
which have been used in the study of birds and
their vegetative substrate. These approaches
vary from intensive studies of single species
to studies where assemblages of bird species
were identified, counted and related to some
component of the vegetative community. Each
of these approaches has its advantages and
disadvantages for the manager in a decision-
making position. The techniques and procedures
outlined above enable one to gather data which
are in a versatile form. Pertinent information
is available for a resource manager interested
in managing for critical habitat of a rare
species as well as for a theoretical ecologist
exploring community organization patterns.

ACKNOWLEDGMENTS

We wish to thank the many field biologists
who have helped us in collecting data. We are
grateful to Jack Gildar for computerizing the
data. The efforts of the secretarial staff

0.8-

0.7-

0.6-

0.5-

0)
>

.§ 0.4

o
I

0.2

0.1

>
o
Q

3
O

a>
>
o
Q
■a

a>
c

4)
-Q
<

(0
3

E
to
O

-O

o
O

■a

0)
■D

ea
a>
.c

c
5

o
i_

Figure 6. — Horizontal niche overlap among

birds in salt cedar April through August 1975
(Anderson and Ohmart 1975).

154

in typing early drafts and of Penny Dunlop and
Katherine Hildebrandt in typing the final
manuscript are greatly appreciated. We thank
Jane Durham for editorial suggestions concerning
the manuscript. Linda Cheney kindly prepared
the illustrations. The research was funded
through grant number 14-06-300-2415 from the
U. S. Bureau of Reclamation.

LITERATURE CITED

Anderson, B. W. and R. D. Ohmart.

1975. Annual report: Vegetation management
studies. U. S. Bur. Rec. 107 pp.

Anderson, B. W. and R. D. Ohmart.

1977. Rodent bait additive which repels
insects. J. Mammal. 58:242.

Balda, R. P.

1975. Vegetation structure and breeding bird
diversity. Proc. of the Symp. on Mgmt. of
Forest and Range Habitats for Nongame Birds,
pp. 59-80.

Cody, M. L.

1974. Competition and the structure of bird
communities. Princeton University Press,
Princeton, N.J. 381pp.
Emlen, J. T.

1971. Population densities of birds derived
from transect count. Auk 88:323-341.
Horn, H. S.

1966. Measurement of "overlap" in comparative
ecological studies. Am. Nat'l. 100:419-424.
MacArthur, R. H. and J. MacArthur.

1961. On bird species diversity. Ecol.
42:594-598.
Shannon, C. E. and W. Weaver.

1949. The mathematical theory of communi-
cation. Univ. of 111. Press, Urbana, 111.

155

The Importance of Riparian Habitat
to Migrating Birds1 •

Lawrence E. Stevens? Bryan T. Brown? James M. Simpson^ and R. Roy Johnson ^

Abstract. — Seven pairs of study sites in riparian and
adjacent, nonriparian habitats were censused for spring
migrant passerines. Riparian plots contained up to 10.6
times the number of migrants per hectare found on adjacent,
nonriparian plots. Stop-over habitat selection is
indicated by differing migrant densities and species
diversities in various habitats. Passerine migration
strategies are discussed.

INTRODUCTION

Field investigators have long noted that
migrating passerines show a decided preference
for riparian habitats over nonriparian habitats;
however, virtually no data have been published
concerning the nature of this preference.
Riparian habitats provide an important source
of food and cover for migrants and these
habitats are being eliminated at such an
alarming rate that the damage to migrant
populations may be significant. The aim of
this paper is to illustrate the importance of
stop-over riparian habitats to migrant passer-
ines in the Southwest. Only a few aspects of
migration of western passerines are mentioned
here but it is hoped that the data presented
will stimulate additional research in this
important field.

Many researchers have contributed to our
knowledge of the timing of migration in
southwestern passerines (Phillips 1951, Phillips
et al. 1964, Hubbard 1971, Johnson and Simpson
1971, and others), but as yet no large-scale
synthesis of migration patterns has been
attempted. A growing concern for improved

1/ Contributed paper, Symposium on the
Importance, Preservation and Management of the
Riparian Habitat, July 9, 1977, Tucson, Arizona.

2/ Research Biologist, Museum of Northern
Arizona, Flagstaff, Arizona 86001.

3/ Research Assistant, Grand Canyon Nation-
al Park, Grand Canyon, Arizona 8602 3.

4/ Associate, Museum of Northern Arizona,
Flagstaff, Arizqna 86001.

5/ Senior Research Scientist, Grand Canyon
National Park, Grand Canyon, Arizona 8602 3.

riparian habitat management practices has
provided the impetus for a number of studies
on riparian habitats by various government
agencies (Johnson et al. 1974, Carothers and
Johnson 1975, Lacey et al. 1975, Smith 1975,
Carothers et al. 1976, Pace 19777, and others).
Nearly all studies to date have ignored migrant
passerines and their relationships to stop-over
habitats in the Southwest (Sprunt 1975) .

Migration, as Emlen (1975) indicated, is
a multifaceted phenomenon. Some aspects of
vernal (spring) migration related to stop-over
habitat selection include migratory strategy,
the influence of weather, and the development
of migration routes. Literature for eastern
North America indicates that passerines generally
migrate nocturnally, resting and foraging during
the day (Helms 1959, Able 1970, Welty 1975,
and others) . Gauthreaux (1972) suggested that
vernal migrant passerines generally fly singly
or in small, unispecific flocks. Vernal noctur-
nal migration in land passerines has been
correlated with atmospheric stability and wind
direction (Raynor 1956) . Pleistocene speciation
of Parulidae was discussed by Mengel (1964) and
contemporary continental migration patterns were
reviewed by Dorst (1962) .

6/ Johnson, R.R., S.W. Carothers, and D.B.
Wertheimer. 1974. The importance of the lower
Gila River, New Mexico, as a refuge for threaten
wildlife: a multiple agency land management
program. Unpublished, 53 pp.

7/ Pace, CP. 1977. Classification,
restoration and management of riparian habitats
in southwestern forests. Rocky Mountain Forest
and Range Experiment Station Study Plan 1710-44.
Tempe . Unpublished, 44 pp.

156

Parnell's (1969) investigation of habitat
selection in migrant eastern Parulidae (the North
American wood warblers) demonstrated some corre-
lation between warbler species and the stop-over
habitat-niche chosen. Unexpectedly, he could not
clearly demonstrate selection of major habitat
types by migrant warblers in eastern forests.

METHODS

During the spring of 1977 a total of seven
pairs of study areas were censused to determine
migrant passerine densities and migrant diversi-
ties in stop-over habitats. The study sites,
ranging in size from 1.6 hectares to 20.0 hec-
tares, were located throughout Arizona. One
site of each pair was situated in mature ,
riparian growth and the other in adjacent,
nonriparian growth. Four pairs of study plots,
those being used in the Rocky Mountain Forest
and Range Experiment Station (RMFRES) riparian
habitats study program, were examined in greater
depth using the spot-map method (Williams 1936,
Kendeigh 1944, and Franzreb 1976). The remain-
ing three pairs of study sites were censused
using a modified Emlen (1971) transect technique
wherein an absolute count of birds was made.
Data on the vegetation of the four paired RMFRES
sites were gathered using the plotless point-
quarter method of Cottam and Curtis (1956) and
are included in Table 1. Tree heights were

measured with a clinometer.

In addition, observation data on the spring
migration of paulids for the Blue Point cotton-
wood stand was gathered from 1969 through 1974.

STUDY SITE DESCRIPTIONS

Terminology follows that of Hubbard (1971)
with modifications. Study site sizes are included
parenthetically.

1. Wet Beaver Creek (WBC) - Sullivan Ranch near
Camp Verde, Yavapai Co., elev. 1250 m. A
heterogeneous riparian forest with Platanus-
Fraxinus overstory (4.1 hectares).

Wet Beaver Creek Adjacent (WBCA) - a mixed
microphyll (valley and slope mesquite) -evergreen
woodland of Prosopis , Juniperus and Canotia
(3.0 hectares) .

2. Ash Creek (AC) - Rincon Mountains, Coronado
Forest, Pima Co., elev. 1200 m. A heavily grazed
heterogenous riparian woodland of Prosopis ,
Fraxinus and Celtis (4.1 hectares).

Ash Creek Adjacent (ACA) - A heavily grazed,
mixed microphyll (valley and slope mesquite) -
evergreen-xeric shrubland of Prosopis , Mimosa,
Quercus and Fouquieria (20.0 hectares).

3. Rucker Canyon (RC) - Chiricahua Mountains,
Coronado National Forest, Cochise Co., elev.
ca. 1600 m. A heterogenous, mixsd riparian and
evergreen forest with Quercus -Platanus overstory
(5 . 0 hectares) .

Table 1. — Vegetation of four RMFRES study sites

STUDY
SITE

TREE SPECIES COMPOSITION

# TREES
/HA

AVERAGE
HT . TREES
(METERS)

SHRUB SPECIES COMPOSITION

# SHRUBS
/HA

AVERAGE
HT.SHRUI
(METERS)

WBC

Platanus (25%)
Juniperus (21%)
Fraxinus (20%)

193

13.8

Mimosa (19%)
Fraxinus (18%)
Rubus (17%)

489

1.8

WBCA

Juniperus (68%)
Prosopis (19%)

3.2

Canotia (28%)
Juniperus (19%)
Prosopis (17%)

1290

1.1

Prosopis (52%)
Fraxinus (23%)
Celtis (17%)

124

5.3

Mimosa (41%)
Baccharis (25%)

1510

0.9

ACA

Prosopis (86%)
Quercus (14%)

59.3

3.6

Mimosa (44%)
Gutierrezia (20%)

2638.5

0.8

Juniperus (45%)
Platanus (20%)
Quercus (20%)

142

9.3

Rhus (43%)
Juniperus (23%)

476

1.4

RCA

Quercus (61%)
Juniperus (34%)

228

5.0

Rhus (36%)
Nolina (33%)

691

1.3

Juglans (31%)
Prosopis (28%)
Platanus (26%)

121

11.1

Juglans (48%)
Prosopis (13%)
Mimosa (12%)

490

1.6

TCA

Prosopis (33%)
Quercus (32%)

29.4

4.4

Arctostaphylos (36%)
Mimosa (27%)
Prosopis (11%)

809

1.5

157

Rucker Canyon Adjacent (RCA) - a heavily
grazed evergreen woodland of Que reus and
Juniperus (6.0 hectares).

4. Turkey Creek (TC) - Rincon Mountains,
Cornado National Forest, Pima Co., elev. 1250 m.
A grazed homogeneous riparian forest with
Platanus-Fraxinus overstory (5.9 hectares).

Turkey Creek Adjacent (TCA) - A grazed,
mixed microphyll (slope mesquite) -evergreen
shrubland of Prosopis , Quercus and Arctostaphylos
(11 . 4 hectares) .

5. Watson Lake (WL) - Near Prescott, Yavapai
Co. , elev. 1600 m. A homogeneous riparian

forest with Populus-Salix overstory (7.5 hectares).

Watson Lake Adjacent (WLA) - A grazed, xeric
grassland (7.1 hectares).

6. Blue Point (BP) - On the Salt River near
Fort McDowell, Maricopa Co., elev. 400 m. A
grazed, homogeneous riparian forest-woodland
of Populus and Prosopis (10 hectares) .

Blue Point Adjacent (BPA) - A grazed, xeric
shrubland Cercidium association (10 hectares) .

7. Indian Gardens (IG) - Grand Canyon National
Park, Coconino Co., elev. 1150 m. An island of
homogeneous riparian forest with Populus over-
story (1.6 hectares).

Indian Gardens Adjacent (IGA) - A xeric,
Coleogyne shrubland association (1.6 hectares).

RESULTS AND DISCUSSION

Tables 2, 3, and 4, and Figures 1 and 2
present the results of our 19 77 spring migrant
passerine survey. Residents and all possible
breeding individuals and species were combined
in the breeding bird category. This constitutes
a significant overestimation of breeding bird
populations and an underestimation of MPDen
(migrant passerine density) and MPSD (migrant
passerine species diversity) because some
migrants were treated as breeding birds. Thus,
in terms of migrants, these data are extremely
conservative .

Table 2 presents the migrant species seen on
the fourteen study plots. Note the generally
higher MPSD of insectivores and the uniformly
higher total MPSD on the riparian plots.

Table 3 and Figure 1 present MPDen data.
The total number of migrant individuals in
riparian habitats is shown to be uniformly
greater (by up to 10.6 times) than the total
number in adjacent, nonriparian habitats, with
one exception. The open understory on the Turkey
Creek plot attracted large flocks of Chipping
Sparrows (Spizella passerina) and White -crowned
Sparrows (Zonotrichia leucophrys) from the
surrounding grasslands. Table 4 and Figure 2
show that the MPSD on all riparian study areas
was distinctly higher than the MPSD on adjacent
plots .

Other trends in these data can be observed.
As in Table 2 insectivorous migrants generally
preferred riparian habitats (WBC, TC, IG) and
a higher insectivorous MPSD was also evident in
most of those habitats . The larger number of

granivorous individuals and smaller species
diversity (WBCA, AC, TCA) of this group is an
indication of flocking behavior in Fringillidae
(the Finches, Sparrows and their allies).
Granivores displayed habitat selection by avoid- -
ing dense riparian forest and woodland situations
(WBC, BP) ; they tended to concentrate in adjacent
shrublands (WBCA, WLA) and open, riparian forests
(TC, IG) . The importance of riparian habitats
to breeding birds is also shown; in general, at
least twice as many breeding individuals and
species occurred in the riparian plots as did
on the nonriparian plots.

Figures 1 and 2 illustrate differences in
migrant use of heterogeneous and homogeneous
habitats. Heterogeneous deciduous riparian
habitats (WBC, AC) had generally higher MPDen
and higher MPSD than did uniform stands of
riparian growth (TC, WL, BP) . Mixed riparian
and evergreen forest habitats (RC) had lower
MPDen and MPSD. Heterogeneous deciduous vege-
tation offers the greatest variety of habitat-
niches for migrants, thus it is not unreasonable
to expect substantial migrant use of these
habitats. While uniform stands of riparian
growth may be expected to support a lower MPSD
the extremely low numbers on the Blue Point plot
reflect inadequate sampling and poor weather
conditions. Inaccessibility to migrants may
account, in part, for the limited usage of the
Rucker Canyon study area. Able (1970) has
shown that 75% of eastern passerines migrate
below an altitude of 920 meters and the same
is probably true in the Southwest. A narrow
canyon at a higher elevation may not be used
by many migrants simply because the birds fly
between mountain ranges rather than over them.

Moderately high MPDen and MPSD occurred
in the only island stand of riparian vegetation
studied (IG) . It is not surprising that high
MPDen and MPSD were found in riparian islands
because these situations provide the only
available food and cover for passage birds.
The high percentage of granivores may reflect
differences in migration patterns between fring-
illids and the insectivorous passerines.

Adjacent habitat depauperacy promotes a
higher concentration of migrants in riparian
habitats. Adjacent, nonriparian habitats which
were not heavily grazed (WBCA and, to a lesser
extent, TCA) supported a higher MPDen and MPSD
than did those areas which were more heavily
grazed (ACA, RC, WLA) .

Patterns of migration in the Southwest
have not been explored in depth. We observed
only one wave migration of parulids in a five-
year study of the Blue Point cottonwood stand;
this concurs with Parnell's (1969) observation
that wave migration is quite uncommon in eastern
North American parulids. More frequently, though
still not commonly, we have observed wide fronts
of single species of parulids and fringillids.
Most spring migration through the Southwest
probably occurs in small, unispecific flocks,

158

Table 2. — Migrant passerine species occurrence and number of censuses per plot

Paired (riparian

and nonriparian habitats)

study sites

Heterogeneous

Homogeneous

Island

MIGRANT PASSERINE SPECIES

WBC WBCA AC ACA

RC RCA TC TCA WL WLA

BP BPA IG IGA

Empidonax spp.

X X X X

Western Wood Pewee

(Contopus sordidulus)

Olive-sided Flycatcher
(Nuttalornis borealis)

Mountain Chickadee
(Parus gambeli)

Hermit Thrush

(Catharus guttata)

Western Bluebird
(Sialia mexicana)

Blue-gray Gnatcatcher
(Polioptila caerulea)

Ruby-crowned Kinglet
(Regulus calendula)

X X

Cedar Waxwing

(Bombycilla cedrorum)

Solitary Vireo

(Vireo solitarius)

X X

Warbling Vireo
(Vireo gilvus)

XXX

XXX X

Virginia's Warbler

(Vermivora virginiae)

X X

Lucy's Warbler

(Vermivora luciae)

Yellow-rumped Warbler
(Dendroica coronata)

X X X X

X X X X X

Black- throated Gray Warbler
(D. nigrescens)

X X

Townsend's Warbler
(D. townsendi)

Hermit Warbler

(D. occidentalis)

159

Table 2. — continued

MIGRANT PASSERINE SPECIES

Paired (riparian

and nonriparian habitats) study

sites

Heterogeneous

Homogeneous

Island

WBC WBCA AC ACA RC RCA TC TCA WL WLA BP BPA IG IGA

MacGillivray 1 s Warbler
(Oporornis tolmiei)

Wilson's Warbler
(Wilsonia pusilla)

Western Tanager

(Piranga ludoviciana)

X X

Black-headed Grosbeak *

(Pheuticus melanocephalus )

Lazuli Bunting

(Passerina amoena)

Pine Siskin

(Carduelis pinus)

Green-tailed Towhee
(Pipilo chlorurus)

XXX

X X

Dark-eyed Junco
( Junco hyemalis )

Chipping Sparrow

(Spizella passerina

X X X X

X X

Brewer 1 s Sparrow
(Spizella breweri )

White-crowned Sparrow

(Zonotrichia leucophrys )

Total # migratory insectivorous
species

Total # migratory granivorous
species

Total # migratory species

Total # breeding passerine
species

Total # passerine species

Total # censuses/plot

* Insectivorous migrant species through Black-headed Grosbeak; granivorous migrant species

below Black -headed Grosbeak.
1 Study sites are of varying sizes and are not comparable.

160

Table 3. — Spring migrant and breeding passerine densities in riparian and adjacent nonriparian
habitats

Heterogeneous

Homogeneous

Island

WBC WBCA AC ACA RC RCA

TC TCA WL WLA BP BPA

IG IGA

Average # migrant
birds/ha

Insectivorous

34.7 5.0 4.3 0.3 1.7 0.3

3.4 0.8 4.1 0 1.5 0

i .8 1 : fo"';

Granivorous

n 7 7n < ^ ns or o ?

VJ . / / . U J • J W.O \J ■ J VJ • -J

S^lsO 04 050 0

8.8 0

Total

35.4 12.0 7.8 1.1 2.2 0.6

8.7 13.8 4.5 0.5 1.5 0

10.6 0

Average # breeding
birds/ha *

12.4 7.7 11.1 2.9 8.4 3.8

18.0 7.0 6.1 0.6 7.0 0.4

9.9 1.3

Average total #
birds/ha

47.8 19.7 18.9 4.0 10.6 4.4

26.7 20.8 10.6 1.1 8.5 0.4

20.5 1.3

* Including all potentially breeding passerine individuals

50 t

INSECTIVOROUS

GRANIVOROUS

BREEDING BIROS

WBClWBCA AC I ACA RClRCA
HETEROGENEOUS
STUDY AREAS

TCI TCA WlJWLA

HOMOGENEOUS
STUDY AREAS

BP I BPA IGMG4

ISLAND
STUDY AREA

Figure 1. — Spring migrant and breeding passerine densities/ha
in riparian and nonriparian habitats.

161

Table 4. — Spring migrant and breeding passerine species diversities/ha in riparian and adjacent,
nonriparian habitats

Heterogeneous

Homogeneous

Island

WriL, WdLA A.U ALA KL- KLA

1L. WLi WLiA ntr dPA

IG IGA

Average # migrant
species/ha

Insectivorous

2.7 1.7 1.2 0.3 0.7 0.2

0.8 0.5 0.8 0 0.9 0

0.6 0

Granivorous

0.2 0.7 0.4 0.1 0.2 0.1

0.2 0.2 0.3 0.1 0 0

1.6 0

Total

2.9 2.4 1.6 0.4 0.9 0.3

1.0 0.7 1.1 0.1 0.9 0

2.2 0

Average # breeding
species/ha *

4.4 3.7 2.3 1.0 2.7 1.2

2.9 2.1 2.6 0.4 1.4 0.5

6.2 1.3

Average total #
species/ha

7.3 6.1 3.9 1.4 3.6 1.5

...

3.9 2.8 3.7 0.5 2.3 0.5

8.4 1.3

* Including all potentially breeding passerine species

V.
W
UJ

o
uj

K
UJ
ffl

D
Z

AC
Ul

INSECTIVOROUS

WBClWBCA

HETEROGENEOUS
STUDY AREAS

TClTCA WLlWLA BP I BPA
HOMOGENEOUS
STUDY AREAS

G 1 IGA
SLAND
STUDY AREA

Figure 2. — Spring migrant and breeding passerine species diversities/ha
in riparian and adjacent, nonriparian habitats.

162

as has been suggested by Gauthreaux (1972) for
passerines migrating across the Gulf of Mexico.

CONCLUSIONS

From the data presented above it is evident
that stop-over habitat selection by migrants
occurs commonly in the Southwest. Riparian habi-
tats support significantly higher MPDen and MPSD
than do adjacent, nonriparian habitats. Insuffi-
cient data have been gathered as yet to substan-
tiate the occurrence of niche selection, but the
likelihood of this phenomenon is great. While
Parnell (1969) could not clearly demonstrate
habitat selection in migrant eastern warblers,
habitat delineation is more distinct in the
Southwest than in eastern deciduous and
coniferous forests.

Parameters influencing migrant passerine
use of riparian habitats include: specific
habitat preferences of the bird (stop-over habitat
selection) ; floral components (niche diversity
and vegetational species composition) ; location
of habitat (island situations and, perhaps,
accessibility) ; and quality of the adjacent
habitat (including the amount of grazing and
other forms of impact) .

The importance of riparian habitats to
migrant passerines is substantial. Riparian
habitat managers should consider the impact of
proposed management not only on breeding species
but also on migratory species. As Balda (1975)
suggests, managers must be concerned with the
quality of the avian populations they are
indirectly managing through habitat manipulation.
Riparian habitat management in vegetational
islands and in heavily-grazed areas may have a
greater effect on migrants and manipulation of
these areas must be carefully evaluated.

ACKNOWLEDGMENTS

Part of the data presented here was gather-
ed in conjunction with the Rocky Mountain Forest
and Range Experiment Station, U.S. Dept. of
Agriculture, Riparian Habitats Study Program
Number 1710-44. We also thank the National
Park Service at Grand Canyon National Park and
the Museum of Northern Arizona for supporting
this project. Special thanks go to S.W.
Carothers, J. A. Downs, C.E. Franz, L.T. Haight,
N.J. Sharber and K.W. Shoemaker for their
assistance in gathering and compiling field
data. We thank J. Scott for all her field
and secretarial assistance.

LITERATURE CITED

Able, K.P. 1970. A radar study of the altitude

of nocturnal passerine migration. Bird

Banding 41:282-290.
Balda, R.P. 1975. Vegetation structure and

breeding bird diversity. In_ Proc. Symp.

on Management of Forest and Range Habitats for

Nongame Birds. U.S.D.A. Forest Service

General Technical Report WO-1. Tucson,

Arizona. Pp. 59-80.
Carothers, S.W. and R.R. Johnson. 1975. The

effects of stream channel modification on

birds in the southwestern United States.

In Proc, Symp. on Stream Channel Modifica-
tion (Corning, R.V. , et al. eds.) Grottoes,

Virginia. PP. 60-76.
Carothers, S.W. and S.A. Aitchison (editors).

1976. An ecological survey of the riparian

zone of the Colorado River between Lees

Ferry and the Grand Wash Cliffs, Arizona.

U.S. Dept. Int. NPS Technical Report No. 10.
Cottam, G. and J.T. Curtis. 1956. The use

of distance measure in phytosociological

sampling. Ecology 37:451-460.
Dorst, J.D. 1962. The Migrations of Birds

(trans, by CD. Sherman). Houghton

Mifflin Co., Boston. 476 pp.
Emlen, J.T. 1971. Population densities of

birds derived from transect counts.

Auk 88:323-342.
Emlen, S.T. 1975. Migration: Orientation and

Navigation In Avian Biology, Vol V.

D.S. Farner and J.R. King, eds. Academic

Press, New York. Pp. 77-128.
Franzreb, K.E. 1976. Comparison of variable

strip transect and spot-map methods for

censusing avian populations in a mixed

coniferous forest. Condor 78:260-262.
Gauthreaux, S.A. 1972. Behavioral responses

of migrating birds to daylight and darkness :

a radar and direct visual study. Wilson

Bulletin 84:136-148.
Helms, C.W. 1959. Song and Tree Sparrow weight

and fat before and after a night of migration.

Wilson Bulletin 71:244-253.
Hubbard, J. P. 1971. The summer birds of the

Gila Valley, New Mexico. Nemouria, 1-35.

Occ. Pap. Delaware Mus . Natur. Hist.

May 13, #2.

Johnson, R.R. and J.M. Simpson. 1971. Impor-
tant birds from Blue Point cottonwoods ,
Maricopa County, Arizona. Condor 73:379-
380.

Kendeigh, S.C. 1944. Measurement of bird
populations. Ecol. Monogr. 14:67-106.

Lacey, J.R. , P.R. Ogden, and K.E. Foster. 1975.
Southern Arizona Riparian Habitat: Spatial
Distribution and Analysis. Univ. of Ariz.,
Tucson. 148 pp.

Mengel, R.M. 1964. The probable history of
species formation in some northern wood
warblers (Parulidae) . T_n The Living Bird
(3rd annual). Cornell. Pp 9-43.

Parnell, J.F. 1969. Habitat relations of
the Parulidae during spring migration.
Auk 86:505-521.

163

Phillips, A.R. 1951. Complexities of migra-
tion: a review. Wilson Bulletin 63:129-136.

Phillips, A.R., J.T. Marshall, and G. Monson.
1964. The Birds of Arizona. Univ. of Ariz.
Press, Tucson. 212 pp.

Raynor, G.S. 1956. Meteorological variables
and the northward movement of nocturnal land
bird migrants. Auk 73:153-175.

Smith, D.R. (coord.) 1975. Proc. Symp. on
Management of Forest and Range Habitats for
Nongame Birds. USDA Forest Service General
Technical Report WO- 1. Tucson, Ariz. 342 pp.

Sprunt, A., IV. 1975. Habitat management
implications of migration. In_ Proc. Symp
on Management of Forest and Range Habitats
for Nongame Birds. USDA Forest Service
General Technical Report WO-1. Tucson,
Ariz. PP. 81-86.

Welty, J.C. 1975. The Life of Birds, 2nd
edition. W.B. Saunders Co. , Philadelphia.
623 pp.

Williams, A.B. 1936. The composition and
dynamics of a beech-maple climax community
Ecol. Monogr. 6:317-408.

164

Significance of
Rio Grande Riparian Systems
Upon the Avifauna1

/ 2
Roland H. Wauer

ABSTRACT. — The Rio Grande corridor in West Texas serves as
a significant migratory and emigration route for avifauna, and
38 species are known to nest within the riparian habitat. A
total of 94 species are known to breed within riparian systems
within the American Southwest. The Rio Grande area provides
suitable habitat for 40% of those. Nine species — great blue
and green herons, peregrine falcon, American kestrel, white-
winged dove, screech owl, Bell's vireo, yellow warbler, and
bronzed cowbird — are discussed as indicators of changes within
the system and the important of the Rio Grande area as a refugium.

Few rivers on the North American continent
stimulate the imagination as does the Rio Grande.
It is that magical line that blends cultures
and links natural environments from the high-
lands of the Rocky Mountains to the lowlands
of the Chihuahuan Desert. It is a legendary
river that joins, rather than separates, two
countries. Both countries contribute to its
flow and utilize its precious ingredients.

The Rio Grande begins in the mountains of
Colorado and New Mexico, but little water is
left by the time it reaches Texas. Irrigation
and channelization claim a good deal of its
cargo above El Paso. For another 100 miles
below El Paso the once proud Rio Bravo del
Norte, as it is known south of the border, is
little more than a trickle. Below Presidio,
Texas, however, the Rio Grande is rejuvenated
by the Rio Conchos that brings water northward
from the slopes of Mexico's Sierra Madre Occi-
dental.

Tamayo (West, 1964) reported that the Rio
Conchos supplies 18 percent of the Rio Grande's
total flow, and that almost one-half of the
Rio Grande's annual discharge is derived from
Mexican tributaries. Since 1964 the flow of
the Rio Conchos has been restricted by Granero
Dam, and so an increasing amount of its water

Contributed paper for Symposium on Impor-
tance, Preservation and Management of the
Riparian Habitat, Tucson, Arizona, July 9, 1977.

^Chief, Division of Natural Resources Man-
agement, National Park Service, Southwest Region,
P.O. Box 728, 1100 Old Santa Fe Trail, Santa Fe,
New Mexico 87501.

is utilized for irrigation in Mexico before
reaching the Rio Grande influence. The least
changed portion of the entire Rio Grande prob-
ably is within a strip of about 250 miles that
lies between Colorado Canyon and Langtry, Texas
There the river makes a great southern swing
into the Mexican states of Chihuahua and Coahui
la. This paper is restricted to the riparian
zone of that "Big Bend" area and eastward
through Langtry to Amistad National Recreation
Area.

Area Description

A satellite view of West Texas provides
a realistic assessment of the contrast of the
Rio Grande to its environment. The river ap-
pears like a green ribbon winding through a
maze of grays, browns and blacks. This ribbon
of greenery truly does meander its way through
the arid desertscape of canyons, arroyos, and
mesas. Elevations adjacent to the riverway
range from 2,000 to 3,500 feet above sea level.
A few mountains that are at least several miles
from the river, such as the Chisos and del
Carmens , may reach 7,800 feet elevation, but
these have little influence upon the riparian
zones along the riverway.

The general physiographic character of the
Lower Canyons (an area from Big Bend National
Park to near Langtry) of the Rio Grande was
discussed by Johnson, et. al. (in press) in
describing the vegetation found there. The
zones include riverbed, riverbank, lower ter-
race or high floodplain, talus slope, canyon
walls, side-canyon, and uplands or mesa-butte-
rim. Although riparian vegetation does not

165

exist in all of these zones, wildlife in each
is influenced by the nearest riparian habitats.
Wauer (in press) described the avifauna within
each of the above seven zones. A good deal of
the following material is taken from that pub-
lication and an earlier one by Wauer (1973) ,
Birds of Big Bend National Park and Vicinity.

Riparian vegetation usually is considered
to include that growing upon the floodplain
and in adjacent arroyos, generally, wherever
periodic flooding occurs. On the Rio Grande,
riparian vegetation may extend from a few feet
to one-half mile from the riverbed, except
where sheer cliffs rise directly out of the
river .

Common reed (Phragmites communis) and
giant reed (Arundo donax) form tall "cane"
stands along isolated or protected places and
often hang out over the waterway. Other mesic
forms that grow on the riverbank include lance-
leaf cottonwood (Populus acuminata) , honey
mesquite (Prosopis j ulif lora) , seepwillow
(Baccharis glutonosa) , the exotic salt cedar
(Tamarisk sps.), and willows; the tall tree-
willow is southwest black willow (Salix good-
dingii) , and black (j>. niger) and sandbar
willows (JS. interior) are also present.

Just beyond this zone is another one that
includes all the above plants and several
others, including screwbean (Prosopis pubescens) ,
catclaw acacia (Acacia greggii) , black brush
(A. rigidula) , huisache (A. f arnesiana) ,
desertwillow (Chilopsis linearis) , tree tobacco
(Nicotiana glauca) , common buttonbush (Cepha-
lanthus occidentalis) , and Texas palo verde
(Cercidium texanum) .

Ponds can be found in a few places where
the Rio Grande has changed its course or where
high water has dredged a deep hole. Seepwillow,
salt cedar, and cowpen daisy (Verbesina encelio-
ides) , are early invaders of the silty soils.
Common cattail (Typha latifolia), lanceleaf
cottonwood, willows and Mexican devilweed
(Aster spinosus) appear soon if there is suf-
ficient moisture.

In places where the river regularly scours
the rocky shoreline, a mat of Bermuda grass
(Cynodom dactylon) may form a luxuriant cover.
In protected flats within the lower canyons,
extensive grassy vegas may result.

Riparian Avifauna of the Southwest

The American Southwest comprises a large
and extremely diverse section of the United
States. Elevations range from below sea level
to over 12,000 feet, and environmental condi-
tions vary from arid, desert lowlands to pine-

clad forests. For the purpose of this discus-
sion, the Southwest is the land mass north of
the Mexican border between southeastern Calif-
ornia (east of the Sierra Nevada foothills) and
west Texas (west of Del Rio) , north through the
southern one-third of Utah and all of New Mexico
but the northern highlands.

Two major river systems drain the Southwest,
the Colorado River on the west and the Rio
Grande on the east. Major tributaries include
the Gila, Virgin, and San Juan on the Colorado,
and the Rio Conchos, Pecos, and Devil's on the
Rio Grande.

Table 1 includes 94 known avifauna that
nest within riparian vegetation below the moun-
tain woodlands and forests, below approximately
5,500 feet elevation. This list includes ground
nesters and social parasites. It does not in-
clude species that nest in habitats that may be
adjacent to riparian zones, such as peregrine
falcon (Falco peregrinus) and cliff swallow
(Petrochelidon pyrrhonota) that are clif f-nesters
and common yellowthroat (Geothlypis trichas)
and red-winged blackbird (Agelaius phoeniceus)
that utilize swamps and marshes. It does in-
clude species such as American kestrel and
great horned owl that usually nest on cliffs
but have been found nesting on riparian vegeta-
tion .

Several earlier publications were utilized
in developing the list of breeding avifauna.
They include Death Valley National Monument,
California, studies by Wauer (1962a, 1962b, and
1964), and Remsen's 1976 breeding bird survey
near Needles, California, (1977); Nevada avi-
faunal summary by Linsdale (1936) and studies
at Las Vegas (Austin, 1970) and the lower
Virgin River (Wauer, 1969); Utah studies on the
upper Virgin River (Wauer, 1967; Wauer, 1969),
Zion National Park (Wauer and Carter, 1965);
the Arizona summary by Phillips, Marshall,
and Monson (1964) ; the New Mexico summary by
Hubbard (1970); and west Texas studies in Big
Bend National Park (Wauer, 1973), the Lower
Canyons on the Rio Grande (Wauer, in press),
and Amistad Recreation Area (LoBello, 1976).

Examination of the 94 riparian species
indicates that only the mourning dove, verdin,
northern oriole, brown-headed cowbird, and house
finch occur within all nine riparian areas.
Additional dominant species include ladder-
backed woodpecker, ash-throated flycatcher,
yellow-breasted chat, hooded oriole, and blue
grosbeak — recorded for eight areas — and white-
winged dove, black-chinned hummingbird, northern
mockingbird, and Bell's vireo recorded for seven
areas. Seventy-nine additional species were
recorded on one to six of the areas, and only
13 species are listed only once: gray hawk in
Arizona, California quail in Death Valley

166

TABLE 1 con't

Breeding Avifauna of Riparian Systems in Southwest U.S.

SPECIES3

SE Calif
1* 2

Nev
3

Utah
4

Ariz
5

NMex
6

W.
7

Tex
8 9

Verdin

(Auriparus flaviceps)

Bushtit

(Psaltriparus minimus)

White-breasted Nuthatch

(Sitta carolinensis)

House Wren

(Troglodytes aedon)

Bewick's Wren

(Thryomanes bewickii)

Carolina Wren

(Thryothorus ludovicianus)

Cactus Wren

(Campy lorhynchus brunneicapillus)

Northern Mocking bird

(Mimus polyglottos)

uray catbird

(Dumetella carolinensis)

Crissal Thrasher

(Toxostoma dorsale)

American Robin

(Turdus migratorius)

Western Bluebird

(Sialia mexicana)

Blue-gray Gnatcatcher

(Polioptila caerulea)

Black-tailed Gnatcatcher

(Polioptila melanura)

Phainopepla

(Phinopepla nitens)

Starling

(Sturnus vulgaris)

Bell s Vireo

(.vireo bellii.)

Solitary Vireo

(Vireo solitarius)

Warbling Vireo

(Vireo gilvus)

Lucy's Warbler

(Vermivora luciae)

Yellow Warbler

(Dendroica petechia)

Yellow-breasted Chat

(Icteria virens)

Orchard Oriole

(Icterus spurius)

Hooded Oriole

(Icterus cucullatus)

Northern Oriole

(Icterus galbula)

Great-tailed Grackle

(Quiscalus mexicanus)

167

TABLE 1 con't

Breeding Avifauna of Riparian Systems in Southwest U.S.

SPECIES

SE Calif Nev

Utah

Ariz

NMex

Tex

1 2. 5

r
0

Broad-billed Hummingbird

(Cynanthus latirostris)

Common Flicker

(Colaptes auratus)

Golden— fronted Woodpecker

Y
A

(Centurus aurifrons)

Gila Woodpecker

v
X

(Centurus uropygialis)

Lewis' Woodpecker

(Melanerpes lewis)

Hairy Woodpecker

Y
A

(Dendrocopos villosus)

Downy Woodpecker

(Dendrocopos pubescens)

Ladder-backed Woodpecker

V V
A A

v
X

V Y
X X

(Dendrocopos scalaris)

Rose-throated Becard

v
X

(Platypsaris aglaiae)

Eastern Kingbird

(Tyrannus tyrannus)

Tropical Kingbird

(Tyrannus melancholicus)

Western Kingbird

X A

v
X

(Tyrannus verticalis)

Cassin's Kingbird

v
X

(Tyrannus vociferans)

Thick-billed Kingbird

(Tyrannus crassirostris)

Wied's Crested Flycatcher

v
X

(Myiarchus tyrannulus)

Ash-throated Flycatcher

A X

v
X

Y Y

X X

(Myiarchus cinerascens)

Olivaceous Flycatcher

Y
A

(Myiarchus tuberculif er)

WlllOW riyCdCCncL

(Empidonax traillii)

Hammond's Flycatcher

(Empidonax hammondii)

Western Flycatcher

Y
A

(Empidonax difficilis)

Western Wood Pewee

Y
A

(Contopus sordidulus)

Vermilion Flycatcher

(Pyrocephalus rubinus)

Beardless Flycatcher

(Camptostoma imberbe)

Violet-green Swallow

(Tachycineta thalassina)

Black-billed Magpie

(Pica pica)

Black-capped Chickadee

(Parus atricapillus)

168

TABLE 1

Breeding Avifauna of Riparian Systems in Southwest U.S.

SPECIES3

SE Calif Nev

Utah

Ariz

NMex

W. Tex

1* 2 3

7 8 9

Great Blue Heron

(Ardea herodias)

Green Heron

h p

(Butorides striatus)

Great Egret

(Casmerodius albus)

Snowy Egret

(Egretta thula)

Black-crowned Night Heron

(Nycticorax nycticorax)

Cooper's Hawk

(Accipiter cooperii)

Gray Hawk

(Buteo nitidus)

Black Hawk

(Buteogallus anthracinus)

American Kestrel

(Falco sparverius)

California Quail

(Lophortyx calif ornicus)
Gambel's Quail

XX X

(Lophortyx gambelii)

White-winged Dove

XXX

(Zenaida asiatica)

Mourning Dove

XX X

XXX

(Zenaida macroura)

Common Ground Dove

XXX

(Columbigna passerina)

Inca Dove

X X

(Scardafella inca)

Yellow-billed Cuckoo

X X p

(Coccyzus americanus)

Greater Roadrunner

X X

(Geococcyx calif ornianus)

Common Screech Owl

X X

(Otus asio)

Great Horned Owl

(Budo virginianus)

Ferruginous Owl

(Glaucidium brasilianum)

Elf Owl

X X

(Micrathene whitneyi)

Long-eared Owl

(Asio otus)

Lesser Nighthawk

X X p

(Chordeiles acutipennis)

Black-chinned Hummingbird

XXX

(Archilochus alexandri)

Costa's Hummingbird

X X

(Calypte costae)

Violet-crowned Hummingbird

(Amazilia verticalis)

169

TABLE 1 con't

Breeding Avifauna of Riparian Systems in Southwest U.S.

SPECIES3

SE Calif
1* 2

Nev
3

Utah
4

Ariz
5

NMex
6

Tex
8 9

Brown-headed Cowbird XX X X X X XXX

(Molothrus ater )

Bronzed Cowbird XXX
(Molothrus aeneus)

Summer Tanager XX X XXX

(Piranga rubra)

Northern Cardinal X X XXX

(Cardinalis £. )

Black-headed Grosbeak X XXX

(Pheucticus melanopephalus)

Blue Grosbeak XX X X XXXX

(Guiraca caerulea)

Indigo Bunting XXX
(Passerina cyanea)

Lazuli Bunting XX XX X X

(Passerina amoena)

Painted Bunting X XXX

(Passerina ciris)

House Finch XX X X X X XXX

(Carpodacus mexicanus)

American Goldfinch X X

(Spinus tristis)

Lesser Goldfinch XXXX X

(Spinus psaltria)

Lawrence's Goldfinch X
(Spinus lawrencei)

Rufous-sided Towhee X XX

(Pipilo erythrophthalmus)

Abert's Towhee XXXX
(Pipilo aberti)

Song Sparrow XXXX
(Melospiza melodia)

a = names follow AOU (1957, 1976)
h = historical only
p = probable

TOTALS 24 16 57 51 59 75 33 27 20

*1 = Wauer, R.H. , 1962a, 1962b, 1964

2 = Remsen, J.V. , 1977

3 = Linsdale, J.M., 1936; Wauer, R.H., 1969; Austin, G.T., 1970

4 = Wauer, R.H. , 1967; Wauer, R.H. and D.L. Cater, 1965

5 = Phillips, A., J. Marshall, G. Monson, 1964

6 = Hubbard, J.P. , 1970

7 = Wauer, R.H. , 1973

8 = Wauer, R.H., (in press)

9 = LoBello, R.L. , 1976

170

(an introduced species there according to
Wauer (1962)), ferruginous owl in Arizona,
Lewis' and downy woodpeckers in New Mexico,
rose-throated becard and tropical kingbird in
Arizona, Cassin's kingbird in west Texas,
eastern kingbird and Hammond's flycatcher in
New Mexico, western flycatcher in Zion National
Park, Utah, Carolina wren in the Lower Canyons
of the Rio Grande, and Lawrence's goldfinch
in southern Arizona.

Significance of Riparian Habitat in West Texas

Before roads, the Rio Grande corridor was
the most sensible route of traveling through
the harsh, arid environment. It provided a
practical passageway between the Chihuahuan
Desert region of west Texas and the semi-
tropical Lower Valley. It was used by man and
wildlife alike. Today, the Rio Grande route
is left to wildlife and plants, to a few re-
creationists, and those of us who need the
"spirit of the river."

The Rio Grande still serves as migratory
and emigration routes for plants and animals.
Plant species washed away by high water from
one place may be deposited in an appropriate
place to take root many miles downriver. Down-
river species may edge slowly along or be trans-
ferred distances by some natural agent. Ex-
tensions of western and eastern forms occur
all along the Rio Grande waterway. This
phenomena, as it pertains to the flora, is
discussed by Johnson, et. al. (in press), but
the natural parameters that restrict plant
species, such as temperatures and soils, do
not always apply to the more mobile avifauna.

Several breeding birds known to occur with-
in the Texas Big Bend Country appear to owe
their presence there to the river corridor.
Good examples of breeding birds of the Rio
Grande area with eastern affinities are orchard
oriole, hooded oriole, and great-tailed grackle.
Orchard oriole frequents deciduous environments
throughout the eastern United States and its
breeding range extends through west Texas to
El Paso along the Rio Grande. Two races of
hooded orioles breed within riparian zones
along the lower Rio Grande. The nominant form
occurs from near the eastern border of Big
Bend National Park downriver to near Laredo
(Oberholser, 1974). The breeding range of
J-.£. nelsoni extends westerly along the river
in Texas through Big Bend National Park. The
breeding race of great-tailed grackle (C.m.
monsoni) extends only within the Trans-Pecos
and along the Rio Grande to about Amistad
Reservoir where C^.m. prosopidicola occurs
(Selander and Giller, 1961).

The Rio Grande corridor further explains
the presence of a number of post-nesting birds
that breed out of the area and wander into the
area afterwards. The green kingfisher (Chloro-
ceryle americana) breeds in central and southern
Texas and disperses outward after the breeding
season. Records exist along the Rio Grande to
the west edge of the Big Bend National Park
(Wauer, 1973; Oberholser, 1974). The great-
crested flycatcher (Myiarchus crinitus) , a
species that nests throughout the eastern two-
thirds of Texas (Oberholser, 1974), has been
recorded only from August 24 through October 29
in Big Bend National Park (Wauer, 1973). The
third example is the Carolina wren that breeds
to the eastern edge of the area of concern, and
all records west of Del Rio exist along the
Rio Grande (Wauer, 1973; Oberholser, 1974).

The use of the Rio Grande corridor for a
migratory route is one of the well recognized
patterns of southwestern avian behavior.
Wauer (1973) thoroughly discussed the spring
and fall migration through Big Bend National
Park, and much of that discussion applied to
the rest of the lower Rio Grande area.

Although the characteristic of northbound
and southbound bird movement along major water-
ways is common elsewhere, it is likely that a
river corridor is more important to migrating
birds in arid parts of the country than in humid,
heavier vegetated areas. In west Texas, hot-
dry desert conditions prevail during the prin-
cipal portion of the spring and fall migration.
Therefore, the availability of food, water,
cover, and suitable north-south routing are
exceptionally important and strongly influence
migrants. A quantitative analysis of bird
movement along the Rio Grande corridor is not
available, although such a study would be ex-
tremely worthwhile.

The migration periods for Big Bend National
Park were analyzed by Wauer (1973) who stated
that "Fall migration in the lowlands is only a
shadow of the spring movement." To understand
the migration patterns for the avifauna it is
necessary to point out that the area is split
into two halves by a mountain range that runs
north-south along the eastern side of Big Bend
National Park. This range — the Sierra del
Carmens — extends north of the Rio Grande for
about 60 miles and south for about 100 miles.
It results in a splitting of the northbound
migrants so that the north-south valleys on
both sides of the del Carmens are utilized.
Fall migrants following the valleys southward
are diverted along the northern boundary of
the park by the Santiago Mountains (an exten-
sion of the Sierra del Carmens that continue
northwesterly) toward the southeast and along
the eastern side of the del Carmens. This

171

route is through the desert north of Persimmon
Gap and into the Black Gap Wildlife Management
Area and then into the Rio Grande canyons and
riparian habitats. Southbound migrants un-
doubtedly find water and food within the lush
riparian vegetation along the river that is
more inviting than that in the adjacent desert
landscapes .

The lower Rio Grande also is a refugium
for nesting avifauna that rely upon the ripar-
ian systems. I have selected nine species to
analyze: great blue heron, green heron, pere-
grine falcon, American kestrel, white-winged
dove, screech owl, Bell's vireo, yellow warbler
and bronzed cowbird.

Great blue heron and green heron are known
to have nested within the riparian vegetation
along the Rio Grande in historic times (Van
Tyne and Sutton, 1937). Wauer (1973), reported
that great blue herons are present in Big Bend
National Park all year but no recent nesting.
He stated that, "Today, between Boquillas and
Presidio, there is only a single grove of
cottonwoods and willows large enough to support
a rookery; it is located on the floodplain
near Santa Elena Crossing, where there probably
is too much human activity for nesting herons."
More recently, LoBello (1976), reported a nest
containing four eggs, on Javelina Bluff near
Rough Canyon Marina, Amistad Reservoir, April
26, 1975. The nest was located on a sheer
cliff about 200 feet above the water. It is
possible that this species can adapt to soli-
tary nesting on cliffs within the Rio Grande
canyons to continue its status as a breeding
bird of the lower Rio Grande. It will un-
doubtedly continue to utilize the riparian
vegetation for perching, cover, and finding
food.

There are no known records of breeding
green herons in recent years. I suspect that
nesting does occur in out of the way places
within the riparian zone. It is present
throughout the summer months at Big Bend
National Park (Wauer, 1973), and LoBello (1976)
found it at Amistad Reservoir consistently in
summer, including an immature bird on August 11.
Riparian vegetation is essential to the sur-
vival of this heron for perching, roosting,
and hunting. It is extremely unlikely that
cliff-nesting is feasible for this species.

Peregrine falcon is an excellent tribute
to the wilderness character of the Rio Grande
canyons and the accessability of relatively
non-polluted food. It nests on high canyon
walls and hunts for food along the riverway.
Mourning and white-winged doves, common nesting
birds of the riparian habitat, provide suffi-
cient food supplies for at least five pairs of

this endangered species from Big Bend National
Park to Amistad Reservoir (Wauer, 1973; Hunt,
1975) .

Peregrines have declined drastically within
the United States since the 1940 's. The popu-
lation east of the Mississippi River and south
of the boreal forests was estimated at 400
pairs, and today not a single breeding wild
pair remains. In the western United States the
population is only ten percent of what it was
in the 1940' s. So, the Rio Grande canyons
serve as a significant refugium for this highly
endangered species. American kestrel is a
common nesting bird along the Rio Grande canyons
It is an adaptable species that obtains its
food from many sources, including the riparian
habitats within the Rio Grande corridor and
adjacent arroyos . Also, I have found it numer-
ous in migration, when up to half-dozen birds
can be expected along a few miles of riparian
habitats. The principal migration occurs from
late September through November and mid-March
through early May (Wauer, 1973).

White-winged dove is resident along the
Rio Grande throughout the Big Bend area, and
flocks at the more extensive riparian zones
during winter; Wauer (1973) recorded 70 indi-
viduals at Rio Grande Village on January 22,
1970. Generally, white-winged dove populations
have been severely depleted throughout much of
their range in recent decades. In the Lower
Valley of the Rio Grande, thousands of acres
of land were cleared of native brushland during
the 1940 's and 1950' s, and the white-winged
population took a drastic nosedive. Since
then, citrus trees have been planted and sup-
port nesting white-wings, although the popula-
tion has not returned to what it was under
natural conditions.

A combination of land clearing, land
flooding by Falcon and Amistad Reservoirs, and
the addition of increasing hunting pressures
upon the white-winged dove populations in
Mexico (several populations from the United
States winter in Mexico) have had severe
negative effects upon the Texas white-wings.
The riparian zones within the Big Bend country
provide some of the most stable known habitat.

Screech owl occurs within the riparian
habitat along the river and adjacent arroyos
and oases throughout the year. The Rio Grande
corridor appears to provide habitat for ex-
tensions of ranges of two races that overlap
within the Big Bend area. Marshall (1967)
found hybridizing Otus asio suttoni and O.a.
mccalli in the riparian vegetation near Rio
Grande Village.

172

Bell's vireo may be regarded as the most
numerous breeding bird within the Rio Grande
riparian systems of west Texas. Its song is
ubiquitous among the floodplain and arroyo
vegetation from late March through early June.
Two races breed within the area; the nominant
form along the eastern edge, and V.b_. medius
through the Big Bend National Park area (AOU,
1957) .

In Arizona, the species is "scarce and
local in at least the Phoenix and Benson areas,
this tiny bird was certainly decimated in the
latter by cowbird parasitism" (Phillips, 1968).
It is strange that the species is so abundant
within the west Texas riparian systems in spite
of the abundance of brown-headed cowbirds.

Summary

The significance of the Rio Grande ripar-
ian zones within west Texas can only be specu-
lated upon, but evidence suggests that it is
of major importance. Several avian species
are present that are absent or rare elsewhere,
and numerous species utilize the river corridor
as routes through inhospitable habitat. So
much of the riparian communities of the South-
west have been destroyed and changed in recent
decades, that one that possesses natural char-
acteristics must be given special protection.
The importance of a relatively unchanged ex-
tensive riparian system becomes more signifi-
cant daily.

Yellow warbler is absent or a rare breed-
ing bird of riparian zones within west Texas.
This apparently was not the case during earlier
years. Wauer (1973) stated that, "Van Tyne
and Sutton reported that it nested at Boquillas,
Hot Springs, and San Vicente during the 1930' s,
but I have searched the floodplain for nesting
birds without success. In five years (1966-
1971) I have found only three summertime
birds." Allan Phillips (1964) believes that
the species has been extirpated by parasitism
of brown-headed cowbirds in some parts of
southern Arizona, and this may well be the
case in the Big Bend; cowbirds have increased
in recent years and are now abundant along the
river floodplain where yellow warblers once
nested .

Bronzed cowbird has been a summer resident
within the Big Bend National Park only since
1969 (Wauer, 1973). There were no summertime
records of the species prior to 1969, but there
are numerous records within the riparian zones
since. These recent sightings include several
cases of parasitism, principally on hooded and
orchard orioles. These orioles are common
within the riparian zones, but long-range
analysis of their status should be undertaken
to determine if the recent impacts pose signi-
ficant threats to their populations.

Phillips (1968) states that the species
was first seen in Arizona in 1909, but spread
through the southern one-half of the state
within 20 years. He suggests that additional
spread was restricted by the species failure
to find enough hooded oriole nests to para-
sitize. Based upon these comments, it is
likely that the bronzed cowbird will readily
spread throughout the Rio Grande at least
north of El Paso where hooded orioles breed.

LITERATURE CITED

American Ornithologists' Union.

1957. Check-list of North American Birds,
5th Ed. Amer. Ornith. Union, Baltimore.
American Ornithologists' Union.

1976. Thirty-third supplement to the

American Ornithologists' Union check-list
or North American birds.
Austin, G.T.

1970. Breeding birds of desert riparian
habitat in southern Nevada. Condor,
72:431-436.
Hubbard, J. P.

1970. Check-list of the birds of New

Mexico. New Mexico Ornithol. Soc. Publ.
No. 3.
Hunt, W.G.

1975. The Chihuahuan Desert peregrine
falcon survey, 1975. Typewritten report
to the National Park Service.

21 pages
Johnson, M.C. et.al.
(in press)

A botanical survey of the lower canyons
of the Rio Grande. Lower canyons of the
Rio Grande, Univ. Texas Press, Austin.

Linsdale, J.M.

1936. The birds of Nevada. Pacific Coast
Avifauna #023.

LoBello, R.L.

1976. Vertebrates of the Lake Amistad
National Recreation Area, Texas. Thesis,
Sul Ross State Univ. , Alpine, Texas.

Marshall, J.T.

1967. Parallel variation in North and
Middle America screech owls. Monog.
West Foun. Vert. 1

Oberholser, H.C.(Ed. E.B. Kincaid , Jr.)

1974. The bird life of Texas. Univ Texas
Press, Austin.
Phillips, A.

1968. The instability of the distribution
of land birds in the Southwest. Papers
of the Arch. Soc. New Mex. 1:129-162.

173

Phillips, A., J. Marshall, and G. Monson.
1964. The birds of Arizona. Univ. of
Arizona Press, Tucson.
Remsen, J.V.

1977. Desert Riparian. American Birds.
31:77

Selander, R.K. , and D.R. Giller.

1961. Analysis of sympatry of great-
tailed and boat-tailed grackles. Condor,
63:29-86.
Van Tyne, J., and G.M. Sutton.

1937. The birds of Brewster County, Texas.
Misc. Publ., Mus. Zool., Univ. Mich.,
37:1-119.
Wauer, R.H.

1962a. Birds of Death Valley National
Monument. Typewritten report to N.P.S.,
Southwest Region. 228 pages.

1962b. A survey of the birds of Death
Valley. Condor, 64:220-233.

1964. Ecological distribution of the birds
of the Panamint Mountains, California.
Condor, 66:287-301.

1969. Recent bird records from the Virgin
River Valley of Utah, Arizona, and Nevada.
Condor, 71:331-335.

1973. Birds of Big Bend National Park and
Vicinity. Univ. Texas Press, Austin.

(in press)

The birds of the Lower Canyons of the
Rio Grande. Lower Canyon of the Rio
Grande. Univ. Texas Press, Austin.

Wauer, R.H. , and L. Carter.

1965. Birds of Zion National Park and
Vicinity. Zion Nat. Hist. Assoc.

Wauer, R.H., and r.c. Russell.

1967. New and additional records in the
Virgin River Valley. Condor, 69:420-423.

West, R.C. (Ed.)

1964. Natural environment and early cultures.
Handbook of Middle American Indians, Vol.
1, Univ. Texas Press, Austin.

174

Some Effects of a Campground
on Breeding Birds in Arizona1

Stewart W. Aitchison_/

Abstract. — Over a three year period, breeding bird den-
sities were found to be similar between a constructed camp-
ground and a relatively natural area when the campground was
closed to campers. However, bird species composition differ-
ed between sites, the campground having relatively heavier
bodied birds {x = 48.5 g) than the control area (x" = 38.2 g) .
Once the campground was opened for human use, the breeding
bird population decreased in density and diversity. On the
control site the population either remained the same or
increased.

INTRODUCTION

It has been well documented that the
human manipulation of Southwestern habitats
greatly affects the configuration of the avian
community that will continue to utilize the
area (e.g., Carothers et al. 1974, Carothers
and Johnson 1975) . These studies have prima-
rily concerned themselves with phreatophyte
control, channelization, and other water
management practices. Very little research
has dealt with the impact caused by the con-
struction of permanent structures and human
occupation of these areas (e.g., subdivisions,
trailer parks, and campgrounds) . The present
study examines the effects of a U.S. Forest
Service improved campground upon breeding birds .

ACKNOWLEDGEMENTS

The author wishes to thank M.E. Theroux
for collecting and processing the vegetation
data, S.W. Carothers and O.J. Reichman for
critically proofing the manuscript, and many
other members of the Museum of Northern Arizona
staff for their help and encouragement. Special
thanks goes to B. Johnston who did nearly all
the field work in 1974 and a substantial amount
in 1975.

V Contributed paper, Symposium on the
Importance, Preservation and Management of the
Riparian Habitat, July 9, 1977, Tucson, Arizona.
_/ Research Assistant, Museum of Northern
Arizona, Route 4, Box 720, Flagstaff, Arizona
86001.

Financial support for this study was par-
tially provided by the U.S. Forest Service,
Coconino National Forest, and to this end I
wish to thank B. Burbridge and W. Finley. Also,
the Sedona Ranger District graciously allowed
us to set up the two study plots in Oak Creek
Canyon .

STUDY AREA

Two sites were chosen which appeared sim-
ilar in vegetative structure, each being com-
posed primarily of ponderosa pine with some
cottonwood, Arizona walnut and other deciduous
trees and shrubs. (Table 1 summarizes the
vegetation analysis of the study areas.) Both
sites are within Oak Creek Canyon, Sec. 27,
T19N, R6E, Coconino County, Arizona, at an
elevation of 1,646 m. One plot was located
within the Cave Springs Campground; the other
plot, control, was slightly north and across
the Oak Creek.

The Cave Springs Campground study site
is 4.0 ha. Extensive timber and shrub removal,
construction of roads, pit toilets, and erection
of tables has occurred here. The campground is
open for public use from approximately Memorial
Day to Labor Day of each year.

Control is approximately 2.0 ha. Relative
to the campground, this area is undisturbed by
human activities.

Though the two sites are small in area, thus
making extrapolation of figures somewhat mis-
leading, they encompass as much homogeneous
habitat as possible. Further, the small size
enabled the investigators to know the avian
components intimately, and therefore we feel a
very accurate count of species was achieved.

175

Table 1 . --Vegetation Analysis

Study

Tree

Basal

Average

Density

Area (m )

Tree

Area

Per

Height (m)

Hectare

Ponderosa Pine

Pinus Ponderosa

437 . 1

2153.6

12 .4

CAMPGROUND

*A11 other species

316.5

415.0

7.3

Ponderosa Pine

347.8

2118.4

14.0

CONTROL

**A11 other species

1010.7

447.8

5.0

* Acer negundo , Alnus oblongifolia , Juniperus scopulorum, Quercus gambelli , Populus lanceolata ,
Salix gooddingii .

** Acer negundo , Alnus oblongifolia, Fraxinus pennsylvanica var. velutina , Juniperus scopulorum,
J. monosperma, Pinus edulis , Populus lanceolata, Quercus spp.

METHODS

The populations of breeding birds were
determined by the spot-map method (Williams
1936, Kendeigh 1944) . On each census the
location of singing males, song posts, and
nest sites was recorded for each census and
information on every species was later recorded
onto species maps. Censusing was carried out
from 4 April to 6 July 1973, 18 February to
1 July 1974, and 9 May to 10 July 1975; these
periods included the entire observable breeding
season. Densities were determined for each
area before and after the date the campground
was opened (e.g., 29 May 1973, 17 May 1974,
and 16 May 1975). [Note: All densities were
extrapolated to 40 ha to make inter-area
comparisons easier.]

Foliage height diversity (FHD) was sampled
along ten 100 m transects established at random
throughout the study areas. Presence or absence
of vegetation at 2 m intervals along the trans-
ects was noted at three layers chosen to
approximate foliage stratification into
herbaceous (0-0.6 m) , shrub (0.6-4.49 m) and
canopy (>4.5 m) layers. A 4.5 m rod marked
at 0.6 m from one end was used to record the
presence or absence of foliage for the herba-
ceous and shrub layers, and the ocular tube
method (Winkworth and Goodall 1962) was used
for the canopy layer. For recording the pres-
sence of vegetation in the herbaceous and
shrub layers, it was necessary for green foliage
to touch the vertically held rod.

All vegetative and avian diversity indices
are computed as H' = - ^P^loggP. based on
the Shannon-Wiener model of information theory

(see Shannon and Weaver 196 3) as it applies to
biological parameters (MacArthur and MacArthur
1961; MacArthur 1965; Pielou 1966a, b; Lloyd
et al. 1968) .

Tree density, species composition and basal
area were determined by the plotless point-
quarter method of Cottam and Curtis (1956) .
Tree heights were determined by use of a clino-
meter. Samplings witha diameter at breast
height (DBH) of less than 7.6 cm were treated
as shrubs .

Avian standing crop biomass (SCB) was
determined by taking the average adult weight
(W) times the number of adults per unit area.
Existence Energy (EE) or the amount of kcal
consumed per ha per 24 hours was calculated from
these two formulae :

Log EE = 0.3581 + 0.5876 Log W (for passerines)
Log EE = 0.0649 + 0.6722 Log W (for non-passer-
ines) .

These formulae give the energy requirements
to maintain a constant weight at rest. To
determine actual community energetics it would
be necessary to include energy requirements of
the immature birds , and the various energy
demanding activities of breeding birds (e.g.,
singing, displaying, nest building) . The
limitations of this procedure notwithstanding,
it is instructive to make inter- and intra-
community comparisons with these low estimates
of avian community energetics (see Karr 1968
and 1971) .

RESULTS AND DISCUSSION
A total of 58 species (Table 2) of birds

176

1 CO

-P

•

i|H

i oi

1 I— 1

1 CO

U-J

1 •

i cn

1 rH

U MH
0)

01 01 CO 01 01 I I I

01 01 01 01 CTN I

I I I I

en o> cri I i i i i i

• • ■ i i i i i i

01 01 Ol I i I I i I

i i i i i i

■

CTl

•

i-4

i I I co i i co oi i i ^o I

ill -ii • • i l • l

I I I 01 I I (II o\ I I 01 I

I I I r— I I I rH I I CO I

I i I r— i i i oi i i r-

i oi

I (N

i in co I co ro in co co in oi

l • • i

I oi oi l oi Oi Ci oi Oi ci oi

I n h I i-HunrOrHHro

I I CO I

I 01 I
I CN I

I CO CO CO CO
I 01 W 01 <Tl

I I rH I I rH rH rH

I I CO I CO CO I CO I

II • I • • I • I
I I 01 I 01 01 I Ol I

I I rH I rH 1/1 I rH I

i co co co oi r- co oi i i

I I I I I

I in i co

I • i

I 01 I 01

I 00 I rH

I • • I I I I I

IO101I I I I 1010101010101011
I I I I I I H H H CN rH I

i I I icor-covor-oiaii

I I I I OI Ol 01 Ol Ol Ol Ol I
I I I I rH (N rH rO CN I

I CO I I I I I I CO I I

I • I I I I I I -II

I 01 I I I I I I 01 I I

I rH I I I I I I rH I I

1 1 1 CO 1 1

1 1 CO

14-t

III 'II

1 1

•

1 1 1 01 1 1

1 1 01

1 1 1 rH 1 1

1 1 rH

1
1

i
i

1
1

1 01 1 1 1
1 • 1 1 1

1 1 01

•

1 01 1 1 1

i i

1 1 01

•

1 III

i i

r- ai oi i i

i i i

I 01 ID I I
I • • I I
I 01 01 I I

I 01 01 I

l co in i i

l • • l l

I 01 01 I I

I rH ro I I

I CO CO I

I 01 01 I

I rH rH I

I lO I

I ■ I

I 01 I

I ro I

I t 01 01
I CN

i oi oi i oi ai

•

CO 1 1

1 1 CO

00 1

1 ro

-3"

• 1 1
01 1 1

1 1

1 1 01

• i

01 1

1 Ol

rH | |

1 1 rH

rH 1

l in

I I I oo I

I <tf I

I 10 I

I co co i co i in i

i • ■ i -i • i

I 01 01 I 01 I 01 I

I rH rH I rH I 00 I

CO 1 1

1 1 CO

m i

1 ro

• 1 1

01 1 1

1 1

1 1 01

• l

01 1

1 01

rH 1 1

1 1 rH

ro 1

I in

I I I I I -I

I I I I I 01 I

I I I I I rH I

I I ro |

II ■ I
I I 01 I

i i in I

■y

+->

■H

-H

rt)

tji

"rO

•H

■a

•H

•rt

-P

•H

tji

rt)

■H

CQ 3

cn mh
a) P

•

, — i

r»

, 1

O
•

•

L/l

l£>

m
■

•

r 1

[ —

•

10
•

•

r-
•

•

r_|

r~-

-tf

o
•

•

Lfl

CN
•

, |

•

<3"

•H

■H

•H

-P

•H

■H

177

■MM«i%fVlW1r1ftMyililMI1Mlff" \?l.^>U*.g-X^H '/WW

Table 2. — Species recorded on or immediately
adjacent to the study area. [Note; Bird
scientific names according to A.O.U. Checklist,
1957; and 32nd A.O.U. Supplement, Auk 90(2):
411-419. ]

Red-tailed Hawk Buteo jamaicensis
Turkey Vulture Cathartes f asciata
Band-tailed Pigeon Columba f asciata
Mourning Dove Zenaidura macroura
Great Horned Owl Bubo virginianus
Flammulated Owl Otus f lammeolus
White-throated Swift Aeronautes saxatalis
Broad-tailed Hummingbird Selasphorus playcercus
Belted Kingfisher Megaceryle alcyon
Red-shafted Flicker Colaptes caf er
Hairy Woodpecker Dendrocopos villosus
Yellow-bellied Sapsucker Sphyrapicus varius
Western Kingbird Tyrannus verticalis
Cassin's Kingbird Tyrannus vociferans
Black Phoebe Sayornis nigricans
Western Flycatcher Empidonax dif f icilis
Western Wood Pewee Contopus sordidulus
Violet-green Swallow Tachycineta thalassina
Steller's Jay Cyanocitta stelleri
Common Raven Corvus corax
Mountain Chickadee Parus gambeli
Dipper Cinclus mexicanus

White-breasted Nuthatch Sitta carolinensis
Pygmy Nuthatch Sitta pygmaea
Brown Creeper Certhia f amiliaris
House Wren Troglodytes aedon
Canyon Wren Catherpes mexicanus
Mockingbird Mimus polyglottos
Robin Turdus migratorius
Hermit Thrush Catharus guttata
Townsend's Solitaire Myadestes townsendi
Ruby-crowned Kinglet Regulus calendula
Solitary Vireo Vireo solitarius
Warbling Vireo Vireo gilvus
Virginia's Warbler Vermivora virginiae
Yellow Warbler Dendroica petechia
Audubon's Warbler Dendroica auduboni
Grace's Warbler Dendroica graciae
MacQillivray ' s Warbler Oporonis tolmiei
American Redstart Setophaga ruticilla
Painted Redstart Setophaga picta
Red-faced Warbler Cardellina rubrif rons
Wilson's Warbler Wilsonia pusilla
Hooded Oriole Icterus cucullatus
Bullock's Oriole Icterus galbula
Brown-headed Cowbird Moluthrus ater
Western Tanager Piranga ludoviciana
Hepatic Tanager Piranga f lava
Summer Tanager Piranga rubra

Rose-breasted Grosbeak Pheucticus ludovicianus
Black-headed Grosbeak Hesperiphona vespertina
Indigo Bunting Passerina cyanea
Pine Siskin Spinus pinus
American Goldfinch Spinus tristis
Lesser Goldfinch Spinus psaltria
Rufous-sided Towhee Pipilo erythrophthalmus
Gray-headed Junco Junco caniceps

were seen on or immediately adjacent to the
study areas. Of these, 2 3 species nested on
one or both of the study areas at least once
during the three years of censusing. The
density and species' richness (Whittaker 1970)
are summarized in Table 3. The changes in
these values and other resultant calculations
prior to and after the opening of the camp-
ground are discussed below as indicators of
human impact upon the breeding bird community.

Avian Density and Species Richness

Every species is apparently adjusted to
breed at the time of year at which it can raise
its young most efficiently (Immelmann 1971) .
For most northern temperate birds this nesting
period extends from late spring to mid-summer
(Lack 1950) . This is certainly true for the
Cave Springs area of Oak Creek Canyon, where
breeding begins about mid-April and lasts
through July. The campground opening date
falls within this period.

1973. --In 1973 a total of 17 species nested on
one or both of the study areas (Table 3) . A
40 percent decrease in density occurred on
the campground after the opening day. Part

of the losses incurred were through direct
human manipulation of the nest site. Forest
Service employees, by removing trees and slash,
destroyed 20 percent of the Steller's Jay nests.
Campers destroyed 30 percent more of the
Steller's Jay nests and 20 percent of the Robin
nests by removing branches for firewood, making
room for tents , and other reasons .

The parulid warblers, Solitary Vireos,
Broad-tailed Hummingbirds, and Hairy Woodpeckers
abandoned their nests but occasionally foraged
within the area. No losses can be attributed
to adults leaving with fledged young prior to
the opening date. Of those actually nesting
on 20 May 1973, breeding had not proceeded
beyond the incubation stage.

The density on the control site increased
12.1 percent after the opening date. This was
not due to individuals emigrating from the
campground but rather to the arrival of mid-
summer breeders, namely, the Red-faced Warbler,
Western Wood Pewee, and Hepatic Tanager (Bent
1968) .

After the opening of the campground, species
richness went from 12 to 8 on the campground
and increased from 9 to 12 on control.

1974 . — In 1974 a total of 17 species nested on
one or both of the study areas (Table 3) .
However, these were not the same 17 of 1973.
There was a change of three species, with
Black Phoebes, Western Flycatchers and Hepatic
Tanagers being replaced by Pygmy Nuthatches ,
Warbling Vireos and Summer Tanagers. This
area in Oak Creek is an ecotonal situation
between confierous forest and a deciduous

178

riparian habitat. The Summer Tanager prefers
cottonwoods along streams for nest sites and
apparently found conditions suitable in 1974.
On the other hand, the Hepatic Tanager prefers
pines and oaks and found Cave Springs acceptable
in 1973. It is probably subtle environmental
differences (i.e., temperature, rainfall, etc.)
that determine which tanager will be present
in what might be considered marginal habitat
for either.

A 25 percent decrease in densities occurred
on the campground, whereas there was a dramatic
87.3 percent increase on control. The campground
was opened 12 days earlier than in 1973 and
may account for the initially low density on
control. It was simply too early for many
species to be breeding. Yet initial densities
on the campground matched 1973 figures. No
satisfactory explanation has been found.

Species richness dropped from 16 to 13
on the campground and climbed from 8 to 12 on
control .

1975. — During 1975 a total of 21 species nested
on one or both of the study areas (Table 3) .
New breeders included Mourning Doves, Cassin's
Kingbirds, Virginia's Warblers, Western Tanagers,
Lesser Goldfinches, and Rufous-sided Towhees.
In previous years all of these had either
appeared as transients or nested within the
canyon but off the study areas.

In 1975 for the first time there was in
increase in density (44.4 percent) on camp-
ground. In 1973 and 1974 breeding activity
was well underway on the campground prior to
the opening date. In 1975, however, colder
temperatures, higher winds, and increased
precipitation postponed breeding. In many
species a positive correlation between tem-
perature and the rate of testicular development
or egg production has been found (Farner and
Wilson 1957). In 1975 the average temperature
two weeks prior to opening was 57.7°F and for
the same period in 1973 and 1974 was 63.7°F
and 62.5°F, respectively. Perusal of the
1975 censuses indicates almost no breeding
activity (i.e., singing, displaying, nest
building) before 16 May.

It is interesting to note that once breed-
ing did commence, the maximum density reached
was still less than the maximum measured in
1973 and 1974.

On control there was in increase of 149.8
percent. This phenomenal climb is also no
doubt related to the later breeding period.
Before the opening day, weather conditions
were too severe for breeding to commence. In
addition, several species that would normally
nest elsewhere (e.g., Rufous-sided Towhees
usually nest in the chaparrel found on the
canyon walls and Lesser Goldfinches usually nest
above the rim) were found on the control.
Perhaps environmental conditions were relatively
less severe within the canyon than elsewhere

and these species chose to accept marginal
habitat under these limitations.

Species richness went from 10 to 12 on the
campground and 7 to 17 on control.

Yearly fluctuations of density on each
area are difficult to explain because of so
many determining factors . Not only local weather
but events on the wintering grounds can play an
important role in predicting a particular year's
breeding population. Attempts to explain avian
population fluctuations have so far led to only
ambiguous conclusions (Von Haartman 1971) . It
is pertinent to note, though, that over the
three-year period there was nearly twice the
range of densities on control as the campground.

Avian Diversity and Habitat Diversity

MacArthur (1964) found a correlation between
BSD and FHD in "tall forests of sycamores and
cottonwoods" in southeastern Arizona. The
relationship in this study between BSD and FHD
was almost identical to what he and others
found in earlier studies in eastern deciduous
forests (MacArthur and MacArthur 1961 , MacArthur
et al. 1962) . Austin (1970) , working in "desert
riparian" habitats in Nevada, plotted his data
against MacArthur 's (1964) regression line for
BSD vs. FHD and found similar results. Carothers
et al. (1974) found that in "desert riparian"
habitats immediately adjacent to areas of
relatively higher productivity but low avian
densities, the BSD and FHD correlation no
longer held. Yet, Carothers found that in
"desert riparian" habitats immediately adjacent
to areas of relatively the same productivity
and having a compliment avian community, the
BSD and FHD relationships did come close to
MacArthur 's regression line.

The BSD's and FHD ' s obtained in Oak Creek
are summarized in Table 4 and graphed in Figure
1. Although the points do cluster around
MacArthur 's line, there is enough deviation to
suggest other forces at work besides foliage
height diversity.

As in Carother's study plots, this is a
riparian system and MacArthur 's line fails to
take into account the added dimension of
permanent water. Also, human disturbance is
not considered. An additional downfall of
FHD is that there has been no stipulation by past
investigators when to measure FHD. As we see
here , BSD and FHD vary through time (or sampling
error). BS D was measured from the first signs
of breeding to the opening date and then from
that date to the end of breeding activity. On
the other hand, FHD was measured once before
and once after. It is possible that a day
could be found during the vegetative growing
season when the FHD would be such that BSD
for the entire period matched MacArthur 's line.
This leads me to question the value of FHD as
a predictor of BSD except in those specific
cases studied by MacArthur and the need for

17 9

Table 4. — Bird species diversity and habitat diversity.

CAMPGROUND

CONTROL

Bird Species
Diversity
(BSD)

Before After

Foliage Height
Diversity
(FHD)
Before After

Bird Species
Diversity
(BSD)

Before After

Foliage Height
Diversity
(FHD)
Before After

1973
1974
1975

2.19
2.62
2.19

1.95
2.42
2.34

.98
.96
.97

1.00
.98
1.01

2.08
2.08
1.83

2.34
2.43
2.71

1.04
1.06
.99

1.08
1.05
.97

BSD

3-|

1974

1975

Before

After

Camp

□

■

Control

•

Camp

Control

♦

Camp

Control

•

FHD

Figure l.--Bird species diversity (SD)
as a function of foliage height diver-
sity (FHD) before and after occupation
of the campground by campers. Regress-
ion line from MacArthur et al. 1966.

Table 5. — Individual bird weights and
individual existence energy.

Species

Weight in

Grams

Existence

Energy

Mourning Dove

137

Broad-tailed Hummingbird

.94

Red-shafted Flicker

125

Hairy Woodpecker

69.

Yellow-bellied Sapsucker

Cassin's Kingbird

.47

Black Phoebe

18.

.71

Western Flycatcher

12.

Western Wood Pewee

14.

Steller's Jay

105.

White-breasted Nuthatch

20.

Pygmy Nuthatch

10.

House Wren

10.

Robin

88.

Solitary Vireo

13.

Warbling Vireo

11.

Virginia's Warbler

Grace's Warbler

Painted Redstart

Red-faced Warbler

Bullock's Oriole

35.

Western Tanager

28.

Hepatic Tanager

40.

Summer Tanager

35.

Black-headed Grosbeak

46.

Lesser Goldfinch

Rufous-sided Towhee

38.

1 From Carothers et al. 1973, Marshall 1972,
and collections of the Museum of Northern Arizona.

180

specific time limitations when these parameters
are to be measured.

Examining BSD, we see that in 1973 and
1974 there was a decrease in diversity on the
campground after it was opened. An increase
occurred on control. In 1975, contol's diversity
again increased but so did the campground's.
The reason for this is, once again, the late
breeding season in 1975 (see previous section) .

Bioenergetics

In order to better understand the energetics
and organization of these avian communities, it
is important to look at standing crop biomess
(SCB) and existence energy (EE) of the birds
(Salt 1957, Karr 1968). The former is the
total weight (in grams) of the entire avian
community. In order to consider community
metabolism, a conversion is made that reflects
the difference in metabolism due to differences
in body weight. This is expressed as existence
energy (or Kcal) consumed by the total avian
community (see Carothers et al. 1974 for
limitations of this measure).

1973 . — The SCB of control decreased slightly
after 29 May, although density increased.
This is possible because the average weight
per individual bird decreased from 40.0 g to
31.7 g. Table 3 shows that several small-
bodied birds, Western Wood Pewees and Red-
faced Warblers, did move onto the area; see
Table 5 for weights. Two larger species,
Steller's Jays and Black-headed Grosbeaks,
moved off the area.

The campground had a drastic SCB decrease;
however, the average weight per individual
remained essentially constant (52.7 g to 54.2 g) .
The decrease can therefore only be attributed
to a general loss of birds of all sizes.

The initial and final differences between
the average weight per individual values on
the control and campground show that relative to
each other light-weight birds inhabited the
control and heavier birds inhabited the camp-
ground .

The existence energy values were initially
the same but after the opening the campground
EE showed a decrease of 37.3 percent.

We see then that before the intrusion of
campers the two areas differed in the average
weight per individual by 12.7 g but the EE was
the same . Following campground occupation ,
the average weight per bird became more
dissimilar (22.5 g) , and the total community
EE was nearly halved on the campground.

1974. — The SCB of control, once again, decreased
slightly after the opening, although density
increased. Again, this was due to a decrease

in the average weight per individual bird
(46.8 g to 2 3.4 g) caused by an influx of
smaller-bodied species (Table 3 and 4 ) .

The campground SCB decreased greatly, as in
1973, and the average weight per individual
remained fairly constant.

In 1974, larger-bodied birds occupied the
control initially, but this changed sharply after
the opening of the campground.

EE values on control changed upwardly 27.8
percent, whereas the campground's was decreased
by 22.3 percent.

Once again, the opening appears to be
detrimental to the birds in the campground.

1975 . — The overall trends remained the same in
1975. Smaller-bodied birds made up a majority of
the population on control. The increase in SCB
and EE on the campground, of course, was related
to the density increase that year.

SUMMARY AND MANAGEMENT ALTERNATIVES

After three breeding seasons, several
phenomena were discerned: 1) although bird
densities on the campground and control are
similar before the campground opening date, the
average weights of an individual bird is greater
on the campground (x~ = 48.5 g) than on control
(x = 38.2 g) and 2), population density and species
diversity (H') decrease when the campground is
occupied by people.

In other words, the presence of the camp-
ground produced a significant shift in the avian
community to heavier bodied birds relative to the
natural control area. This is probably a response
to the "opening" of the habitat during camp-
ground construction. Inhabitation of the camp-
ground by people causes a direct reduction in
the numbers and kinds of breeding birds.

Those in managerial posistions might
consider the following suggestions:

1 . Locations for new campgrounds should
be carefully scrutinized in terms of usage by
wildlife. In this specific case, riparian habi-
tats are very important to birds and of such a
limited extent in the Southwest that further
destruction of habitat needs to be discouraged.

2 . Existing campgrounds should be period-
ically closed to allow regeneration of vegetation
and reduce stress on resident wildlife. This is
being done in Oak Creek Canyon, but much too
often the reason behind the closure is financial
rather than ecological.

3. Opening the campground before or after
the height of the breeding season may be better
for the avifauna. If people are already present
when birds arrive to nest, the birds may be able
to find suitable habitat elsewhere instead of
"wasting energy" by attempting to breed and then
being disrupted during incubation. Of course,
not opening the campground until after breeding
season would be ideal for the birds but probably
very impractical for the campers.

4. Habitat manipulation should be carefully
controlled. This includes breaking off branches

181

for firewood, trenching for tents, running of
noisy equipment, and even clearing of snags,
slash, and brush by USFS crews.

5. Educational programs may be the only
effectual solution of human recreation and
wildlife problems. Government agencies have
had very good results in some public educational
canpaigns (e.g., Smokey the Bear). There is
no reason the general public could not be
exposed to broad ecological concepts such as
camping with less impact.

Finally, it is hoped that studies of this
type and education of the public will lead to a
happy medium between preserving native wildlife
and also allowing human enjoyment of an area.

LITERATURE CITED

American Ornithologist Union. 1957. Checklist
of North American Birds, Fifth Ed. Lord
Baltimore Press, Baltimore.

American Ornithologists Union, 32nd Supplement.
1973. Auk 90:411-419.

Austin, G.T. 1970. Breeding birds of desert
riparian habitat in southern Nevada.
Condor 72:431-436.

Bent, A.C. 1968. Life histories of North
American birds. U.S. Govt. Print. Off.,
Washington, D.C.

Carothers, S.W. and R.R. Johnson. 1975. Water
management practices and their effects on
nongame birds in range habitats. Proc. of
the Sym. on Management of Forest and Range
Habitat for Nongame Birds. U.S.D.A. Forest
Service General Tech. Report WO-1, July.

Carothers, S.W. , J.R. Haldeman, and R.P. Balda
1973. Breeding birds of the San Francisco
Mountain Area and the White Mountains,
Arizona. Museum of Northern Arizona Tech.
Ser. Bull. No. 12.

Carothers, S.W., R.R. Johnson, and S.W.

Aitchison. 1974. Population structure and
social organization of southwestern riparian
birds. Amer. Zool. 14:97-108.

Cottom, G. and J.T. Curtis. 1956. Use of
distance measurements in phytological
sampling. Ecology 37:451-460.

Farner, D.S. and A.C. Wilson. 1957. A quanti-
tative examination of testicular growth in the
White-crowned Sparrow. Biol. Bull. 113:254-267.

Immelmann, K. 1971. Ecological aspects of
periodic reproduction. In D.S. Farner and
J.R. King, eds . , Avian Biology, Vol. 1.
Academic Press, N.Y. and London.

Karr, J.R. 1968. Habitat and avian diversity
on strip-mined land in east-central Illinois.
Condor 70:348-357.

Karr, J.R. 1971. Structure of avian communities

in selected Panama and Illinois habitats.

Ecol. Monogr. 41:207-233.
Kendeigh, S.C. 1944. Measurement of bird

populations. Ecol. Monogr. 14:67-106.
Lack, D. 1950. The breeding seasons of

European birds. Ibis 92:288-316.
Lloyd, M. , J.H. Zar , and J.R. Karr. 1968.

On the calculation of information-theoretical

measures of diversity. Amer. Midi. Nat. 79:

257-272.

MacArthur, R.H. 1964. Environmental factors

affecting birds species diversity. Amer.

Nat. 98:387-397.
MacArthur, R.H. 1965. Patterns of species

diversity. Biol. Rev. 40:510-533.
MacArthur, R.H. and J.W. MacArthur. 1961.

On bird species diversity. Ecology 42:594-

598.

MacArthur, R.H., J.W. MacArthur, and J. Preer.
1962. On bird species diversity, II
Prediction of birds census from habitat meas-
urements. Amer. Nat. 96:167-174.

MacArthur, R.H., H. Recher, and M. Cody. 1966.
On the relation between habitat selection and
species diversity. Amer. Nat. 100:319-332.

Marshall, J. and R.P. Balda. 1972. The breed-
ing biology of the Painted Redstart. Unpubl .
manuscript, Northern Arizona Univ. , Flagstaff.

Morisita, M. 1959. Measuring of interspecific
association and similarity between communities.
Mem. Fac. Sci. Kyushu Univ., Ser. E., 3:65-80.

Odum, E.P. 1950. Bird populations of the

highlands (North Carolina) plateau in relation
to plant succession and avian invasion.
Ecology 31:587-605.

Pielou, E.C. 1966a. Species-diversity and

pattern diversity in the study of ecological
succession. J. Theoret. Biol. 10:370-383.

Peilou, E.C. 1966b. Shannon's formulas as a
measure of species diversity; its use and
misuse. Amer. Nat. 100:463-465.

Salt, G.W. 1957. An analysis of avifaunas in
the Teton Mountains and Jackson Hole, Wyoming.
Condor 50:373-393.

Shannon, C.E. and W. Weaver. 1963. The mathe-
matical theory of communication. Univ. of
111. Press, Urbana, 111.

Von Haartman, L. 1971. Population dynamics.
In D.S. Farner and J.R. King, eds.,
Avian Biology, Vol. 1. Academic Press,
N.Y. and London .

Whittaker, R.H. 1970. Communities and eco-
systems. MacMillan Co., N.Y.

Williams, A.B. 1936. The composition and

dynamics of a beech-maple climax community.
Ecol. Monogr. 6:317-408.

Winkworth, R.E. and D.W. Goodall. 1962. A
crosswire sighting tube for point quadrat
analysis. Ecology 43:342.

182

Population Fluctuations in
Nocturnal Rodents
in the Lower Colorado River Valley1

[ 2/
Bertin W. 1 Anderson, Jeff Drake, and Robert D. Ohmart—

Abstract. — An examination of population fluctuations in
a sample of over 10,000 rodents comprising five species
collected along the lower Colorado River revealed distinct
seasonal (annual) cycles in Perognathus penicillatus and
Dipodomys merriami . Overall rodent populations were
decreasing for the 3.5 year period for which data are
presented. This was most pronounced in Peromyscus eremicus.
Although these populations were declining, there was
significant intraspecif ic asynchrony among the populations
in different vegetation types. There was also a significant
degree of interspecific asynchrony in population fluctuations
which renders the task of evaluating habitat difficult and
subject to error unless carried out for several years in
various vegetation types.

INTRODUCTION

Populations are said to be cyclic when
they alternately irrupt and subside in a more
or less uniform manner between high and low
levels of density. These population fluctua-
tions sometimes follow a general pattern with
respect to time (annual, seasonal, monthly).
Cyclic events broken into finer detail follow
four fundamental phases: increase, peak,
decline, and low density. Several studies
have investigated cyclic events exhibited by
microtine rodents, which experience cycles of
three-to-four and nine-to-ten year intervals
as well as annual fluctuations (Dymond 1947,
Elton 1942, Keith 1963, Pearson 1966, Speirs
1939, Wing 1961), while McClosky (1972) has
studied temporal changes in densities and
diversities over short periods of time.

Designating fluctuations as cyclic implies
considerable regularity. However, the cause
of cyclic populations can usually be expected

1/ Contributed paper, Symposium on the
Importance, Preservation and Management of the
Riparian Habitat, July 9, 1977, Tucson, Arizona.
2/ Respectively, Faculty Research
Associate, Field Biologist, and Associate
Professor of Zoology, Arizona State University,
Dept. Zoology and Center for Environmental
Studies, Tempe, Arizona 85281.

to contain random components inextricably mixed
with any systematic ones. If the random
components outweigh the systematic ones or if
different substrates respond differently to
the systematic components, the occurrence of
noncyclic or asynchronous populations becomes
predominant .

The causal mechanism of cycles include
biotic as well as abiotic factors. Biotic
factors affecting animal cycles are those
inherent in the populations themselves and in
the interrelations of different species.
These include disease, predation, food, and
physiological mechanisms. Abiotic factors are
the physico-chemical element of the environment
such as organic compounds, moisture, winds,
solar radiation and others.

The purpose of this report is to examine
the annual population cycles of nocturnal
rodents in the lower Colorado River Valley.
Coupled with this examination are observations
dealing with general population trends over the
course of the study, as well as an evaluation
of fluctuations occurring in various community
types. The study was initiated in September
1973 and is on-going; data will be presented
through March 1977.

183

CLIMATE AND VEGETATION
Climate

The study area includes a major part of
the riparian vegetation located along the
Colorado River from the Mexican boundary north
to the Nevada border at Davis Dam. Climatic
data were supplied by the U.S. Bureau of
Reclamation and the Palo Verde Irrigation
District. Rainfall is highly irregular, with
annual amounts rarely exceeding 26 cm
(Table 1) . Drought conditions prevail through-
out most of the year. Over the 3.5 years of
this study, the highest monthly rainfall was

Table 1. — Rainfall on the lower Colorado River.

recorded in September 1976 (6.45 cm). The
highest annual rainfall (12.97 cm) was also
recorded in 1976. Although the area receives
little rain, the water table is high and
established vegetation with roots at least
3 m deep rarely shows ill effects. Annuals,
a staple food source for rodents, are very
dependent upon rainfall.

Temperatures in the desert are highly
variable and show a wide range between daily
and annual extremes (Table 2) . The highest
official temperature was 50.0°C in July 1973;
the lowest was -5°C in January 1973, although
we recorded a temperature of -14°C near

Precipitation (cm)

Month

1972

1973

1974

1975

1976

1977

Twenty-year
average

Jan.

0.00

0.12

Feb.

0.00

3.07

Mar .

0.00

1.85

Apr .

0.00

May

0.00

0.12

June

1.75

0.00

July

0.00

Aug.

2.18

5.84

Sept .

0.00

Oct.

6.63

0.00

Nov.

1.55

0.28

Dec.

0.10

0.00

Total

12.2

11.3

Table 2. — Temperatures for the lower Colorado River.

Mean High and Low Temperatures (°C)

Month

1973

1974

1975

1976

1977

Twenty-year
average

Jan.

19.8

Feb.

25.4

Mar .

23.6

Apr.

31.9

May

June

July

Aug.

Sept.

Oct.

Nov .

Dec.

184

Winter Summer Winter Summer Winter Summer Winter
1974 1974 1975 1975 1976 1976 1977

Figure 1. — Densities (N/270tn) of Perognathus penicillatus and Dipodomys merriami along the
lower Colorado River Valley.

Winter Summer Winter Summer Winter Summer Winter

1974 1974 1975 1975 1976 1976 1977

Figure 2. — Densities (N/270tn) of Peromyscus maniculatus and Neotoma albigula along the lower
Colorado River Valley.

185

Needles, California in January 1975.

Vegetation

In this study we recognized six basic
community types in the riparian vegetation.
Data from four of these (cottonwood-willow,
Populus f remontii-Salix gooddingii; salt cedar,
Tamar ix chinensis ; honey mesquite, Prosopis
julif lora; and screwbean mesquite, P_. pubescens)
are presented in this report. The criteria
used in defining these communities are included
elsewhere in these proceedings (Anderson,
Engel-Wilson, Wells, and Ohmart) and will not
be repeated here.

METHODS AND MATERIALS

Within the study area approximately 100
transects (the actual number varied from year
to year) were established throughout the various
community types. Rodent populations were
sampled in transected areas with a snap trap
grid consisting of two parallel lines 15 m
apart. Each line included 15 stations that
were each 15 m apart. At each station two
museum special traps and one Victor rat trap
were set and baited with rolled oats and peanut
butter treated with a chemical to repel insects
(Anderson and Ohmart, 1977). Trap lines were
run for three consecutive nights. Catches
are expressed as numbers per 270 trap nights.
(For further details see Anderson, Engel-Wilson,
Wells and Ohmart, these proceedings.)

All community types were sampled on an
approximately equal basis throughout the study.
Trap nights for the 3.5 year period totaled
150,930. The number of grids per season
ranged from 44 (11 per community type) to 116
(29 per community type) . For this study we
recognized two seasons: summer, April through
October, and winter, November through March.

The species which were most frequently
caught and which were analyzed for this report
were the cactus mouse (Peromyscus eremicus) ,
deer mouse (P. maniculatus) , desert pocket mouse
(Perognathus penicillatus) , white-throated
woodrat (Neotoma albigula) , and Merriam's
kangaroo rat (Dipodomys merr iami) . During the
study we caught 6,178 Peromyscus eremicus ,
829 _P. maniculatus , 881 Neotoma albigula ,
1,736 Perognathus penicillatus , and 1,317
D. merriami . Other species such as the grass-
hopper mouse (Onychomys torridus) , desert
kangaroo rat (Dipodomys deserti) and several
others were caught regularly but numbers were
too small for analysis.

ANNUAL CYCLES AND POPULATION TRENDS

The presence of annual population cycles
for the five numerically dominant rodent species !
was investigated by capture data, within each
season, and from the four sampled community
types. In the combined analysis we are consid-
ering seasonal population fluctuations within
the total riparian habitat. The presence of
an annual cycle should be evident if there are
regular and predictable population fluctuations
between seasons. Analysis of annual cycles
should reveal general population trends, but
obscure intercommunity trends. Therefore,
overall trends are analyzed at the community
level as well as the species level.

Perognathus penicillatus and Dipodomys
merriami exhibited regular annual cycles
(fig . 1) . Peak population levels were reached
in summer and low levels in the winter. The
pattern was more developed in P_. penicillatus
than in D_. merriami or Neotoma albigula.
N^. albigula maintained a nearly constant
population level from winter 1974 to summer
197 4 at which time the population began to
fluctuate in a manner similar to that shown
by P. penicillatus and I), merriami .
The seasonal fluctuations shown by N. albigula
and Peromyscus maniculatus (fig. 2) are not
readily interpreted. JP. maniculatus showed
high populations in winter and low populations
for summer 1974 through summer 1975, but in
winter 1976 the population declined precipitously.

Peromyscus eremicus demonstrated a complete
lack of any annual cycle (fig. 3). The pattern
was almost a linear progression of declining
population for nearly the entire study period
with only a slight increase in winter 1977.
This overall population decline was found with
some degree of consistency in each species
examined. In all cases 1976 winter populations
were lower than 1974 winter populations, but
most 1977 winter populations showed a slight
increase. In Perognathus penicillatus and
Neotoma albigula the summer populations also
showed a gradual decline. Summer populations
of Peromyscus maniculatus increased substantially
from 1974 to 1975 whereas summer populations
of Dipodomys merriami increased only slightly.

The general decline in rodent numbers
appears to be related to precipitation. With
the exception of September 1975, dry conditions
prevailed from September 1973 through January
1976 (Table 1), but in 1972 and in spring 1973
wet conditions existed long enough for seed-
producing annuals to flower abundantly (pers.
obs., R. D. Ohmart). Seed production continued
for a while after a return to dry conditions
and species that cache seeds can probably
survive on these stores for some time. There

186

is a lag in the effect of changing conditions
on the population densities. One might have
expected populations to have peaked in 1974,
following the second breeding season after
the favorable wet conditions of 1973. This
is seen to be the case in three of the five
species (figs. 1-3). Subsequent dry conditions
resulted in almost no flowering of annuals in
1974 and 1975. One might have expected popu-
lations to be lower in 1975 and 1976, following
the second breeding season after the onset of
poor conditions. There was some flowering in
spring 1976 and rather profuse flowering in

winter and spring 1977. These favorable
conditions would be expected to lead to an
increase in rodent populations beginning with
the onset of reproduction. However, another
factor must be considered. The precipitation
in September 1976 probably was responsible for
an initial decrease in rodent populations, be-
cause the relatively heavy rains caused extensive
flooding in much of the prime rodent habitat
in honey mesquite. Standing water, a few cm
to a meter deep, remained in many places from
a few days to two weeks or more and many addi-
tional days were required for the mud to dry.

187

During the rains there was running water which,
in many areas, carried away most of the screw-
bean and honey mesquite pod crop. Trapping in
these areas yielded few rodents through May 1977

Low temperatures may also have negative
effects on rodent populations . The spring of
1975 was unusually cool with locally occurring
freezing temperatures in March, April, and
early May. These frosts may have been respon-
sible for an almost total lack of mesquite pod
production the following fall.

In summary, decreasing rodent populations
from 1974 through early 1977 were probably the
result of (1) dry conditions in which annual
plants were unable to reproduce, (2) freezing
spring temperatures in 1975 which may have
affected mesquite seed productivity and
(3) flooding of prime habitat in September
and October 1976.

RODENT POPULATION TRENDS IN
DIFFERENT PLANT COMMUNITIES

One of our purposes was to study the extent
of synchrony in fluctuations between species
both within and between community types. To
do this we studied population trends of the
five most abundant rodents in each community
separately. Considerable asynchrony in popu-
lation fluctuations between species at a
locality would introduce additional difficulties
in evaluating the rodent diversities and
densities of different habitat types. If these
interspecific fluctuations in population size
also were asynchronous between localities,
evaluation of habitats would be further compli-
cated. In this section we explore these points
and offer suggestions for the collection of
useful data.

Intraspecif ic Fluctuations between
Community Types

Fluctuations in populations of Perognathus
penicillatus in the four community types studied
were highly synchronous (fig. 4) . Dipodomys
merriami populations were synchronous in screw-
bean (SM) and honey mesquite (HM) and in salt
cedar (SC) and cottonwood-willow (CW) , but
these two sets were asynchronous with respect
to each other (fig. 5). Peromyscus maniculatus
showed moderately asynchronous fluctuations
(fig. 6); populations were increasing in
cottonwood-willow at a time when they remained
about the same or increased slightly in other
community types. Neotoma albigula population
fluctuations tertded to be quite asynchronous
(fig. 7). Peak populations were reached at a
different time in all community types.

Peromyscus eremicus populations also showed a
marked degree of asynchrony (fig. 8) . We also
noted some evidence of reproduction in October
and November of 1976 (juveniles, scrotal males,
females with enlarged mammae) in screwbean
mesquite. This may be the first indication
that the overall population will be increasing
again .

Interspecific Fluctuations between
Community Types

Interspecific variations in population
were not of the same magnitude for all species,
i.e. peaks may not be as high nor valleys as
low in some species as in others. In addition
highs and lows do not occur at the same time
in all species. Finally, there were inter-
community differences. McClosky (1972) has
shown that there were significant changes in
densities and diversities in a relatively brief
period of time within a locality. Our data
tend to support his findings, but since we did
not trap continuously in the same area over a
long period of time, our data are not well-
suited for analysis on a local level.

Because changes in densities of different
species occur at different rates from season
to season, species diversities also change
from season to season (fig. 9). From summer
1974 to summer 1975 the diversity in cottonwood-
willow nearly doubled, due primarily to a re-
duction in numbers of Peromyscus eremicus .
In general, the greatest diversities were reached
in the warm months when Perognathus penicillatus
was present in the greatest numbers. Ranked
from greatest to least diversity the communities
would be ordered SM-HM-SC-CW, SM-CW-SC-HM, and
SM-HM-SC-CW in the three summers, respectively.
In the four winters the order would be SM-HM-
SC-CW, HM-SM-SC-CW, SM-HM-CW-SC, and SM-SC-CW-HM,
respectively. Had trapping only been conducted
in 1974, the results would have indicated that
rodent densities were greater and diversities
lower than they are on the average, whereas
1976 trap results would have indicated lower
populations and greater diversities than the
average of all years. Obviously, for an accurate
assessment of rodent population densities and
diversities one should trap extensively in
several community types over several years.
There are, however, a number of important con-
clusions that could have been drawn had a con-
certed trapping effort been made for a relatively
short time, e.g., three months in the summer and
three months in the winter.

First, trapping at almost any time would
reveal Peromyscus eremicus as the most abundant
species and Perognathus penicillatus as having
greater densities in the summer. Second,

188

Winter Summer Winter Summer Winter Summer Winter
1974 1974 1975 1975 1976 1976 1977

Figure 4. — Densities (N/270tn) of Perognathus penicillatus in four community types in the lower
Colorado River Valley.

Winter Summer Winter Summer Winter Summer Winter
1974 1974 1975 1975 1976 1976 1977

Figure 5. — Densities (N/270tn) of Dipodomys merriami in four community types in the lower
Colorado River Valley.

diversities would be greatest in the summer.
Caution should be exercised in determining
the relative value of the various community
types to rodents in terms of densities and
diversities. Similarly, caution should be
used in identifying community preferences of
each species, but even with limited trapping

Perognathus penicillatus and Neotoma albigula
should be most numerous in screwbean mesquite,
Dipodomys merriami and I\ penicillatus should
be scarcest in cottonwood-willow and Peromyscus
maniculatus and _P. eremicus should be scarcest
in honey mesquite.

189

5-1

£ 3-

Winter Summer Winter Summer ' Winter ^Summer Winter
1974 1974 1975 1975 1976 1976 1977

Figure 6. — Densities (N/270tn) of Peromyscus maniculatus in four community types in the lower
Colorado River Valley.

£ 4H
'E

Q.
(0

1 «

SBM
HM

Winter
1974

Summer
1974

Winter
1975

Summer
1975

Winter
1976

Summer
1976

Winter
1977

Figure 7. — Densities (N/270tn) of Neotoma albigula in four community types in the lower Colorado
River Valley.

CONCLUSIONS

Perognathus penicillatus and Dipodomys
merriami displayed distinct annual cycles
which were synchronous with respect to each
other and between the various community types.
Longer term fluctuations displayed a high degree
of asynchrony between community types within a
given species as well as between species. The
major factors which seem to have caused a steady
decline in populations since 1974 are lack of
rainfall up to September 1976, flooding in
September 1976, and unusually cold temperatures
in the spring of 1974. Because abiotic factors
such as these occur in a more or less random
pattern it seems possible that rodent populations
in the lower Colorado River Valley do not display
regular cycles although we have not trapped
over a long enough period of time to draw a
definite conclusion on this. The marked

asynchrony in population fluctuations between
community types within a given species and
between species makes evaluation of habitat
in terms of densities and diversities very
difficult unless trapping is done over a period
of two or three years in several community types.

ACKNOWLEDGMENTS

190

32-
31-

Winter ' Summer ' Winter ' Summer ' Winter ' Summer ' Winter
1974 1974 1975 1975 1976 1976 1977

gure 8. — Densities (N/270tn) of Peromyscus eremicus in four community types in the lower
Colorado River Valley.

191

1.5-

'74 '74 '75 '75 '76 '76 '77

We thank Jane Durham, Jake Rice, James Bays
and Kathleen Conine for critically reading
the manuscript. The research was funded
through grant number 14-06-300-2415 from the
U.S. Bureau of Reclamation.

LITERATURE CITED

Anderson, B. W. and R. D. Ohmart.

1977. Rodent bait additive which repels
insects. J. Mamm. 58:242.
Dymond, J. R.

1947. Fluctuations in animal populations
with species reference to those of Canada.
Trans. Royal Soc. Canada 41(5):l-34.
Elton, C.

1942. Voles, mice and lemmings. Clarendon
Press, Oxford.
Keith, L. B.

1963. Wildlife's ten-year cycle. Univ.
Wisconsin Press, Madison.
McClosky, R. T.

1972. Temporal changes in populations and
species diversity in a California rodent
community. J. Mamm. 53:657-676.
Pearson, 0. P.

1966. The prey of carnivores during one
cycle of mouse abundance. J. Anim. Ecol.
35:217-233.
Speirs, J. M.

1939. Fluctuations in numbers of birds in
the Toronto region. Auk 56:411-419.
Wing, L. W.

1961. The 3.864 year cycle and latitudinal
passage in temperature. J. Cycle Res.
10(2) :59-70.

Figure 9. — Rodent species diversities in four
community types in the lower Colorado River
Valley. The diversities are calculated from
the average densities (N/270tn) of Peromyscus
eremicus, P_. maniculatus , Perognathus
penicillatus, Dipodomys merriami, and
Neotoma albigula caught in a given community
type for a given time of year.

192

Climatological and Physical

Characteristics Affecting
Avian Population Estimates in
Southwestern Riparian Communities
Using Transect Counts1

\ 2/
Bertin W. Anderson and Robert D. Ohmart—

Abstract. — Comparative data from about 10,000 censuses
of line transects on the lower Colorado River show that strong
winds (20 to 50 kmph) may reduce censusing accuracy but winds
below 20 kmph appear not to strongly influence avian
estimates. In winter, optimum censusing time is from 1 hour
after sunrise to 2.5 hours after sunrise, whereas in summer
the optimum period is 0.25 hours before sunrise to 1 hour
after sunup. Consecutive censuses from the same area by
highly experienced observers (>5 years of birding) are more
consistent than less experienced personnel (10 to 16 months
of birding) .

Each transect should be censused at least twice monthly
for minimum best density estimates and three times for
maximum best density estimates and three times for minimum
best avian species diversity estimates. Four censuses
reveal a greater number of species than with two or three
censuses. Number of transects needed for a minimum
adequate sample of 3,200 ha of mature honey mesquite
(Prosopis julif lora) habitat is four (sampled three times
monthly) but for more precise population data for each
species six to nine are required.

INTRODUCTION

Since 1973 we have been censusing birds
in about 80,000 ha of riparian vegetation in
the Colorado River Valley from Davis Dam to
the Mexican boundary using the transect count
technique (Emlen 1971) . Censusing is conducted
by 11 full-time observers; the turnover rate
among observers being about 25 percent per
year. Since the beginning of our study, we
have made about 10,000 censuses each 0.8 to
1.6 km in length. With the increased need to
determine avian population levels, it is
important to quantify the effects of such
factors as time of day and wind on censuses.
Furthermore, if more than one observer is

1/ Contributed paper, Symposium on the
Importance, Preservation and Management of the
Riparian Habitat, July 9, 1977, Tucson, Arizona.

2/ Respectively, Faculty Research
Associate and Associate Professor of Zoology,
Arizona State University, Dept. Zoology and
Center for Environmental Studies, Tempe,
Arizona 85281.

involved, it is important to know how results
from the same area compare. Since the inter-
related factors of available time, financial
support, and manpower limit the extent of
censusing, it is important to know the minimum
area in which censusing must be done in order
to get reasonably accurate results. Under-
sampling can lead to erroneous conclusions,
while over-sampling is inefficient. This
report, although not providing definitive
answers, addresses these points and may be
helpful to those planning bird censuses in
large areas, especially riparian communities
in the Southwest.

CLIMATIC FACTORS AFFECTING CENSUSES
Wind

Intuitively it seems that strong wind
(20 to 50 kmph) would decrease bird activity
and in general interfere with bird detection.
In order to quantify this we censused 22
transects on days with winds 24 to 50 kmph
and compared the number of detections with the

193

number made on the same transects on calm days
immediately preceding or following windy days.
The results (Table 1) are equivocal in that in
three of five comparisons involving twelve
transects, the number of detections on windy
versus calm days were similar; on two occasions
involving ten transects, the counts on windy
and calm days were dissimilar with significantly
greater (p<0.01) numbers of detections having
been made on calm days. Visual and auditory
detections will probably be reduced because
birds tend to seek shelter and sounds may be
drowned out by wind noises.

be initiated earlier on days which are relatively
mild .

In order to census all of our transects
monthly, we frequently must begin before and
continue beyond the time when birds are most
active. To compensate for this we rotate the
order in which a set of transects is censused.
Shields (1977) indicated that this may be
inadequate and significant differences in
densities may be obscured. He provided a
method for including time of day as a factor
in the analysis of variance.

Table 1.-

-Number of

bird

detections on windy

(20-50

kmph) and

calm days.

Number

Detections

Probability of

Difference Being

Transects

Windy

Calm

Due to Chance

>0.05

203

<0.005

132

136

>0.05

258

259

>0.05

196

255

<0.01

DIFFERENCES BETWEEN OBSERVERS

The relatively large area being studied
has dictated the use of 8 to 12 observers.
The experience of these individuals is quite
varied. Highly experienced observers were
those who had been observing birds for several
years — sometimes for most of their lives.
Others, which we shall refer to as less experi-
enced observers, had not done much bird watching
before joining our staff. On the average, the
less experienced observers, after training, had
been on the job for about ten months.

In general, if there is a strong wind
during the prime census time, we do not begin
censusing; but if the winds develop after
censusing has begun, we ordinarily finish the
transect or set of transects but do not initiate
censusing new ones. Winds less than 20 kmph
apparently interfere very little with accurate
censusing .

Time of Day

Winter censuses (December through January)
initiated earlier than one hour after sunrise
or later than 2.5 hours after sunrise resulted
in significantly (p<0.05) fewer detections than
those begun 1.0 to 1.5 hours after sunrise.
Summer censuses begun between 0.25 hours before
sunrise to sunup were significantly (p<0.05)
greater than censuses begun later than 0.75
hours after sunup (Table 2) .

Birds are active earlier and for shorter
periods of time in summer than in winter. The
often-expressed statement that birds are more
active earlier than later at all times of the
year and are active for longer periods of time
with increasing photoperiod is not supported
by data from the lower Colorado River Valley
and appears to be an over-generalization.
Time of day is, of course, related to temper-
ature (Shields 1977) . In summer the censusing
can be continued longer on days which are
relatively cool, and in winter censusing can

To determine the extent to which differences
in avian densities and diversities vary because
of the way observers detect birds, we first
evaluated the extent to which observers could
reproduce their own results. To make this
evaluation three different observers censused
a set of transects (4 to 5 km) for four
consecutive days. Following this we had five
groups of four observers and one group of three
observers census about 16 km of transects.
Within each 16 km group of transects each
observer censused one fourth of the area each
day so that by the end of four days each had
censused the entire area.

The first part of the test was conducted
in April 1977; a difficult time to try to
reproduce census results because of the potential
for encountering transient individuals of newly
arriving summer residents (strictly transient
species were not included in the census results) .
For the three different highly experienced
observers, nonsignificant differences in bird
density were obtained in 12 of 18 possible
comparisons (Table 3) . The largest and smallest
number of species tabulated differed by four in
area 2 and by one in the other areas (Table 3) .
The largest BSD was 8.1, 9.2, and 11.9 percent
above the lowest value in the three areas
(Table 3). Thus, the same observer is not
always able to reproduce his own results,
suggesting day-to-day variation within a given
population in an area.

194

Table 2. — Results of bird censuses taken at different times of the morning in winter and summer in
riparian vegetation in the lower Colorado River Valley.

Totals

Censusing begun Excluding Passerines

(hours after Number Phaino- Phaino- Non- Excluding Number

sunrise) Transects Detections peplas peplas Passerines Phainopeplas Species

Dec. 1974

1.00

175

1.25

191

33*

Z. . uu

ft1;

1 AO*

O J

Dec. 1974

1.50

234

149

Z . jU

1 97
S.L /

Jan. 1975

1.00

194

1 . DU

119

1 O C

1 JD*

111*

i ^n
1 . jU

1 7Q

1 oi
1 Jl

1U1

1 C

z . UU

loo

11^:

llo

1 o

Feb. 1975

1.50

113

1.75

121

July 1974

-0.25

449

292

157

1.00

308*

199*

92*

July 1975

-0.25

294

166

128

1.00

187*

85*

102

June 1975

0.00

328

183

145

0.75

242*

137*

105*

*Signif icantly different from earlier count at p<0.01.

Table 3. — Density estimates for three different areas
with about 3.2 km of transects each. Each area was
censused by the same observer for four consecutive days.

Area

Density:

Day

396

271**

455

411

306**

449

405

374

365***

352*

365

421

Species :

Day

BSD: Day

2.502

2.616

2.475

2.405

2.801

2.302

2.480

2.648

2.534

2.205

2.574

2.399

*Signif icantly smaller than on days 2 and 3 (p<0.05)
**Signif icantly smaller than on days 3 and 4 (p<0.005).
***Signif icantly smaller than on other days (p<0.05).

195

When different observers censused the
same transects, significantly different (p<0.05)
densities were obtained in 18 (60 percent) of
30 possible comparisons between observers of
the same transect (Table 4) . Test 1 involved
two highly experienced and one less experienced
observer. The main source of difference was
a flock of Chipping Sparrows (Spizella passer ina)
recorded by one of the highly experienced
people but not by subsequent recorders. Test 2
involved four highly experienced recorders.
The main difference was in density estimates
of White-crowned Sparrows (Zonotrichia
leucophrys) , a flocking species. Tests 3 and
6 involved three less experienced and only one
highly experienced observer and the results
varied widely. Results of test 4, involving
two highly experienced and two less experienced
observers, were fairly similar. Test 5,
conducted in desert washes by three highly
experienced observers, revealed wide differences
in population estimates; but all recorded the
same number of species. Significant differences
in tests involving two or more highly experi-

enced observers were found in 8 (44 percent)
of 18 comparisons. This is significantly
fewer than expected based on the results of
tests where at least three of the observers
were not highly experienced (test 3).

Although some of the differences between
observers are significant, they represent
population estimates based on a single census.
As we shall demonstrate in the following section,
differences between observers may be due to
this factor more than as a result of different
abilities to detect birds. Furthermore, we
have already pointed out that the same observer
censusing the same area for several consecutive
days was not always able to reproduce his own
results. With this in mind, the real differ-
ences obtained by highly experienced observers
in the same area appear quite small, with
larger differences resulting when the same
area is censused by observers with less experi-
ence. The conclusion to be drawn from this is
obvious, but in practice it is not always

Table 4. — Six sets of census data obtained by different observers in the same area in the
Colorado River Valley.

Date/

Test

Vegetation Type/

Total

No.

Area Length

Observer

Density

BSD

Species

21-23 Mar 1975/

259

2.66

Honey Mesquite/

209

2.69

8.9 km

216

2.73

6-9 Nov 1976/

291

2.42

Honey Mesquite/

364

2.41

11.3 km

336

2.67

305

2.71

6-9 Nov 1976/

163

2.39

Salt Cedar, Arrowweed,

1.97

Screwbean, Willow mix/

229

2.36

8.0 km

127

2.51

6-9 Feb 1977/

255

2.56

Desert Washes/

180

2.38

9.7 km

129

2.70

6-9 Feb 1977/

243

2.29

Honey Mesquite/

286

2.34

11.3 km

248

2.31

213

2.51

6-9 Feb 1977/

2.59

Salt Cedar, Arrowweed,

2.26

Screwbean, Willow mix/

2.13

8.0 km

2.46

*highly experienced observer

196

possible to find highly experienced observers
when they are needed.

To minimize the possibility that differ-
ences between two areas are due to the differ-
ences among observers, we schedule censuses
so that no individual observer censuses the
same transect twice in one month. If this is
not possible, only highly experienced personnel
census the same area twice.

NUMBER OF CENSUSES REQUIRED

The number of times a transect 0.8 to 1.6
km long must be censused monthly to obtain an
accurate population estimate is of crucial
importance both for scientific accuracy and to
be as economically practical as possible. To
acquire insight into this question we censused
single transects, groups of three transects
(about 3.6 km), and groups of eleven transects
(about 13 km) on four consecutive days. The
results of the first day are the extrapolated
detections obtained on the first day. The
totals for the second day represent the mean
of the first two days and so on. Species
occurring at densities less than 0.5 per 40 ha
were dropped from the analysis, thus the number
of species sometimes decreased with the addition
of a second or third census. Rationale for
this procedure is presented elsewhere in these
proceedings (Anderson, Engel-Wilson, Wells and
Ohmart) .

Data for 55 single transects indicate that
after the first census there was very little
change in the estimated density (Table 5) .
BSD increased steadily through the fourth
census of an area, but the increase (4 percent)
with addition of a fourth census was small
(Table 5) . The number of species increased
through the fourth census. Sixteen groups of
three transects showed little change in esti-
mated density after the first census; the
inclusion of a third and fourth census resulted
in very small changes in BSD and number of
species (Table 5) . Four groups of eleven
transects were similarly censused. Once again
a third and fourth census of the same area
modified the results very little (Table 5) .
We conclude that if a community type is repre-
sented by only one or two transects, censusing
should be done at least four times per month.
If several transects have been established,
two or three censuses appear to yield data
as accurate as four censuses. It might be
possible to census each area once or twice and
to adjust the results by correction factors
determined from a larger number of censuses.
Such correction factors should be based on a
larger data set than we have presented in this
report. We have sufficient data to make such

an assessment; but for the sake of brevity and
because of time constraints, it is not presented
here .

NUMBER OF TRANSECTS TO ESTABLISH IN AN AREA

Once it has been determined that there
are, for example, 3,200 ha of honey mesquite,
the next step is to determine how many transects
are needed to obtain an accurate estimate of the
population of each species in that area. We
established nine transects (11 km) in 3,200 ha
of structure type IV honey mesquite. To deter-
mine the number of transects necessary to
accurately estimate avian populations, a single
transect was selected at random, the density
and BSD were computed and number of species
counted. A second transect was added and new
parameters calculated; the new density being
the mean of the two transects. From this a
new BSD was computed. This randomized process
was repeated until all nine transects had been
included. If the area was not adequately
sampled, the density, BSD, and number of species
should have increased or decreased through the
ninth census. If a total of nine transects was
more than adequate, these parameters should
reach a plateau before the addition of the ninth
transect. The point at which they level off
indicates the number of transects necessary to
adequately sample the area. Total density
appeared to level off after inclusion of six
transects (fig. 1) and BSD and the number of
species after the random combination of four
transects (fig. 2) . On this basis we concluded
that four transects rather than nine would have
been sufficient in this 3,200 ha area. Three
censuses for three transects are apparently
as informative as four censuses (see above).
For this reason we used data after three
censuses from four transects selected randomly
from the original nine transects to see if that
number was adequate for reproducing the results
obtained from all nine transects. Four observers
censused the nine transects with each observer
having censused all nine once each. This was
done in November 1976 and was repeated in
February and April 1977.

In November three censuses of four randomly
selected transects revealed a density about 19
percent higher and a BSD 2.8 percent lower than
that obtained when all nine transects were
censused four times each. The two lists con-
tained the same number of species: 24 (89
percent) species were common to both lists;
three species occurred on one list but not on
the other. These constituted a very small
percent of the total density on either list
(2.8 percent and 0.7 percent, respectively).
Twelve species (44 percent) varied within a
range of one individual on the two lists (not

197

Table 5. — Changes in estimates of bird population parameters with four censuses of single transects,
groups of three transects, and groups of 11 transects.

Number Number of Mean
of Transects Number of

Groups Per Group Censuses Density (2SE) * BSD (2SE) Number of Species (2SE)

249

(42)

14)

243

(40)

12)

240

(19)

11)

239

(19)

09)

222

(68)

17)

247

(62)

14)

246

(60)

11)

242

(60)

10)

189

196

195

200

*Not calculated for groups less than 5.

500-

480-

460-

440-

-C
O

420-

400-

3
Z

380-

360-

340-

320-

1 23456789

Number of censuses

Figure 1. — Changes in avian density estimates
with the random additions of transects in
3200 ha of honey mesquite woodland along the
lower Colorado River.

counting those cases of a zero (0) on one list
and a one (1) on the other) . Among the species
occurring in densities of 10 or more on at
least one list, the average of the four randomly
selected transects was about 50 percent higher
or lower than when all nine transects were
included. Three of the five species varying
by over 40 percent were flocking species.

In February three censuses of four randomly
selected transects revealed a density 11.2
percent higher and a BSD 4.9 percent lower

1 23456789

Number of censuses

Figure 2. — Changes in the estimates" of bird
species diversity and number of species with
the random addition of transects in 3200 ha
of honey mesquite woodland along the lower
Colorado River.

than that obtained on all nine transects
(Table 6). The two lists contained 28 species,
26 of which were common to both lists. Seven-
teen species (63 percent) varied by only one
individual. Those species occurring in densities
of ten or more on at least one list varied in
density by an average of 31 percent on the
randomly selected transects. Both species
varying by over 40 percent were flocking species.

In April three censuses of four randomly
selected transects revealed a density 5.5
percent smaller and a BSD 2.6 percent larger
than that obtained on all nine transects
(Table 6) . Both lists contained 31 species
of which 29 (94 percent) were in common.
Twelve species (39 percent) varied by only

198

Table 6. — Bird densities and diversities in honey mesquite for November, 1976 and February and
April, 1977 in the lower Colorado River Valley.

Density (N/40 ha)

November February April

Species

4-)

I— 1

4-1

•o

4-1

4-)

-a-

t-i

4-1

udlUUC 1 o y Ud 1 ± , Lj U \J l L U L L y A ^dlllUClll

Mm i T"n "i n o* Hnup 7pnai Ha marTnnrs

t 1U U L L 1 1 1 1 UU VC ) 4t_.Cl.lClXvJ.Cl Utdl_J.lJU.JLCl

Road runner , Geo co ccyx cal i f ornianus

! !

bUlllillU LI 1 1 J.LI\.CL a b U 1 a U L C O a U J. d L U

P. i 1 a T*J rtor1n£3ir'Lr£i"r" Mo 1 anpfnpc a 1 1 1" ~f f rnn c

Ijlla WUUUUcLtXCL ^ LlCXaLlCL UCO alii 1L LUllo

J-idLl UCL U cl V— tvCU nUUUUCL rvt. L , I ILUluco ol_o.icll.-LO

Yd 1 1 ni.i— V\ p 1 1 i p H Qancnrlf pi* Qnhuyani nic uavi nc
ICllUW UclXlcU JdpbULtVCl j I dpiLUb vdi. lub

AcVi— 'f"ViT*oat"0/H T-T" H \r a f~ ("■• Vi o t~ M\7 T a t* i*"- Vi 1 1 c PinPTacf pne
t\s 1 1 LULUdLcU r lyLdLLllcL , 1 ly -L cl JL *_ llUo Llllci doLCUb

^ a\7 ' c Phnphp ^auprm' c cava
JdV o jl 11UCU C , Od y U L lli.b bdyd

T-1 1 a r* If Phnprip Qav/rrrn' c ni oti ranc

D J. d L. r>. jl 11UCU C , OdyUl Lllb LllgL ILdLlb

Rough- winged Swallow , Stelgidopteryx ruficollis

PI "i t T ^ua 1 1 nu PptrnrVipl iHnri nvri*hnnnfa

Llll 1 OWdllUW j rci.l.ULllc±lUUU iJy L L UUllU Ld

VptH i n Ann' naniQ f 1 a\n' pphq

VC1U1L1 j nUL 1 pd J. Ub J. XdV ±LC|Jb

Hnncp TaIt"p>ti Trnol r\ r\ \7 1~ & c a pn nn
nuubc i\ l t: 1 1 . i iujjjiuuy Lea acuuii

Bewick's Wren, Thryomanes bewiclcii

Cactus Wren, Canipylorhynchus brunneicapillus

Mockingbird, Mimus polyglottos

fn' QQ3 1 ThraQhpT* Tnvn q f~ nina HnrQfl 1 p

ViL ISOal 1 LI L dOllCl , 1 UAUo LUIIItl UU1 OulC

Attipt* i ran RnhiTi TiitHiiq ttii oyafnn'nc

milC L 1 Lu 11 111 j i U L U Uo UlJ-gLdLUL -LUO

Western Bluebird, Sialia mexicana

Mountain Bluebird, Sialia currucoides

T4, 1 1 1 0 — CTT*a\7 ('^T"iaf~r':af~r,VioT" P/il l nnf l 1 a 0307*111 0 a
□ J.UC gLciy Cjrilct L C- d L t_. 1 1 c L , CUliUpLlld LacI Ulcd

R 1 3 Plf — fai 1 fZT"iat~r*ait"i~*V"ioT* Pnl 1 nnf i 1 a mol anni"a
U-LdUtv Lul 1LU UUaLLd CL11C1 , rUllUp Llla 1 Itr 1 d 1 1 LI L <.l

1 1

X -L

Ruby- crowned Kinglet, Regulus calendula

Phainopepla, Phainopepla nitens

Loggerhead Shrike, Lanius ludovicianus

Orange- crowned Warbler, Vermivora celata

Lucy's Warbler, Vermivora luciae

Yellow-rumped Warbler, Dendroica coronata

Brown-headed Cowbird, Molothrus ater

Northern Oriole, Icterus galbula

Painted Bunting, Passerina ciris

House Finch, Carpodacus mexicanus

Lesser Goldfinch, Carduelis psaltria

Abert's Towhee, Pipilo aberti

Savannah Sparrow, Passerculus sandwichensis

Sage Sparrow, Amphispiza belli

Dark-eyed Junco, Junco hyemalis

Brewer's Sparrow, Spizella breweri

White-crowned Sparrow, Zonotrichia leucophrys

Lincoln's Sparrow, Melospiza lincolnii

Song Sparrow, Melospiza melodia

Total

327

389

249

277

381

360

BSD

2.54

2.47

2.43

2.31

2.74

Total Species

199

one individual. Those species occurring in
densities of 10 or more on at least one list
varied by an average of 28 percent.

Censuses of the smaller number of transects
most frequently included species which did not
really use the vegetation type but which were
merely occasional visitors for brief periods
of time or occurred in some edaphic situation.
This was true for the Song Sparrow (Melospiza
melodia) , Painted Bunting (Passerina ciris) ,
Mountain Bluebird ( Sialia currucoides) in
November, and the Lincoln's Sparrow (Melospiza
lincolnii) in April. On the other hand, scarce
species which occurred regularly in a small
part of the area being studied tended to be
missed with a small number of transects —
Loggerhead Shrike (Lanius ludovicianus ) in
November, February and April; Western Bluebird
(Sialia mexicana) and American Robin (Turdus
migratorius) in November; Cactus Wren
( Campy lorhynchus brunneicapillus) in February;
and the Say Phoebe (Sayornis saya) in April.
It is also clear that species which occurred
in densities greater than 10 per 40 ha and
which are not evenly distributed were 20 to 50
percent over- or under-represented. The total
population estimates, BSD's, and number of
species were very similar when either four or
all nine transects were used. Four transects
censused three times are probably adequate for
making general comparisons of overall density,
BSD, and number of species in different vegeta-
tive types. However, if more precise population
data for individual species is desired, more
transects are necessary; but two censuses
instead of three or more are probably adequate.
If availability of time and manpower are
severely limited, the smaller number of transects
(3 to 6 km) will yield valuable data if censused
at least three times monthly.

CONCLUSIONS

Time of day was found to be important in
detectability of birds which became active
earlier but which have a shorter activity period

with increasing photoperiod. If censusing
must begin before and last beyond the optimum
time, the order in which transects are censused
should be arranged so that the same transect is
not censused first or last consistently.

Wind potentially reduces detectability and
censusing should not begin if there are strong
winds (20 to AO kmph) . If winds develop after
censusing has begun, the transect or transect
set can probably be completed without serious
bias; but individual judgment should be exer-
cised in such situations.

For relatively homogeneous areas, up to
at least 3,200 ha, four transects totaling
about 5 km are probably adequate for useful
comparative work. Three censuses per transect
per month are adequate. If more precise species
by species population data are required, more
transects should be established.

ACKNOWLEDGMENTS

We wish to thank the many field biologists
who have helped in collecting data. We are
grateful to Jack Gildar for computerizing the
data. The efforts of the secretarial staff
in typing early drafts and of Penny Dunlop and
Katherine Hildebrandt in typing the final
manuscript are appreciated. Linda Cheney kindly
prepared the illustrations. We are grateful
to Russell Balda, Jane Durham, Jake Rice, and
Kenneth Rosenberg for critically reading the
manuscript. The research was funded through
a grant from the U.S. Bureau of Reclamation.

LITERATURE CITED

Emlen, J. T.

1971. Population densities of birds derived
from transect counts. Auk 88:323-342.
Shields, W. M.

1977. The effect of time of day on avian
census results. Auk 94:380-383.

200

Southwestern Riparian
Communities: Their Biotic
Importance and Management
in Arizona

1 2
David E. Brown ^

Charles H. Lowe .

V 4
Janet F. Hauslsr

Abstract. — The various riparian communities occuring in
Arizona and the Southwest are described and their biotic impor-
tance discussed. Recommendations are made concerning the manage-
ment of streamside environments and their watersheds. These
include recommendations pertaining to the classification and
inventory of riparian habitats; the determination of limiting
factors for key riparian species; the establishment of study
areas; the regulation and elimination of livestock grazing;
tne greater consideration of streamside vegetation in author-
izing water management projects; and the more conservative use
of our watersheds.

INTRODUCTION

No report on riparian habitats would be
complete without a discussion of the char-
acteristics and limiting factors of South-
western riparian vegetation and its asso-
ciated fauna. These biotic communities have
an importance to wildlife and outdoor recrea-
tion greatly disproportionate to their limited
linear acreage. While man's various manipu-
lations and alterations have resulted in enor-
mous changes in the riparian vegetation, so
have his watershed practices affected riparian
environments. The long-term effects of past
and present land management practices are
imperfectly known, but the current situation
for many of our riparian communities cannot
be termed less than disastrous when compared
to conditions of even a short time ago (Freeman
1930, Phillips et al. 1964, Lowe 1964, Jordan
and Maynard 1970, Hubbard 1971, Davis 1973,
Minckley 1973, Turner 1974 and others). Some

Paper contributed to the Importance,
Preservation and Management of the Riparian
I Habitat, July 9, 1977, Tucson, Arizona

. 2,

Arizona Game and Fish Department, Phoenix

University of Arizona, Tucson.

Rocky Mountain Forest and Range Experi-
ment Station, Forestry Sciences Laboratory,
Arizona State University, Tempe.

understanding of our remaining riparian commu-
nities is therefore necessary if we are to
make intelligent judgments about the desir-
ability of future watershed projects in Arizona.

The various riparian communities of Arizona
may be represented as formations or vegetation
types of forest, woodland, marshland, and even
grassland and scrub. A riparian community or
association is one that occurs in or adjacent
to a drainageway and/or its floodplain and
which is further characterized by species and/or
life forms different from those of the immedi-
ately surrounding non-riparian climax (Lowe 1964) .
A riparian community may be composed either of
constituents peculiar to the riparian situation,
or an extension of a higher, climax association
fingering downward into the drainageway; the
latter type has been termed "pseudo-riparian"
(Campbell and Green 1968) to distinguish its
faculative nature from the obligate nature of
purely riparian species. Examples of pseudo-
riparian communities are (1) ponderosa pine
(Pinus ponderosa) forests above the Mogollon
Rim that follow canyons into the pinyon-juniper
woodland, and (2) extensions of some Arizona
upland desert scrub species such as palo verde
(Cercidium f loridum) , ironwood (Olneya tesota)
and saguaros (Cereus giganteus) in arroyos and
along washes within creosote communities in
Yuma County. Another regularly observed
riparian community of this kind is the extension
of encinal or oak woodlands along creeks into
plains and desert grasslands in southwestern
Arizona .

201

It is the riparian communities proper,
commonly with distinctive plant and animal com-
ponents not found elsewhere, that are of great-
est concern here. This concern stems from their
unique character and the resulting changes
brought about by modern man, who has reduced
and eliminated them at an alarmingly rapid
rate. Hopefully an increased awareness and
enlightened attitude on the part of public -
spirited citizens will prevail and many of
the more interesting riparian communities
remaining will be available for future enjoy-
ment and study. The following discussion and
summary of these riparian communities generally
follows the classification outlined in Brown
and Lowe (1974).

I. Temperate Deciduous Forests and Woodlands

Warm-temperate, winter-deciduous gallery
forest and woodlands, where they still occur,
are the most interesting and spectacular ripar-
ian communities in Arizona. Originally,
interior riparian forests occupied most of the
major drainages in the Southwest from the Mohave
and Sonoran Deserts through submogollon Arizona,
northeastern Sonora, southern New Mexico,
northern and eastern Chihuahua to the Rio
Grande and its tributaries in southwest Texas.
Other, cold-temperate deciduous forests occupy
streamsides in montane habitats and in the
Great Plains and Great Basin. These forests
are maintained along perennial or seasonally
intermittent streams and springs and can be
divided into two major communities: mix
broadleaf and cottonwood-willow. Today only
a few drainage systems, such as the undammed Rio
Magdalena in Sonora and (to a lesser extent)
the San Pedro River in Arizona, present extensive
linear riparian forest development. Where
streamflows are seasonally intermittent, riparian
deciduous forests can be expected only where
surface runoff occurs from November through
March (Zimmerman 1969, Hibbert et al. 1974)
and where the advent of the spring growing
season can be expected prior to April 15
(warm-temperate) . After mid-April increased
evapotranspiration and phytotranspiration may
result in only subsurface flow, especially
during daytime hours. Summer precipitation
usually does not result in sustained streamflow
(Zimmerman 1969, Hibbert 1971, Hibbert et al.
1974) , and riparian deciduous forests in the
Southwest are vernal adapted. As such, Arizona's
warm-temperate forests require abundant water
during March and April, when most species set
seed and germinate (Zimmerman 1969) . Probably for
this reason, these forests are poorly represented
or largely absent from the westerm pediments of
the Sierra Madres in southeastern Sonora and
Sinaloa, where winter-spring precipitation is
less than 25 percent of the total.

Interior mixed broadleaf communities are
usually found in Arizona between about 3,500
and 6,500 feet along rubble-bottomed perennial
and semiperennial streams (fig. 1). They are

Figure 1. — Interior riparian deciduous

forest; mixed broadleaf series along Beaver
Creek, Coconino National Forest, Yavapai
County, Arizona; ca. 3850 ft., July, 1971.
Arboreal associates at this locality in this
warm-temperate "gallery" forest are alder,
walnut, ash, cottonwoods and willows. Note
the luxuriant understory and streamside
vegetation without the presence of livestock.

represented in the western portions of the state
along Trout, Francis and Burro Creeks in Mohave
and Yavapai Counties, through the submogollon
region to Rucker and Guadalupe Canyons in south-
eastern Cochise County. Arboreal constituents
may be admixtures or stands of a variety of
Holarctic genera consisting of sycamore
(Platanus wrighti) , ash (Fraxinus pennsylvanica
velutina) , cottonwood (Populus f remontii,
P_. angustif olia) , boxelder (Acer negundo) ,
alder (Alnus oblongif olia) , bigtooth maple
(Acer grandidentatum) , willow (Salix spp.),
walnut ( Juglans major) , mulberry (Morus
microphylla) , bitter cherry (Prunus emarginata) ,
and other deciduous species intermingled with
oaks and, to a lesser extent, conifers from
the adjacent mountains. Arizona cypress
(Cupressus ar izonica) is not infrequent.
Characteristic understory species include
brackenfern (Pteridium aquilinum) , scarlet
sumac (Rhus glabra) , poison ivy (Rhus radicans)
and the deciduous vines, Virginia creeper
(Pathenocissus quinquef olia) and canyon grape
(Vitus arizonica) .

202

Several species of wildlife are totally
or largely dependent on this community. Among
these are the Arizona grey squirrel (Sciurus
arizonensis) , otter (Lutra canadensis) ,
zone-tailed hawk (Buteo albonotatus) , black
hawk (Buteogallus anthracinus) , water ouzel or
dipper (Cincius mexicanus) , sulphur-bellied
flycatcher (My iodyr caster luteiventris) , summer
tanager (Piranga rubra) , Bullock oriole (Icterus
bullocki) , yellow warbler (Dendroica petechial) ,
Arizona alligator lizard (Gerrhonotus kingi) ,
Sonoran mud turtle (Klinosternon sonoriense) ,
and canyon tree frog (Hyla arenicolor) . These
communities also provide major habitat types
for white-tailed deer (Odocoileus virginianus) ,
black bear (Ursus americanus) , turkey (Meleagris
gallopavo) , as well as a myriad of nesting and
migrating raptors and songbirds. Unfortunately,
intensive investigations of the populations and
nesting densities are lacking for most species
of wildlife in this habi,tat type. An important
exception is the lower Gila River in New Mexico
where the biota has been inventoried by Hubbard
(1977) . Lowered streamf low has reduced a
number of forests to scattered, individual
constituents (woodlands) , opening the canopy
and presumably reducing its desirability to
| wildlife dependent on this type. Flash floods,
I such as the notorious Labor Day flood of
September, 1970, have affected many miles of
this beautiful streamside forest, and grazing
by livestock has reduced the quality of the
forest understory almost everywhere, cur-
|; tailing or eliminating reproduction of some
forest species.

Excellent examples of mixed broadleaf
forests are still found in Arizona along Wet
Beaver Creek above Rim Rock, along Oak Creek in
Oak Creek Canyon, along Ash, Redfield, Eagle
and Aravaipa Creeks and the San Francisco River.
A revitalized forest along Rock Creek on the
Three Bar Wildlife Area in the Mazatzal Mountains
is especially worthy of mention. In 1959, after
the elimination of grazing about 15 years before,
the majority of the chaparral watershed burned;
subsequent herbicide treatment prevented the

i rejuvenation of the nonriparian, climax
chaparral community, and the sparsely forested
vegetation along the drainage was transformed
into a dense, excellent representative of mixed
broadleaf deciduous forest. The area now pro-
vides habitats of importance to black bear
and turkey, neither of which had utilized the
area before the transformation (Gallizioli
1974) . Since the streamf low was transformed

' from ephemeral to almost perennial prior to

ji the application of herbicides (Pase and Ingebo
1965), the determining roles of fire and range

• restoration need further consideration.

Forests and woodlands in Arizona dominated
by cottonwood and willow (Populus f remonti
Salix gooddingii, j>. bonplandiana and others)
are confined primarily to riparian environments
below 3,500 feet on clay or other fine soil and
rock deposits —'(fig. 2). Streamf lows are
perennial or nearly so. The understory may be
a tangle of riparian trees or shrubs or rela-
tively open and parklike. Once extensive,
these forests have diminished greatly over the
past 100 years with the diversion, interruption
and elimination of streamf lows. Descriptions
taken from accounts telling of the extent of
these forests along the Santa Cruz, Gila and
Colorado Rivers prior to 1900 are indeed
difficult to envision today (Davis 1973). Up-
stream impoundments, channel cutting, channel-
ization, increased water salinity, irrigation
diversions, and ground water pumping have made
and continue to make massive inroads on these
now relict communities. As in the mixed
broadleaf community upstream, cattle grazing
has negatively influenced the understory and
the quality of remaining stands. Many remaining

Figure 2. — Interior riparian deciduous

forest; Cottonwood-willow series along Aravaipa
Creek, Pinal County, Arizona; ca. 2800 ft.,
September, 1968. Willows, principally Salix
gooddingii, outnumber cottonwood in this
warm-temperate forest and woodland. The prin-
cipal shrub is seep-willow and because of
grazing, the understory vegetation is scant as
oppossed to that shown in Figure 1. Photo
by Richard L. Todd.

iThe limited woodlands of cottonwoods
(Populus acuminata and others) willows (Salix
lasiandra, J5. lutea and others) and other
deciduous trees north of the Mogollon Rim above
6,000 feet in northeastern Arizona are here
considered extreme fasciations of riparian forest
other than warm-temperate interior riparian
deciduous forest.

203

mixed broadleaf riparian forests are under the
jurisdiction of the U.S. Forest Service, where
it is hoped future management of grazing and
timber resources will give added consideration
to these valuable environments (USFS 1969) .

Interrupted examples of cottonwood-willow
forests are still found along the Verde,
Hassayampa, San Pedro, Bill Williams, Colorado
and other rivers. Indications are that these
communities are maintained through periodic
winter-spring flooding. Stabilized water flows
result in decadent stands, in which the dominant
species are lacking in reproduction. Cottonwood
regenerates itself principally from seed, unlike
sycamore and other broadleaf riparian species
that reproduce by sprouting, forming clones
(Horton et al 1960) . Further indications of
the subclimax nature of this community are
the "new" stands adjacent to portions of the
Verde River and Santa Cruz Rivers, which were
generated after heavy winter-spring runoffs
on these drainages in 1965 and 1967 respectively.
The presence of similar fasciations in California
also indicates that these forests are vernal-
adapted, and that late summer runoff is of little
or no benefit to their regeneration.

Studies by Carothers and Johnson (1970)
on the Verde River in Arizona have shown the
importance of cottonwood-willow forests to
breeding birds. More species are recorded as
nesting in this vegetation type than any other;
in Arizona several species such as the yellow-
billed cuckoo (Coccyzus amer icanus) and blue-
throated hummingbird (Lampornis clemenciae) are,
for all practical purposes, restricted to it.
A comparable study of the nesting birds of a
cottonwood-willow community in California showed
a similar importance to nesting birdlife (Ingles
1950). The importance of the cottonwood-willow
community to avian species including raptors,
particularly the black hawk (Buteogallus
anthracinus) , grey hawk (Buteo nitidus) , and
bald eagle (Haliaeetus leucocephhalus) is dis-
cussed by Todd (1969, 1970, 1971, 1972; Hubbard
1971) and otherso The Sonoita Creek Natural
Area retained by The Nature Conservancy along
Sonoita Creek in Santa Cruz County is an over-
mature example of the cottonwood-willow asso-
ciation and a mecca for observers of songbirds
and other wildlife. Because of its proximity to
Mexico, several peripheral species of birds such
as the sub-tropical becard (Pachyramphus agaiae)
are regularly observed here. The importance of
these communities in maintaining environments
for the Southwest's aquatic biota is imperfectly
known, but studies by Minckley (1969) on Sonoita
Creek and other drainages indicate that they may
be of great consequence (also see Miller 1961) .

II . Subtropical Deciduous Woodland

The famous mesquite bosques of pre-settle-
ment Arizona are discussed by Brandt (1951) ,
Phillips et al (1964), Lowe (1964), Davis (1973)
and others. Unfortunately, the major bosques
such as the ones at San Xavier, Komatke (New
York Thicket) , and Texas Hill are now mostly
of historical interest (Brown 1970, 1974; Wigal
1973) (fig. 3). Remnants, some of which are
nonetheless excellent examples, still occur
along the San Pedro, Santa Maria and Verde
Rivers, on the Robbins Butte Wildlife Area
adjacent to the Gila River, along the upper
middle Gila, and in scattered patches along
other Lower Sonoran water courses (fig. 4).
While winter deciduous, these bosques are very
much subtropical and in Arizona are largely
restricted to below 3,500 feet elevation within
the Sonoran Desert, where they attain maximum
development on the alluvium of old dissected
flood plains laid down between the intersection
of major watercourses and their larger tribu-
taries (fig. 5).

Figure 3. — Subtropical riparian deciduous

woodland; remnant of the recently great mes-
quite bosque at Komatke (New York thicket)
near confluence of the Gila and Santa Cruz
Rivers, Gila River Indian Reservation, Mari-
copa County, ca. 1,050 ft., July, 1972. The
rapidly dropping ground water table has re-
sulted in this scene of dead and dying mes-
quites.

204

Figure 4. — Subtropical riparian deciduous wood-
land; interior view of mesquite bosque along
San Pedro Rivers between Cascabel and
Redington, Cochise County, Arizona; May, 1977.
The thrifty appearance and abundant repro-
duction of the mesquites here is in marked
contrast to most of the other bosques in
Arizona. These bosques are being rapidly
cleared for agriculture, however.

Figure 5. — Subtropical riparian deciduous wood-
land: mesquite bosque community along Gila
River below its confluence with Bonita Creek,
Graham County, ca. 3,100 ft., December, 1970.
Note the sharp contrast between the riparian
bosque and the nonriparian Sonoran desert-
scrub .

In the past these subtropic woodlands were
almost completely dominated by mesquite (Proso-
pis julif lora velutina) , once containing indi-
viduals of great size (see e.g., Brandt 1951).
Hackberry (Celtis reticulata) , screwbean
(Prosopis pubescens) , and increasingly the
deciduous saltcedar or tamarisk (Tamarix
chinensis) may now share dominance in local
situations (Bowser 1957, Robinson 1965, Turner
1974) . As in areas of former cottonwood-willow
forest, riparian scrub and marshland, the intro-
duced saltcedar now often exclusively consti-
tutes a disclimax community (fig. 6) at the
expense of native plant and animal diversity
(see e.g., Phillips et al. 1964, Ohmart 1973).

Figure 6. — Riparian deciduous scrubland; a sub-
tropical disclimax consociation along the
Salt River in south-central Arizona; Septem-
ber 1958. Scrublands and woodlands of the
hybrid saltcedar (Tamarix chinensis) now
occupy hundreds of miles of stream channels
in the Southwest where they provide important
nesting habitats for mourning doves, and in
subtropical areas, mourning and white-winged
doves .

Historically, saltbushes (Atriplex poly-
carpa , A. lentif ormis) , or annual and perennial
grasses and forbs formed the ground cover in
mature mesquite bosques; the understory was
relatively open. Today, introduced annual forbs
such as f ilaree (Erodium cicutartium) , mustards
(Crucif erae) and grasses, e.g. Cynadon dactylon,
Bromis rubens , Schismus barbatus and others,
are frequently encountered as understory species.
Vines such as janusia (Janusia gracilis) , canyon
grape (Vitis arizonica) , gourds (Cucurbita
palmata) and others were, and still may be,
conspicuous constituents. Individual cotton-
woods, velvet ash and Goodding willow may be
interspersed in more mesic sites within the
bosque. Grey thorn (Condalia lycioides) or a
blue palo verde (Cercidium f loridum) may occupy
an occasional opening or sunny place.

205

The importance of this woodland type to
colonial nesting white-winged (Zenaida asiatica)
and mourning (Zenaidura macroura) doves is well
documented (Neff 1940, Arnold 1943, Wigal 1973,
Carr 1960 and others). Its importance to other
avian species is discussed by Brandt (1951),
Phillips et al. (1964), Gavin (1972) and others.
This community too has suffered greatly from a
variety of man-related causes including water
diversion, flood control, agricultural clearing
programs, and, principally, dropping water
tables. This llast factor, including interrupted
subsurface flow, has been responsible for the
almost total destruction of the mesquite
"forests" at San Xavier, Casa Grande Ruins
National Monument, Komatke and Texas Hill
(Phillips et al. 1964, Brown 1970, Judd et al.
1971) .

The continued clearing of other bosques
along the Gila and Colorado Rivers has resulted
in their replacement by agricultural crops and
other type conversions. It has been noted that
where intermittent flooding and/or slowly re-
ceeding summer surface flow occurs, saltcedar
tends to replace mesquite. This is particularly
prevalent after the woodlands have been cleared
or burned and ground water is close to the sur-
face and water storage facilities and agricul-
tural tracts are present upstream. Whether this
replacement is partially due to irreversible
changes in water quality and soil chemistry,
or is entirely due to the inherent ability of
tamarisk to repopulate floodplains rapidly, is
a matter for some discussion.

Saltcedar in Arizona has hybridized; it
sets seed and germinates throughout the long
Southwestern growing season (Horton 1960, Horton
et al. 1960), and it is hypothesized that stor-
age facilities which hold back winter-spring
runoff and release water irregularly during
the summer months favor the establishment of this
adventive at the expense of native riparian
communties. The aggressive ability of salt-
cedar to outcompete native riparian species
after summer flooding has been well demonstrated
by Turner (1974) and Warren and Turner (1975).
Nonetheless, saltcedar now provides satisfactory
and important nesting sites for mourning and
white-winged doves (Carr 1960, Shaw 1961,
Wigal 1973 and others) . Several thousand acres
of federal land along the Gila River, much of
which is saltcedar and mesquite, have been
withdrawn for these species under Public Law
1015 as the "Fred Weiler Greenbelt". Other
areas receiving some degree of protection include
the mesquite bosques on the Black Butte Wildlife
Management Area, maintained by the Arizona Game
and Fish Department, and Tonto National Forest

For a discussion of the salt secretion
abilities of saltcedar see Decker 1961.

lands along the Verde River. The high demands
placed on both mesquite wood and ground water,
however threaten all remaining bosques (see
e.g., Lacey et al. 1975).

Ill . Subtropical Evergreen Forest

This complex tropic-subtropic formation
has its northern terminus in moist canyons
and warm springs in and adjacent to the Sonoran
Desert in Arizona and California, where it is
represented by stands of California fan palm
(Washingtonia f ilif era) . In Arizona native
groves — -but not all individuals — are limited
to two canyons in the Kofa Mountains (Benson
and Darrow 1954, Smith 1974), to three sites at
end near Alkali Springs in the Hieroglyphic
Mountains (Brown et al. 1976) and possibly
Cienega Springs near Parker (fig. 7). Because
of their miniscule acreage and disjunct
occurrence, these communities lack the charac-
teristic vegetational and faunal associates of
more southerly subtropic evergreen forests and
possess instead distinctive Sonoran oasis
associates (Vogl and McHargue 1966, Brown et
al. 1976). That these relics of the Miocene
and Pliocene remained at all in Arizona was due
to the continual presence of abundant sub-
surface waters in favored tropic-subtropic
microenvironments . One also suspects that the
adaptibility of this species to alkaline waters
may have been a competitive advantage with cer-
tain warm temperate forms.

Figure 7. — Subtropical riparian evergreen
forest; California fan palm series at
Cienega Springs, Yuma County, Arizona.
Abundant reproduction frequently characterizes
native palm groves in Arizona; the fan palms,
tolerant of alkaline waters, have outcom-
peted their cottonwood-willow competitors
over the years at this and other sites.

206

California fan palms are attractive trees,
and their adaptibility to cultivation has made
them an ubiquitous ornamental landscape feature
throughout the Southwest. The few native com-
munities are considered botantical phenomena
to be maintained with a minimum of disturbance.
The palms in Palm Canyon, Hidden Canyon and
elsewhere have had their shag of dead fronds
burned but otherwise appear in good condition,
with some reproduction noted. Palm groves and
individuals in the Kofa Mountains are within
the Kofa Game Range and are under the juris-
diction of the United States Fish and Wildlife
Service. The palms at Alkali Springs and
Cienega Springs are privately owned.

IV. Riparian Scrublands

While riparian scrub communities cover
extensive areas of stream channels and flood
plains, scientific investigations and resource
managers have generally ignored them and con-
centrated on the more interesting and diverse
communities upstream and downstream. They are,
nonetheless, both interesting and important.

Above 8500 feet, a boreal riparian scrub is
usually present along subalpine streams and in
some wetlands. These scrublands are dominated
by scrub willows (Salix bebbiana, j>. scouleriana) ,
although red-osier dogwood (Cornus stolonif era) ,
blueberry elder (Sambucus glauca) , rocky moun-
tain maple (Acer glabrum) and thin-leaf alder
(Alnus tenuifolia) may be locally important,
particularly downstream as one approaches and
enters more cold temperate conditions (fig. 8).
Occasional trees such as blue spruce (Picea
pungens) and aspen (Populus tremuloides) may
stand out within the scrub. These streamside
scrublands are nesting habitat for dusky fly-
catchers (Empidonax oberholseri) , MacGillivary
warblers (Oporornis tolmiei) , orange-crowned
warblers (Helminthophila celata) , broad-tailed
hummingbirds (Selasphorus platycercus) , white-
crowned sparrows (Zonotrichia leucophrys) and
Lincoln sparrows (Melospiza lincolni) . The
perennial streams are themselves the habitat of
the native Arizona trout (Salmo apache) and the
now ubiquitous rainbow (Salmo gairdneri) . These
stream habitats are subject during the summer
months to extensive and intensive livestock
grazing, including use by sheep. Stream quality
has also been altered by logging activity on
adjacent watersheds, a situation which can be
expected to increase with the demand for timber
resources.

In temperate and subtropic situations in
intermittent and perennial stream channels and
in and along flood channels one also encounters
riparian scrublands (fig. 9). Stream flows in
these types are irregular and often occur in
the form of flash floods. Dominant species are

frequently but not necessarily seepwillow or
batamote (Baccharis glutinosa) , broom (Baccharis
sarothroides or 13. emoryi) , arroweed (Pluchea
sericea) , and, increasingly, saltcedar. The
reasons for the increase in saltcedar at the
expense of the native seepwillow since 1940
have been discussed earlier and are well
documented by Horton et al. 1960, Zimmerman
1969, Turner 1974, and Warren and Turner 1975.
Riparian scrub may exhibit a dense "chaparral"
aspect — scrubland — or present a very open
one — desertscrub . Desert willow (Chilopsis
linearis) , mesquite, catclaw (Acacia greggi)
and other arboreal species are frequent asso-
ciates and may share aspect dominance. These
trees as well as those of the riparian deciduous
forest, if present, provide less than 15 percent
of the ground cover. Faunal relationships with-
in these riparian communities are poorly in-
vestigated, but there appears to be a consid-
erable interaction with greater or lesser
populations of adjacent or upslope nonriparian
species. Bird species particularly well
represented in riparian scrub include the Say's
phoebe (Sayornis saya) , crissal thrasher
(Toxostoma dorsale) , black-tailed gnatcatcher
(Polioptila melanura) , phainopepla (Phainopepla
nitens) and the brown towhee (Pipilo fuscus) .
To date, little attempt has been made to "manage"
these habitats.

Figure 8. — Montane riparian deciduous scrubland;
Mixed series along the North Fort of White
River, Fort Apache Indian Reservation; ca,
7500 ft., July, 1977. Prevalent and dominant
plants here include two willows, thin-leaf
alder, blueberry elder, and hawthorn
(Crataegus erythropoda) .

207

Figure 9. — Evergreen riparian scrubland in the
channel of the San Carlos River, San Carlos
Indian Reservation; ca. 3200 ft., March, 1975.
The thick "Chaparral" in foreground is largely
seep-willow or batamote. The deciduous scrub
is mostly saltcedar. Note the decadent stand
of cottonwood along the bank in the distance.

Figure 10. — Saltwater marshland; Saltgrass

series at Obed Meadows, Navajo County, Arizona,
Saltgrass occupies wetland and riparian areas
throughout Arizona's subtropic and temperate
zones wherever alkaline habitats exist. The
deciduous trees in background are the now
ubiquitous saltcedar.

V. Marshlands

These wetland formations may be comprised
if any of several boreal, temperate or sub-
tropical emergent communities and are defined
as aquatic communities, the principal plant
constituents of which are emergents not trees,
woody shrubs, or nonhalophy tic grasses ^ , and
which normally or regularly have their basal
portions annually, periodically or continually
submerged. In the Southwest these include
communities in both fresh or brackish water
environments. They range from the more xeric
and alkali communities of salt grass (Distichlis
stricta) , and alkali bulrush (Scirpus paludosus)
through the carrizo or reed communities
(Phragmites communis) of the Colorado River and
elsewhere to mesic freshwater communities of
rushes ( Juncus spp.), sedges (Carex spp.).
bulrushes (Scirpus spp.) and cattail (Typha
spp.) (fig. 10, 11).

J Riparian grasslands of sacaton
(Sporobolus wr ight ii) , tobosa (Hilaria mutica)
and other communities, while not discussed,
occur in Arizona and the Southwest. See Lowe
(1964) for a discussion of tobosa swales. Salt-
grass communities are treated here as part of
the marshland formation.

Figure 11. — Freshwater marshland; Topock Marsh
looking north from north dike, Mohave County,
ca. 550 ft. Bullrush and cattail are the
principal vegetational constituents in fore-
ground. This famous marsh is one of the few
remaining on the Colorado River and is an
important breeding area for the Yuma clapper
rail.' Photo by Richard L. Todd

208

These rapidly disappearing communities
are found in riparian and littoral situations
only where streamflow is turgid, shallow and
dependable enough to permit their establishment.
Since they are the most mesic of Arizona's
vegetational and biotic communities, they have
suffered most from the resultant desiccation of
the state's natural environment through water
diversions and water "management" projects
(see e.g., Ohmart ca. 1974). The few riparian
marshland communities that remain are habitats
for a number of species of Arizona's rare and
vanishing wildlife, such as the Yuma clapper
rail (Rallus longirostris) , black rail
(Laterallus jamaicensis) , bitterns (Ixobrychus
exilis , Botaurus lentiginosus) , and Mexican
duck (Anas diazi) (Todd 1972a) . Numerous other
rails, shorebirds, and waterfowl are highly de-
pendent on these diverse environments, both
during nesting and migration (Todd 1972a). These
marshland oases are now frequently dependent on
stored and/or recycled argricultural and indus-
trail waste waters from diverted upstream flow.
Examples in Arizona are Picacho Lake in Pinal
County and Quigley Pond on the Gila River in
Yuma County (see also Brown et al. 1977).
Exceptions are a few sloughs and old oxbows
of the San Pedro, lower Salt, Verde and Colorado
Rivers, almost all of which are threatened by
existing or planned projects. It is also an
ironic fact that Arizona's most valuable
wildlife habitats are too frequently subjected
to trampling and grazing by livestock, in
addition to hydrological limitations.

VI. Recommendations

It has become increasingly evident that
the most valuable and interesting of Arizona's
streamside environments are greatly in need of
more enlightened management of both the actual
riparian communities and the watersheds upon
which they depend. Their present limited
acreage and importance to endangered, threatened,
and peripheral wildlife species have prompted
a growing concern by wildlife-oriented groups
and individuals in addition to the concern long
voiced by professional biologists. This concern
has now manifested itself in the political arena
and requires that our riparian environments
receive greater consideration from resource
management agencies.

The following recommendations are suggested
to perpetuate and enhance those riparian com-
munities of greatest value to wildlife and
public interest:

I. Identify and classify Arizona's riparian
environments. Identification and mapping of
streamside vegetation is presently either
being considered or in the process of inven-
tory by land management agencies, other public

agencies, academic groups and ad hoc consul-
tants. These efforts should be coordinated
and classifications of the various types
determined. A statewide inventory, including
maps, of remaining habitats should be pre-
pared and published.

Investigate factors d
specific riparian const
The environmental requi
many of the major ripar
must be determined, at
would of necessity be 1
uous studies to provide
the factors controlling
and their constituents,
to preserve and manage
ents through regulated
from reservoirs, select
techniques .

etermining the limiting
ituents and communities,
sities and limits of
ian plant species
least in part. These
ong-range and contin-
an understanding of
the various communities
Only then can we hope
our riparian constitu-
discharges of water
ive cutting and other

3. Establish representative study areas con-
taining all major riparian communities and
their surface and groundwater requirements.
In addition, as reserves these areas would
provide "bench marks" and controls for com-
parison with "managed" or other "modified"
ecosystems .

4. Grazing and other distruptive influences
should be eliminated or controlled in ripar-
ian forests, woodlands and marshlands. Many
of these have had their public values com-
promised through the degradation of their
flora and fauna. Areas presently supporting
little or no understory and showing no repro-
duction of major riparian constituents should
be restored where still possible.

5. Riparian and watershed management project
planners should reconsider the values both
actual and potential of streamside vegetation
before irreversible alterations. Several
"phreatophyte clearing" projects have resulted
in unwarranted destruction of native riparian
associations with little or no documented
water "salvage" or other claimed conservation
measures accomplished (Campbell 1970, Horton
1972, Patrick 1971).

6. Increase the effort to avoid torrential
summer and fall flooding through more conserva^
tive use of grazing and timbering watershed
resources. Shrub invasions of Southwestern
watersheds, due to livestock grazing pressures
and suppression of fire, have long been
documented (see e.g., Leopold 1924, Humphrey
1958). Through proper management, streamflows
can be stabilized and increased to the
benefit of our riparian resources. These
management techniques should be applied

now throughout our rapidly deteriorating
Southwest riparian wonderland.

209

LITERATURE CITED

Arnold, L.W. 1943. A study of the factors
influencing the management of and a sug-
gested management plan for the western
white-winged dove in Arizona. Ariz. Game
and Fish Comm. 103 p.

Benson, L. and R. A. Darrow. 1954. The trees
and shrubs of the Southwestern Deserts. 2nd
ed. Univ. of Ariz. Press, Tucson, and Univ.
New Mexico Press, Albuquerque. 437 p.

Brandt, H. 1951. Arizona and its birdlife.
Bird Research Foundation, Cleveland, Ohio.
723 p.

Brown, D.E. 1970. Summary of white-winged
dove banding and hunt information. Proj .
W-53-R-20: WP3, J2. Spec. Rep. 1961-1969.
Ariz. Game and Fish Dep. 28 p.

Brown, D.E. and G.C. Sanderson. 1974. White-
winged dove (Zenaida asiatica) In Management
of Migratory Shore and Upland Game Brids in
North America. The Intern. Assoc. of Fish
and Wildl. Agencies, Washington, D.C. p. 246-
272.

Brown, D.E. and C.H. Lowe. 1974. A proposed
classification for natural and potential vege-
tation in the Southwest with particular refer-
ence to Arizona. J. Ariz. Acad. Sci. 9(2).
11 p.

Brown, D.E., N.B. Carmony, C.H. Lowe, and R.M.
Turner. 19760 A second locality for native
California fan palms (Washintonia f ilif era)
in Arizona. J. Ariz. Acad. Sci. 11(1): 37-41.

Brown, D.E., N.B. Carmony, and R.M. Turner.
1977. Drainage map of Arizona showing peren-
nial streams and some important wetlands.
Map. Proj. W-53-R, Ariz. Game and Fish Dep.

Bowser, C.W. 1957. Introduction and spread of
the undesirable tamarisks in the Pacific
southwestern section of the United States and
comments concerning the plants influence upon
the indigenous vegetation. Trans, of Amer.
Geophy. Union. 38(3). 7 p.

Campbell, C.J. 1970. Ecological implications
of riparian vegetation management. J. of Soil
and Water Conserv. 25(2):49-52.

Campbell, C.J., and W. Green. 1968. Perpetual
succession of stream-channel vegetation in a
semi-arid region. J. Ariz. Acad. Sci. 5(2)
96-98.

Carothers, S.W. and R.R. Johnson. 1970. A

summary of the Verde Valley breeding bird sur-
vey, 1970. Mus. of Northern Ariz. 13 p.

Carr, J.N. 1960. Mourning dove whitewinged
dove nest surveys during the summer of 1960.
Proj. W-53-R-11; WP3, 12. 6 p.

Davis, G.P. 1973. Man and wildlife in Arizona:
the pre-settlement era, 1823-1864. Master
of Sci. Thesis, Univ. of Ariz.

Decker, J. P. 1961. Salt secretion by Tamarix
pentandra Pall. Forest Sci. 7;214-217.

Freeman, L.R. 1930. Down the Grand Canyon.

Dodd, Mead and Co., N.Y. 371 p.
Gallizioli, S. 1974. Personal communication.

Ariz. Game and Fish Dep.
Gavin, T.A. 1972. Avian and mammalian species

composition in a mesquite bosque. Proj. 73,

Quart. Rep. Ariz. Coop. Wildl. Res. Unit,

Univ. of Ariz, unpubl.
Hibbert, A. R. 1971. Increases in stream-
flow after converting chaparral to grass.

Water Resour. Res. 7(l):71-80.
Hibbert, A.R. , E.A. Davis and D.G. Scholl. 1974.

Chaparral coversion potential in Arizona, Part

lrwater yield response and effects on other

resources. USDA For. Serv. Res. Pap. RM-126,

Rocky Mt. For. and rAnge Exp. Sta. , Fort

Collins. 36 p.
Horton, J.S. 1960. Ecology of salt cedar. In

Watershed and Related Water Management Problems.

Proc. 4th An. Ariz. Watershed Symp., Phoenix.
Horton, J.S., F.C. Mounts and J.M. Kraft. 1960.

Seed germination and seedling establishment of

phreatophyte species. USDA For. Serv. Sta. Pap.

48. Rocky Mt. For. and Range Exp. Sta., Fort

Collins. 26 p.
Horton, J.S. 1972. Management problems in the

phreatophyte and riparian zones of the Southwest.

J. of Soil and Water Conserv. 27(2): 57-61.
Horton, J.S. and C.J. Campbell. 1974. Manage-
ment of phreatophyte and riparian vegetation

for maximum multiple use values. USDA For.

Serv. Res. Pap. RM-117. Rocky Mt. For. and

Range Exp. Sta., Fort Collins. 23 p.
Hubbard, J. P. 1971. The summer birds of the

Gila Valley, N.M. Occa. Pap. of the Delaware

Mus. of Nat. His., Nemouria 2:1-35.
Hubbard, J. P. 1977. A biological inventory of

the lower Gila River, N.M. Bur. of Land Manage.

Bur. of Rec, New Mexico Dep. Game and Fish,

Soil Conserv. Serv., U.S. Fish and Wildl. Serv.

and U.S. For. Serv. 56 p.
Humphrey, R.R. 1958. The desert grassland. A

history of vegetational change and an analysis

of causes. Bot. Rev. 24(4) :193-252.
Ingles, L.G. 1950. Nesting birds of the willow-

cottonwood community in California. Auk 67(3).
Jordan, G.L. and M.L. Maynard . 1970. The San

Simon watershed, historical review. Prog.

Agric. in Ariz. 22 (4): 6-9.
Judd, I.B., J.M. Laughlin, H.R. Guenther and

R. Handegarde. 1971. The lethal decline

of mesquite on the Casa Grande Ruins National

Monument. Great Basin Nat. 31(3).
Leopold, A. 1924. Grass, brush, timber and fire

in southern Arizona. J. of For. 22(6):1-10.
Lacy, J.R., P.R. Ogden and K.E. Foster. 1975.

Southern Arizona riparian habitat : spatial

distribution and analysis. School of Renew.

Nat. Resour. and Office of Arid Lands Studies.

OALS Bull. 8:1-148.
Lowe, C.H. 1964. Arizona's natural environment;

landscape and habitats. Univ. of Ariz. Press,

Tucson. 270 p.

210

Miller, R.R. 1961. Man and the changing fish
fauna of the American Southwest. Papers
Michigan Acad. Sci., Art. Lett. 46:365-404.

Minckley, W.L. 1969. Aquatic biota of the
Sonoita Creek Basin, Santa Cruz County, Ariz.
Ecol. Studies Leaflet, No. 15. The Nature
Conservancy. 8 p.

Minckley, W.L. 1973. Fishes of Arizona. Ariz.
Game and Fish Dep., Phoenix. 293 p.

Neff, J. A. 1940. Range, population and game
status of the western whitewinged dove in
Arizona. J. Wildl. Manage. 4 (2) : 117-127 .

Ohmart, R.D. 1973. Observations on the winter-
ing birds and mammals in the proposed Buttes
Dam site, Central Ariz. Pro j . Bur. Reclam. ,
Boulder City, Nev. 16 p.

Ohmart, R.D. 1974 (ca.) Dynamics of marsh
land formation and succession along the lower
Colorado River and their importance and
management problems as related to wildlife
in the arid Southwest. Bur. of Reclam.
Preprint 1-21.

Pase, CP. and P. A. Ingebo . 1965. Burned
chaparral to grass: early effects on water
and sediment yields from two granite soil
watershed in Arizona. Proc. Ariz. Watershed
Symp. 9.

Patrick, R. 1971. The effects of channelization

on the aquatic life of streams. Acad, of Nat.

Sci. of Philadelphia. 12 p.
Phillips, A., J. Marshall and G. Monson. 1964.

The birds of Arizona. Univ. of Ariz. Press,

Tucson 212 p.
Robinson, T.W. 1965. Introduction, spread and

areal extent of salt cedar (Tamarix) in the

western states. Studies of Evapotranspiration

Geol. Surv. Prof. Pap. 491-A.
Shaw, H. 1961. Influence of salt cedar on

whitewinged doves in the Gila Valley. Spec.

Rep., Ariz. Game and Fish Dep. 9 p.

Smith, E.L. 1974. Established natural areas

in Arizona — A Guidebook for scientists and

educators. Office of Econ. Plan, and Develop.

Phoenix, 300 p.
Todd, R.L. 1969. Nongame investigations. Proj .

W-53-R-19; WP5, Jl. Prog. Rep. Ariz. Game and

Fish Dep. 26 p.
Todd, R.L. 1970. Nongame investigations. Proj.

W-53-R-20; WP5, Jl. Prog. Rep., Ariz. Game and

Fish Dep. 16 p.
Todd, R.L. 1971. Nongame investigations. Proj.

W-53-R-21; WP5, Jl. Prog. Rep., Ariz. Game and

Fish Dep. 11 p.
Todd, R.L. 1972. Nongame investigations. Proj.

W-53-R-22; WP5, Jl. Prog. Rep. Ariz. Game and

Fish Dep. 22 p.
Todd, R.L. 1972a. Biological report on a marsh

near Tuzigoot National Monument. Proj. W-53-R-

23; WP5, Jl. Spe. Rep., Ariz. Game and Fish

Dep., 4 p.

Turner, R.M. 1974. Quantitative and historical

evidence of vegetation changes along the upper

Gila River, Arizona. U.S. Geol. Surv. Prof.

Pap. 655-H. 20 p.
U.S. Forest Service. 1969. Wildlife Habitat

improvment handbook. USDA For. Serv.
Vogl, R.J. and L.T. McHargue. 1966. Vegetation

of California fan palm oases on the San Andreas

fault. Ecol. 47:532-540.
Warren, D.K. and R.M. Turner. 1975. Saltcedar

(Tamarix chinensis) seed production, seedling

establishment, and response to inundation.

J. Ariz. Acad. Sci. 10 (3) : 117-119 .
Wigal, D.D. 1973. A survey of the nesting

habitats of the whitewinged dove in Arizona.

Spec. Rep. No. 2, Ariz. Game and Fish Dep.

37 p.

Zimmerman, R.C. 1969. Plant ecology of an

arid basin Tres Alamos — Redington areas. U.S.
Geol. Surv. Prof. Pap. 485-D. 51 p.

211

Terrestrial Mammals
of the Riparian Corridor
in Big Bend National Park1

s , 2

William J. Boeer and David J. Schmidly

Abstract. — Thirty species of terrestrial mammals inhabit
riparian habitats in Big Bend National Park (BBNP) , but only
one species (the beaver, Castor canadensis) is restricted to
these areas. Major changes in the vegetation during the past
30 years, involving an increase in basal and canopy cover,
have resulted in the elimination of at least one species
(Di-podomys ordii) from the river corridor as well as
increased abundance and distribution for two other species
(Sigmodon hispidus and Peromysaus leuaopus) . Compared to
the other major plant communities in BBNP, the rodent fauna
of the riparian community has lower evenness, richness, and
diversity indices (based on the Shannon-Weaver Index) .
Human use and trespass livestock grazing are the major
impacts acting upon the natural riparian communities in
BBNP today.

INTRODUCTION

Mammalian studies of the Big Bend area began
with general surveys (Bailey 1905; Johnson 1936;
Borell and Bryant 1942; and Taylor et at. 1944)
designed to identify and document the varied
fauna of the area. After the park was esta-
blished, the perspective of mammalian research
changed somewhat and in recent years studies
have concentrated on mammalian autecology and
synecology (Porter 1962; Dixon 1958; and Easter-
la 1973) . Most mammalian studies have focused
on the mammals of the montane woodland and
desert grassland habitats. There have been
no comprehensive studies of riparian mammals.
Baccus (1971) investigated the distribution
of rodents in the park with respect to the major
physical features, focusing on the effects of
the elimination of grazing on the vegetation
and the rodent populations. He also described
the similarities and. dissimilarities of the
rodent faunas of the woodland, grassland, and
desert shrub communities; however, he divided
the fauna of the Rio Grande floodplain between
the desert shrub and grassland communities and
did not consider the riparian corridor as a
unique habitat.

Contributed paper, Symposium on the
Importance, Preservation and Management of the
Riparian Habitat, July 9, 1977, Tucson, Arizona.
2

Texas Agricultural Experiment Station,
Department of Wildlife and Fisheries Sciences,
Texas A&M University, College Station, Texas.

DESCRIPTION OF THE RIPARIAN CORRIDOR

Floodplain or riparian vegetation exists
wherever periodic flooding occurs along the
Rio Grande in BBNP. These riparian communities
vary from areas a few meters (m) wide to areas
extending inland a distance of one kilometer
(km); furthermore, adjacent arroyos and creeks
may carry enough surface or ground water to
produce a similar floodplain environment. Top-
ography along the river includes (1) sheer wall
canyons (i.e., Santa Elena and Mariscal canyons
which rise to elevations of 366 m) with few
areas of alluvial deposits; (2) long deep canyons
(i.e., Boquillas Canyon) where the walls do not
rise abruptly and where larger areas of alluvial
deposits occur; and (3) areas of broad flat
floodplain with extensive alluvial deposits.

Denyes (1956) recognized three plant associ-
ations along the Rio Grande floodplain: (1) the
riverbank association, consisting of mesquite
(Prosopis juli flora) , seep willow (Bacoharis
sp.), willow (Salix gooddingi) , or cottonwood
(Populus palmeri) , located adjacent to areas of
exposed silt and coarse gravel at the water's
edge; (2) the Baeoharis association, composed
of dense stands of seep willow; and (3) the
mesquite association, consisting of a thin line
of extensive mesquite trees or an extensive area
of several different plant forms. We have found
these three associations to be somewhat altered
from Denyes' description and, although difficult
to document, our general impression is that
significant vegetation changes have occurred

212

in the riparian habitats of BBNP over the past
30 years. The major change is associated with
the tremendous increase of salt cedar (Tamarix
chinensis) along the river. The Baccharis asso-
ciation, mentioned by Denyes (1956) as common in
the fine sandy loam soils along the river, is
recognizable today only at a few places (for
example, Black Dike) and appears to have been
replaced by a mixed mesquite-salt cedar-bermuda
grass (Cynodon daotylon) association. Similarly,
salt cedar also appears to be replacing native
cottonwood and willow trees at many places.

J - MAMMALIAN FAUNA OF THE RIPARIAN CORRIDOR

Thirty species of terrestrial mammals have
been either collected or observed in the rip-
arian habitats of BBNP. These are listed below
in checklist fashion with their current status
(C = common; U = uncommon; R = rare; E = prev-
iously present, but no longer occurs; P = pos-
sibly occurs) in the park.

Pouched Mammals - Order Marsupialia
Opossum - Family Didelphidae
Virginia Opossum Didelphis
virginiana - P
Lagomorphs - Order Lagomorpha

Hares and Rabbits - Family Leporidae
Desert Cottontail Sylvilagus

audubonii - C
Black-tailed Jack Rabbit Lepus
californieus - U
Rodents - Order Rodentia

Squirrels - Family Sciuridae

Texas Antelope Squirrel Ammospermophilus

interpres - R
Mexican Ground Squirrel Spermophilus

mexicanus - R
Spotted Ground Squirrel Spermophilus
spilosoma - U
Pocket Gophers - Family Geomyidae
Yellow-faced Pocket Gopher
Pappogeomys aastanops - C
Pocket Mice - Family Heteromyidae
Silky Pocket Mouse Perognathus

flavus - U
Desert Pocket Mouse Perognathus

penicillatus - C
Nelson's Pocket Mouse Perognathus

nelsoni - R
Ord ' s Kangaroo Rat Dipodomys

ordii - E
Merriam's Kangaroo Rat Dipodomys
merriami - C
Beaver - Family Castoridae

Beaver Castor canadensis - U
New World Rats and Mice - Family
Cricetidae
Cactus Mouse Peromyscus eremicus - U
White-footed Mouse Peromyscus

leucopus - C
Deer Mouse Peromyscus maniculatus - R
Hispid Cotton Rat Sigmodon hispidus - C

Southern Plains Woodrat Neotoma
micropus - C
New World Porcupines - Family
Erethizontidae
Porcupine Erethizon dorsatum - R
Carnivores - Order Carnivora

Dogs and Relatives - Family Canidae
Coyote Canis latrans - C
Gray Fox Urocyon cinereoargenteus - U
Raccoons - Family Procyonidae

Ringtail Bassariscus astutus - U
Raccoon Procyon lotor - C
Weasels and Relatives - Family Mustelidae
Striped Skunk Mephitis mephitis - U
Western Spotted Skunk Spilogale

gracilis - R
Hog-nosed Skunk Conepatus mesoleucus - R
Cats - Family Felidae

Mountain Lion Felis concolor - R
Bobcat Felis rufus - U
Even-toed Ungulates - Order Artiodactyla
Peccaries - Family Tayassuidae

Collared Peccary (Javelina) Dicotyles
tajacu - U
Deer - Family Cervidae

Mule Deer Odocoileus hemionus - U

During 1975-1976, we sampled small rodents
at 18 different sites along the riparian corri-
dor. Each site was trapped (using Sherman live
traps) a total of 720 trap nights resulting in
12,960 trap nights for the entire river corridor.
A total of 1,292 rodents representing two fami-
lies (Heteromyidae and Cricetidae) were captured
as follows (number trapped in parentheses) :
Family Heteromyidae: Perognathus penicillatus
(896); Perognathus nelsoni (2); Perognathus
flavus (5); Dipodomys merriami (65). Family
Cricetidae: Peromyscus leucopus (162); Peromyscus
eremicus (19) ; Sigmodon hispidus (70) ; Neotoma
micropus (73). Perognathus penicillatus was
overwhelmingly the most abundant small rodent
in the riparian habitats and, for this reason,
the total density of heteromyid rodents was
greater than that of cricetid rodents. The
three other heteromyid rodents were relatively
rare along the river, although D. merriami was
common at a few sites. Densities of the four
species of cricetid rodents were more similar
to one another than the densities of the heter-
omyid species. Peromyscus leucopus was the
most common cricetid and P. eremicus the least
common; Sigmodon hispidus and Neotoma micropus
occurred in about equal numbers •

Borell and Bryant (1942) also found
Perognathus penicillatus to be the most abundant
rodent in the riparian corridor. However, com-
paring our data with that of Borell and Bryant
(1942) for three other species (Dipodomys
ordii, Peromyscus leucopus , and Sigmodon
hispidus) reveals that significant changes in
abundance and distribution have occurred in
these species over the past 30 years. These

213

differences correlate with major vegetative
changes associated with the cessation of exten-
sive livestock grazing. Early accounts (Taylor
et at. 1944; Sperry 1938) describe the vegeta-
tion along the river as open and severely
over-grazed. However, since ranching activities
ceased at the inception of the park, plant den-
sities seem to have increased greatly so that
at several places (e.g., Johnson Ranch) mes-
quite forests now occur where the river bottom
was once open and sparsely vegetated. Exten-
sive fields of grass also occur today at sites
(e.g., Smoky Creek and Coyote) which formerly
were cultivated and farmed.

Generally, cricetid rodents prefer habitats
with considerable ground cover. Thus, the
increased density of grass and cane (Phragmites
communis) along the riparian corridor, as a
result of the elimination of grazing, has
served to substantially increase suitable habi-
tat for these rodents. Two cricetines
(Sigmodon hispidus and Peromysaus leucopus)
exemplify this trend. Borell and Bryant (1942)
collected only four specimens of Sigmodon
hispidus along the river among the cane and
cultivated fields around the Johnson Ranch.
We recorded 70 cotton rats from 12 different
localities along the river in areas where thick
bermuda grass, cane, and fleabane (Erigeron sp.)
were present. Similarly, Borell and Bryant
(1942) reported taking a few Peromysaus
leucopus along the river from one mile SW
Boquillas and the Johnson Ranch. Our trapping
records indicate that P. leucopus is now one
of the most common rodents of the riparian
corridor and this mouse occurs all along the
river from the mouth of Santa Elena Canyon to
Rio Grande Village.

Ord's kangaroo rat (Dipodomys ordii) is a
species which apparently has completely dis-
appeared from the riparian corridor during the
past 30 years. This species was first reported
from BBNP in 1939 by M. D. Bryant who described
a distinct subspecies (D. o. attenuatus) from
the mouth of Santa Elena Canyon. In 1944, Dr.
William B. Davis (pers. comm.) collected two
specimens from the type locality and another
from the Johnson Ranch. There have been no
additional specimens captured along the Rio
Grande since then, although Baccus (1971)
trapped at the mouth of Santa Elena Canyon,
the Johnson Ranch, and other sites along the
river. In over 13,000 trap nights along the
river, including efforts at the type locality
and the Johnson Ranch, we failed to capture a
single D. ordii. Baccus (1971), however, did
obtain a few specimens from Upper Tornillo
Creek Bridge (16 km NE Panther Junction) , and
this apparently represents the only remaining
population of this subspecies in BBNP.

In order to ascertain the status of D. o.

attenuatus, we spent eight days (from 4 April
1976 to 12 April 1976) trapping at Upper
Tornillo Creek Bridge and other places where
this species had been previously collected.
Initially, 70 traps were set on both the east
and west side of Upper Tornillo Creek. Later,
the number of traps was increased to 110 on
the west side and 160 on the east side. The
traps were set out in various soil and vege-
tation types ranging from deep sand-sparse
burro-brush (Eymenoclea monogyra) , to packed
sand-mesquite and gravelly-creosote (Larrea
divaricata) flats. A total of 18 D. ordii
and 21 D. merriami were caught during the
first two nights of trapping. Thirty traps
were also set at Lower Tornillo Creek Bridge
and 120 were placed along Terlingua Creek
where it enters the mouth of Santa Elena
Canyon. Most of the Lower Tornillo Creek area
was a creosote flat with clumps of catclaw
{Acacia greggii) and lechuguilla {Agave
lechuguitla) next to the creek bed. The dry
creek bed itself was very rocky and surrounded
a small knoll of deep sand covered with little
vegetation. Eight traps were placed on this
knoll and 22 in the surrounding flats adjacent
to the creek bed. A single D. ordii was
captured on the knoll and nine D. merriami
were captured on the flats. At Terlingua
Creek, the traps were placed in a sandy area
and 10 D. merriami were captured.

A trapping grid established at Upper Tor-
nillo Creek consisted of 10 lines of 40 traps
per line with each trap 18 m apart; each line
was 36 m apart. With regard to vegetation and
soil, three distinct habitats (designated A, B,
and C) were delineated on the grid. Habitat A
was on the first floodplain stage adjacent to
the creek and consisted of a very open area of
deep sandy soil with burro-brush and a few
desert willow (Chilopsis linearis) comprising
the dominant vegetation. Habitat B included
the second floodplain stage and consisted of a
sandy but more compact soil with a moderate
cover of vegetation including creosote, white-
thorn (Acacia constricta) , mesquite, burro-
brush, grass, and prickly pear (Opuntia sp.).
Habitat C was located on a bench about 3-4 m
above habitats A and B. The soil was very
compact and the vegetation moderate to thick.
Typical desert vegetation, consisting of clumps
of mesquite, prickly pear, allthorn
(Koeberlinia spinosa) , creosote, and . tasaj illo
(Opuntia leptocaulis) were interspersed through-
out habitat C.

The total number of captures of D. ordii
and D. merriami for each of the 10 trap lines
is presented in Table 1. The percentage of
captures of D. ordii for each of the three
habitat types was as follows: habitat A,
66.1%; habitat B, 30.6%; habitat C, 3.2%.
Thus, D. ordii was most common in the deep,

214

Table 1. — Four day capture totals by trap line for Dipodomys ordii and Dipodomys merriami at
Upper Tornillo Creek Bridge.

Habitat Trap Total D. ordii D. merriami Percent Percent

type line captures captures captures D. ordii D. merriami

(40

3)1

8)1

(25

(11

(15

(14

(27

100

Represent percent of total catch of D. ordii or D. merriami in a particular trap line.

sandy and sparsely vegetated areas of the first
floodplain stage and decreased in number in
habitats away from the creek bottom. The per-
centage of captures of D. merriami for each of
the three habitat types was: habitat A, 11.6%;
habitat B, 50.7%; and habitat C, 37.7%. Thus,
D. merriami was more generally distributed
throughout the three habitats but was much less
common in habitat A where D. ordii dominated.
D. merriami seemed to prefer areas where the
soil was more compact or gravelly and the vege-
tative cover was greater.

Dr. William B. Davis trapped at the
Johnson Ranch and the mouth of Santa Elena
Canyon in the early 1940' s and his description
of the vegetation there is completely different
from what these areas are like today. Accord-
ing to Davis (pers. comm.), the river bank at
the Johnson Ranch was a very open sandy area
and was used as a river crossing point for
Mexicans and cattle. The mouth of Santa Elena
Canyon, according to Davis, was also a sandy,
open area with a considerable growth of
Baoaharis. Davis collected Ord's kangaroo rat
at both of these locations. Today, the Johnson
Ranch and the mouth of Santa Elena Canyon are
more like mesquite forests with very little
open terrain. In over 1,500 trap nights at
these two locations, not a single D. ordii was
captured, although 15 D. merriami (two at
Johnson Ranch and 13 at the mouth of Santa
Elena Canyon) were collected. In fact, after
examining the entire riparian corridor in BBNP,
the only place which seeminly had suitable
habitat for D. ordii was the Gaughing Station.
Trapping at this site (720 trap nights), however,
produced 15 D. merriami and no D. ordii. D.
ordii attenuates now seems to be confined to
the first and second floodplain stages of

Tornillo Creek in BBNP and no longer occurs
along the river or at the type locality.

Another mammal affected by vegetative
changes in the riparian corridor is the beaver
which, more than any other mammal in BBNP, is
dependent on the riparian corridor for food
and shelter. Beaver along the Rio Grande
utilize a variety of plants including cane,
seepwillow, willow, and cottonwood. Cotton-
woods occur today only in park service nurseries
at Rio Grande Village and Cottonwood Campground;
salt cedars are rapidly replacing cottonwoods
and willows at other sites. For example, the
Gauging Station is one of the few areas where
extensive stands of willow still exist, and
these are currently being used as a food
source by beavers. Nowhere along the river
corridor is there any evidence of beaver using
salt cedar. As a result, beavers are literally
eating themselves out of "house and home"
because they utilize willow saplings for food
and leave only salt cedar saplings which they
will not use. Taylor et at. (1944) reported a
beaver population of approximately 100 indivi-
duals for the river corridor. However, con-
versations with park personnel and the evident
lack of beaver sign along most of the river
indicate the beaver population today is well
below the figure reported by Taylor et at.
(1944).

The Shannon-Weaver Index of Diversity
(Odum 1971) was used to compare the rodent
fauna of the riparian community with that of
the woodland, grassland, and desert shrub
communities in BBNP. This index reveals infor-
mation concerning the stability of a community
in terms of its fauna. Compared to the other
plant communities, the riparian community has
the lowest evenness, richness, and diversity

215

Table 2. — Shannon-Weaver Index of diversity for the terrestrial rodent fauna of the four major
plant communities in BBNP.

Communities

1 2 2 2

Parameters Riparian Desert-shrub Grassland Woodland

Diversity (H)

1.157

2.008

2.249

849

Evenness (e)

.465

.783

.793

771

Richness (d)

1.523

1.854

2.912

596

No. of Species

Data from Schmidly et al. (1976a, table 20, p. 94).
Data from Baccus (1971, table 10, p. 49).

indices (Table 2). In particular, the evenness
value (0.465) for the riparian community is
considerably lower than that of the other
communities, indicating that one or two species
tend to dominate the rodent fauna of this com-
munity. This is evident when examining the
total catch figures along the riparian corridor.
The two dominant species of the riparian com-
munity are Perognathus penicillatus, with a
total of 924 individuals or 67.7% of the total
catch, and Peromyscus leucopus, with a total of
162 individuals or 11.9% of the total catch.
The grassland is the most diverse community,
having the highest diversity, evenness, and
richness indices as well as the greatest number
of species (17). The desert-shrub, although it
only has 13 species (one more than the riparian
community) , is a more diverse community because
it has a more even distribution, which is indi-
cated by the fact that the dominant species
(Perognathus penicillatus) in this community
accounts for only 38.6% of the total catch as
compared to 67.7% for the riparian community.

IMPACTS IN THE RIPARIAN CORRIDOR

In recant years, many riparian areas
along the Rio Grande have been impacted by human
activity. Around El Paso and Presidio, man
has destroyed or greatly altered natural riparian
natural habitats through water salvage, cultiva-
tion and grazing. The International Boundary
and Water Commission is presently considering
a boundary restoration project along the Rio
Grande from Fort Quitman (Hudspeth County) to
Presidio (Presidio County). This project would
straighten the channel of the river and result
in the virtual destruction of riparian habitats
along this stretch of the Rio Grande.

Human use (floating and camping) and tres-
pass livestock grazing are the major impacts
acting upon the natural riparian communities in
BBNP today. In 1975 the Rio Grande accounted

for 49% of the total backcountry use (in man-
days) in BBNP (Ditton et al. 1976). Twenty-
five percent of this use was float trips on the
Rio Grande and 24% involved camping at primitive
sites along the River Road. Schmidly et al.
(1976b) used correlation analysis to investigate
the relationship among human use, impacts, and
biological parameters (i.e., rodent fauna and
vegetation) at 18 riparian sites in BBNP. Their
results revealed a positive and significant
relationship between total subjective human
impact ratings and annual camping use by site
(man-days). However, the extent of human impact
did not correlate significantly with rodent den-
sities or vegetative parameters at the 18 samp-
ling sites. Thus, correlation analysis revealed
that site impacts have occurred as a result of
recreational use, but not to the point where
ecological conditions, as indicated by the bio-
logical health of the rodent fauna and vegeta-
tion, are in jeopardy (Schmidly et al. 1976b) .

Domestic mammals also occur in the riparian
corridor and pose a major problem. The increase
in grasses over the past 30 years has provided
forage that is not available in the same quan-
tity or quality across the river in Mexico. As
a result, trespass livestock from Mexico are
invading the riparian corridor in increasing
numbers. Grazing by trespass livestock is a
constant feature of almost all riparian sites
and is not confined to one particular region or
section of the river. Should this grazing activ-
ity continue to increase, it could have dangerous
repercussions on the existing vegetation of the
riparian corridor. Hence, dealing with the
livestock problem may prove more difficult for
park managers than dealing with human use and
impacts which tend to be concentrated in some
areas and virtually absent in others.

CONCLUSIONS
Analysis of small mammals, vegetation,

216

and impacts along the Rio Grande in BBNP has
led to five important conclusions: (1) Major
vegetative changes (including the replacement
of cottonwoods and willows by salt cedar as
well as a tremendous increase in basal and
canopy cover) have occurred over the past 30
years. (2) These vegetational changes have
resulted in an alteration of the rodent fauna
so that certain species which were once rare
in riparian habitats (i.e., cricetids such as
Sigmodon hispidus and Peromyscus leuaopus) have
increased their numbers and ranges along the
river, whereas other rodents which were once
common (i.e, Dipodomys ordii) no longer exist
in the riparian corridor. (3) The increase in
vegetative cover, especially grasses, has
caused a reinvasion of domestic livestock (tres-
pass livestock from Mexico) into the riparian
corridor and this may have potentially serious
repercussions on the vegetation. (4) Impacts
at certain riparian sites have occurred as a
result of recreational use, but not to the point
where ecological conditions are in jeopardy.
Human impacts seem to be confined to areas of
convenient access. (5) The riparian community
(as revealed by Shannon-Weaver Index) is less
stable than the other major communities in BBNP
and possibly would be more susceptible to great-
er oscillations resulting from increased impacts
(either human or livestock).

ACKNOWLEDGMENT

This paper is based on research conducted
by the authors as part of research contract No.
CX70050442 for the Office of Natural Resources,
Southwest Region, National Park Service, Santa
Fe, New Mexico.

LITERATURE CITED

Baccus , J. T. 1971. The influence of a return
of native grasslands upon the ecology and
distribution of small rodents in Big Bend
National Park. Unpublished Ph.D. disser-
tation. North Texas State Univ. , Denton,
114 pp.

Bailey, V. 1905. Biological survey of Texas.
N. Amer. Fauna, 25:1-222.

Borell, A. E. and M. D. Bryant. 1942. Mammals
of the Big Bend region of Texas. Univ.
California Publ., Zool. , 48:1-62.

Bryant, M. D. 1939. A new kangaroo rat of

the D. ordii group from the Big Bend region
of Texas. Occas. Papers Mus. Zool., Louis-
iana State Univ., 5:65.

Denyes, H. 1956. Natural terrestrial communi-
ties of Brewster Co., Texas, with special
reference to the distribution of mammals.
Amer. Midland Nat., 55:289-320.
Ditton, R. B., D. J. Schmidly, W. J. Boeer, and
A. R. Graefe. 1976. A survey and analysis
of recreational and livestock impact on
the riparian zone of the Rio Grande in Big
Bend National Park. Proceedings: River
Recreation Management and Research Sympo-
sium, U.S.D.A. Forest Service General
Technical Report NC-28 : 256-266 .

Dixon, K. L. 1958. Spatial organization in
a population of Nelson pocket mouse.
Southwestern Nat., 3:107-113.

Easterla, D. A. 1973. Ecology of the 18

species of chiroptera at Big Bend National
Park, Texas. The Northwest Missouri State
University Studies. Northwest Missouri
State Univ. , 34:1-165.

Johnson, M. S. 1936. Preliminary report on
wildlife survey of Big Bend National Park
(proposed), Texas. Ms., p. 1-46 (National
Park Service) .

Odum, E. P. 1971. Fundamentals of Ecology.
W. B. Saunders Co., Philadelphia, ^63 pp.

Porter, R. D. 1962. Movement, populations,
and habitat preferences of three species
of pocket mice (Perognathus) in the Big
Bend region of Texas. Unpublished Ph.D.
dissertation. Texas A&M Univ. , College
Station, Texas, 255 pp.

Schmidly, D. J., and R. B. Ditton. 1976a.

A Survey and analysis of recreational and
livestock impacts on the riparian zone of
the Rio Grande in Big Bend National Park.
Report prepared for the Office of Natural
Resources, Southwest Region, National Park
Service, Santa Fe , New Mexico (Contract
No. CX70050442) , 160 pp.

Schmidly, D. J., R. B. Ditton, W. J. Boeer, and
A. R. Graefe. 1976b. Inter-relationships
among visitor usage, human impact, and the
biotic resources of the riparian ecosystem
in Big Bend National Park. Paper presented
at the Proc. of First Conference on Scien-
tific Research in the National Parks, New
Orleans, Louisiana, November.

Sperry, 0. E. 1938. A checklist of the ferns,
gymnosperms, and flowering plants of the
proposed Big Bend National Park. Sul
Ross State Teachers College Bull., 19(4):
10-98.

Taylor, W. P., W. B. McDougall, and W. B. Davis.
1944. Preliminary report of an ecological
survey of Big Bend National Park. Report
of work accomplished March-June, 1944.
Administrative report submitted by the
Fish and Wildlife Service to the National
Park Service, 9:55 pp (mimeo) .

217

h tI Qi H

O co H
CX O «0

(0 E-t H

!>• < -fa

01 -H XI

e 3 c

0) C CO

0) C C u-.

CO ,C -H
4J w 3

W E D -H

N 3 p£ H

O tH <-t

a. to ■ o

E o .c o
h & a

B QJ 4-1

>* H )-i

• to o

^ a) 6 3 C

a oj to w

« >. V3 s

h U 3 W Q)

O M 4-1

a. ai * oi to

cd > co ^, ^

ex -h 3 to to

o w a

J3 CO tn 60

to a OS
a. to

•H S3 X)

jC QJ 4-1 *

a. oi to co

« >, co B

H 4-j 3 co oi

01 M 4J

a. a> * o> to

ffl > to >>

cx -h 3 to to

4J 3 0) O •

■o « to a u ui

CO -O >-i C

J-l 3 4-1

O C tH CX rH

DS O U Oi

•H < r- ir\

• 4J t— I O

p5 nj * (N 00

o co u <r o

CO 01 3 I rH

"2 E

3 -H
cO T3

CO CD CO cO

0 n a

CD 01 -rH CJ

W k £ -H

0) o> -o c o

■m 3 w c o 4-j

C O* CO -H CO

01 -H C 4J -H

CX CO CO > J-i 3

0) 01 - 4J
H MO S
CO CO

t3 S ID 01 tO
3 -H S 0) a
tO T3 01 0> "H

0) 0) >n XI 3 O
J-! 3 CO 3 O 4J

Q* to td > U C

O C "H CXrH 0) -H C

■ 4J rH O

OS CO * CM CO

> c

" u o -

c m w n •

o to u -j- o

tO CD 3 I H

3 P H 2 O

j= a— ■ 2 u

CO C O - rfl X i-l
01 3 -H CO CtJ

4J xt QJ

0) 4-1 CO (h -H 4-1

3 3 CO O) 01

H O CO 4J xt 00

£ H H C HI

CO O CO CO >

o *

c • to

co . a c

4J s 0) -h

U 3 Di rH

O -H rH

a- co • o

S o ,c o

H O O

S CD 4J

3 4-1

0) -H T3

B 3 C

QJ 3 cO

00 'H

CO 4J QJ

3 3 MH

cO O -H

B O i— I

• a 3

S 0) -h

3 a: rH

co B 3 c

CJ QJ 3 fO

CO -H 60 -H

•H 4-1 flj 4J Q)

co 3 3 ^

4J >H (0 O -iH

MH tO QJ

3 CO X

O 01 4-1

a u

•H ■• U U

01 V4 < CO

4J co o Ptf
CX CO

'H D O

•"5 r-» u

u r>- o
■ (!'0\ W

<C 0) rH

0) OJ " 4J
rH 60 0> S

ffl tO
P 3 >»

' 0) « 0) CO

> cO >,

"H C n) M

4-1 3 QJ O

I XJ 3 3

OJ . cO

4-> CO -H 4J

tO Xt U 3

I -H OJ CO 0)

u oi cx i-i

O 3 *H "

i-i 3

H <C co

CO O PS

a to

■H 3 O

OJ OJ * 4-1

rH 60 ON S

cO (0

a 3 ^

U 4J 3 CO 0)

01 J-i 4J

a oi - qj co

fO > CO

Di'H C ffl ffl

> 4J CD -rl

CO XI
•H 0) CO 0)

o qj a. n

rH ^

o c -h a r

• h o »

3 oi to cn

O (0 CJ «st

(0 0) 3 I ■

0) >n XI C O

BO * ffl rC

0< to CO > 3
„. «H O -H U O

tO U (0 tO >

CO 01 3

b i-i h ;

43 CX — I

01 01 >* XI 3 O *H
4J 3 CO 3 O 4-1 U

to c o » cd j:

QJ 3 *H tO rH

l_l 4-1 QJ 01 *

cx to CO > J-i c

QJ 4J CO U -H 4-1

^U. S. GOVERNMENT PRINTING OFFICE 1977-781-058/96 Reg. 8

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "Importance, preservation and management of riparian habitat: a symposium, Tucson, Arizona, July 9, 1977"

See other formats