USGS Circular 1282 C1282

User Manual: 1282

Open the PDF directly: View PDF PDF.
Page Count: 232 [warning: Documents this large are best viewed by clicking the View PDF Link!]

53$EPARTMENTOFTHE)NTERIOR
53'EOLOGICAL3URVEY
86*6&LUFXODU
7KH6WDWHRIWKH
&RORUDGR5LYHU
(FRV\VWHPLQ
*UDQG&DQ\RQ
$5HSRUWRIWKH
*UDQG&DQ\RQ
0RQLWRULQJDQG
5HVHDUFK&HQWHU
²
6RXWKZHVW%LRORJLFDO
6FLHQFH&HQWHU
The State of the Colorado River
Ecosystem in Grand Canyon
A Report of the Grand Canyon
Monitoring and Research Center
1991–2004
Edited by Steven P. Gloss, Jeffrey E. Lovich, and Theodore S. Melis
USGS Circular 1282
U.S. Department of the Interior
U.S. Geological Survey
U.S. Department of the Interior
Gale A. Norton, Secretary
U.S. Geological Survey
Patrick Leahy, Acting Director
U.S. Geological Survey, Reston, Virginia: 2005
For product and ordering information:
World Wide Web: http://www.usgs.gov/pubprod
Telephone: 1-888-ASK-USGS
For more information on the USGS--the Federal source for science about the Earth, its natural and living resources,
natural hazards, and the environment:
World Wide Web: http://www.usgs.gov
Telephone: 1-888-ASK-USGS
Any use of trade, product, or firm names is for descriptive purposes only and does not imply endorsement by the
U.S. Government.
Although this report is in the public domain, permission must be secured from the individual copyright owners to
reproduce any copyrighted materials contained within this report.
Suggested citation:
Gloss, S.P., Lovich, J.E., and Melis, T.S., eds., 2005, The state of the Colorado River ecosystem in Grand Canyon: U.S.
Geological Survey Circular 1282, 220 p.
Library of Congress Cataloging-in-Publication Data
The state of the Colorado River ecosystem in Grand Canyon : a report of the Grand Canyon Monitoring and Research
Center / Steven P. Gloss, Jeffrey E. Lovich, Theodore S. Melis, editors.
p. cm.
Includes bibliographical references and index.
1. Stream ecology--Arizona--Grand Canyon. 2. Stream ecology--Colorado River (Colo.-Mexico) I. Gloss, Steven. II.
Lovich, Jeffrey E. III. Melis, Theodore S. IV. Grand Canyon Monitoring and Research Center.
QH105.A65S73 2005
577.6’4’0979132--dc22
2005018333
This report is a scientific product of the U.S. Geological Survey. As such, it will be an important ele-
ment in informing the policy dialogue for decisionmakers and stakeholders involved with or interested
in operations of Glen Canyon Dam and the protection of downstream resources of Grand Canyon
National Park. Like all scientific documents, however, it will be only one element of the policy dialogue.
Ultimately, many other factors will also be considered by decisionmakers when they formulate official
policy governing the operation of Glen Canyon Dam.
iii
As a “larger than life” Director of the
U.S. Geological Survey and the first
person known to have successfully rafted
the Colorado River through Grand
Canyon, John Wesley Powell sent out a
call to raise science aloft, a call that has
particular resonance for the Glen
Canyon Dam Adaptive
M
anagement Program. In
Grand Canyon, science
offers a means of under-
standing and predict-
ing the relationships
between the opera-
tions o
f Glen Canyon
Dam and downstream
resources of
concern.
This fact was recog-
nized by both the Grand
Canyon Protection A
ct of
1992 and the final environ-
mental impact statement that proposed the
Adaptive Management Program. Moni-
toring and research were selected as the
tools to allow scientists to unravel the
many uncertainties that existed, and con-
tinue to exist, about downstream impacts
from dam operations.
Significantly, science within the context
of adaptive management is intended to
serve management and policy. Scientists
are responsible for developing relevant
information, and river managers are
responsible for making resource decisions
by using the best information available.
When scientists and managers work
together, science can be the olive branch
of peace and emblem of hope needed to
mitigate the adverse effects of dam opera-
tions and improve the values for which
Glen Canyon National Recreation Area
and Grand Canyon National Park were
established. These are the wishes of the
American people as expressed in the
Grand Canyon Protection Act of 1992.
The following chapters summa-
rize a decade of monitoring
and research activities for
many key resources in
the Colorado River
corridor below Glen
C
anyon Dam. Where
possible, scientists assess
the effects of dam opera-
tions, par
ticularly the
modified low fluctuat-
ing flow alternative,
on given resources and
highlight the linkages
among system features that
managers identified as important.
The role that John Wesley Powell envi-
sioned for science in 1882 reflects the
highest goals of the scientists and other
professionals of the U.S. Geological
Survey today. In keeping with this vision,
The State of the Colorado River Ecosystem in
Grand Canyon is emblematic of the high
quality science that the U.S. Geological
Survey is committed to providing to
its customers. Science of the type
reported here, which can be used to
make informed decisions, is the return
on investment that American taxpayers
deserve and appreciate.
P. Patrick Leahy, Ph.D.
Acting Director
U.S. Geological Survey
Foreword
“Let us not
gird science to our
loins as the warrior
buckles on his sword. Let
us raise science aloft as the
olive branch of peace and the
emblem of hope.”
—John Wesley Powell,
1882, p. 70
Powell, J.W., 1882, Darwin’s contribution to philosophy, in Addresses delivered on the occasion of the
Darwin memorial meeting, May 12, 1882: Washington, D.C., Biological Society of Washington, p. 60–70.
iv
v
Project Staff
Editors
Steven P. Gloss, Ph.D.
U.S. Geological Survey
Southwest Biological Science Center
Tucson, Arizona
Jeffrey E. Lovich, Ph.D.
U.S. Geological Survey
Southwest Biological Science Center
Flagstaff, Arizona
Theodore S. Melis, Ph.D.
U.S. Geological Survey
Southwest Biological Science Center
Flagstaff, Arizona
Project Coordinator
Lara M. Schmit
Northern Arizona University
Center for Sustainable Environments
Flagstaff, Arizona
Research Assistant
Christopher N. Updike
Northern Arizona University
Center for Sustainable Environments
Flagstaff, Arizona
USGS National Wetlands
Research Center
Production Staff
Tammy Charron, Victoria Chachere Jenkins,
Natalie Gormanous Trahan, Ann Gaygan:
IAP World Services;
Beth Vairin: USGS
Additional production support
Gaye Farris and Rhonda Davis: USGS;
Connie Herndon and Jarita Davis:
IAP World Services
J
e
ff
S
o
r
e
n
s
e
n
,
A
r
i
z
o
n
a
G
a
m
e
a
n
d
F
i
s
h
D
e
p
a
rt
m
e
n
t
Special Acknowledgment
The assistance of Stephanie Wyse Mietz,
librarian, U.S. Geological Survey, Grand
Canyon Monitoring and Research Center;
Dale Blank, network administrator,
U.S. Geological Survey, Grand Canyon
Monitoring and Research Center; David
J. Topping, research hydrologist, U.S.
Geological Survey, National Research
Program, Water Resources Discipline; Scott
A. Wright, hydrologist, U.S. Geological
Survey, Grand Canyon Monitoring and
Research Center; and Thomas M. Gushue,
GIS coordinator, U.S. Geological Survey,
Grand Canyon Monitoring and Research
Center, is gratefully acknowledged.
Peer and Technical
Reviewers
The content of this report was enhanced by
the wisdom and research of many colleagues
and cooperators. In particular, we gratefully
recognize the indepth and timely technical
reviews by the Glen Canyon Dam Adaptive
Management Program Science Advisors and
individual peer reviewers.
vi
Ted Angradi, Ph.D.
U.S. Environmental Protection Agency
Denver, Colorado
Jill Baron, Ph.D. (Science Advisor)
U.S. Geological Survey
Fort Collins, Colorado
Bryan Brown, Ph.D.
SWCA Environmental Consultants, Inc.
Salt Lake City, Utah
David E. Busch, Ph.D.
U.S. Geological Survey
Portland, Oregon
Steven W. Carothers, Ph.D.
SWCA Environmental Consultants, Inc.
Flagstaff, Arizona
Bonnie G. Colby, Ph.D.
University of Arizona
Tucson, Arizona
David Cole, Ph.D.
Aldo Leopold Wilderness Research Institute
Missoula, Montana
Chip Colwell-Chanthaphonh, Ph.D.
Center for Desert Archaeology
Tucson, Arizona
Kenton R. Corum, Ph.D.
Northwest Power and Conservation Council
Portland, Oregon
Virginia Dale, Ph.D. (Science Advisor)
Oak Ridge National Laboratory
Oak Ridge, Tennessee
Gregg Garfin, Ph.D.
University of Arizona
Tucson, Arizona
L.D. Garrett, Ph.D.
(Executive Director of Science Advisors)
M3 Research
Olathe, Colorado
Randy Gimblett, Ph.D.
University of Arizona
Tucson, Arizona
Brian Graeb
South Dakota State University
Brookings, South Dakota
Gordon E. Grant, Ph.D.
U.S. Department of Agriculture
Forest Service
Corvallis, Oregon
Al Groeger, Ph.D.
Texas State University
San Marcos, Texas
Lance Gunderson, Ph.D.
(Science Advisor)
Emory University
Atlanta, Georgia
Steve Gutreuter, Ph.D.
U.S. Geological Survey
La Crosse, Wisconsin
Joel R. Hamilton, Ph.D.
University of Idaho (emeritus)
Moscow, Idaho
Judson W. Harvey, Ph.D.
U.S. Geological Survey
Reston, Virginia
Alan Howard, Ph.D. (Science Advisor)
University of Virginia
Charlottesville, Virginia
W. Carter Johnson, Ph.D.
South Dakota State University
Brookings, South Dakota
Pierre Y. Julien, Ph.D.
Colorado State University
Fort Collins, Colorado
James Kitchell, Ph.D. (Science Advisor)
University of Wisconsin
Madison, Wisconsin
G. Richard Marzolf, Ph.D.
U.S. Geological Survey (retired)
Berryville, Virginia
Barbara J. Mills, Ph.D.
University of Arizona
Tucson, Arizona
vii
Craig J. Palmer, Ph.D.
University of Nevada, Las Vegas
Las Vegas, Nevada
Margaret Palmer, Ph.D. (Science Advisor)
University of Maryland
College Park, Maryland
Craig Paukert, Ph.D.
U.S. Geological Survey
Manhattan, Kansas
Bruce Peacock, Ph.D.
National Park Service
Fort Collins, Colorado
Roger Pulwarty, Ph.D.
National Oceanic and Atmospheric
Administration–Cooperative Institute for
Research in Environmental Sciences
Climate Diagnostics Center
Boulder, Colorado
J. Jefferson Reid, Ph.D.
University of Arizona
Tucson, Arizona
Bruce L. Rhoads, Ph.D.
University of Illinois
Urbana, Illinois
Catherine A. Roberts, Ph.D.
College of the Holy Cross
Worcester, Massachusetts
Dale Robertson, Ph.D. (Science Advisor)
U.S. Geological Survey
Middleton, Wisconsin
Douglas Schwartz, Ph.D. (Science Advisor)
School of American Research
Santa Fe, New Mexico
Michael L. Scott, Ph.D.
U.S. Geological Survey
Fort Collins, Colorado
Francisco J.M. Simões, Ph.D.
U.S. Geological Survey
Denver, Colorado
Kristin E. Skrabis, Ph.D.
U.S. Department of the Interior
Washington, D.C.
Charles R. Smith, Ph.D.
Cornell University
Ithaca, New York
Alan P. Sullivan, Ph.D.
University of Cincinnati
Cincinnati, Ohio
David Tarboton, Ph.D.
Utah State University
Logan, Utah
Todd Tietjen, Ph.D.
Mississippi State University
Mississippi State, Mississippi
Mark R. Vinson, Ph.D.
Utah State University
Logan, Utah
Joe Watkins, Ph.D. (Science Advisor)
University of New Mexico
Albuquerque, New Mexico
Michael Welsh, Ph.D.
Monona, Wisconsin
Ellen Wohl, Ph.D.
Colorado State University
Fort Collins, Colorado
viii
Contents
Overview.........................................................................................................................................................1
Lara M. Schmit, Steven P. Gloss, and Christopher N. Updike
Chapter 1. Influence of Glen Canyon Dam Operations on
Downstream Sand Resources of the Colorado River in Grand Canyon ...........................................17
Scott A. Wright, Theodore S. Melis, David J. Topping,
and David M. Rubin
Chapter 2. Fishes of Grand Canyon ...................................................................................................... 33
Steven P. Gloss and Lewis G. Coggins
With text box by Jeffrey E. Lovich ....................................................................................................50
Chapter 3. Climatic Fluctuations, Drought, and Flow
in the Colorado River .................................................................................................................................. 57
Robert H. Webb, Richard Hereford, and Gregory J. McCabe
Chapter 4. Water Quality in Lake Powell and the Colorado River ...................................................69
William S. Vernieu, Susan J. Hueftle, and Steven P. Gloss
Chapter 5. Aquatic Ecology: the Role of Organic Matter
and Invertebrates........................................................................................................................................87
Theodore A. Kennedy and Steven P. Gloss
Chapter 6. Riparian Vegetation and Associated Wildlife ................................................................103
Barbara E. Ralston
With text box by Charles Drost .......................................................................................................116
Chapter 7. Birds of the Colorado River in Grand Canyon:
a Synthesis of Status, Trends, and Dam Operation Effects ............................................................... 123
Jennifer A. Holmes, John R. Spence, and Mark K. Sogge
Chapter 8. Debris Flows in Grand Canyon and the Rapids
of the Colorado River ................................................................................................................................ 139
Robert H. Webb, Peter G. Griffiths, Christopher S. Magirl,
and Thomas C. Hanks
Chapter 9. Recreation Use Values and Nonuse Values
of Glen and Grand Canyons.....................................................................................................................153
John Loomis, Aaron J. Douglas, and David A. Harpman
ix
Chapter 10. Status and Trends of Hydropower Production
at Glen Canyon Dam ................................................................................................................................. 165
David A. Harpman and Aaron J. Douglas
Chapter 11. Cultural Resources in the Colorado River Corridor ....................................................177
Helen C. Fairley
Chapter 12. Recreational Values and Campsites
in the Colorado River Ecosystem ...........................................................................................................193
Matt Kaplinski, Jeff Behan, Joseph E. Hazel, Jr.,
Roderic A. Parnell, and Helen C. Fairley
Chapter 13. Lessons from 10 Years of
Adaptive Management in Grand Canyon .............................................................................................. 207
Jeffrey E. Lovich and Theodore S. Melis
J
e
f
f
S
o
r
e
n
s
e
n
,
A
r
i
z
o
n
a
G
a
m
e
a
n
d
F
i
s
h
D
e
p
a
r
t
m
e
n
t
Overview
Lara M. Schmit
Steven P. Gloss
Christopher N. Updike
Introduction
This report is an important milestone in the effort
by the Secretary of the Interior to implement the Grand
Canyon Protection Act of 1992 (GCPA; title XVIII, secs.
1801–1809, of Public Law 102-575), the most recent
authorizing legislation for Federal efforts to protect
resources downstream from Glen Canyon Dam. The
chapters that follow are intended to provide decision
makers and the American public with relevant scientific
information about the status and recent trends of the
natural, cultural, and recreational resources of those
portions of Grand Canyon National Park and Glen
Canyon National Recreation Area affected by Glen
Canyon Dam operations. Glen Canyon Dam is one of
the last major dams that was built on the Colorado River
and is located just south of the Arizona-Utah border
in the lower reaches of Glen Canyon National Recre-
ation Area, approximately 15 mi (24 km) upriver from
Grand Canyon National Park (fig. 1). The information
presented here is a product of the Glen Canyon Dam
Adaptive Management Program (GCDAMP), a federally
authorized initiative to ensure that the primary mandate
of the GCPA is met through advances in information
and resource management. The U.S. Geological Survey’s
(USGS) Grand Canyon Monitoring and Research
Center (GCMRC) has responsibility for the scientific
monitoring and research efforts for the program, includ-
ing the preparation of reports such as this one.
The Study Area
Carved from the Earth by the Colorado River,
Grand Canyon is a natural wonder that is “absolutely
unparalleled throughout the rest of the world,” as
President Theodore Roosevelt said upon seeing it for the
first time in 1903 (Roosevelt, ca. 1905, p. 369). Consid-
ered one of the world’s most spectacular gorges, Grand
Canyon exhibits a depth of more than 6,720 ft (2,048
m) at its most extreme in Granite Gorge (Annerino,
2000). The colorful strata of the canyon’s walls also
reveal an invaluable record of the Earth’s geologic his-
tory dating back to the 1.84-billion-yr-old rock forma-
tions found at Elves Chasm, which are the oldest rocks
known in the Southwestern United States (Beus and
Morales, 2003). President Woodrow Wilson signed the
2 The State of the Colorado River Ecosystem in Grand Canyon
bill that established Grand Canyon as a national park
on February 26, 1919, in recognition of its exceptional
natural beauty and geologic wonders. Grand Canyon
National Park is also of cultural and spiritual significance
to many of the region’s Native Americans and contains
more than 2,600 documented prehistoric ruins, which
span thousands of years and provide an important record
of human adaptation to an arid environment. In addi-
tion to its geologic and cultural significance, the Grand
Canyon ecosystem is home to a diverse array of plants
and animals such as the humpback chub (Gila cypha) and
the southwestern willow flycatcher (Empidonax traillii exti-
mus), both of which are species that are federally listed as
endangered. Because of its global significance as a natural
and cultural treasure, Grand Canyon National Park was
inscribed by the United Nations Educational, Scien-
tific and Cultural Organization (UNESCO) as a World
Heritage Site in 1979.
The GCPA (see timeline) directs the Secretary of the
Interior to operate Glen Canyon Dam and exercise other
authorities “in such a manner as to protect, mitigate
adverse impacts to, and improve the values for which
Grand Canyon National Park and Glen Canyon National
Recreation Area were established, including, but not
limited to natural and cultural resources and visitor use”
(GCPA, sec. 1802(a)). As a result, the Glen Canyon Dam
Adaptive Management Program, created by the 1996
Record of Decision (ROD) for the operation of Glen
Canyon Dam, focuses on a study area that encompasses
the Colorado River corridor from Glen Canyon Dam to
the western boundary of Grand Canyon National Park.
The study area includes the approximately 15 river miles
(RM) of river from the dam to Lees Ferry within Glen
Canyon National Recreation Area and the entire 277-
RM river corridor below Lees Ferry and within Grand
Canyon National Park. In total, the study area includes
some 293 RM of the Colorado River (fig. 1).
Administrative History
The Colorado River is the most important water
resource in the American West, serving as the main
source of drinking water for more than 25 million people
(Water Education Foundation, 2001). The Colorado
River has been extensively engineered to meet the
demands placed upon it (see timeline). There are 22
major storage reservoirs in the Colorado River Basin
and 8 major out-of-basin diversions (Pontius, 1997).
The two largest storage projects—Hoover and Glen
Canyon Dams—are located on either end of Grand
Canyon National Park. Glen Canyon Dam is located
just north of the Grand Canyon National Park bound-
ary, where it creates Lake Powell. At full capacity, Lake
Powell was designed to hold 27 million acre-feet (maf)
(>33,000 million m3) of water and is the key storage unit
within the Colorado River Storage Project (CRSP) (U.S.
Department of the Interior, 1970).
Signed into law by President Dwight D. Eisenhower
in 1956, the Colorado River Storage Project Act
authorized four mainstem water-storage units, includ-
ing Glen Canyon Dam. Construction of Glen Canyon
Dam began on September 29, 1956, and the last bucket
of concrete was poured on September 13, 1963 (U.S.
Department of the Interior, 1970). The regulation
of the Colorado River by Glen Canyon Dam began
with the closure of the dam in 1963 and when Lake
Powell began filling. The CRSP reservoirs allow the
upper basin States—Utah, Colorado, Wyoming, and
New Mexico—to store water in wet years and release
water in times of shortages, thereby enabling the upper
basin to meet its obligations under the 1922 Colorado
River Compact while also maximizing future water uses
(Ingram and others, 1991). To repay Federal expendi-
tures for the water-storage units and supplement the costs
of related irrigation units, CRSP dams were equipped
with hydroelectric generators to produce salable power.
Glen Canyon Dam operates eight electric generators,
which produce 78% of the total power generated by the
CRSP (Hughes, 1991). In 2004, Glen Canyon Dam gen-
erated approximately 3.3 million megawatthours (MWh).
The power is sold to approximately 200 wholesale
customers—municipal and county utilities, rural electric
cooperatives, U.S. Government installations, and other
nonprofit organizations—located primarily in six States:
Arizona, Colorado, Utah, Wyoming, New Mexico, and
Nevada (National Research Council, 1996).
Natural History
Before the dam, the Colorado River was a sediment-
rich river that when swelled with snowmelt from the
Rocky Mountains transported large quantities of sedi-
ment during spring and early summer and commonly
produced flood events. Peak discharge typically reached
85,000 cubic feet per second (cfs) at 2-yr intervals and
120,000 cfs at 6-yr intervals during these seasonal flood
events (Topping and others, 2003). By contrast, flows of
less than 3,000 cfs were typical during late summer, fall,
and winter. Prior to the dam, water temperature also
Overview 3
Figure 1. Study area.
4 The State of the Colorado River Ecosystem in Grand Canyon
fluctuated seasonally from 32°F to 80°F (0–29°C) (U.S.
Department of the Interior, 1995).
Glen Canyon Dam has changed the seasonal flow,
sediment-carrying capacity, and temperature of the
Colorado River. Operation of the dam has altered the
frequency of floods on the Colorado River and increased
median discharge rates at Lees Ferry, whereas managing
for hydroelectric power generation has introduced wide-
ranging daily fluctuations (Topping and others, 2003).
For example, from 1963 to 1991 (the no action period
or historical operations), when the dam was managed
primarily to maximize hydroelectric power revenue, it
was not uncommon for daily flows to vary from 5,000
to 30,000 cfs (U.S. Department of the Interior, 1988).
Release patterns of this type caused the river level below
the dam to change 7–13 ft (2–4 m) per day, creating pub-
lic concerns about the quality and safety of fishing and
boating and about adverse impacts to natural resources
(U.S. Department of the Interior, 1988). Because the
sediment load of the Colorado River is deposited in
Lake Powell, water released from Glen Canyon Dam is
essentially clear. Furthermore, because the penstocks of
the dam are well below the surface of Lake Powell, the
water released from the dam is cold, with an average
temperature of 46°F (8°C) (Webb and others, 1999).
The construction of Glen Canyon Dam also
affected a number of aquatic and terrestrial resources
downstream in lower Glen and Grand Canyons. Dam-
induced changes in the Colorado River’s flow, tempera-
ture, and sediment-carrying capacity are blamed for
narrowing rapids, beach erosion, invasion of nonnative
riparian vegetation, and losses of native fishes (Webb and
others, 1999). These same changes are also associated
with an increase in total species richness within Grand
Canyon National Park; however, the increases are pri-
marily for species not originally found in Grand Canyon.
Some changes to the ecosystem of the Colorado River,
such as the introduction of nonnative fish, were already
taking place before the construction of Glen Canyon
Dam (Wieringa and Morton, 1996).
It is important to note that Glen Canyon Dam
was completed before the enactment of the National
Environmental Policy Act of 1969 and the Endangered
Species Act of 1973 (see timeline). At the time of Glen
Canyon Dam’s construction (1956–63), little consider-
ation was given to how dam operations might affect the
downstream environment in Grand Canyon National
Park (Babbitt, 1990). Nevertheless, public values were
undergoing a shift: at the same time that Congress autho-
rized Glen Canyon Dam in 1956, authorization of Echo
Park Dam on the Green River was defeated because of
environmental reasons (Ingram and others, 1991).
Federal Efforts to Protect
Grand Canyon
The international prominence of Grand Canyon
National Park and public concern about the impacts of
Glen Canyon Dam caused the Bureau of Reclamation
in 1982 to undertake a science program, Glen Canyon
11,000 BP Paleo-Indian peoples
occupy Grand Canyon region
1869 Major John Wesley
Powell leads first recorded
expedition to traverse
Grand Canyon
1893 President Benjamin
Harrison creates Grand
Canyon Forest Reserve
1908 President
Theodore Roosevelt
creates Grand Canyon
National Monument
1916 National Park
Service Organic Act passed
1902 Reclamation Act creates
the Bureau of Reclamation
Overview 5
Environmental Studies, to examine the effects of dam
operations on downstream resources. Glen Canyon
Environmental Studies, the USGS Grand Canyon
Monitoring and Research Center’s predecessor, issued
a final report in 1988 concluding that changes in dam
operations “could reduce the resource losses occur-
ring under current operations and, in some cases, even
improve the status of the resources” (U.S. Department
of the Interior, 1988, p. xvi). In 1989, in response to
these findings, Secretary of the Interior Manuel Lujan,
Jr., ordered the Bureau of Reclamation to complete an
environmental impact statement on the operation of
Glen Canyon Dam. To further ensure the protection of
downstream resources, Secretary Lujan adopted interim
operating criteria for the dam in 1991, which restricted
dam operations and remained in effect until the end of
the environmental impact statement process.
Congress passed the Grand Canyon Protection Act
of 1992 to provide guidance and legal support to the
Secretary of the Interior in his efforts to protect Grand
Canyon. In addition to directing the Secretary to operate
Glen Canyon Dam to protect and improve downstream
resources, the act also validated the interim operating
criteria, provided a deadline for the completion of the
environmental impact statement, required the creation
of a long-term monitoring and research program, and
allocated program costs. The act clearly stated that it
was to be implemented in accordance with existing laws,
treaties, and institutional agreements that govern alloca-
tion, appropriation, development, and exportation of the
waters of the Colorado River Basin (GCPA, sec. 1802(b)).
The Operation of Glen Canyon Dam Final Envi-
ronmental Impact Statement (hereafter EIS) was filed in
March 1995, and the Record of Decision was signed by
Bruce Babbitt, Secretary of the Interior, in October 1996.
The Record of Decision noted that the goal “was not to
maximize benefits for the most resources, but rather to
find an alternative dam operating plan that would permit
recovery and long-term sustainability of downstream
resources while limiting hydropower capacity and flex-
ibility only to the extent necessary to achieve recovery
and long-term sustainability” (U.S. Department of the
Interior, 1996, p. G-11). Having established this goal, the
Secretary’s decision was to implement the modified low
fluctuating flow (MLFF) alternative (the preferred alter-
native in the EIS) as described in the EIS but with minor
changes in the upramp rate, maximum release rate, and
the timing of beach/habitat-building flows (BHBF; see
below). The document also formally established the Glen
Canyon Dam Adaptive Management Program.
Glen Canyon Dam Adaptive
Management Program
The creation of an adaptive management program
was a common element for all alternatives considered
in the EIS, and its implementation was subsequently
mandated by the Record of Decision. Adaptive man-
agement was selected to create a process whereby “the
effects of dam operations on downstream resources
1919 Grand Canyon
National Park created
1921-23 U.S. Geological Survey’s
Birdseye Expedition surveys possible
dam sites along the Colorado River
1922 Colorado River Compact signed allocating the water of the
Colorado River between the upper and lower basins. Upper basin
States have the right to use 7.5 maf/yr only if that quantity is available
after meeting delivery requirements of 7.5 maf/yr to the lower basin
plus the amount required to satisfy anticipated claims by Mexico
1928 Boulder Canyon Project Act
passed authorizing Hoover Dam
6 The State of the Colorado River Ecosystem in Grand Canyon
would be assessed and the results of those assessments
would form the basis of future modifications of dam
operations” (U.S. Department of the Interior, 1995, p.
34). The selection of adaptive management and the
focus on the effects of dam operations on downstream
resources have significant implications. First, the promi-
nence of Grand Canyon National Park elevates adaptive
management and the GCDAMP to national significance.
Second, the program’s focus on the effects of dam opera-
tions on downstream resources constrains the range of
management options and creates a relatively well-defined
geographic area within which to operate.
Envisioned as a new paradigm for addressing com-
plex environmental management problems through a
dynamic interplay of ecosystem science, management,
and policy, adaptive management has gained attention
and has been tested in various contexts in the last several
decades (National Research Council, 1999). Although
concepts and methods continue to evolve, adaptive
management is generally understood to be a systematic
process for continually improving management practices
by emphasizing learning through experimentation. Also,
adaptive management incorporates collaboration among
stakeholders, managers, and scientists as a means of social
learning that can prevent policy gridlock. In Downstream,
the National Research Council (1999, p. 53) noted that
the key components of adaptive management include
(1) commitment to ongoing management adjust-
ments based, in part, upon scientific experimen-
tation, (2) shift from “trial and error” to formal
experimentation with management actions and
alternatives, (3) shift from fragmented scientific
investigations to integrated ecosystem science,
(4) explicit attention to scientific uncertainties in
ecosystem processes and effects of management
alternatives, (5) formal experimental design and
hypothesis testing to reduce those uncertainties
and help guide management adjustments, (6)
careful monitoring of ecological and social effects
and of responses to management operations, (7)
analysis of experimental outcomes in ways that
guide future management decisions, and (8) close
collaboration among stakeholders, managers,
and scientists in all phases of these processes.
The Role of Science
The Colorado River provides many benefits to
society including numerous natural processes; habitat
for unique organisms such as native fishes; water for
humans, agriculture, and recreational purposes; and
hydroelectric power generation. Science-based status
and trends information is increasingly valuable as soci-
ety attempts to balance the competing uses of natural
resources. The need for credible scientific information
that can serve as a feedback loop between management
actions and the effects of those actions is of critical
importance in adaptive management.
The role of science in the GCDAMP is fourfold:
(1) to provide the aforementioned credible scientific
information about management actions deemed appro-
1935 Hoover
Dam completed
1944 Treaty with Mexico
obligating the United States
to provide 1.5 maf of Colorado
River water to Mexico annually
1946 Robert R. Miller
describes humpback chub
(Gila cypha) from specimens
taken in Grand Canyon
1948 Upper Colorado River
Basin Compact signed
1956 Colorado River
Storage Project Act
passed authorizing
Glen Canyon Dam
2
YHUYLHZ  
ULDWHWRLP
OHPHQWDVH[
HULPHQWVE
WKHVWDNHKROG
HUVDQ
G
PDQDJHUVWRFRQ
G
XFWDQ
G
FRPPXQ
L
FDWH
S
HHUUHYLHZHGUHVHDUFKUHOHYDQWWRPDQD
J
HPHQWGHFL
V
L
RQQHH
G
VDQ
G
WR
E
HWWHUXQ
G
HUVWDQ
G
I
DFWRUVJRYHUQ
L
QJ
S
RWHQWLDOUHV
S
RQVHVWRPDQD
J
HPHQWDFWLRQVWR
S
URYLGH
VF
L
HQW
LIL
FD
OO\
VRXQ
G
DQ
G
G
H
I
HQV
LEO
HH[
S
HU
L
PHQWD
O
G
HV
LJ
QV
IRUPDQDJHPHQWH[SHULPHQWVDQGWRHQVXUHWKDWPRQL
WRU
L
Q
J
S
UR
J
UDPV
\L
H
OG
XVH
I
X
O
L
Q
I
RUPDW
L
RQDQ
G
WR
VWUXFWXUHW
K
HW
L
P
L
QJR
I
PRQ
L
WRU
L
QJDQ
G
UHVHDUF
K
UHVX
O
WV
WRWKHH[WHQW
S
RVVLEOHLQDZD
\
WKDWDIIRUGVWKH$GD
S
WLYH
0DQDJHPHQW:RUN*URXSVHHVHFWLRQRQ&ROODERUDWLRQ
S
DQGWKH
6
HFUHWDU
\
RI WKH,QWHULRUWKHEHVWDYDLODEOH
VF
L
HQW
LIL
F
L
Q
I
RUPDW
L
RQWRFRQV
LG
HU
L
QW
K
H
L
UPDQDJHPHQ
W
GHFLVLRQPDNLQJ
$
VVHVV
L
Q
J
W
K
HVWDWHR
I
N
QRZ
O
H
GJ
HD
E
RXWW
K
HHFR
V\VWHPWREHPDQDJHGDGDSWLYHO\LVDNH\HDUO\DFWLYLW\
R
I
PRVWD
G
D
S
W
L
YHPDQD
J
HPHQW
L
Q
L
W
L
DW
L
YHV7
KL
VW
\S
HR
I
D
VVHVVPHQW
L
VPRVWH
II
HFW
L
YH
O
\
G
RQH
E
\XV
L
QJDFRQFHSWXD
O
P
RGHOWKDWFROODERUDWLYHO
\
HQ
J
D
J
HVVFLHQWLVWVDQGVWDNH
K
R
OG
HUVWRDJUHHRQZ
K
DW
L
V
N
QRZQD
E
RXWSURFHVVHVW
K
DW
R
S
HUDWHZLWKLQD
J
LYHQHFRV
\
VWHPDQGWRH[DPLQH
S
RVVLEOH
L
QWHUDFW
L
RQV
E
\XV
L
QJDFRPSXWHUPR
G
H
O
7
K
HFRPSXWHU
P
RGHOSURYLGHVDFRQFHSWXDOEXWQRWQHFHVVDULO\SUHGLF
W
L
YHFD
S
D
ELOL
W
\
WRFRQV
LG
HUHFRV
\
VWHPUHV
S
RQVHVWRH[
S
HU
L
P
HQWDOPDQDJHPHQWDFWLRQVVHHWH[WER[RQS²
0
R
G
LILH
G
/RZ)
O
XFWXDWLQJ)
O
RZ
$O
WHUQDWLYHDQ
G
([SHULPHQWDWLRQ
,QD
GGL
W
L
RQWRH[DP
L
Q
L
QJW
K
HVWDWXVDQ
G
UHFHQW
WUHQGVRI NH
\
ELROR
J
LFDOFXOWXUDODQGUHFUHDWLRQDO
 'LEN#AN
Y
ON
$
A
M
CO
N
S
TR
UC
T
ED

GALLONSO
F
POISONAPPLIED
T
OMIOFTHE
'
REEN2IVERTOKILLNATIVE
FISHANDESTABLISHATROUTFISHERYRESULTING
INTHEUNINTENTIONALKILLINGO
F
F
ISHESAS
F
AR
D
OWNSTREAMAS$INOSAUR.ATIONAL-ONUMEN
T

5
3

3
UPREME
#
OURTHELDIN !RIZON
A
V
#ALIFORNI
A
THATASARESULTOFTHE"OULDER
#
ANYON0ROJECT!CT
#
ALIFORNIAHELDANALLOCATION
OF
MA
F
!RIZONAMA
F
AND.EVADA
ACREFEETOF#OLORADO2IVERWATER

'
LEN
#
ANYON$AMPOWERGENERATIONSTARTS
.
ATIONAL0ARK
3
ERVICEENDSAYRPRO
G
RAMOF
P
LANTINGRAINBOWTROUT
/
NCORH
Y
NCHUSM
Y
KIS
S
AND
BROWNTROUT
3
ALMOTRUTT
A
EGGSANDFINGERLINGSIN
T
RIBUTARIESOFTHE
#
OLORADO2IVERWITHIN
'
RAND
#
ANYON
UHVRXUFHVWKLVUH
S
RUWH[
S
ORUHVWKHHIIHFWVRI WKHLP
S
OH
PHQWDW
L
RQR
I
W
K
H0/))D
O
WHUQDW
L
YHZ
KL
F
K
VSHF
LIL
HV
GDPR
S
HUDWLRQVXQGHUQRUPDOFRQGLWLRQVDQGLQFOXGHV
H
[SHU
L
PHQWD
O
K
D
EL
WDWPD
L
QWHQDQFH
IO
RZVDQ
G
%+%)V
Z
KHQFHUWDLQFRQGLWLRQVDUHPHW8QGHUQRUPDOFRQGL
W
L
RQV
W
K
H0/))D
O
WHUQDW
L
YHD
OO
RZV
I
RU
G
DPUH
O
HDVHVWR
IOXFWXDWHQRPRUHWKDQFIVSHUGD\DQGJHQHUDOO\
Q
RWWRH[FHH
G
F
I
VH[FH
S
W
G
XU
L
Q
J
S
HU
L
R
G
VR
I
KLJK
UHJ
L
RQD
O
UXQR
II
RU
I
RUH[SHU
L
PHQWD
O
IO
RZVWD
EO
H,Q
DGGLWLRQWKH0/))DOWHUQDWLYHFRQVWUDLQVWKHKRXUO
\
UDWHDWZ
KL
F
K
IO
RZF
K
DQJHVFDQ
E
HPD
G
H
N
QRZQDV
X
S
UDP
S
LQ
J
DQGGRZQUDP
S
LQ
J
+D
E
LWDWPDLQW
H
QDQ
FH
IO
RZV
D
V
GHVF
UL
EHG
LQWK
H
(,6
DUHKLJKVWHDG\GDPUHOHDVHVZLWKLQSRZHUSODQWFDSDF
L
W
\
F
I
VDW
I
X
OO
UHVHUYR
L
UH
O
HYDW
L
RQ
I
RUWRZHH
N
V
L
Q0DUFKDOWKRXJKRWKHUPRQWKVFRXOGEHFRQVLGHUHG
XQGHUWKH
*&
'$03%
\
FRQWUDVW%+%)VDUHLQIUH
TXHQW
KL
J
K
UH
O
HDVHVW
K
DWDUHDW
O
HDVWF
I
VJUHDWHU
WKDQDOORZDEOH
S
HDNGLVFKDU
J
HEXWQRW
J
UHDWHUWKDQ
F
I
V
$O
VR
%+%)UH
O
HDVHVDUHW
L
PH
G
WRRFFXUZ
K
HQ
UHOHDVHVLQH[FHVVRI 
S
RZHU
S
ODQWFD
S
DFLW
\
DUHUH
T
XLUHGIRU
G
DPVD
I
HW\SXUSRVHV0RUHUHFHQW
O
\W
K
HWHUP´H[SHU
L
PHQWDOKLJKIORZVµKDVEHHQXVHGWRGHVFULEHH[SHULPHQ
WD
O
IO
RZVW
K
DWH[FHH
G
S
RZHU
SO
DQWFD
S
DF
L
W
\
DQ
G
UDQ
J
H
I
URP
WRFIV+DELWDWPDLQWHQDQFHIORZVGLIIHU
I
URP%+%)VDQ
G
H[
S
HU
L
PHQWD
O
KLJK
IO
RZV
E
HFDXVHW
K
H
\
R
FFXUZ
L
W
KL
QSRZHUS
O
DQWFDSDF
L
W\DQ
G
ZHUHDQW
L
F
L
SDWH
G
WRRFFXULQPRVW
\
HDUV7KHWZRW
\S
HVRI UHOHDVHVZKLFK
K
D
G
V
L
P
LO
DUSXUSRVHVR
I
UH
I
RUP
L
QJ
E
DF
N
ZDWHUVDQ
G
PDLQWDLQLQ
J
VDQGEDUVZHUHQRWWREHVFKHGXOHGLQWKH
VDPH\HDUDQ
G
QH
L
W
K
HUZDVWRRFFXU
L
QD\HDUZ
K
HQW
K
HUH
8 The State of the Colorado River Ecosystem in Grand Canyon
was concern for the effects on sensitive resources such as
sediment or endangered species.
On the basis of significant scientific research since
1995, some of the assumptions about how Colorado
River resources would respond to ROD operations have
been modified or rejected. As a result, several additional
experimental flows that temporarily modified Glen
Canyon Dam ROD operations have been implemented
since 2000. Additional experimental flows discussed else-
where in this report include the 2000 low summer steady
flow (LSSF) test, the 2003–05 experimental fluctuating
nonnative fish suppression flows, and the November
2004 experimental high flow. The LSSF test included
1966 National Historic
Preservation Act passed
1967 Humpback chub and Colorado
pikeminnow (Ptychocheilus lucius)
federally listed as endangered
1968 Colorado River
Basin Project Act passed
1969 National Environmental Policy Act of 1969
passed requiring Federal agencies to consider the
environmental impacts of their proposed actions
and reasonable alternatives to those actions
1970 Long-range Operating
Criteria developed for Glen
Canyon Dam operations
Table 1. Glen Canyon Dam release prescriptions under the modified low fluctuating flow alternative (cfs = cubic feet
per second).
Monthly release
volume
(acre-feet)
Minimum
release (cfs)1
Maximum
release (cfs)
Allowable daily
fluctuation (cfs)
Upramp/
downramp (cfs/hr)
<600,000 8,000/5,000 25,000 5,000 4,000/1,500
600,000–800,000 8,000/5,000 25,000 6,000 4,000/1,500
>800,000 8,000/5,000 25,000 8,000 4,000/1,500
1 8,000 cfs between 7 a.m. and 7 p.m. and 5,000 cfs at night; releases each weekday during the recreation season (Easter to Labor Day) would
average not less than 8,000 cfs for the period from 8 a.m. to midnight.
two habitat maintenance flows (31,000 cfs for 4 d) in
spring and late summer, with June through August flows
held constant at 8,000 cfs. Fluctuating nonnative fish
suppression releases allowed the flow of the river to fluc-
tuate daily between 5,000 cfs and 20,000 cfs with relaxed
hourly upramp and downramp rates of 5,000 and 2,500
cfs/h, respectively, from January to March. In summer
and fall 2004, fine-sediment inputs from the Paria River
(15 mi below the dam) reached the agreed-upon levels for
triggering an experimental high flow of 41,000 cfs for 2.5
d (see chapter 1, this report).
Experimentation has largely focused on experimental
flows of the type described above to achieve downstream
Overview 9
benefits, with a particular focus on improving fine-
sediment resources and conditions for endangered native
fish. Another experimental effort underway is the manual
removal of nonnative fishes in order to protect native fish,
particularly humpback chub (see chapter 2, this report).
Collaboration
As for collaboration, the EIS outlined an innovative
organizational structure for pursuing the GCDAMP. The
program is administered by a senior Department of the
Interior official (designee) and facilitated by the Adaptive
Management Work Group (AMWG), which is organized
as a Federal Advisory Committee. The AMWG makes
recommendations to the Secretary of the Interior on
how to best alter the operating criteria at Glen Canyon
Dam or other management actions to protect down-
stream resources in order to fulfill the Department of the
Interior’s obligations under the GCPA (U.S. Department
of the Interior, 1995). The Secretary of the Interior
appoints the group’s 25 members, who include repre-
sentatives from Federal and State resource management
agencies, the seven Colorado River Basin States, Native
American tribes, environmental groups, recreation
interests, and contractors of Federal power from Glen
Canyon Dam (fig. 2). The GCDAMP also includes a
monitoring and research center (USGS Grand Canyon
Monitoring and Research Center), the Technical Work
Group, and independent scientific review panels.
As directed thus far by the AMWG, monitoring and
research on sediment dynamics, cultural resources, native
1972 Last verified record of
Colorado pikeminnow caught in
Grand Canyon at Havasu Creek
1973 Endangered Species Act of 1973 passed to protect and promote
the recovery of animals and plants that are in danger of becoming extinct
because of the activities of people. The act is administered by the U.S. Fish
and Wildlife Service (terrestrial and freshwater species) and the National
Oceanic and Atmospheric Administration–Fisheries (marine species)
1974 First lawsuit filed over Glen Canyon Dam
operations by commercial raft operators contending
that the disruption of normal flows was interfering
with their ability to conduct river trips
and nonnative fish, and endangered species have been
emphasized. Monitoring and research of these resources
have resulted in better understanding of their condition
and behavior.
For example, recent studies suggest that, contrary
to expectations under current dam operations, sand
contributed from Colorado River tributaries is rapidly
exported downstream and does not remain available
over multiyear timescales for restoration floods imple-
mented between January and July, which is the current
implementation schedule. Restoration floods are likely
to be more effective if they are carried out in the same
year that sand deliveries occur, before the new sand is
lost downstream. Progress has also been made in under-
standing the dynamics of fish populations and the value
of mechanical removal of nonnative fish for enhancing
native fish populations.
Report Organization
The chapters that follow provide status and trend
data for the natural, cultural, and recreational resources
of the Colorado River ecosystem in Grand Canyon. The
report deals first with the aspects of the natural environ-
ment that have been most emphasized in monitoring
and research—sediment and native fishes—followed by
other important environmental factors including climate
and drought, water quality, aquatic ecology, debris flows,
birds, and shoreline ecology and its associated wildlife.
The report then shifts emphasis to various human uses
10 The State of the Colorado River Ecosystem in Grand Canyon
1975 Grand Canyon National
Park Enlargement Act passed
1978 U.S. Fish and Wildlife
Service files jeopardy opinion
on the effects of Glen Canyon
Dam on endangered fishes
1979 Grand Canyon National Park
designated a UNESCO World Heritage
Site; Bureau of Reclamation proposes an
upgrade of Glen Canyon Dam’s generators
1980 Lake Powell reaches full pool
(3,700 ft); bonytail chub (Gila elegans)
federally listed as endangered
Figure 2. Adaptive Management Work Group committee members.
Interior Secretary’s Designee
Tribes
Hopi Tribe
Hualapai Tribe
Navajo Nation
Pueblo of Zuni
San Juan Southern Paiute Tribe
Southern Paiute Consortium
State and Federal Cooperating Agencies
Arizona Game and Fish Department
Bureau of Indian Affairs
Bureau of Reclamation
National Park Service
U.S. Department of Energy, Western Area Power
Administration
U.S. Fish and Wildlife Service
Colorado River Basin States
Arizona: Arizona Department of Water Resources
California: Colorado River Board of California
Colorado: Colorado Water Conservation Board
Nevada: Colorado River Commission of Nevada
New Mexico: New Mexico Office of the State Engineer
Utah: Water Resources Agency
Wyoming: State Engineer’s Office
Nongovernmental Groups
Environmental:
Grand Canyon Trust
Grand Canyon Wildlands Council
Recreation:
Federation of Fly Fishers/Northern Arizona Flycasters
Grand Canyon River Guides
Contractors for Federal Power from Glen Canyon Dam:
Colorado River Energy Distributors Association
Utah Associated Municipal Power Systems
Overview 11
of the ecosystem, including the economic importance of
the ecosystem, hydroelectric power generation, cultural
resources, and camping beaches. In each case, the infor-
mation is then used to discuss the management options
available to decision makers and the public based on the
best scientific information available. In large measure,
this report represents the first comprehensive assessment
of how effectively the MLFF alternative is allowing the
Secretary of the Interior to meet the resource manage-
ment goals of the Grand Canyon Protection Act of 1992.
Place Names and Units
Throughout the report, “Grand Canyon” is used
broadly to refer to the Colorado River corridor between
Glen Canyon Dam and the western boundary of Grand
Canyon National Park, including Glen, Marble, and
Grand Canyons. The study area is referred to as the
“Grand Canyon ecosystem.” The Colorado River is
discussed in terms of four distinct sections: Lees Ferry
1982 Glen Canyon Environmental
Studies created to study effects
of Glen Canyon Dam operations
1983 Glen Canyon Dam releases more
than 92,000 cfs to stop Lake Powell
from overtopping Glen Canyon Dam
1984 One of the last razorback
suckers (Xyrauchen texanus)
seen in Grand Canyon is caught
and released at Bass Rapids
1987 National Research Council completes
review of Glen Canyon Environmental Studies,
publishing River and Dam Management: a
Review of the Bureau of Reclamation’s Glen
Canyon Environmental Studies
reach, Marble Canyon, upper Grand Canyon, and lower
Grand Canyon. The “Lees Ferry reach” extends from the
downstream end of Glen Canyon Dam to Lees Ferry, and
“Marble Canyon” extends from Lees Ferry to the mouth
of the Little Colorado River. For this report, “upper
Grand Canyon” refers to the river corridor that extends
from the mouth of the Little Colorado River to the Grand
Canyon gaging station (Topping and others, 2003), while
“lower Grand Canyon” extends from the Grand Canyon
gaging station to the western boundary of the park.
In this report, U.S. customary units are used for all
measurements to facilitate understanding by the general
reader. Metric equivalents are provided in parentheses
after the U.S. customary units for all measurements except
for river flow, the standard measure of which is cubic
feet per second, and river mile, which is used to describe
distances along the Colorado River in Grand Canyon
(Stevens, 1990). The use of the river mile has a histori-
cal precedent and provides a reproducible method for
describing location: Lees Ferry is the starting point, as
12 The State of the Colorado River Ecosystem in Grand Canyon
1988 Glen Canyon Environmental Studies
issues Glen Canyon Environmental Studies
Final Report, completing Phase I and starting
Phase II, which would be accelerated to support
environmental impact statement process
1989 Secretary of the Interior Lujan orders
an environmental impact statement on dam
operations, and National Research Council
sponsors symposium that reviews existing
knowledge on Colorado River ecosystem
1990-91 Research flows
used to evaluate a variety
of discharge patterns
1991 Interim operating criteria for
Glen Canyon Dam implemented;
razorback sucker and Kanab ambersnail
(Oxyloma haydeni ssp. kanabensis)
federally listed as endangered
One challenge following completion of the 1995 Operation of Glen Canyon Dam Final Environmental Impact
Statement (EIS) was to identify and implement monitoring efforts that would produce scientific data suitable for
evaluating the new operating policy at Glen Canyon Dam. At that time, there was also a sense among managers and
scientists that additional, comprehensive syntheses of available data needed to be undertaken with respect to major
resource categories, such as sediment and fisheries. In addition, the need for development of a conceptual model
for the Colorado River ecosystem, consistent with the adaptive environmental assessment and management process
(now popularly called “adaptive management”), was also identified by the USGS Grand Canyon Monitoring and
Research Center (GCMRC) and its cooperators. This modeling effort began in 1998 and was continued concur-
rently with the establishment of the stakeholder-based, Federal Advisory Committee—the Adaptive Management
Work Group—and the development of the group’s strategic goals for the Colorado River ecosystem (1998–2002).
Key objectives for the conceptual modeling exercise were to (1) conduct an exhaustive knowledge assessment of
the various elements of the ecosystem on the basis of existing data and hypotheses posed in the EIS and within
the context of workshops that supported stakeholder and scientist interactions; (2) identify, through this process
of modeling and simulation, key areas where data or knowledge did not exist and therefore were impediments to
developing realistic simulations (by using historical data as a means of verification); and (3) identify future research
directives (both experimental or otherwise) that would effectively fill knowledge gaps in the program related to
management needs.
Development of the physical elements of the conceptual model (the Grand Canyon Model or GCM) proceeded
relatively quickly, mostly because there were abundant data in some key areas (hydrology, sediment, and river flow)
and an operational model for the Colorado River Basin (RiverWare™) had already been developed by the Bureau
of Reclamation. Other critical areas of the model development, however, were limited by the paucity of available
data related to biology and sociocultural resource areas (Walters and others, 2000). By 2000, it became clearer that
The Role of Conceptual Modeling in Support of Adaptive
Management in Grand Canyon
Overview 13
1992 Grand Canyon
Protection Act of
1992 passed
1994 Programmatic Agreement on Cultural Resources signed between the State of Arizona,
Department of the Interior agencies, and six tribes over protection of cultural resources in the river
corridor below Glen Canyon Dam; U.S. Fish and Wildlife Service designates critical habitat for four
species of endangered Colorado River fish and completes Biological Opinion outlining reasonable
and prudent alternatives that must be evaluated for dam operation
1995 Operation of Glen Canyon Dam Final Environmental Impact Statement
completed; Transition Work Group and Grand Canyon Monitoring and
Research Center begin formulating strategic plan; southwestern willow
flycatcher (Empidonax traillii extimus) federally listed as endangered;
Department of the Interior constitutes the Grand Canyon Monitoring and
Research Center and locates it in Flagstaff, Arizona
certain critical modules of the model could not even reliably predict the general direction of ecosystem response,
such as response of native fishes to warmer water conditions through implementation of a proposed temperature
control device. While water could be routed through the ecosystem with confidence, there was considerably less
confidence about the longer term relationship of flows to fine-sediment flux and beaches on the basis of remain-
ing downstream sand supplies alone. Although the inability of the GCM to accurately simulate higher level trophic
(e.g., fishes) responses in critical areas was cause for concern among managers, the goal of systematically identifying
gaps in data and knowledge so that future research (including experimentation) and monitoring could be designed
and implemented to fill the gaps was an acknowledged objective of the modeling effort.
In a sense, the largest contribution made by the conceptual modeling project was the identification of vari-
ous experimental flow and nonflow treatments that would need to be tested (presumably, within some longer term
design) to provide managers with scientifically based options for most effectively meeting the proposed management
goals. Experimentation has long been identified as a sign of “active” adaptive management and has been shown to
be an efficient means of resolving the uncertainty associated with various alternative management policies (Walters
and Holling, 1990). Simultaneously, the modeling project helped identify additional monitoring data that would be
required to more fully evaluate the influence of the modified low fluctuating flow policy on downstream resources
of concern. Although evaluation of all the resources outlined in the EIS has not been possible because of pro-
gram funding limitations, the GCM identified the general linkages between the varied resources as related to dam
operation. The experimental designs proposed and implemented in the Glen Canyon Dam Adaptive Management
Program have been a direct and logical outcome of conceptual modeling activities. Though still not complete, to
date, the experimental results have greatly advanced ecosystem understanding. Ultimately, the knowledge gained
through these scientific activities in the Colorado River ecosystem should lead to improved management options for
Glen Canyon Dam that will benefit society.
14 The State of the Colorado River Ecosystem in Grand Canyon
1996 Experimental controlled flood of 45,000
cfs conducted at Glen Canyon Dam; Record of
Decision for the operation of Glen Canyon Dam
signed by Secretary of the Interior Bruce Babbitt
1997 Interior Secretary Bruce Babbitt signed a Notice
of Establishment of the Adaptive Management Work
Group, a Federal Advisory Committee with first meeting
of the group in September; first test of the concept of
the habitat maintenance flows conducted in November
2000 Test of low summer steady flows for the
possible benefit of endangered species of fish,
second and third tests of the habitat maintenance
flows concept conducted in spring and summer
2001 Draft strategic plan for the Glen
Canyon Dam Adaptive Management Program
developed by program members
RM 0, with mileage measured for both upstream and
downstream directions.
References
Annerino, J., 2000, Canyons of the Southwest: a tour of
the great canyon country from Colorado to northern
Mexico: Tucson, University of Arizona Press, 144 p.
Babbitt, B., 1990, Introduction: down the imperiled
Colorado: Land and Water Law Review, v. 25, no. 1,
p. 1–9.
Beus, S.S., and Morales, M., eds., 2003, Grand
Canyon geology (2d ed.): New York, Oxford
University Press, 432 p.
Hughes, T.C., 1991, Reservoir operations, in Commit-
tee to Review the Glen Canyon Environmental Stud-
ies, Commission on Geosciences, Environment, and
Resources, eds., Colorado River ecology and dam man-
agement: proceedings of a symposium: Washington,
D.C., National Academy Press, p. 207–225.
Ingram, H., Tarlock, D.A., and Oggins, C.R., 1991,
The law and politics of the operation of Glen Canyon
Dam, in Committee to Review the Glen Canyon Envi-
ronmental Studies, Commission on Geosciences, Envi-
ronment, and Resources, eds., Colorado River ecology
and dam management: proceedings of a symposium:
Washington, D.C., National Academy Press, p. 10–27.
National Research Council, 1996, River resource man-
agement in Grand Canyon: Committee to Review
the Glen Canyon Environmental Studies, Commis-
sion on Geosciences, Environment, and Resources:
Washington, D.C., National Academy Press, 226 p.
National Research Council, 1999, Downstream: adaptive
management of Glen Canyon Dam and the Colorado
River ecosystem: Committee on Grand Canyon
Monitoring and Research, Water Science and Tech-
nology Board, Commission on Geosciences, Environ-
ment, and Resources: Washington, D.C., National
Academy Press, 230 p.
Pontius, D., 1997, Colorado River Basin study: Report
to the Western Water Policy Review Advisory
Commission, 126 p.
Roosevelt, T., ca. 1905, The works of Theodore
Roosevelt, presidential addresses and state papers,
pt. 1: New York, P.F. Collier and Sons, Publishers, p.
369–370.
Stevens, L., 1990, The Colorado River in Grand Canyon:
a guide: Flagstaff, Ariz., Red Lake Books, 115 p.
Topping, D.J., Schmidt, J.C., and Vierra, L.E., Jr., 2003,
Computation and analysis of the instantaneous-dis-
charge record for the Colorado River at Lees Ferry,
Arizona—May 8, 1921, through September 30, 2000:
Reston, Va., U.S. Geological Survey Professional Paper
1677, 118 p.
Overview 15
2004 Drought conditions cause water level at Lake
Powell to drop to lowest level since the dam began
filling; triggering thresholds based on sand inputs from
the Paria River and lesser Marble Canyon tributaries met;
and high flow experiment initiated on Sunday, Nov. 21
2002 U.S. Fish and Wildlife Service announces recovery goals
for endangered fishes of the Colorado River Basin; Adaptive
Management Work Group recommends implementation of
the first 2 yr of an experimental design proposed by the Grand
Canyon Monitoring and Research Center
2003 Experiment begun
to remove nonnative fish
from the Colorado River
in Grand Canyon
2003-05 Fluctuating nonnative fish
suppression releases from January
through March implemented and
continued through 2005
U.S. Department of the Interior, 1970, Glen
Canyon Dam and powerplant: technical record of the
design and construction: Denver, Colo., Bureau of
Reclamation, Technical Center, 658 p.
U.S. Department of the Interior, 1988, Glen Canyon
Environmental Studies final report: Salt Lake City,
Utah, Bureau of Reclamation, 84 p.
U.S. Department of the Interior, 1995, Operation of
Glen Canyon Dam Final Environmental Impact
Statement, Colorado River Storage Project, Coconino
County, Arizona: Salt Lake City, Utah, Bureau of
Reclamation, Upper Colorado Regional Office, 337 p.
U.S. Department of the Interior, 1996, Record of Deci-
sion, Operation of Glen Canyon Dam Final Environ-
mental Impact Statement: Washington, D.C., Office
of the Secretary of the Interior, 15 p.
Walters, C., Korman, J., Stevens, L.E., and Gold,
B., 2000, Ecosystem modeling for evaluation
of adaptive management policies in the Grand
Canyon: Journal of Conservation Ecology, v. 4, no.
2, http://www.consecol.org/vol4/iss2/art1, accessed
July 14, 2005.
Walters, C.J., and Holling, C.S., 1990, Large-scale man-
agement experiments and learning by doing: Ecology,
v. 71, no. 6, p. 2060–2068.
Water Education Foundation, 2001, Layperson’s guide to
the Colorado River: Sacramento, Calif., 28 p.
Webb, R.H., Wegner, D.L., Andrews, E.D., Valdez, R.A.,
and Patten, D.T., 1999, Downstream effects of Glen
Canyon Dam in Grand Canyon: a review, in Webb,
R.H., Schmidt, J.C., Marzolf, G.R., and Valdez, R.A.,
eds., The controlled flood in Grand Canyon: scientific
experiment and management demonstration: Wash-
ington, D.C., American Geophysical Union, Geophysi-
cal Monograph Series, v. 110, p. 1–21.
Wieringa, M.J., and Morton, A.G., 1996, Hydropower,
adaptive management, and biodiversity: Environmen-
tal Management, v. 20, no. 6, p. 831–840.
16 The State of the Colorado River Ecosystem in Grand Canyon
Timeline photograph credits:
Page 4, Major John Wesley Powell: U.S. Geological Survey Photo Library Archive
Page 5, Colorado River in Grand Canyon: U.S. Geological Survey Photo Library Archive
Page 6, Hoover Dam: Bureau of Reclamation
Page 7, Glen Canyon Dam: Bureau of Reclamation
Page 8, Humpback chub art: Randall D. Babb, Arizona Game and Fish Department
Page 9, Raft and rafters: Jeff Sorensen, Arizona Game and Fish Department
Page 10, Glen Canyon Dam: © 2005 Christopher Taesali; used with permission
Page 11, Benthic sampling fieldwork: Jeff Sorensen, Arizona Game and Fish Department
Page 12, Kanab ambersnail: Roy Averill-Murray, Arizona Game and Fish Department
Page 13, Willow flycatcher: Suzanne Langridge, U.S. Geological Survey
Page 14, Glen Canyon: © 2005 Christopher Taesali; used with permission
Page 15, USGS staff seining fish: © 2005 Dawn Kish; used with permission
First page photograph credit: see front matter for credits
Contact Information:
Lara M. Schmit
Associate Editor
Northern Arizona University
Center for Sustainable Environments
Flagstaff, AZ
lschmit@usgs.gov
Steven P. Gloss
Ecologist
U.S. Department of the Interior
U.S. Geological Survey
Southwest Biological Science Center
Tucson, AZ
sgloss@usgs.gov
Christopher N. Updike
Research Assistant
Northern Arizona University
Center for Sustainable Environments
Flagstaff, AZ
Chris_Updike@nau.edu
Chapter 1
Influence
of Glen
Canyon Dam
Operations on
Downstream
Sand Resources
of the Colorado
River in
Grand Canyon
Scott A. Wright
Theodore S. Melis
David J. Topping
David M. Rubin
Introduction
The closure of Glen Canyon Dam and the begin-
ning of flow regulation of the Colorado River through
Grand Canyon in 1963 all but eliminated the mainstem
sand supply to Grand Canyon and substantially altered
the seasonal pattern of flows in the Colorado River.
Dam-induced changes in both sand supply and flow have
altered the sedimentary processes that create and main-
tain sandbars and related habitats, resulting in smaller
and coarser grained deposits throughout the ecosystem.
From the perspective of river management, the
ecological implications associated with such changes
are not well understood and are the focus of ongoing
integrated science studies. The effects of Glen Canyon
Dam operations on fine-sediment resources (i.e., sand
and finer material), particularly the erosion and restora-
tion of sandbars, are of interest because sandbars are a
fundamental element of the Colorado River’s geomor-
phic framework and the landscape of Grand Canyon
(see Webb, 1996; Webb and others, 2002). Sandbars
are also of interest in terms of the essential role fine-
sediment resources play in other ecosystem processes
(U.S. Department of the Interior, 1995). For example,
emergent sandbars create terrestrial habitats for ripar-
ian vegetation and associated fauna. Similarly, sandbars
create areas of stagnant or low-velocity flow that may
be used as rearing habitat by the endangered humpback
chub (Gila cypha) and other native fish. Recreational river
runners and other backcountry visitors frequently use
sandbars as campsites. Finally, abundant sand and silt
deposits near and above the elevation of typical predam
floods contain archeological resources and protect those
resources from weathering and erosion.
Conservation of Grand Canyon’s fine-sediment
resources is a primary environmental goal of the Glen
Canyon Dam Adaptive Management Program. Despite
this fact, the dam’s hydroelectric powerplant operation
under the Record of Decision (U.S. Department of
the Interior, 1996) continues to erode the limited fine-
sediment deposits that exist downstream. Changes in
the abundance, distribution, size, and composition of
sandbars began to occur under the no action period (his-
torical operations) of dam operation from 1963 through
1991. Sandbar erosion continued despite changes in
the operation of the dam that resulted from the imple-
mentation of the interim operating criteria in 1991 and
the modified low fluctuating flow (MLFF) alternative in
18 The State of the Colorado River Ecosystem in Grand Canyon
1996. The MLFF was the preferred alternative identi-
fied in the 1995 Operation of Glen Canyon Dam Final
Environmental Impact Statement (EIS) and was selected
in the Record of Decision (U.S. Department of the
Interior, 1996).
The U.S. Geological Survey’s (USGS) Grand
Canyon Monitoring and Research Center and its
cooperators have conducted extensive monitoring and
research on fine-sediment transport and sandbar evolu-
tion in Grand Canyon. This chapter presents a sum-
mary of the results of studies since the 1970s, as well as
conclusions derived from recent syntheses of streamflow,
sediment transport, and geomorphic data from 1921 to
2004, including recent sediment budgets. The effects of
the MLFF operating alternative at Glen Canyon Dam
(1996–2004) on fine-sediment transport and sandbars are
examined in the context of these historical data. Finally,
options identified by sediment scientists for testing alter-
native operations aimed at more effective conservation
of fine-sediment resources are discussed.
Background
Predam Sediment-
transport Processes
As described by Rubin and others (2002), sandbars
below Glen Canyon Dam in Marble and Grand Canyons
are maintained by fine sediment that is transported by the
Colorado River through the ecosystem. As sand is car-
ried through these bedrock canyons by the river, some of
it is deposited along channel margins and along shore-
lines within hundreds of eddies, thus building sandbars.
The eddy areas, which are typically located immediately
downstream from channel constrictions created by tribu-
tary debris fans, are susceptible to fine-sediment deposi-
tion because the flow tends to recirculate and be of lower
velocity than the flow in the main channel. Using histori-
cal sediment-transport records from the Lees Ferry (RM
0) and Grand Canyon (RM 87) gages, Laursen and others
(1976) and later Topping and others (2000b) identified
that before closure of Glen Canyon Dam, sand would
accumulate in the Colorado River channel during late
summer, fall, and winter. Annual accumulation of sand
in the channel during predam years apparently resulted
from large sediment inputs from tributaries that occurred
during periods of seasonal low flows in the main channel
of the Colorado River. Following these periods of sand
enrichment in the main channel, spring snowmelt floods
would erode the accumulated sand from the channel and
transport it out of the canyon, along the way depositing
some of the sand in the low-energy eddy areas and thus
leading to the building of the high-elevation sandbars.
Following the spring replenishment of sandbars, some of
this sand would in turn be redistributed to even higher
elevations by winds (Topping and others, 2000b). On
an annual basis, the inputs of sand to the system would
approximately balance the export, maintaining equilib-
rium in background sand storage in the eddies.
Effects of Lake Powell on
Sand Transport
Before the closure of Glen Canyon Dam in 1963,
approximately 25 million tons (23 million Mg) of sand
passed the Lees Ferry stream gage annually. With the
addition of 1.7 million tons (1.5 million Mg) of sand
from the Paria River, which joins the Colorado River just
downstream from Lees Ferry, the total predam annual
sand supply to Marble Canyon reached about 27 million
tons (24 million Mg). At the end of Marble Canyon, the
Little Colorado River joins the Colorado River and con-
tributed, on average, about 1.9 million tons (1.7 million
Mg) to the annual sand supply. Thus, the total predam
sand supply to Grand Canyon, from the Colorado River
upstream from Lees Ferry and with the Paria and Little
Colorado Rivers combined, was approximately 29 million
tons (26 million Mg).
Today, because Lake Powell traps all of the sediment
upstream from Glen Canyon Dam, the Paria River is the
primary source of sand to Marble Canyon, supplying
approximately 6% of predam sand levels. In the case
of Grand Canyon, Glen Canyon Dam has reduced its
sand supply to primarily the contributions of the Paria
and Little Colorado Rivers. Other lesser tributaries also
contribute a small amount of sand to Grand Canyon,
with an estimated cumulative supply that is approxi-
mately 10% to 20% of the mean annual load provided
by the Paria River. Taken together, the contributions of
sand from various sources provide Grand Canyon with
approximately 16% of its predam sand levels. The find-
ings presented here are drawn from Topping and others
(2000b) and Webb and others (2000); readers interested
in more details on the predam and postdam sediment
budgets for Marble and Grand Canyons should consult
these reports.
Influence of Glen Canyon Dam Operations on Downstream Sand Resources 19
Effects of Dam Operations on
Flow Frequency and Duration
Changes in the flow regime of the Colorado River
since construction of Glen Canyon Dam have also been
dramatic in terms of seasonal variability, as well as in
terms of daily fluctuations that occur because of “peak-
ing” hydroelectric power generation. Dam operations
have altered seasonal variability by eliminating long-
duration flood flows that occurred during the spring
snowmelt and short-duration flood flows that occurred
during the late summer and early fall thunderstorm
season, as well as the very low flows that occurred dur-
ing summer, fall, and winter. With regard to the highest
flows, dam operations have reduced the 2-yr recurrence
interval flood (i.e., the flood that occurs every other
year on average) from 85,000 cubic feet per second
(cfs) during the predam period to 31,500 cfs during the
postdam period. In the predam era, discharge exceeded
9,000 cfs only 44.3% of the time, while in the postdam
era this percentage has gradually increased by decade,
from 52.7% in the 1960s to 82.6% in the 1990s. This
decrease in the duration of low flows has important
implications for sediment transport because Topping and
others (2000b) showed that flows less than about 9,000
cfs result in accumulation of tributary sand inputs in
the Marble Canyon and Grand Canyon reaches of the
river, whereas flows above this generally lead to transport
of new sand inputs through these reaches or erosion of
sand from these reaches.
Dam operations have introduced large daily varia-
tions in discharge to generate hydroelectric power that
tracks daily peaks in demand throughout the Western
United States. Also, because peak energy demand varies
seasonally in the West, with peak demand occurring in
midsummer and winter, the month-to-month flow pattern
related to dam operation is substantially different from
natural, predam, seasonal patterns. Highest discharges in
the river now occur during the two seasons when predam
discharge had typically been the lowest, midsummer and
winter. Furthermore, daily patterns of flow in the river
have been altered by dam operations. For example, dur-
ing the predam period the median daily range in dis-
charge was only 524 cfs, whereas in the postdam era the
median daily range increased to 8,580 cfs, a value greater
than the predam median discharge. Before dam opera-
tion, the daily range in discharge exceeded 10,000 cfs
only about 1% of all days; postdam, the daily discharge
range exceeded 10,000 cfs on 43% of all days.
Initially, operation of the dam’s powerplant was
characterized mostly by unconstrained daily fluctua-
tions that were designed to optimize electrical generation
around peak daily demand, which had patterns that also
varied on a monthly timescale related to seasonal changes
in energy demand. From 1963 through 1991, these oper-
ations typically caused the Colorado River’s discharge to
fluctuate on a daily basis from less than 5,000 cfs to near
powerplant capacity of about 31,000 cfs. These so-called
“no action” daily operations (because they were consid-
ered the no action alternative in the EIS) were first altered
in 1990 to facilitate experimental release patterns imple-
mented through July 1991 as part of field investigations
associated with the EIS on dam operations. The experi-
mental flows of 1990–91 were then followed by “interim
operating criteria” from August 1991 until October
1996, when Secretary of the Interior Bruce Babbitt
implemented current Record of Decision dam operations.
Implementation of the interim operating criteria in 1991,
as well as the MLFF in 1996, constrained the change in
discharge over any 24-h period to 5,000; 6,000; or 8,000
cfs, depending on the monthly volume-release schedule
specified in the annual operating plan for the Colorado
River Storage Project. The flow history of the Colorado
River into Grand Canyon as measured at the Lees Ferry
gaging station is shown in figure 1. These flow data
illustrate a transformation of the Colorado River from a
fluvial ecosystem with significant seasonal variability in
the predam era to a postdam river ecosystem with little
seasonal variability and substantial daily fluctuations.
Another important aspect of the MLFF operation
is the schedule of monthly release volumes in relation to
the seasonality of sediment inputs. Because of energy
demand and hydropower economics, monthly release
volumes are highest during months with high demand,
including those in late summer. Historically, however, the
late summer months were characterized by low mainstem
flows and the highest tributary inputs, leading to sediment
accumulation during the predam era. Postdam, high
summer releases coincide with tributary inputs, leading
to rapid export instead of accumulation. Therefore, not
only has the sand supply been drastically reduced through
the impoundment of Lake Powell, but the seasonal timing
of low and high flows has also been both highly com-
pressed and significantly shifted to later periods of the
year that coincide with tributary sand inputs. The infor-
mation in this section was taken from Topping and others
(2003); readers with further interest in the Colorado
River’s hydrology, both before and after the dam was
closed, should consult this report.
20 The State of the Colorado River Ecosystem in Grand Canyon
Figure 1. Instantaneous discharge (A) and daily range in discharge (B) in cubic feet per second of the Colorado River at Lees Ferry
(RM 0) between 1921 and 2004 (modified from Topping and others, 2003). Before construction of Glen Canyon Dam, the annual peak
flow routinely exceeded 100,000 cfs. Dam operations during the period from 1963 through 1990 were characterized by daily fluctuations
from typically less than 5,000 cfs to near powerplant capacity, or about 31,000 cfs, and included the record wet period of the mid-1980s,
which resulted in the use of the spillways in 1983 for emergency releases exceeding about 90,000 cfs. Interim operating criteria, which
constrained daily release fluctuations, began in 1991 and were followed by the modified low fluctuating flow operating alternative that
was implemented as part of the Secretary of the Interior’s Record of Decision (ROD) in 1996 (BHBF = beach/habitat-building flow).
Status and Trends of
Fine Sediment Below
Glen Canyon Dam
Changes in sand supply and flow regime down-
stream from a dam affect the geomorphology of the
downstream channel. When a dam traps sand and
releases clear water, this clear water is often termed
“hungry” because it still has the capacity to transport an
amount of sand and gravel proportional to the flow and
will erode the downstream channel and banks in order
to satisfy its appetite with respect to sediment transport.
On the basis of resurveys of historical cross-sections
upstream from Lees Ferry, approximately 20 million tons
(18 million Mg) of material—gravel and fine sediment,
including sand—have been eroded from the first 15 mi
(24 km) of the Colorado River downstream from the
dam, an area referred to in this report as the Lees Ferry
reach (Grams and others, 2004). The amount of mate-
rial removed is equivalent to a 6 to 10 ft (2–3 m) drop in
channel elevation averaged over the entire reach. Most
of this sediment was removed by daily, high-release
dam operations designed to scour the channel of the
Colorado River below the powerplant during April–June
B.A.
1965 (fig. 1). Daily suspended-sediment measurements
made by the USGS at the Lees Ferry and Grand Canyon
gaging stations indicated that these high flows in 1965
eroded 4.4 million tons (4.0 million Mg) of fine sediment
(mostly sand) from the Lees Ferry reach and 18 million
tons (16 million Mg) of fine sediment (mostly sand) from
Marble and upper Grand Canyons. Channel scour was
anticipated below the dam during its design and was
later needed to optimize energy generation within the
operating range of the hydroelectric powerplant (Grams
and others, 2004). Typical dam releases today do not
result in much erosion from the Lees Ferry reach, and
as a result very little fine sediment is transported down-
stream to Marble and upper Grand Canyons.
Despite the fact that its contributing drainage area
is approximately 18 times smaller than that of the Little
Colorado River, the single largest sand supplier to the
reaches below Glen Canyon from 1990 through 2004
was the Paria River. Farther downstream in Marble and
upper Grand Canyons, the fate of fine-sediment depos-
its is dependent upon the long-term balance between
inputs to the system (i.e., tributary supply) and exports
from the system (i.e., mainstem sediment-transport rates).
Although sand inputs have been greatly reduced by the
closure and operation of Glen Canyon Dam, the annual
Influence of Glen Canyon Dam Operations on Downstream Sand Resources 21
mainstem transport—and thus export—has also most
likely been reduced because of the elimination of the
highest flood flows. As a result, two possibilities exist for
the postdam fine-sediment balance downstream from
the Paria River. First, if the supply from the Paria River
and other lesser Marble Canyon tributaries exceeds the
postdam transport rate on an annual basis, then new
sand inputs would accumulate in the channel and in low-
elevation portions of eddies over multiple years. Such
accumulated sand supplies would then be available at
any time for redistribution to higher elevation sandbars
through release of periodic controlled floods (i.e., beach/
habitat-building flows in the EIS; hereafter BHBF) from
Glen Canyon Dam. This scenario was the conclusion
reached by Howard and Dolan (1981), Andrews (1990,
1991), Smillie and others (1993), and the EIS study
team (U.S. Department of the Interior, 1995) for the
MLFF alternative, leading to its implementation in 1996.
Howard and Dolan (1981) reached their conclusion by
using an estimate for the sand contribution from the lesser
tributaries that is now regarded to be about a factor of
four too high (Topping and others, 2000b; Webb and
others, 2000). Andrews (1990, 1991) and Smillie and
others (1993) reached their conclusions by using stable
sand-transport relationships, also called “rating curves.”
A stable sand-transport rating curve exists where there is
a unique value for sand concentration for any given flow.
This approach invokes the assumption that the upstream
sand supply is in equilibrium with transport capacity.
The methods and data used to reach the conclusion in
the EIS are discussed further in the following section.
Alternatively, if the annual mainstem transport rate
(export) exceeds tributary supply (input), then systematic
long-term erosion of fine sediment from the channel
would be expected. In fact, this second scenario was
originally predicted by Dolan and others (1974) and
Laursen and others (1976) on the basis of their early
sediment-transport studies related to effects of Glen
Canyon Dam on downstream resources. In order for
high-flow releases to be effective at restoring and main-
taining sandbars under this second scenario, controlled
floods would need to be strategically timed to coincide
with or immediately follow tributary sand inputs. These
early studies predated the concept of using controlled
floods to restore eroded sandbars; hence, their estimates
of sand transport in the postdam era could only result
in net export of new sand inputs and continued erosion
of existing sandbars of predam origin. More recent evi-
dence presented in the following section further supports
the conclusion that this second scenario prevails under
the current reoperating strategy and that this situation is
leading to systematic, long-term erosion of fine sediment
from the channel bed and eddies of Marble and Grand
Canyons. On the basis of existing data, it is still uncer-
tain whether or not strategically timed managed floods
can restore and maintain eroded sandbars by using only
the limited and infrequent tributary-derived sand that
enters the river below the dam.
Recent Findings
The Paradigm of Sand Transport and
Storage Used in the 1995 Environmental
Impact Statement
The EIS concluded that sand would accumulate
over multiyear timescales in the channel of the Colorado
River in Marble and upper Grand Canyons during MLFF
powerplant releases in all but the highest release years
(U.S. Department of the Interior, 1995). The basis for
this conclusion was the assumption that the relationship
between the water discharge and sand transport in the
Colorado River did not change substantially over time.
This assumption was used because sediment-transport
data collected in the postdam Colorado River were sparse.
Prior to the early 1970s, suspended-sediment con-
centration was measured on a daily basis at the three
USGS gaging stations that are critical to constructing
a sand budget for Marble and Grand Canyons: the
Paria River at Lees Ferry, the Little Colorado River at
Cameron, and the Colorado River near Grand Canyon.
The sediment sampling program at the Colorado River
near Grand Canyon gaging station began in October
1925; the daily sediment sampling programs at the Paria
and Little Colorado Rivers began in October 1947. The
Little Colorado River sediment record was discontinued
on September 30, 1970; the Colorado River sediment
record at the Grand Canyon gaging station was discon-
tinued on September 30, 1972; and the Paria River sedi-
ment record was discontinued on September 30, 1976.
Thus, the only postdam period of overlap between these
stations that could be used to construct a sand budget
was the period from closure of the dam in March 1963
through September 30, 1970. Furthermore, no post-
dam sand-transport data were collected within Marble
Canyon during this early period.
To fill this data gap, the USGS began a program of
quasi-daily sediment sampling on the major tributaries
to the Colorado River (that is, the Paria River, the Little
Colorado River, and Kanab Creek) and at five locations
on the mainstem Colorado River in Marble and Grand
Canyons (Garrett and others, 1993). On the tributar-
22 The State of the Colorado River Ecosystem in Grand Canyon
Figure 2. Reproduction of figure III-15 from the final
environmental impact statement (EIS) (U.S. Department of the
Interior, 1995), which shows the sand budget as computed
by Randle and Pemberton (1987). Recent studies refute the
conclusion of the EIS that sand accumulates on the bed of the
Colorado River over multiple years under normal dam operations.
(Phantom Ranch is the location of the Grand Canyon gage.)
ies, this program extended from July through December
1983. On the mainstem, this program included the
periods from July through December 1983 and October
1985 through January 1986. All suspended-sediment
samples collected under this program were analyzed for
grain size to allow use in constructing sand budgets.
The sand budget for the Colorado River in Marble
and Grand Canyons used in the EIS was constructed by
Randle and Pemberton (1987) and Pemberton (1987).
For tributary sand input, they constructed stable sand-
rating curves by using all of the historical and 1983
data from the Paria River, the Little Colorado River,
and Kanab Creek. They also included an estimate for
the sand supply from the lesser tributaries. Pemberton
(1987) developed stable sand-transport rating curves at
the five mainstem locations based on the USGS 1983–86
data, and the EIS states, “The sand transport equations
of Randle and Pemberton (1987) and Pemberton (1987)
were used for these computations” (U.S. Department of
the Interior, 1995, p. 95) in reference to the sediment
budget presented in figure III-15 of the EIS (and repro-
duced here as fig. 2). Therefore, the EIS sediment bud-
get was based on the assumption of stable sand-transport
rating curves. Results of recent studies presented in the
following section suggest that this assumption is incorrect
for the Colorado River below Glen Canyon Dam.
Studies Since 1996 That Refute
the Environmental Impact
Statement Findings
Research and monitoring conducted during and
after the 1996 BHBF experiment, also known as the
1996 controlled flood, have led to several findings that
refute the EIS predictions for sand conservation and
suggest that the implementation of this strategy has
not led to sustainable restoration and maintenance of
sandbars in either Marble or Grand Canyon. Instead,
the canyons’ sandbars continue to erode (figs. 3–6). The
primary results of several of these studies are briefly
summarized below:
Rubin and others (1998) and Topping and oth-
ers (1999) showed that the sand supply during
the 1996 BHBF was not as great as was assumed
before the experiment and that the sand on the
bed of the river and in suspension coarsened
dramatically as the upstream supply of sand
decreased over time during this flood. This pro-
cess led to flood deposits that coarsened vertically
upward (i.e., inversely graded deposits).
Topping and others (2000a) demonstrated that
the grain size of sand on the bed of the Colorado
River can change by over a factor of four as func-
tions of tributary resupply of finer sand and higher
dam releases that winnow the bed and that this
factor-of-four change in bed-sand grain size cor-
responds to a change of two orders of magnitude
in the concentration of sand in suspension (for the
same discharge of water). Identification of this
dynamic process precludes the use of stable sand-
transport relationships in the Colorado River,
thus invalidating the approach used to construct
the sand budget in the EIS. Topping and others
(2000a) also showed that Randle and Pemberton
(1987) incorrectly predicted sand accumulation
in the Colorado River because the data they used
to verify their modeled stable sand-export rela-
tionships were from periods in the mid-1980s,
when sand in the river was anomalously coarse
and sand-transport rates were anomalously low
following prolonged releases above powerplant
capacity between 1983 and 1986.
Rubin and Topping (2001) showed that sand
transport in the postdam Colorado River in
Grand Canyon is regulated by both the discharge
of water and the grain size of the sand available
for transport in suspension. This information also
•
•
•
Influence of Glen Canyon Dam Operations on Downstream Sand Resources 23
Figure 3. Repeat photographs of Tapeats Creek at the Colorado River, Grand Canyon (RM 133.8, right shore). A. (July 1952) This view
downstream from below the mouth of Tapeats Creek shows a large sandbar with few rocks or boulders exposed. This sandbar was
frequently used for layovers during river trips in the 1950s (Kent Frost, courtesy of the photographer). B. (March 27, 2003) Large rocks
and boulders are now exposed because of severe beach erosion. New sand was deposited here during the 1996 beach/habitat-building
flow but was quickly removed. This camp is no longer used, which creates a problem for river runners who want to visit Tapeats Creek
(J. Janssen, stake 2676, courtesy of the Desert Laboratory Collection of Repeat Photography). (Figure after Webb and others, 2002.)
B.
Figure 4. Time series of repeat photographs of sandbars along the left shore of the Colorado River near RM 44.5 (Eminence Break)
illustrating deposition on the sandbar during the 1996 beach/habitat-building flow (March 26–April 2; high flow occurred between
photographs B and C) and subsequent erosion since April 1996. Images provided by Northern Arizona University, Department of
Geology in cooperation with the U.S. Geological Survey.
A.
A.
B.
C.
D.
E.
F.
March 13, 1994
March 25, 1996
April 4, 1996
April 19, 1998
June 17, 2000
September 11, 2000
24 The State of the Colorado River Ecosystem in Grand Canyon
Figure 5. A decrease in elevation of the sandbar surface is
seen at Jackass Creek camp located along the left shore of the
Colorado River, 23 mi (37 km) downstream of Glen Canyon Dam.
Elevations were determined by examining oblique and aerial
photographs of the site and by field survey of the elevation and
the former sand surface at its contact with large talus blocks. This
graph shows the elevations near one prominent talus block that
was inundated by predam mean annual floods, but since the dam
was completed, the talus block has been inundated infrequently
(modified from Rubin and others, 2002).
Figure 6. Changes in sandbar size (total surface area) are shown
for 14 long-term sandbar study sites between the Lees Ferry and
Grand Canyon gages (RM 0 to RM 87). Area of bars exposed
above water discharges of 8,000 cfs decreased by 22% from 1991
to 2004. The 1996 beach/habitat-building flow resulted in a net
transfer of sand from mid elevations to high elevations (modified
from Rubin and others, 2002).
contradicts the approach of the EIS, where it was
assumed that sand transport was regulated only by
the discharge of water.
Topping and others (2000b) showed through
their analysis of the 1965–70 daily sediment-
transport data collected by USGS that, under
normal powerplant flows, newly input tributary
sand is exported past the Grand Canyon gaging
station within several months. Their analysis of
predam data indicated that, prior to closure of
Glen Canyon Dam, sand would accumulate in
Marble and upper Grand Canyons only during the
9 mo of the year when discharges were typically
lower than about 9,000 cfs.
Measurements of the channel bed indicate that
tributary sand, which is typically much finer than
the sand on the bed of the Colorado River, accu-
mulates on the bed for only a short time before
being eroded and transported out of the canyon
under normal MLFF dam operations (Topping
and others, 2000a).
Since August 1999, detailed suspended-sediment
transport measurements have been collected at
the Paria and Little Colorado Rivers to document
•
•
•
inputs and at the USGS gaging stations above
the mouth of the Little Colorado River and near
Grand Canyon to document export. Initially,
these quasi-daily measurements were made by
using only conventional USGS methodologies
to obtain cross-sectionally integrated samples
of suspended-sediment concentration and grain
size (methods described in Edwards and Glysson,
1999). Because substantial and rapid (within a
day) changes that are due to tributary inputs can
occur in suspended-sediment concentration and
grain size, emerging technologies for continuous
monitoring of suspended-sediment concentra-
tion and grain size were tested and implemented
beginning in 2001. These technologies include
acoustic backscatter and laser-diffraction methods
and are described in detail in Melis and oth-
ers (2004) and Topping and others (2004). The
detailed sediment-transport measurements allow
for the ability to construct sediment budgets
based on continuous data instead of on rating
curves, a very important distinction from the
EIS approach of using a limited data set. These
data show that the overall mass balance of sand
(input minus export) continues to be negative
(fig. 7), as originally predicted by Laursen and
Influence of Glen Canyon Dam Operations on Downstream Sand Resources 25
Figure 7. Mass balance of sand between Lees Ferry and
Grand Canyon gages from August 1999 through July 2004 (A) and
separately for sediment years (July–June) 2003 (B) and 2004 (C).
Mass balance is computed by subtracting measured, mainstem
suspended-sand export (10% uncertainty) from estimated and
measured sand inputs from the Paria River (20% uncertainty) and
Little Colorado River (30% uncertainty), as well as from estimated
inputs from numerous lesser tributaries (50% uncertainty). The
measurements illustrate the rapid export of tributary inputs by high
dam releases and the continued overall loss of sand from Grand
Canyon under the modified low fluctuating flow (MLFF) alternative,
even during the drought-hydrology, minimum-volume release years
of 2003 and 2004 (modified and updated from Rubin and others, 2002).
others (1976). Most significantly, the sand mass
balance remained negative during water years
2000 through 2004, despite 5 consecutive years
in which minimal release volumes (8.23 million
acre-feet (10,148 million m³)) from Lake Powell
occurred during prolonged drought in the upper
Colorado River Basin. These measurements
and calculations of sand transport also show
that tributary inputs are typically transported
downstream and out of the canyon within a few
months under typical Record of Decision opera-
tions (Rubin and others, 2002).
Repeat topographic mapping of sandbars (Hazel
and others, 1999) showed that the 1996 BHBF
did increase the surface area of high-elevation
sandbars, but more than half of the sand depos-
ited at higher elevations was taken from the lower
portions of the sandbars (Schmidt, 1999) rather
than being derived from tributary sand supplies
accumulated on the channel bed, as originally
hypothesized in the 1995 EIS.
Repeated surveys of channel cross-sections (Flynn
and Hornewer, 2003) revealed erosion at 55 of the
57 locations between 1991 and 1999, even though
daily operations were constrained during the time
series of repeat measurements.
Schmidt and others (2004) conducted geomorphic
mapping from air photos and land surveys for the
predam and postdam periods. They estimated the
loss of sand to be about 25% of the area typically
exposed at base flow in predam photographs, but
estimates range from 0% to 55% depending on
study reach and method of analysis. Their studies
further suggested that loss of the sandbar area
continued at a relatively steady rate between 1983
and 2002, despite constraints on daily operations
imposed after 1991.
Importance of Continuous Long-term
Sediment-transport Data
Because of a lack of continuous data on sediment
inputs and export that would have allowed for a sedi-
ment budget based on measured data, the EIS study
team used stable sand-transport rating curves. Stable
rating curves assume that for any given flow there is a
single value for the corresponding sand concentration
and, therefore, a predictable sand-transport rate related
to flows released from Glen Canyon Dam. The recent
•
•
•
A.
C.
B.
26 The State of the Colorado River Ecosystem in Grand Canyon
Figure 8. Looking upstream into Glen Canyon from the
Paria River confluence with the main channel Colorado River
during a Paria River flood. Tributary inputs of sand, such as
the one pictured, now encounter clear Colorado River water
because Lake Powell traps incoming fine sediment. The
Paria River is the primary source of sand to Marble Canyon
but is only about 6% of the predam sand supply (photograph
by Scott A. Wright, U.S. Geological Survey).
studies reported above, however, have demonstrated that
in the postdam Colorado River the relationship between
flow and sand transport is not stable but instead shifts
quickly and substantially relative to the grain size of sand
on the bed of the river (which is controlled by tributary
inputs and mainstem flows). Rubin and Topping (2001)
and Rubin and others (2002) showed that the grain size
of the sand in the regulated Colorado River ecosystem
depends greatly on the recent history of tributary activ-
ity. For example, during low tributary flow periods the
only source of sand to the mainstem Colorado River
is that on the channel bed and in eddies, and that sand
tends to be much coarser than tributary-delivered sand
because of the winnowing of the finer sizes. When
tributaries are flooding and delivering large quantities of
fine sand (fig. 8), however, the supply is no longer lim-
ited to the coarser channel bed sand, resulting in much
higher mainstem sand concentrations and, hence, greatly
increased suspended-sediment export for any given flow
released from the dam.
Because sand transport cannot be predicted based
on discharge alone, sediment budgets for the Colorado
River in Grand Canyon can only be constructed based
on measurements of sand transport at a frequency great
enough to capture changes in concentration and grain
size resulting from tributary inputs. Fundamentally,
the conclusions drawn by the EIS team, which are not
supported by the more recent data, resulted from a lack
of continuous data in the postdam era; that is, if daily
records had been continued beyond 1972 and into the
EIS period, then the fine-sediment budget would have
been constructed based on these data rather than on
stable rating curves. Recent sediment budgets suggest
that under this scenario the conclusions of the EIS would
have been different and possibly would have led to a
different strategy for operation of Glen Canyon Dam in
1996. Though it is somewhat costly to collect long-term,
high-frequency sediment-transport records, in this case it
may have prevented 13 yr of dam operations that have
continued to erode sandbars from Grand Canyon.
Current Experimental
Plan for Fine Sediment
Because recent research has shown that sand does
not accumulate on the river bed in Marble and Grand
Canyons under normal Record of Decision dam opera-
tions, scientists have recently proposed two possible field
tests of dam operating options that might more effec-
tively conserve limited, downstream sand resources. One
approach is to implement floods immediately following
large tributary inputs that commonly occur in late sum-
mer and early fall. A second approach is to follow tribu-
tary sand-input events with low flows, in order to limit
export and retain most of the sand input, until flooding
can be implemented. This approach would require a
change in the pattern of monthly release volumes and
associated dam operations because July and August
releases of recent drought years still resulted in half of
the sand introduced by a tributary flood being exported
within days or weeks (Rubin and others, 2002).
In September 2002, the U.S. Department of the
Interior (2002) approved implementation of the second
approach described above. Under this plan, changes
in dam operations and restoration floods are linked to
triggering thresholds based on sand inputs from the Paria
River and lesser Marble Canyon tributaries and retention
of sand in Marble and Grand Canyons. For example,
the “autumn sediment input” scenario described in the
2002 environmental assessment (EA) (U.S. Department of
Influence of Glen Canyon Dam Operations on Downstream Sand Resources 27
Figure 9. Sequence of events established in the autumn sediment
input scenario in an environmental assessment by U.S. Department
of the Interior (2002) related to fine-sediment inputs and retention
to trigger a 2-d, 42,000–45,000-cfs experimental high flow in
January. If fine-sediment inputs do not reach specified levels, then
modified low fluctuating flow (MLFF) operations, as specified in the
Record of Decision (ROD) (U.S. Department of the Interior, 1996),
are continued.
the Interior, 2002) defined a sequence of events related
to sand inputs and retention that would trigger a 2-d,
42,000–45,000-cfs experimental high flow in the follow-
ing January (fig. 9). Significant sand inputs to Marble
Canyon that exceeded the triggering threshold for an
experimental high flow occurred during September–
November 2004. Instead of constraining operations
through December (a winter, peak-demand month) in
order to retain sand in Marble Canyon as laid out in the
2002 EA, a supplemental EA was prepared that allowed
for a hybrid of the first and second approaches to be
tested and evaluated. Approval of the supplemental EA
paved the way for the experimental high flow that began
on Sunday, November 21, 2004, when the Bureau of
Reclamation opened the bypass tubes of Glen Canyon
Dam for 90 h. The peak high flows ran for 2.5 d (60 h)
at about 41,000 cfs. Scientists will evaluate data col-
lected during and after the high-flow event to determine
whether or not this strategy succeeded in enlarging exist-
ing beaches and sandbars.
Other dam operation scenarios may be more effec-
tive at retaining tributary inputs, such as Record of
Decision operations modified such that equal volumes
of water are released from the dam each month. Alter-
natively, a scenario of seasonally adjusted steady flows,
which was an alternative in the EIS process, may be
effective. Because of the severely reduced sand sup-
ply, however, even during periods of minimum release
requirements of 8.23 million acre-feet (10,148 million
m³) per year the possibility exists that no operational
scenario will result in management objectives being
achieved for restoring sandbars, simply because of the
volume of water that must be released on an annual
basis. If so, other, more effective alternatives for restor-
ing and maintaining sandbars and related habitats may
need to be evaluated.
Sediment augmentation, one possible alterna-
tive, was eliminated during the development of the
EIS, partly because of the belief that sandbars could
be restored and maintained by constraining the hourly
ramping rates and range of daily dam operations and
partly because of concerns about contamination of sedi-
ment upstream in Lake Powell (Graf, 1985). Addition of
sediment—continuously, seasonally, or perhaps only dur-
ing floods—may offer greater powerplant operating flexi-
bility and therefore may cost less than further restrictions
on annual dam operations. To this end, the feasibility
of mechanically transporting fine sediment around Glen
Canyon Dam and introducing it into the Colorado River
below the dam is currently being investigated.
28 The State of the Colorado River Ecosystem in Grand Canyon
Discussion and Future
Research Needs
Extensive research and monitoring of fine-sediment
transport and sandbars since the completion of the EIS
have resulted in a better understanding of the geomor-
phology of the Colorado River in Marble and Grand
Canyons and of the effects of the operations of Glen
Canyon Dam on the river’s downstream resources. Prob-
ably the single most important finding of this research
and monitoring is that postdam mainstem sand transport
exceeds the postdam supply of sand from tributaries on
a seasonal to annual basis, such that the postdam river
is in an annual fine-sediment deficit (i.e., export exceeds
input). This sediment deficit has resulted in a consistent
downstream pattern of erosion of channel and sandbar
deposits from Marble and Grand Canyons despite restric-
tions on daily powerplant fluctuations required by the
implementation of the MLFF alternative.
The finding of an annual sediment deficit directly
contradicts the critical EIS assumption that sand will
accumulate on the bed of the Colorado River over mul-
tiple years under the MLFF operating alternative (and
minimum annual volume releases) and has important
implications for the potential success of managing tribu-
tary sediment inputs. It is also worth noting that the
EIS conclusion resulted fundamentally from a lack of
long-term records for tributary sand supply and main-
stem sand-transport rates, illustrating the importance of
long-term data sets in river management. A continu-
ous sediment budget for the Colorado River in Grand
Canyon since construction of Glen Canyon Dam,
based on high-frequency measurements, likely would
have resulted in a different EIS conclusion about fine-
sediment dynamics below the dam, one that may have
prevented the continued erosion of sandbars between
1991 and 2004.
A second important finding of recent research
and monitoring efforts is that during the 1996 BHBF
the primary source of sand for building high-elevation
sandbars was the low-elevation portion of the sandbars
instead of the channel bed as hypothesized in the EIS.
This scenario of building high-elevation sandbars at
the expense of the low-elevation portions was repeated
during the powerplant capacity flow in September 2000
(Hazel and others, in press). This process of sandbar
building is supported by the finding of an absence
of multiyear accumulation on the channel bed: sand
cannot be transported from the bed to high-elevation
sandbars because there is typically little sand available
on the channel bed.
Neither of these two findings supports the EIS
hypotheses, but they have led scientists and managers
to reassess the management strategy for sand resources
within Grand Canyon. An emerging paradigm is the
need to strategically time high-flow releases in order to
take advantage of sporadic tributary sediment inputs,
a scenario that requires greater flexibility in the annual
operating plan for the dam with respect to both hydro-
electric power generation and economic cost. Only
immediately after these inputs is significant sand avail-
able on the channel bed for transfer to high-elevation
sandbars through high-flow releases. Alternatively, dam
releases may be constrained following inputs for a period
of time until a high flow can be released from the dam;
however, during extended periods of above-average
upper Colorado River Basin hydrology and high storage
in Lake Powell, constraining daily operations may not be
possible (see fig. 1, 1995 through 1998). In the absence
of high-flow releases strategically timed to redistribute
tributary inputs to high-elevation sandbars, the inputs
are exported from Grand Canyon in a period of weeks
or months under normal dam operations, leading to
continued long-term erosion of sandbars.
In November 2004, this paradigm of strategically
timed, high-flow releases was tested for the first time
on the Colorado River. Scientists are in the process of
evaluating the results of this experiment. The findings
will be critical for the long-term management of fine-
sediment resources and sandbars in Grand Canyon. If a
management approach of strategically timed, high-flow
releases, triggered by tributary inputs, is to be followed,
then further research will be required to define the
appropriate triggering criteria and to develop high-flow
hydrographs (peaks and durations) that may optimize
deposition of tributary sand inputs within eddies while
minimizing export during controlled flood peaks.
If strategically timed, high-flow releases are deemed
inadequate for meeting the management objectives for
Grand Canyon sandbars, then alternative approaches
must be considered, such as further restraints on daily
powerplant operations, changes in monthly volume
release patterns, or sediment augmentation.
Influence of Glen Canyon Dam Operations on Downstream Sand Resources 29
References
Andrews, E.D., 1990, The Colorado River: a perspec-
tive from Lees Ferry, Arizona, in Wolman, M.G.,
Church, M., Newbury, R., Lapointe, M., Frenette, M.,
Andrews, E.D., Lisle, T.E., Buchanan, J.P., Schumm,
S.A., and Winkley, B.R., The riverscape, in Wolman,
M.G., and Riggs, H.C., eds., Surface water hydrol-
ogy, The geology of North America, v. O-1: Boulder,
Colo., Geological Society of America, p. 281–328.
Andrews, E.D., 1991, Sediment transport in the Colorado
River basin, in Committee to Review the Glen Canyon
Environmental Studies, Water Science and Technol-
ogy Board, Commission on Geosciences, Environment,
and Resources, eds., Colorado River ecology and dam
management: Washington, D.C., National Academy
Press, p. 54–74.
Dolan, R., Howard, A., and Gallenson, A., 1974, Man’s
impact on the Colorado River in the Grand Canyon:
American Scientist, v. 62, p. 392–401.
Edwards, T.K., and Glysson, G.D., 1999, Field methods
for measurement of fluvial sediment: U.S. Geological
Survey Techniques of Water Resources Investigations,
book 3, chap. C2, 89 p.
Flynn, M.E., and Hornewer, N.J., 2003, Variations in
sand storage measured at monumented cross sections
in the Colorado River between Glen Canyon Dam
and Lava Falls Rapid, northern Arizona, 1992–99:
U.S. Geological Survey Water Resources Investigations
Report 03-4104, 39 p.
Garrett, W.B., Van De Vanter, E.K., and Graf, J.B., 1993,
Streamflow and sediment-transport data, Colorado
River and three tributaries in Grand Canyon, Arizona,
1983 and 1985–86: U.S. Geological Survey Open-File
Report 93-174, 624 p.
Graf, W.L., 1985, Mercury transport in stream sediments
of the Colorado Plateau: Annals of the Association of
American Geography, v. 75, p. 552–565.
Grams, P.E., Schmidt, J.C., Topping, D.J., and Goeking,
S., 2004, The degraded reach: rate and pattern of bed
and bank adjustment of the Colorado River in the 25
km immediately downstream from Glen Canyon Dam:
final report to the Grand Canyon Monitoring and
Research Center: Logan, Utah State University, 111 p.
Hazel, J.E., Topping, D.J., Schmidt, J.C., Kaplinski, M.,
and Melis, T.S., in press, Downstream effects of a dam
on sediment storage in a bedrock canyon: the rela-
tive roles of eddy and channel storage in the sediment
budget for the Colorado River in Marble Canyon, AZ:
Journal of Geophysical Research, Earth Surface.
Hazel, J.E., Jr., Kaplinski, M., Parnell, R., Manone, M.,
and Dale, A., 1999, Topographic and bathymetric
changes at thirty-three long-term study sites, in Webb,
R.H., Schmidt, J.C., Marzolf, G.R., and Valdez,
R.A., eds., The controlled flood in Grand Canyon:
Washington, D.C., American Geophysical Union,
Geophysical Monograph Series, v. 110, p. 161–184.
Howard, A., and Dolan, R., 1981, Geomorphology of
the Colorado River in the Grand Canyon: Journal of
Geology, v. 89, p. 269–298.
Laursen, E., Ince, S., and Pollack, J., 1976, On sediment
transport through the Grand Canyon, in Proceedings
of the Seventh Federal Interagency Sediment Confer-
ence, Denver, Colo., p. 4-76–4-87.
Melis, T.S., Topping, D.J., and Rubin, D.M., 2004, Test-
ing laser-based sensors for continuous in situ moni-
toring of suspended sediment in the Colorado River,
Arizona, in Bogen, J., Fergus, T., and Walling, D.E.,
eds., Erosion and sediment transport measurement in
rivers: technological and methodological advances:
International Association of Hydrological Sciences
Publication 283, p. 21–27.
Pemberton, E.L., 1987, Sediment data collection and
analysis for five stations on the Colorado River from
Lees Ferry to Diamond Creek: Glen Canyon Envi-
ronmental Studies, Salt Lake City, Utah, Bureau of
Reclamation, Upper Colorado Region, 156 p.
Randle, T.J., and Pemberton, E.L., 1987, Results and
analysis of STARS modeling efforts of the Colorado
River in Grand Canyon: Glen Canyon Environmental
Studies, Salt Lake City, Utah, Bureau of Reclamation,
Upper Colorado Region, 190 p.
Rubin, D.M., Nelson, J.M., and Topping, D.J., 1998,
Relation of inversely graded deposits to suspended-
sediment grain-size evolution during the 1996 flood
experiment in Grand Canyon: Geology, v. 26, p.
99–102.
30 The State of the Colorado River Ecosystem in Grand Canyon
Rubin, D.M., and Topping, D.J., 2001, Quantifying the
relative importance of flow regulation and grain-size
regulation of suspended-sediment transport (α), and
tracking changes in bed-sediment grain size (β): Water
Resources Research, v. 37, p. 133–146.
Rubin, D.M., Topping, D.J., Schmidt, J.C., Hazel, J.,
Kaplinski, M., and Melis, T.S., 2002, Recent sediment
studies refute Glen Canyon Dam hypothesis: Eos,
Transactions, American Geophysical Union, v. 83, no.
25, p. 273, 277–278.
Schmidt, J.C., 1999, Summary and synthesis of geo-
morphic studies conducted during the 1996 controlled
flood in Grand Canyon, in Webb, R.H., Schmidt, J.C.,
Marzolf, G.R., and Valdez, R.A., eds., The controlled
flood in Grand Canyon: Washington, D.C., American
Geophysical Union, Geophysical Monograph Series, v.
110, p. 329–341.
Schmidt, J.C., Topping, D.J., and Grams, P.E., 2004, Sys-
tem-wide changes in the distribution of fine-grained
alluvium in the Colorado River corridor between Glen
Canyon Dam and Bright Angel Creek, Arizona: final
report to the Grand Canyon Monitoring and Research
Center: Logan, Utah State University, 117 p.
Smillie, G.M., Jackson, W.L., and Tucker, D., 1993,
Colorado River sand budget: Lees Ferry to Little
Colorado River: National Park Service Technical
Report NPS/NRWRD/NRTR-92/12, 11 p.
Topping, D.J., Melis, T.S., Rubin, D.M., and Wright,
S.A., 2004, High-resolution monitoring of suspended-
sediment concentration and grain size in the Colorado
River in Grand Canyon using a laser-acoustic system,
in Chunhong Hu, Ying Tan, and Cheng Liu, eds.,
Proceedings of the Ninth International Symposium on
River Sedimentation, October 18–21, 2004, Yichang,
China, v. 4, p. 2507–2514.
Topping, D.J., Rubin, D.M., Nelson, J.M., Kinzel, P.J.,
III, and Bennett, J.P., 1999, Linkage between grain-
size evolution and sediment depletion during Colorado
River floods, in Webb, R.H., Schmidt, J.C., Marzolf,
G.R., and Valdez, R.A., eds., The 1996 controlled
flood in Grand Canyon: Washington, D.C., American
Geophysical Union, Geophysical Monograph Series, v.
110, p. 71–98.
Topping, D.J., Rubin, D.M., Nelson, J.M., Kinzel, P.J.,
III, and Corson, I.C., 2000a, Colorado River sedi-
ment transport: pt. 2: systematic bed-elevation and
grain-size effects of supply limitation: Water Resources
Research, v. 36, p. 543–570.
Topping, D.J., Rubin, D.M., and Vierra, L.E., Jr., 2000b,
Colorado River sediment transport: pt. 1: natural
sediment supply limitation and the influence of Glen
Canyon Dam: Water Resources Research, v. 36, p.
515–542.
Topping, D.J., Schmidt, J.C., and Vierra, L.E., Jr., 2003,
Computation and analysis of the instantaneous-dis-
charge record for the Colorado River at Lees Ferry,
Arizona—May 8, 1921, through September 30, 2000:
U.S. Geological Survey Professional Paper 1677, 118 p.
U.S. Department of the Interior, 1995, Operation of
Glen Canyon Dam Final Environmental Impact State-
ment: Salt Lake City, Utah, Bureau of Reclamation,
Upper Colorado Region, 337 p., appendices.
U.S. Department of the Interior, 1996, Record of Deci-
sion, operation of Glen Canyon Dam: Washington,
D.C., Office of the Secretary of the Interior, 13 p.
U.S. Department of the Interior, 2002, Proposed experi-
mental releases from Glen Canyon Dam and removal
of nonnative fish: environmental assessment: Salt Lake
City, Utah, Bureau of Reclamation, Upper Colorado
Region, 112 p., appendices.
Webb, R.H., 1996, Grand Canyon, a century of change:
Tucson, University of Arizona Press, 290 p.
Webb, R.H., Griffiths, P.G., Melis, T.S., and Hartley,
D.R., 2000, Sediment delivery by ungaged tributar-
ies of the Colorado River in Grand Canyon, Arizona:
U.S. Geological Survey Water Resources Investigation
Report 00-4055, 67 p.
Webb, R.H., Melis, T.S., and Valdez, R.A., 2002, Obser-
vations of environmental change in Grand Canyon:
U.S. Geological Survey Water Resources Investigation
Report 02-4080, 33 p.
Influence of Glen Canyon Dam Operations on Downstream Sand Resources 31
Contact Information:
Scott A. Wright
Hydrologist
U.S. Department of the Interior
U.S. Geological Survey
Southwest Biological Science Center
Flagstaff, AZ
sawright@usgs.gov
Theodore S. Melis
Physical Scientist
U.S. Department of the Interior
U.S. Geological Survey
Southwest Biological Science Center
Flagstaff, AZ
tmelis@usgs.gov
David J. Topping
Research Hydrologist
U.S. Department of the Interior
U.S. Geological Survey
Water Resources Discipline, National Research Program
Denver, CO
dtopping@usgs.gov
David M. Rubin
Geologist
U.S. Department of the Interior
U.S. Geological Survey
Pacific Science Center
Santa Cruz, CA
drubin@usgs.gov
J
e
f
f
S
o
r
e
n
s
e
n
,
A
r
i
z
o
n
a
G
a
m
e
a
n
d
F
i
s
h
D
e
p
a
r
t
m
e
n
t
First page photograph credit: Jeff Sorensen, Arizona Game and Fish Department
Chapter 2
Fishes of
Grand Canyon
Steven P. Gloss
Lewis G. Coggins
Introduction
Fishes of the Colorado River vary from coldwater
trout species found in the river’s mountainous headwa-
ters to uniquely adapted desert river species found at
lower elevations. Within the study area, the Colorado
River corridor between Glen Canyon Dam and the west-
ern boundary of Grand Canyon National Park (hereafter
Grand Canyon), the Colorado River was a seasonally
warm and turbid river characterized by large seasonal
variations in flow before it was altered by the closure of
Glen Canyon Dam in 1963 (Topping and others, 2003).
Although water temperatures fluctuated between 32°F
(0°C) during winter to a high approaching 86°F (30°C)
during late summer, several warmwater native fish spe-
cies successfully inhabited this stretch of the river (Cole
and Kubly, 1976). Because of the harsh environment
created by dramatic seasonal fluctuations in the river’s
predam flow and temperature, only 8 of the 32 species
of native fish historically found in the Colorado River
were common in the Grand Canyon reach of the river.
Other native fishes within the study area were restricted
to small tributary streams or occurred only in transient
or seasonal numbers. Of the eight fish species that were
originally common to the study area, only four species
are known to persist today.
The number of species that made up the original
fish community of the Colorado River was altered well
before the construction of mainstem dams because
of the introduction of nonnative fishes by early Euro-
pean settlers. Nonnative fishes, from sport fishes to
escapees from aquaria, have been intentionally and
inadvertently stocked in the Colorado River for more
than 100 yr (Mueller and Marsh, 2002). Today, non-
native fishes originating in many parts of the world
are found in the Colorado River. Table 1 contains a
list of the native and nonnative fishes of the Colorado
River in Grand Canyon.
This chapter examines the status, trends, and recent
condition of Grand Canyon fishes, focusing particular
attention on the endangered humpback chub (Gila cypha)
because of its prominence within the Glen Canyon
Dam Adaptive Management Program (see Overview,
this report). The chapter begins with a discussion of the
conditions that led to the development of the Grand
Canyon’s unique native fish populations and then moves
on to the reasons for their decline. The effects of the
modified low fluctuating flow (MLFF) alternative on fish
34 The State of the Colorado River Ecosystem in Grand Canyon
Table 1. Historical and present relative abundance of fish species in the Colorado River from Glen Canyon to Separation
Canyon. P = present, abundance unknown; A = abundant; C = common; LC = locally common; R = rare; and - = not encountered.
[Modified from Valdez and Ryel, 1995. Species that are federally listed as endangered are indicated by an asterisk (*). Species that are endemic
to the lower basin of the Colorado River but occurred almost exclusively in smaller streams or rivers tributary to the mainstem Colorado River
are indicated by a plus sign (+)]
Species Pre-1850 1958–59 1970–73 1984–86 1990–93
Family: Clupeidae, shads (introduced)
Threadfin shad - - R - C
Family: Cyprinidae, minnows
Native
*Humpback chub P - R R LC
*Bonytail chub P - - - -
Roundtail chub P R - - -
*Colorado pikeminnow P R - - -
Speckled dace P A A A C
Virgin spinedace+ P - R - -
Woundfin+ P - - - -
Introduced
Red shiner - - R - A
Common carp - C A A A
Utah chub - R - R -
Golden shiner - - R R R
Fathead minnow - A C A LC
Family: Catostomidae, suckers (all native)
Bluehead sucker P C C C C
Flannelmouth sucker P C C C C
*Razorback sucker P R - R -
Family: Ictaluridae, bullhead catfishes (all introduced)
Black bullhead - C - R R
Yellow bullhead - - - R -
Channel catfish - A C R LC
Family: Salmonidae, salmon and trout (all introduced)
Cutthroat trout - - - R -
Coho salmon - - R - -
Rainbow trout - - C A A
Brown trout - - - C C
Brook trout - - - C R
Family: Cyprinodontidae, killifishes (introduced)
Plains killifish - R C R LC
Family: Poeciliidae, livebearers (introduced)
Mosquitofish - R R - LC
Family: Percichthyidae, temperate basses (introduced)
Striped bass - - - R R
Family: Centrarchidae, sunfishes (all introduced)
Green sunfish - C R R R
Bluegill - R R - R
Largemouth bass - R R R R
Black crappie - - - - R
Family: Percidae, perch (all introduced)
Yellow perch - R - - -
Walleye - - - - R
Total number of species 10 17 18 20 22
Fishes of Grand Canyon 35
populations are also examined. The chapter concludes
with a discussion of possible management options to
slow or reverse the decline of humpback chub numbers.
Background
The Colorado River was one of the last areas of the
continental United States to be explored by Europeans; it
was first traversed during the expedition headed by John
Wesley Powell in 1869. For this reason, it is not surpris-
ing that scientific descriptions of many of the organisms
in the Colorado River corridor, especially the fishes,
did not begin until the 1930s and 1940s; earlier expe-
ditions collected and described fishes generally rather
than specifically. Emery and Ellsworth Kolb, explorers
and photographers of the Colorado River in the early
1900s, reported that fishes were very abundant (Kolb
and Kolb, 1914). The humpback chub was the last of
the native fishes in Grand Canyon to be described in
1946 by Robert R. Miller from specimens taken from the
Colorado River in Grand Canyon (Miller, 1946).
Scientific description of the native fishes of Grand
Canyon showed that these species were unique in at least
two ways. Most noticeably, several of the species share
unusual body shapes, including large adult body size,
small depressed skulls, large predorsal humps or keels,
and small eyes, which presumably developed as adapta-
tions to life in a large, turbid, and seasonably variable
riverine environment. These features are perhaps best
observed in the razorback sucker (Xyrauchen texanus) and
the humpback chub (see accompanying text box, p. 51).
A second, and perhaps more important, measure
of the uniqueness of Grand Canyon native fishes is that
most of these species are not found elsewhere in the
world. Organisms that are native to a certain location
and do not occur anywhere else are called endemic spe-
cies. Of the eight native species common to the Grand
Canyon, six are species endemic to the Colorado River
Basin. As early as 1895, scientists recognized the special
nature of Colorado River fishes and the high rates of
endemism (Minckley, 1991). Later research did not alter
this conclusion, and despite a relatively low number
of species compared to other drainages in the United
States, the Colorado River Basin has a recognized ende-
mism at the species level of approximately 75% and sup-
ports the most distinctive ichthyofauna in North America
(Minckley, 1991).
Before European settlement, the native fishes found
in the Grand Canyon portion of the Colorado River
were exclusively minnows and suckers. The biggest of
these fish was the Colorado pikeminnow (Ptychocheilus
lucius), which is also the largest of all native minnow
(cyprinid) species in North America and was found
only in the Colorado River Basin (fig. 1). Called a white
salmon by early settlers, the Colorado pikeminnow
reached up to 6 ft (2 m) in length and had a weight of up
to 80 lb (36 kg) (Mueller and Marsh, 2002).
Today, three of the eight native fish species have
been eliminated from the Colorado River in Glen and
Grand Canyons (roundtail chub (Gila robusta), bonytail
chub (Gila elegans), and Colorado pikeminnow), and two
are federally listed as endangered (humpback chub and
razorback sucker) under the Endangered Species Act.
Although listed as an endangered species with designated
critical habitat in Grand Canyon, the razorback sucker
has rarely been collected (Minckley, 1991; Valdez and
Carothers, 1998) and is widely thought to no longer be
found in Grand Canyon. The status of the flannelmouth
sucker (Catostomus latipinnis) is common, and the species
persists in the study area and throughout much of the
upper Colorado River Basin. The remaining two fish
(bluehead sucker (Catostomus discobolus) and speckled dace
(Rhinichthys osculus)) are relatively common. Brief descrip-
tions of the life histories of all the Grand Canyon native
fishes can be found in Minckley (1991); this chapter
provides text boxes (see p. 50) with summary information
for the four native fishes that continue to inhabit Grand
Figure 1. Historical photograph (date unknown) of someone
identified as James Fagen holding a large Colorado pikeminnow in
lower Granite Gorge (courtesy of the Kolb Collection, Cline Library,
Northern Arizona University, NAU.PH.568.5737).
36 The State of the Colorado River Ecosystem in Grand Canyon
Canyon, as well as for the two most common nonnative
species, rainbow trout (Oncorhynchus mykiss) and brown
trout (Salmo trutta).
Decline of Native Fish
Introductions of Nonnative Fishes
There are a number of reasons for the decline of
native fishes, including the potential effects of nonna-
tive fish species. Nonnative fish have been found in the
Colorado River since the 1800s (Minckley, 1991). These
species are potential predators of and competitors with
native fish and include common carp (Cyprinus carpio),
fathead minnow (Pimephales promelas), plains killifish
(Fundulus zebrinus), rainbow trout, brown trout, red shiner
(Cyprinella lutrensis), and channel catfish (Ictalurus punctatus).
Nonnative species may share rearing habitats used by
native fish, habitats which include complex shorelines,
tributaries, backwater areas, and eddies. The presence
of warmwater, coolwater, and coldwater nonnative fish
species in the Colorado River is an issue of consider-
able importance (U.S. Department of the Interior,
1995) because there are now nonnative fishes that may
negatively interact with native fishes under virtually any
temperature regime and in any habitat of the river.
Today, the Colorado River has nearly twice as many
nonnative species (60) as native species (32); in the Grand
Canyon reach of the river the situation is even more
extreme, where the ratio of native to nonnative spe-
cies is more than 4 to 1 (Valdez and Carothers, 1998).
The introduction of nonnative species to the Colorado
River, both intentionally and unintentionally, was well
underway before 1900. As such, the ratio of nonnative
to native fishes was high in Grand Canyon before the
construction of Glen Canyon Dam. For example, the
National Park Service introduced both brown trout and
rainbow trout to tributaries like Bright Angel Creek in
the 1920s to provide sport fishing opportunities (Valdez
and Carothers, 1998). Because of the continuous nature
of the river and its tributaries before dam building, spe-
cies introduced almost anywhere in the basin had the
potential to find their way to the Grand Canyon por-
tion of the river, and many did. Before Glen Canyon
Dam, the Grand Canyon reach was dominated by a
single introduced species, the channel catfish (Valdez
and Carothers, 1998). Following construction of the dam
in 1963, Federal and State agencies again introduced
rainbow trout below Glen Canyon Dam to establish and
maintain a sport fishery in the 15-RM reach between the
dam and Lees Ferry. This stocking continued for more
than 30 yr, until the mid-1990s. Numerous other spe-
cies of nonnative fishes were also introduced into Lake
Powell and Lake Mead to create or enhance recreational
fishing (Mueller and Marsh, 2002).
The effects of nonnative fish on native species,
including predation and competition, are important
considerations when evaluating any management action
intended to benefit native fishes. These considerations
are particularly important given the proximity of Lake
Powell and Lake Mead, reservoirs with diverse nonna-
tive fish populations, to Grand Canyon. Any manage-
ment action intended to improve habitat conditions for
native warmwater fishes also runs the risk of providing
additional habitat that is suitable for nonnative predators
and competitors. Nonnative fish predators currently in
the Grand Canyon reach of the Colorado River include
striped bass (Morone saxatilis), channel catfish, largemouth
bass (Micropterus salmoides), green sunfish (Lepomis cyanellus),
brown trout, and rainbow trout. Currently, nonnative
coldwater species (trout) are abundant, while the nonna-
tive warmwater species exist in relatively low numbers.
Glen Canyon Dam Effects
The predam success of nonnative species was, in
part, due to the fact that the river was generally what
fishery biologists term a “warmwater habitat.” The
annual temperature cycle of the Colorado River through
Grand Canyon was similar to temperate lakes and
streams at lower elevations, where temperatures ranged
from cold or cool in winter to warm in summer. Native
species require warmer temperatures to spawn and
reproduce successfully. This seasonal pattern also allowed
many of the introduced species to complete their life
cycle. One of the major impacts of Glen Canyon Dam
on the Colorado River was the change in water tempera-
ture to a relatively cold, steady temperature that favored
coldwater species like trout over native fishes and intro-
duced, warmwater species. While most of the warmwa-
ter species can survive in these colder waters, they cannot
reproduce and do not grow well, having been adapted to
at least seasonally warmer temperatures.
Other possible effects of dam operations on the
riverine environment that may affect fishes include
increased water clarity, altered flow patterns, and
reduced sediment. All species that are native to Grand
Canyon evolved in highly turbid environments, so the
clear water released from the dam may favor nonnative
Fishes of Grand Canyon 37
predators like trout, which are adapted to hunting in
clear water (Valdez and Ryel, 1995). Similarly, the post-
dam river hydrology is different from the predam river
with respect to daily flow variation, flood frequency, and
seasonal pattern and magnitude of maximum and mini-
mum flows (Topping and others, 2003). These alterations
in flow patterns potentially affect the spawning cues,
habitat use, and distribution of native fish, as well as the
suitability of mainstem Colorado River rearing habitat,
in ways that are largely unknown and potentially com-
plex (Korman and others, 2004). Finally, as Glen Canyon
Dam blocks the majority of sediment transported by the
Colorado River to the upstream portions of Lake Powell,
the nearshore physical habitat available to native fish is
fundamentally different from the predam river (Goeking
and others, 2003; also see chapter 1, this report). Except
for temperature, the other potential effects of the dam
that are mentioned here are based on inferences about
what is known regarding fishes from other river systems.
Little direct scientific evidence from the Colorado River
itself exists regarding these effects, and there remains
considerable uncertainty regarding the potential effects
of management actions associated with these factors
(Walters and others, 2000).
Other Factors
New fish parasites in the system, changes in tribu-
tary hydrology, and alterations in the food base that
support fish populations are additional environmental
factors that may be affecting native and nonnative fish
species in Grand Canyon. Asian tapeworm (Bothriocepha-
lus acheilognathi), a parasitic cestode, is a prominent exam-
ple of a recently introduced parasite. Introduced into
the United States in the 1970s with imported grass carp
(Ctenopharyngodon idella) from China, the Asian tapeworm
was discovered in 1990 in the Little Colorado River,
which is an important spawning area for humpback chub
(Choudhury and others, 2004). The tapeworm can cause
mortality, but most often it is responsible for reduced
growth and poor condition of infected fish. This para-
site is currently restricted to the Little Colorado River
because cold mainstem temperatures preclude comple-
tion of its life cycle. The Little Colorado River is also
an example of a tributary system in which upstream
water use and development have changed the amount
and timing of flows reaching the Colorado River.
Such changes could affect fishes in the Little Colorado
River and throughout Grand Canyon, especially below
the tributary.
Status and Trends
Until the 1990s, there were few attempts to monitor
the status and trends of fishes in Grand Canyon. Infor-
mation before the mid- to late-1980s was anecdotal and
was provided by explorers, river runners, and occasional
scientific expeditions. As a result, few data are available
for the first 20 yr after Glen Canyon Dam was closed.
Early fish collection efforts were reviewed by Valdez and
Carothers (1998), and where appropriate these earlier
data are used in comparison to current data for fishes in
Grand Canyon.
Efforts to estimate population size or relative abun-
dance of fishes in Grand Canyon began under Glen
Canyon Environmental Studies Phase II when private
consulting firms, university researchers, the U.S. Fish
and Wildlife Service (USFWS), and the Arizona Game
and Fish Department conducted surveys and under-
took population estimates in the mainstem Colorado
River and in the Little Colorado River. Beginning in
1997, these efforts became the responsibility of the U.S.
Geological Survey’s (USGS) Grand Canyon Monitoring
and Research Center, which has worked cooperatively
on monitoring activities with the U.S. Fish and Wildlife
Service, the Arizona Game and Fish Department, and
consulting firms (SWCA Environmental Consultants,
Inc., and Ecometric Research). For the purposes of
monitoring, the study area is divided into three seg-
ments: the Lees Ferry reach (15 RM of Colorado River
corridor from Glen Canyon Dam to Lees Ferry); the
mainstem Colorado River (downstream of Lees Ferry,
RM 0, and the Paria River to RM 226 at the conflu-
ence of Diamond Creek); and the Little Colorado River
(the 8.7 mi (14 km) of the tributary upstream from the
mainstem). The status and trends of fish found in each
of these reaches will be discussed separately. Humpback
chub are discussed in a separate section.
Lees Ferry
The Lees Ferry reach of the river is managed pri-
marily as a rainbow trout sport fishery. The Lees Ferry
reach is known as a tailwater trout fishery because it
occurs downstream from a large dam where deepwater
discharges afford cooler water temperatures that allow
coldwater species like trout to survive. In fact, trout not
only survived in the Lees Ferry reach following their
initial stocking in 1964 but also flourished in the new
habitat created by Glen Canyon Dam. The Lees Ferry
38 The State of the Colorado River Ecosystem in Grand Canyon
rainbow trout fishery gained a reputation by the mid-
1970s as a world class, blue ribbon fishery famous for its
scenic grandeur and large, trophy-sized trout. Monitor-
ing in this reach is primarily done through electrofish-
ing and surveys of anglers by the Arizona Game and
Fish Department in cooperation with the USGS Grand
Canyon Monitoring and Research Center. The fishery
was initiated with stocking efforts and was maintained
primarily by stocking until the late 1990s.1 Since closure
of Glen Canyon Dam in 1963, however, this fishery has
experienced variable success rates by anglers, and the
trout populations have changed in response to stocking,
dam releases, and food availability (McKinney and oth-
ers, 1999, 2001).
Recently, more stable river flows, which are the
result of the interim flows in 1991 and subsequent
implementation of the MLFF alternative in 1996, have
encouraged natural reproduction and made stocking
unnecessary. Stable flows and increased natural repro-
duction resulted in an expanding number of fish (fig.
2), but the larger number of fish was offset by smaller
average size and decreasing condition (plumpness) of the
fish (fig. 3). Because the overall carrying capacity of the
river remains relatively constant, the Lees Ferry reach
is able to produce a smaller number of large fish or a
greater number of small fish, a principle that is known as
conservation of biomass. As early as 1996, the Arizona
Game and Fish Department recognized the declining
size of trout in this fishery and recommended changes in
angling regulations to increase the size of fish; however,
anglers appeared unwilling to accept lower catch rates
of larger fish (Niccum and others, 1998). Average fish
condition continued to decline for several more years but
finally rebounded in 2002 (fig. 3).
As part of the Glen Canyon Dam Adaptive
Management Program, fluctuating nonnative fish sup-
pression flows were initiated beginning in 2003 and
continued through 2005 in an effort to reduce the
number of trout and increase their average size. The
experimental flow treatment involved increased diur-
nal flow fluctuations of 5,000 to 20,000 cubic feet per
second (cfs) from January through March of each year.
Overall, these fluctuating flows were intended to disrupt
spawning activity, to reduce egg survival, and to disad-
vantage young-of-year (YOY) trout that did survive. Early
indications suggest that these experimental flows have
had only minimal effects on the recruitment dynamics of
rainbow trout. The total egg deposition loss because of
Glen Canyon Dam operations in 2003 ranged from 30%
to 40% in the Lees Ferry reach, with about half of this
mortality being a direct consequence of the enhanced
fluctuating flows in January through March (Korman
and others, 2005); however, electrofishing catch rates
began to increase in 2003 (fig. 2). There also appears to
be a corresponding increase in angler use associated with
1 Stocking of fingerling rainbow trout was reduced in the mid-
1990s to about 20,000 fish per year and ended completely in 1999
when it was apparent that natural reproduction under the modified
low fluctuating flow alternative was producing more than enough
recruitment to sustain the fishery (William R. Persons, Arizona Game
and Fish Department, oral commun., 2005).
Figure 2. The average number of rainbow trout caught by using
electrofishing at several fixed sampling locations in the Lees Ferry
reach of the Colorado River from 1991 to 2003. Increasing catch-
per-unit effort is thought to be indicative of an increasing number
of fish in the population (Arizona Game and Fish Department and
U.S. Geological Survey, unpub. data, 2005).
Figure 3. Condition factor, or relative weight, of Lees Ferry
trout from 1991 to 2003. Condition factor expresses the length-
to-weight relationship and is an attribute that reflects the health
of individual fish as well as affects angler satisfaction. Relative
weight declined with the increase in fish density in the late 1990s
but increased in 2002–03. Present condition seems acceptable to
anglers (Arizona Game and Fish Department and U.S. Geological
Survey, unpub. data, 2005).
Fishes of Grand Canyon 39
the increased electrofishing catch rate and the implemen-
tation of fluctuating flows (see chapter 9, this report).
Otoliths (minute boney structures found in the inner
ear) of young rainbow trout (fig. 4) were examined in
2003 and 2004 to infer growth rate patterns during the
late spring and summer months following the end of
fluctuating nonnative fish suppression flows. Microscopic
examination of these bony structures allows research-
ers to determine daily growth patterns. Results of these
examinations suggest that YOY rainbow trout experi-
enced more growth on Sundays than on other days of
the week in 2003; however, otoliths collected in 2004
do not display increased growth on Sundays. Korman
and others (2005) hypothesized that this difference was
related to less severe flow fluctuations on Sundays during
2003 as compared to 2004.
Mainstem Colorado River
Management objectives of the Glen Canyon Dam
Adaptive Management Program call for managing the
mainstem Colorado River and its tributaries below the
Paria River for the benefit of native fishes (GCDAMP,
2001, http://www.usbr.gov/uc/rm/amp/amwg/mtgs/
02jan17/Attach_06.pdf, accessed July 14, 2005). Fish
monitoring in the mainstem Colorado River is primarily
conducted by electrofishing or with trammel nets, hoop
nets, and beach seines. Each of these methods is “selec-
tive,” or has higher efficiency for particular species or
fish sizes. For instance, electrofishing is very effective in
catching rainbow and brown trout and common carp
but is inefficient in capturing adult humpback chub.
Alternatively, trammel and hoop nets are more efficient
than electrofishing in capturing humpback chub. These
differences in sampling gear efficiency, coupled with
differences in abundance, influence the ability of the
monitoring program to detect differences in abundance
over time and space.
The current monitoring program, which uses elec-
trofishing for rainbow trout, brown trout, and common
carp, is able to show trends in the abundance of these
species over time and space (fig. 5 a, b, c). The abun-
dance of rainbow trout declines as a function of distance
Figure 4. Photomicrograph of an otolith cross-section of young-of-year rainbow trout sampled from Glen Canyon in April 2003. Otoliths
are minute boney structures found in the inner ear that show daily growth patterns in many fishes. The image shows the weekly striping
pattern (identified by white arrows and shown at magnifications of 16x (A) and 400x (B)) caused by increased growth during lower peak
Sunday flows (8,000 cfs) during April 2003 when normal weekday operations ranged from 7,000–13,000 cfs on a 24-h cycle (photographs
courtesy of Steven Campana, Bedford Institute of Oceanography, Canada).
A. B.
40 The State of the Colorado River Ecosystem in Grand Canyon
downstream of Glen Canyon Dam, but common carp
increase downstream. Brown trout abundance is cen-
tered near RM 88 and declines with distance upstream
or downstream of this location. This pattern is explained
most readily by the occurrence of several tributaries in
this reach that are suitable for spawning by this species.
Monitoring efforts in the mainstem Colorado River
for both native and nonnative species have generally
resulted in an adequate description of species distribu-
tion. In general, humpback chub distribution is centered
near the Little Colorado River where successful spawn-
ing and rearing is known to occur (Douglas and Marsh,
1996; Gorman and Stone, 1999). Also, humpback chub
occur in several other smaller aggregations throughout
the river corridor (see below). Flannelmouth sucker,
bluehead sucker, and speckled dace abundance typi-
cally increases with distance downstream of the Little
Colorado River and is generally high near major tribu-
tary confluences (e.g., Little Colorado River, Paria River,
Kanab Creek, and Bright Angel Creek) (Gorman and
Coggins, 2000; Johnstone and others, 2003; Johnstone
and Lauretta, 2004). Warmwater nonnative species such
as channel catfish and striped bass increase in abundance
with distance from Glen Canyon Dam, particularly
below RM 160. Small-bodied, nonnative fish such as
fathead minnow, red shiner, and plains killifish are found
almost exclusively downstream of the Little Colorado
River confluence, and all evidence suggests that this
tributary is the dominant source of these fishes in the
Colorado River ecosystem (Johnstone and others, 2003;
Johnstone and Lauretta, 2004).
Although the current monitoring program is suffi-
cient to describe these general patterns in distribution of
native and certain nonnative fishes, it cannot provide a
specific measure of trends in relative abundance. Despite
sampling efforts that are randomly distributed over the
226 mi (364 km) of river from Lees Ferry to Diamond
Creek, the monitoring program is unable to measure
with any certainty the spatial or temporal trends in the
relative abundance of native or nonnative fishes in the
mainstem Colorado River. An exception is the abun-
dance and distribution of rainbow trout, brown trout,
common carp, and the Little Colorado River population
of humpback chub previously discussed. Low abundance
of these fishes coupled with the very poor sampling
efficiency of current sampling gear make measuring
trends in relative abundance difficult. Typically, monitor-
ing efforts include over 600 trammel net sets each year
and between 100 and 200 seining sites. Several examples
of the low and highly variable catch rate experienced
with trammel nets are illustrated for select species and
sites in figure 6.
Figure 5. Relative abundance (mean catch-per-unit efforts, or
fish/hour) of rainbow trout (A), brown trout (B), and common carp
(C) as indicated by electrofishing catch rates from Lees Ferry (RM
0) to Diamond Creek (RM 226) (Arizona Game and Fish Department
and U.S. Geological Survey, unpub. data, 2005). Note inverse
abundance of coldwater trout to warmwater carp as distance
from Lees Ferry and Diamond Creek increases. Increase in brown
trout abundance in the middle of Grand Canyon is thought to
be caused by spawning, which occurs in Bright Angel Creek, a
tributary at RM 88. The National Park Service is trying to reduce
spawning in Bright Angel Creek.
A.
B.
C.
Fishes of Grand Canyon 41
Figure 6. Trends in the relative abundance (trammel net catch rate, fish/hour) of selected species near the confluences of several tributaries where native fishes, particularly
the suckers, attempt to reproduce. These figures illustrate the inability of the current monitoring program to detect all but extremely large changes in the relative abundance
of key native and nonnative species in most areas of the Colorado River. Error bars depict 95% confidence intervals for mean catch rate. Note that catch rate estimates with
overlapping confidence intervals are statistically insignificant and represent years of no statistically apparent difference in relative abundance (U.S. Geological Survey, unpub.
data, 2005).
42 The State of the Colorado River Ecosystem in Grand Canyon
The presence of many nonnative fish in the system
has created a substantial management challenge. It is
known that some of these nonnative species, particu-
larly brown trout, prey upon native fishes (Valdez and
Carothers, 1998). Furthermore, nonnative species may
compete for habitat and food with native species in ways
that are difficult to document. Monitoring the relative
abundance of nonnative fish in this part of the river
provides some insight into the potential severity of the
problem. Both coldwater nonnative species such as trout
and warmwater fishes such as carp inhabit the river.
Coldwater species dominate the upstream reaches of
Grand Canyon, whereas warmwater species are more
prominent further downstream because the tempera-
ture of the river water gradually increases after leaving
the dam.
Little Colorado River
The Little Colorado River, which flows into the
Colorado River at RM 61, represents perhaps the best
remaining native fish habitat in Grand Canyon under
the current temperature and flow management regimes
in the Colorado River. Because native fish are abundant
and the sampling gear is efficient in the Little Colorado
River, relative abundance of native fish and some non-
native fish can be well determined in this tributary.
Two kinds of fish sampling are conducted in the Little
Colorado River: spring and fall hoop netting aimed pri-
marily at collecting humpback chub to estimate popula-
tion size and hoop netting conducted in April and May
at fixed sites in the lower 0.75 mi (1,200 m) of the river.
The humpback chub data are discussed separately below.
Despite the presence of several nonnative fishes in the
Little Colorado River, the catch in hoop nets suggests
that native fish (>80%) dominate the fish community
in most years (fig. 7). The data from the lower 0.75 mi
(1,200 m) sampling depict trend information for the rela-
tive abundance of three native species: humpback chub
(fig. 8), bluehead sucker (fig. 9), and flannelmouth sucker
(fig. 10). These data represent the best time series regard-
ing status and trends of flannelmouth and bluehead
suckers in the Little Colorado River.
Humpback Chub
The life history and ecology of humpback chub in
Grand Canyon have been intensively studied (Suttkus
and Clemmer, 1979; Carothers and Minckley, 1981;
Kaeding and Zimmerman, 1983; Maddux and others,
1987; Gorman, 1994; Valdez and Ryel, 1995; Valdez
and Carothers, 1998). The humpback chub population
in Grand Canyon is centered near the confluence of
the Colorado and Little Colorado Rivers (Kaeding and
Zimmerman, 1983; Douglas and Marsh, 1996; Gorman
and Stone, 1999). Valdez and Ryel (1995) defined the
humpback chub distribution as occurring in nine aggre-
gations throughout Glen and Grand Canyons. Only the
aggregation near the confluence of the Little Colorado
and Colorado Rivers (hereafter referred to as the LCR
population) is known to successfully reproduce. The
other eight aggregations are much smaller in abundance,
averaging from a few dozen to a few hundred fish. Most
likely these eight aggregations are not supported from
local reproduction but primarily from the emigration
of juveniles and limited numbers of subadult and adult
fish from the LCR population (Valdez and Ryel, 1995).
Additionally, because of abiotic and biotic changes in
the Colorado River following the construction of Glen
Canyon Dam, the LCR population relies on the Little
Colorado River as the primary spawning and juvenile-
rearing habitat (Gorman and Stone, 1999).
Reproduction and Early Life History
Adult fish in the LCR population initially stage for
spawning runs in large eddies near the confluence of
the Little Colorado River in February and March and
make spawning runs into the tributary that average 17 d
from March through May. As the Little Colorado River’s
spring flows decrease and the water warms and clears,
reproduction increases and larval fish appear (Valdez and
Ryel, 1995). Spawning has not been observed, primarily
because of the turbid water, but ripe males have been
seen gathering in areas of complex habitat structure
(boulders and travertine masses near gravel deposits); it
is thought that ripe females move to these areas to spawn
(Gorman and Stone, 1999). After spawning, some adult
chub return to specific locations in the mainstem, while
others remain in the Little Colorado River for unknown
periods of time.
Humpback chub require warm water to reproduce
successfully. Perennially cold mainstem water tempera-
tures are thought to be the reason for unsuccessful main-
stem reproduction. The minimum water temperature for
successful reproduction is 61ºF (16ºC) (Hamman, 1982;
Marsh, 1985), which is well above the summer mainstem
temperatures commonly observed of 50°F–54ºF (10°C–
12ºC). Mortality of larval and postlarval humpback chub
emerging from the warm waters of the Little Colorado
River has been attributed to thermal shock and their
enhanced susceptibility to predation caused by the more
protracted debilitating effects of cold water on swim-
Fishes of Grand Canyon 43
Figure 7. Observed species composition of all fish captured in hoop nets in the Little Colorado River, 1988–2004 (U.S. Fish and Wildlife
Service, Arizona Game and Fish Department, Arizona State University, and U.S. Geological Survey, unpub. data, 2005). The top panel
(A) includes species composition of the four native species and a pooled nonnative category. The bottom panel (B) displays the annual
species composition of the nonnative catch. Dominant species of minnows include fathead minnow, red shiner, and common carp.
Dominant species of catfishes include channel catfish and black and yellow bullheads.
A.
B.
44 The State of the Colorado River Ecosystem in Grand Canyon
ming ability and growth (Lupher and Clarkson, 1994;
Clarkson and Childs, 2000; Robinson and Childs, 2001;
Ward and others, 2002).
A key issue associated with humpback chub is lack
of recruitment to the adult population because of the
low survivorship of young fish (Valdez and Ryel, 1995).
Young humpback chub remain in the Little Colorado
River or drift and swim into the mainstem (Robinson
and others, 1998). The lack of recruitment and docu-
mented predation indicate that mortality is extremely
high in the mainstem (Lupher and Clarkson, 1994;
Valdez and Ryel, 1995; Marsh and Douglas, 1997;
Clarkson and Childs, 2000; Robinson and Childs, 2001).
During summer, the young humpback chub that survive
in the mainstem occupy low-velocity, talus, and vegetated
shoreline habitats, including backwaters; however, low
survivorship over the year virtually eliminates the YOY
humpback chub in the mainstem. As a result, few if any
humpback chub spawned during the previous year are
present in the mainstem in March. Those YOY hump-
back chub that do survive, and ultimately recruit to the
adult population, are fish that remain resident in the
Little Colorado River during their early life history.
Limited breeding of humpback chub occurs among
other subpopulations in the Colorado River. Valdez
and Ryel (1995) documented limited spawning suc-
cess at a warm underwater spring near RM 30, known
locally as 30-Mile Spring, in upper Marble Canyon.
YOY humpback chub in the size range of 0.4–1.2
inches (10–30 mm) have been sporadically collected at
considerable distances below the Little Colorado River,
usually beginning in June (Kubly, 1990; Arizona Game
and Fish Department, 1996; Brouder and others, 1997).
Some limited reproduction may occur in other smaller
tributaries. Young humpback chub have been collected
in or near Bright Angel Creek, Shinumo Creek, Kanab
Creek, and Havasu Creek, but spawning success has
not been well documented (Maddux and others, 1987;
Kubly, 1990; Arizona Game and Fish Department, 1996;
Brouder and others, 1997). These limited observations of
spawning success among subpopulations outside of the
Little Colorado have not been shown to lead to successful
recruitment, likely because of the factors mentioned above.
Food Habits and Diseases
Dietary analyses reveal humpback chub to be
opportunistic feeders, selectively feeding on algae,
aquatic and terrestrial invertebrates, and small fish
(Kaeding and Zimmerman, 1983; Kubly, 1990; Valdez
and Ryel, 1995; Stone, 2004). Humpback chub diet
changes over the course of the year in response to food
Figure 8. Humpback chub catch-per-unit effort (fish/hour) with
95% confidence intervals in the lower 0.75 mi (1,200 m) of the
Little Colorado River using hoop nets, 1987–2003 (no sampling
conducted 2000–01). Solid squares are for fish between 5.9 and
7.8 inches (151–199 mm) total length (TL) and open diamonds are
for fish more than 7.9 inches (200 mm) total length (modified from
Coggins and others, in press).
Figure 9. Hoop net catch (fish/hour) of adult bluehead sucker
more than 7.5 inches (190 mm) in total length in the lower 0.75
mi (1,200 m) of the Little Colorado River (Arizona Game and Fish
Department and U.S. Geological Survey, unpub. data, 2005).
Figure 10. Hoop net catch (fish/hour) of adult flannelmouth
sucker more than 13.8 inches (350 mm) in total length in the lower
0.75 mi (1,200 m) of the Little Colorado River, 1987–2004 (Arizona
Game and Fish Department and U.S. Geological Survey, unpub.
data, 2005).
Fishes of Grand Canyon 45
availability and turbidity-related decreases in benthic-
standing biomass over distance downstream from Glen
Canyon Dam (Blinn and others, 1992). Nonnative scuds
(Gammarus lacustris) and simuliid (black fly) larvae occa-
sionally make up a large proportion of humpback chub
diet. Gammarus lacustris selectively feeds on epiphytes (i.e.,
diatoms) associated with Cladophora glomerata, the domi-
nant algae species in the upper reaches where clear water
conditions often prevail. Chironomid (midge fly) larvae
are also important in all areas of the river. As the river
becomes more turbid downstream, simuliids become the
dominant food source (see chapter 5, this report).
Kaeding and Zimmerman (1983) identified 13
species of bacteria, 6 protozoans, and 1 fungus from
humpback chub in Grand Canyon. The role of these
organisms in the life history of humpback chub is not
known. In 1990, the Asian tapeworm was first identi-
fied from humpback chub in the Little Colorado River
(Clarkson and others, 1997; Choudhury and others,
2004). This cestode is particularly worrisome because
it infects humpback chub at a high rate and has been
reported to be pathogenic and potentially fatal in a
variety of other fish (Hoffman and Schubert, 1984;
Hoffnagle and others, 2000).
Population Dynamics
Very large numbers of humpback chub, as well as
of flannelmouth sucker and bluehead sucker, have been
tagged in Grand Canyon since 1989. As a result, today
most of the older humpback chub have been tagged.
Previous analyses of the recapture data of tagged fish
indicate that there is likely strong age-dependence in
survival rates and that recruitment of humpback chub
has likely declined considerably since the early 1990s
(Coggins and others, in press). The USGS Grand
Canyon Monitoring and Research Center uses an
analysis method for the mark and recapture data that
reinforces these results and allows recovery of informa-
tion about likely recruitment changes that date back
to the early 1980s. The mark and recapture data are
analyzed by assigning each marked fish an age at first
capture based on its size at that time and then perform-
ing mark-recapture analysis on the resulting age-struc-
tured data on first captures and later recaptures (Coggins
and others, in press). Results of this open population
mark-recapture model, known as age-structured mark
recapture (ASMR), show decreases in the recruitment
of young humpback chub into the adult population and
as a consequence an overall decline in numbers of adult
humpback chub (figs. 11 and 12).
Figure 11. Age-2 humpback chub recruitment estimated by
using the three formulations of the annual age-structured mark
recapture (ASMR) model (from Coggins and others, in press).
Figure 12. Adult (age-4+) humpback chub population estimates
for 1989–2001 made by using the age-structured Jolly-Seber
model and the three formulations of the annual age-structured
mark recapture (ASMR) model (from Coggins and others, in press).
Overall, about 15%–20% of the adult humpback
chub are dying each year. If this mortality rate and the
dramatically reduced recruitment rate of young chub
experienced since the early 1990s remain unchanged,
there will be a decline in the adult population of
humpback chub from the present 3,000–5,000 fish to a
level of 1,500–2,000 adult fish over the next 10–15 yr.
Cause and Effect Relationships
The Glen Canyon Dam Adaptive Management
Program has a goal of maintaining a self-sustain-
ing population of humpback chub in Grand Canyon
(GCDAMP, 2001, http://www.usbr.gov/uc/rm/amp/
amwg/mtgs/02jan17/Attach_06.pdf, accessed July
14, 2005); however, this goal is qualitative and has no
46 The State of the Colorado River Ecosystem in Grand Canyon
defined target population abundance levels. The U.S.
Fish and Wildlife Service, which has jurisdiction over
the humpback chub as a federally endangered species,
promulgated recovery goals based on the known distribu-
tion of the species (U.S. Fish and Wildlife Service, 2002).
These goals recognize the Grand Canyon population of
humpback chub as the only potentially viable population
in the lower Colorado River Basin and include it, along
with at least one population from the upper Colorado
River Basin, as having to attain certain population num-
bers before the species can be considered for downlisting
or delisting under the Endangered Species Act. Briefly,
these goals require that a viable population be attained
and maintained for a period of at least 5 yr, with a mini-
mum of 2,100 sexually mature individuals in each popu-
lation. Furthermore, the recruitment of new individu-
als into the population must meet or exceed the adult
mortality rate, thereby providing a stable or increasing
adult abundance trend. In the case of the Grand Canyon
population, sexually mature fish are assumed to be 4 yr
old or older.
Of paramount importance in conserving the
population of the federally endangered humpback chub
is determining the factors contributing to population
decline and implementing management actions designed
to minimize or eliminate the effect of those factors.
Not all of the factors that may be responsible for the
recruitment decline of humpback chub beginning in
the early 1990s are clear, but a list of likely factors that
could be acting either singly or in combination include
(1) Colorado River and Little Colorado River hydrology
(discharge and temperature), (2) infestation of juvenile
humpback chub by Asian tapeworm, (3) predation by
or competition with warmwater native cyprinids and
catostomids and nonnative cyprinids and ictalurids
within the Little Colorado River, and (4) predation by or
competition with coldwater nonnative salmonids within
the Colorado River.
The body of evidence available to evaluate spe-
cific questions varies among these postulated factors.
For instance, beginning in 1990 the operation of Glen
Canyon Dam was changed through the implementation
of research flows (a series of discharges and data collec-
tion programs conducted from June 1990 through July
1991) and the interim operating criteria. This hydrology,
and the subsequent MLFF alternative that continues to
present, can generally be characterized as having less
severe daily flow fluctuations than the previous 28 yr
of the no action period when the dam was managed
primarily to maximize hydroelectric power revenue.
This major change in Colorado River hydrology cor-
relates closely to the decline in humpback chub recruit-
ment. Also, it is possible that the decline in humpback
chub recruitment in the early 1990s was caused by the
nearly continuous flooding in the Little Colorado River
that occurred during the summer of 1992, particularly
during the early summer when larval humpback chub
emerge (Robinson and others, 1998). It is also possible
that the high infestation rate of juvenile humpback chub
by the Asian tapeworm is a factor. Humpback chub
infected with Asian tapeworm were first found in 1990,
and infestation rates in 2001 exceeded 90% (Choudhury
and others, 2004). Finally, predation and competition
by nonnative fishes either in the Little Colorado River
or in the Colorado River may be driving the humpback
chub recruitment trend. Although robust relative abun-
dance data do not exist for common carp and channel
catfish within the Little Colorado River, there was a large
increase in the abundance of nonnative salmonids in
the Colorado River documented near the confluence of
the Little Colorado River (RM 56.6–68.3) (Gorman and
Coggins, 2000).
Recent Management Actions
Undertaken or Proposed
While it is difficult to determine the factor most
responsible for the humpback chub recruitment decline,
a likely primary factor is negative interactions (predation
and competition) with nonnative fish. Interaction with
nonnative fish is implicated in the decline and extinction
of native fishes throughout the Colorado River Basin. In
response to the need to address this factor, a program of
selective removal of nonnative fishes (known as mechani-
cal removal) was implemented in 2003 near the conflu-
ence of the Little Colorado River and in other tributar-
ies (work in Bright Angel Creek and other tributaries
has been undertaken by the National Park Service). To
complement these efforts, the work group also approved
initiation of a multiobjective study to evaluate the poten-
tial effect of rainbow trout and brown trout predation on
humpback chub recruitment and the efficacy of mechan-
ical removal of nonnative fishes from the Colorado River
near the confluence of the Little Colorado River.
In early 2003, a major effort was begun by the
Glen Canyon Dam Adaptive Management Program to
remove nonnative fish from the area of the river near the
confluence of the Little Colorado River (RM 61), which
is considered important habitat for native fish, especially
humpback chub. A total of 16,045 rainbow trout and
many other nonnative fish (fig. 13a) were removed from
this river reach during 2003–04. While native fish con-
tributed only approximately 5% of the overall catch in
January 2003, native fish contributed greater than 35%
Fishes of Grand Canyon 47
in September 2004, generally reflecting a reduction in
nonnative fish abundance. Also, the overall abundance
of rainbow trout has been reduced by more than 60%
in this river reach (fig. 13b). Whether this reduction in
nonnative fish density will benefit native fish is unknown
at this time.
Moreover, an experimental program to move YOY
humpback chub upstream of an impassable barrier
(where few nonnative fish live) in the Little Colorado
River was initiated and has shown some early signs of
success (Stone and Sponholtz, 2005). Future introduc-
tions of humpback chub into other Grand Canyon
tributary streams may help augment the population in
Grand Canyon. Additional management options include
potential hatchery rearing of humpback chub as a refu-
gium population or for stocking in the river.
Other management options include the installation
of a multilevel water intake structure(s) for Glen Canyon
Dam to warm the water in Grand Canyon. The Bureau
of Reclamation has developed preliminary plans and
is scoping out the possible installation of a temperature
control device, which would provide flexibility to release
warmer water into the river. Warmer water could cre-
ate more favorable habitat conditions for native fishes
in general; however, its operation could also improve
habitat conditions for nonnative, warmwater species
and degrade habitat quality for trout inhabiting the Lees
Ferry reach. Obviously the operation of such a device,
if built, will need to be carefully considered and imple-
mented experimentally.
Discussion and Future
Research Needs
The salient findings of the research and monitor-
ing programs undertaken by the Glen Canyon Dam
Adaptive Management Program regarding fishes are
twofold. First, there has been a dramatic and continuing
decline in the number of adult humpback chub in the
Grand Canyon ecosystem since at least the late 1980s.
This decrease in adult fish is due to a steady decline in
the recruitment of young fish into the population begin-
ning in the early 1980s, with an additional reduction in
the early 1990s. This decline in recruitment results in a
dwindling population of adults as older age fish die off
and are not replaced. It is currently estimated that if
recruitment remains stable at this level, the adult popula-
tion of humpback chub in the Grand Canyon ecosystem
will stabilize at approximately 1,500–2,000 fish over
the next decade or so. The current population decline
combined with the low recruitment in this population
relative to adult mortality indicates that this population
will attain neither positive trends nor sufficient numbers
of fish to meet USFWS recovery goals in the foreseeable
future.
The second major result regarding fishes is the
proliferation of rainbow trout in both the Lees Ferry
reach and downstream as far as RM 75. Numbers of
brown trout have also increased dramatically in the area
around Bright Angel Creek and upstream to above the
Little Colorado River confluence. Both of these species
are known to prey on native fishes, and their substantial
increase in abundance near the principal remaining
native-fish habitat in Grand Canyon remains a concern.
It has yet to be determined whether the experimental
management action to reduce the numbers of nonnative
fish in the area around the Little Colorado River conflu-
ence has resulted in any increase in survival and recruit-
ment of the federally endangered humpback chub.
Dam Operations
It is not possible to say conclusively that the decline
in humpback chub recruitment that began to occur in
the early 1990s is because of the implementation of the
MLFF regime; however, the flow regime has not reversed
the decline in recruitment and adult abundance either.
Approximately 15%–20% of the adult humpback chub
population is dying each year. These fish are most likely
being replaced, albeit at a lower rate, predominately by
young humpback chub that have spent the first 3 to 4
yr of their lives in the Little Colorado River. In other
words, the MLFF alternative had either a negative effect
or no effect at all, but it has not had a measurable benefi-
cial effect on humpback chub.
The MLFF alternative has not improved condi-
tions for other native fishes as indicated by their stable or
declining numbers. Different daily, seasonal, or annual
changes in river flows could be considered on an experi-
mental basis. Such flow options could include reduced
daily fluctuations and equalized monthly volumes to pro-
vide a more stable environment for young native fishes.
There is a good chance that juvenile humpback chub dis-
persing into the mainstem in summer and fall would be
able to grow, survive, and return to the Little Colorado
River for extended rearing if they were to encounter (1)
reduced predation and competition by nonnative trout
(trout would have to be removed by mechanical removal
treatments) and (2) relatively warm refuges in nearshore
locations (these locations could be created by steady flow
conditions in late summer and fall). The low summer
48 The State of the Colorado River Ecosystem in Grand Canyon
Figure 13. A. Total catch and percent contribution by species and month during mechanical removal efforts in the Little Colorado River
removal reach, 2003–04. B. Estimated abundance of rainbow trout in the Little Colorado River removal reach before and after each
mechanical removal effort, 2003–04 (U.S. Geological Survey, unpub. data, 2005).
A.
B.
Fishes of Grand Canyon 49
steady flow (LSSF) experiment demonstrated that such
lateral warming of backwater areas can be quite dra-
matic. A summer-fall steady flow experiment would need
to maintain conditions for backwater warming from the
time of the first summer high flow that disperses juve-
niles into the mainstem until around November 1, when
the equilibrium temperature in standing backwaters
decreases (because of nighttime cooling) to about the
same as the mainstem temperature.
Three additional flow possibilities for Glen Canyon
Dam could be made based on recommendations from
the 2003 YOY rainbow trout surveys and analyses of
otoliths: (1) fluctuating flows targeting YOY rainbow
trout could be implemented from April through July to
coincide with the timing of hatch, (2) summer steady
flows could likely improve the growth of YOY rain-
bow trout, and (3) sudden reductions in the minimum
daily flow could have the potential to strand or displace
many YOY rainbow trout in the Lees Ferry reach.
The latter recommendation was based on an almost
complete absence of fry from low-angle shorelines
after the reduction in the minimum flow from 10,000
cfs to 5,000 cfs in early September 2003 and a similar
but less dramatic reduction in September 2004 (Kor-
man and others, 2005). An event-based approach—in
which flows are increased to approximately 20,000 cfs
for 2 d, followed by a reduction to 5,000 cfs for 1 d, and
implemented on a monthly basis from May through
September—would almost certainly be much more effec-
tive at reducing recruitment in the Lees Ferry reach than
the January–March fluctuating nonnative fish suppres-
sion flows implemented beginning in 2003. Steady flows
could be conducted between events to increase water
temperatures for native fish downstream and would not
have beneficial effects for YOY rainbow trout, as their
densities would be controlled through the temporary
reductions in minimum flow.
Researchers have been unable to identify or imple-
ment an effective mainstem monitoring program for
native fishes or most nonnative species (the exceptions
are rainbow trout, brown trout, and carp). Because of
this situation, the USGS Grand Canyon Monitoring
and Research Center has called for a research initia-
tive to investigate the utility of alternative sampling
methods such as acoustic devices that may assist in
providing better measures of relative abundance and
change detection.
The most important research task associated with
humpback chub conservation is determining the fac-
tors controlling the recruitment dynamics of this spe-
cies. These factors can only be determined through an
appropriately designed experiment that addresses the
multiple important biotic and abiotic factors likely influ-
encing humpback chub. As stated by Korman and others
(2004, p. 395–396) in summary of an intensive model-
ing effort aimed at characterizing changes in nearshore
humpback chub habitat with changes in Glen Canyon
Dam operation,
The interaction between habitat and ecosystem
processes like competition and predation remain
highly uncertain. Ultimately, questions regarding
the effects of dam operations on juvenile hump-
back chub must be addressed by monitoring the
response of critical population parameters to
flow manipulations conducted within a sound
experimental design.
50 The State of the Colorado River Ecosystem in Grand Canyon
Speckled dace
(native)
Size–
rarely exceeds 3 inches (7.6 cm).
Distribution–
extensively distributed
throughout Western
United States.
Status–
abundant in some areas and widely distributed. This
species is represented by several subspecies.
Natural history–
The speckled dace (Rhinichthys osculus) is the only
native dace in Arizona, although the genus is widely
Bluehead sucker
(native)
Size–
maximum of about 20 inches
(51 cm).
Distribution–
found in fast-flowing river
systems in Arizona, Colorado,
New Mexico, Utah, and Wyoming.
distributed elsewhere. Dace are widely distributed
in the Colorado River, with many inhabiting
backwaters in western Grand Canyon. Diet includes
algae, insect larvae, small crustaceans, and small
snails. Spawning occurs in spring and late summer.
Large schools of dace congregate over gravel
bottoms to spawn. Populations appear to be stable
in Grand Canyon.
Randall D. Babb, Arizona Game and Fish Department
Randall D. Babb, Arizona Game and Fish Department
Profiles of Selected Fish Species Found
in the Grand Canyon Ecosystem
Information compiled by Jeffrey E. Lovich
Fishes of Grand Canyon 51
Status–
not uncommon in some areas.
Natural history–
This species (Catostomus discobolus) occurs in the
Colorado River upstream from Lake Mead. Diet
includes algae, diatoms, insects, amphipods, and
organic debris that it scrapes from rocks with
Randall D. Babb, Arizona Game and Fish Department
specialized cartilage lips. In Grand Canyon,
spawning occurs over gravel, sand, and cobbles in
April and May, when water temperatures exceed
61oF (16oC). Young inhabit backwaters in Grand
Canyon. Bluehead suckers are known to hybridize
with other sucker species. Populations appear to be
stable in Grand Canyon. Individuals can live for
more than 20 yr.
Humpback chub
(native)
Size–
maximum of about 20 inches
(51 cm).
Distribution–
found only in the Colorado
River system.
Status–
federally endangered.
Natural history–
The humpback chub (Gila cypha) formerly
ranged downstream to the area now occupied by
Lake Mohave, but it is now confined to several
aggregations in Grand Canyon and isolated
populations in various deep canyon stretches of
the Colorado River and its major tributaries above
Lake Powell. Most humpback chub in Grand
Canyon are found in the vicinity of the Little
Colorado River (LCR) and its confluence with
the mainstem. Humpback chubs are omnivorous,
and their diet includes a diversity of aquatic and
terrestrial invertebrates, small fish, algae, and
other plant material. In Grand Canyon the diet
of the nonnative rainbow trout is almost identical,
setting the stage for possible resource competition
between the species. Spawning occurs in spring in
the LCR, and young enter the mainstem during
floods associated with storm events, most commonly
in spring and late summer/fall. Aggregations of
humpback chub, well upstream and downstream of
the LCR population, may result from (1) emigration
of juveniles, subadults, or adults from the LCR;
(2) survival of relict fish from before the dam; or
(3) mainstem spawning. The latter has not been
documented in the postdam era, so additional
research is needed to resolve this issue. The
estimated adult population in Grand Canyon has
declined sharply from about 10,000 a decade ago to
about 3,000–5,000 today.
52 The State of the Colorado River Ecosystem in Grand Canyon
Flannelmouth
sucker (native)
Size–
can exceed about 20 inches
(51 cm).
Distribution–
Colorado River Basin in
Arizona, California, Colorado,
New Mexico, Nevada, Utah,
and Wyoming. Extirpated from the Gila River
Basin of Arizona.
Status–
not uncommon in some areas.
Natural history–
This species (Catostomus latipinnis) occurs in the
Colorado River upstream from Lake Mead.
Flannelmouth suckers below Lake Mead exist
because of the success of reintroduction from the
Paria River in the mid-1970s. Diet varies with age
class and size but includes algae, insects, plankton,
ostracod, crustaceans, plant materials, and detritus.
This species likely makes spawning runs in most
of the major tributaries in Grand Canyon before
returning to the mainstem. Spawning occurs
from March to July, when water temperatures are
between 43°F and 68oF (6°C and 20oC). Spawning
occurs in shallow water over sand and gravel
bottoms. Females lay from 4,000 to 40,000 eggs.
Juveniles are frequently captured in the mainstem
from lower Marble Canyon downstream to Lake
Mead. Juveniles are also frequently captured in
the Little Colorado River and other tributaries
downstream. Known to hybridize with the
razorback sucker, a species that is presumed to be
gone from the Grand Canyon region. Populations
appear to be stable in Grand Canyon.
Rainbow trout
(nonnative)
Size–
up to 47 inches (120 cm).
Arizona State record was 32.25
inches (81.9 cm).
Distribution–
extensively distributed
throughout Western North
America in river systems
Randall D. Babb, Arizona Game and Fish DepartmentRandall D. Babb, Arizona Game and Fish Department
Fishes of Grand Canyon 53
draining into the Pacific Ocean. Widely introduced
worldwide, including into the Colorado River.
Status–
common.
Natural history–
Rainbow trout (Oncorhynchus mykiss) were
introduced into the Grand Canyon area in the
1920s for sport fishing. Originally confined to
clear tributary streams, the construction of Glen
Canyon Dam created cold, clear conditions that
allowed trout to colonize the mainstem. Trout
were also stocked in the tailwaters of the dam by
the State of Arizona shortly after construction
Brown trout
(nonnative)
Size–
Arizona State record is 36
inches (91.4 cm). The world
record is a 40 lb, 4 oz (18.3 kg)
specimen caught in Arkansas.
Distribution–
widely introduced worldwide,
including into the Colorado River.
Status–
common.
Natural history–
Native to Europe and Asia, brown trout (Salmo
trutta) were introduced into the Grand Canyon area
in the 1920s for sport fishing. Originally confined
to clear tributary streams, brown trout were able
to colonize the mainstem of the Colorado River
when the construction of Glen Canyon Dam
created cold, clear conditions. Brown trout eat
a variety of aquatic and terrestrial insects and
other invertebrates. Large specimens are highly
predaceous on other fish, including smaller trout.
Reproduction is as in other species of trout (see text
box for rainbow trout). Bright Angel Creek is an
important spawning stream for mainstem trout that
move into the smaller tributary for this purpose in
winter and early spring. Brown trout are capable
of tolerating slightly higher water temperatures
than most other trout. Most brown trout in Grand
Canyon today occur near the confluence with
Bright Angel Creek.
was completed in the 1960s. The diet consists
mainly of both aquatic and terrestrial insects and
other aquatic invertebrates including amphipods.
Spawning in Grand Canyon occurs in winter
and early spring. After fertilization by males,
females excavate a depression, or redd, in gravelly
bottoms, and the eggs are buried in the substrate to
hatch unattended. Rainbow trout like cold water
temperatures and rarely live in water above about
77oF (25oC). The Lees Ferry reach of the Colorado
River is where most spawning occurs in the Grand
Canyon area and is managed as a “blue ribbon”
trout fishery. Trout numbers have been increasing
in recent years, possibly to the detriment of the
endangered humpback chub.
Randall D. Babb, Arizona Game and Fish Department
54 The State of the Colorado River Ecosystem in Grand Canyon
References
Arizona Game and Fish Department, 1996, Ecology of
Grand Canyon backwaters: Flagstaff, Ariz., report to
Bureau of Reclamation, Glen Canyon Environmental
Studies, 155 p.
Blinn, D.W., Stevens, L.E., and Shannon, J.P., 1992, The
effects of Glen Canyon Dam on the aquatic food base
in the Colorado River corridor in Grand Canyon,
Arizona: Flagstaff, Ariz., Bureau of Reclamation, Glen
Canyon Environmental Studies Report 11-02.
Brouder, M.J., Hoffnagle,T.L., Tuegel, M.A., and
Persons, W.R., 1997, Glen Canyon Environmental
Studies Colorado River native fish study, 1995 annual
report: Arizona Game and Fish Department report,
Grand Canyon Monitoring and Research Center
cooperative agreement no. 9-FC-4007940, 101 p.
Carothers, S.W., and Minckley, C.O., 1981, A survey of
the fishes, aquatic invertebrates and aquatic plants of
the Colorado River and selected tributaries from Lees
Ferry to Separation Rapids, final report: Salt Lake
City, Utah, Water and Power Resources Service.
Choudhury, A., Hoffnagle, T.L., and Cole, R.A., 2004,
Parasites of native and non-native fishes of the Little
Colorado River, Grand Canyon, Arizona: Journal of
Parasitology, v. 90, no. 5, p. 1042–1053.
Clarkson, R.W., and Childs, M.R., 2000, Temperature
effects of hypolimnial-release dams on early life stages
of Colorado River Basin big-river fishes: Copeia, v.
2000, no. 2, p. 402–412.
Clarkson, R.W., Robinson, A.T., and Hoffnagle, T.L.,
1997, Asian tapeworm (Bothriocephalus acheilognathi) in
native fishes from the Little Colorado River, Grand
Canyon, Arizona: Great Basin Naturalist, v. 57, no. 1,
p. 66–69.
Coggins, L.G., Jr., Pine, W.E., III, Walters, C.J., Van
Haverbeke, D.R., Ward, D., and Johnstone, H., in
press, Abundance trends and status of the Little
Colorado River population of humpback chub: North
American Journal of Fisheries Management.
Cole, G.A., and Kubly, D.M., 1976, Limnologic studies
on the Colorado River from Lees Ferry to Diamond
Creek: Colorado River Research Program Technical
Report No. 6, 88 p.
Douglas, M.E., and Marsh, P.C., 1996, Population
estimates/population movements of Gila cypha, an
endangered cyprinid fish in the Grand Canyon region
of Arizona: Copeia, v. 1996, no. 1, p. 15–28.
Goeking, S.A., Schmidt, J.C., and Webb, M.K., 2003,
Spatial and temporal trends in the size and number
of backwaters between 1935 and 2000, Marble and
Grand Canyons, Arizona: Grand Canyon Monitoring
and Research Center final report, cooperative agree-
ment no. 01WRAG0059, 26 p.
Gorman, O.T., 1994, Habitat use by humpback chub,
Gila cypha, in the Little Colorado River and other
tributaries of the Colorado River: Glen Canyon Envi-
ronmental Studies, Phase II, final report: Flagstaff,
Ariz., U.S. Fish and Wildlife Service Arizona Fishery
Resources Office.
Gorman, O.T., and Coggins, L.G., 2000, Status and
trends of native and non-native fishes of the Colorado
River in Grand Canyon 1990–2000: Grand Canyon
Monitoring and Research Center final report, inter-
agency acquisition no. 98-AA-40-0040, 78 p.
Gorman, O.T., and Stone, D.M., 1999, Ecology of
spawning humpback chub, Gila cypha, in the Little
Colorado River near Grand Canyon, Arizona:
Environmental Biology of Fishes, v. 55, p. 115–133.
Hamman, R.L., 1982, Spawning and culture of
humpback chub: Progressive Fish-Culturist, v. 44, no.
4, p. 213–216.
Hoffman, G.L., and Schubert, G., 1984, Some parasites
of exotic fishes, in Courtney, W.R., Jr., and Stauffer,
J.R., Jr., eds., Distribution, biology, and manage-
ment of exotic fishes: Baltimore, Md., Johns Hopkins
University Press, p. 233–261.
Hoffnagle, T.L., Choudhury, A., and Cole, R.A., 2000,
Parasites of native and non-native fishes of the lower
Little Colorado River, Arizona. 2000 annual report:
Phoenix, Arizona Game and Fish Department, 14 p.
Johnstone, H.C., and Lauretta, M., 2004, Native fish
monitoring activities in the Colorado River within
Grand Canyon during 2003: Grand Canyon Moni-
toring and Research Center final report, contract no.
01WRAG0046, 80 p.
Fishes of Grand Canyon 55
Johnstone, H.C., Lauretta, M., and Trammel, M., 2003,
Native fish monitoring activities in the Colorado River
within Grand Canyon during 2002: Grand Canyon
Monitoring and Research Center final report, contract
no. 01WRAG0046, 56 p.
Kaeding, L.R., and Zimmerman, M.A., 1983, Life his-
tory and ecology of the humpback chub in the Little
Colorado and Colorado Rivers of the Grand Canyon:
Transactions of the American Fisheries Society, v. 112,
p. 577–594.
Kolb, E., and Kolb, E., 1914, Experiences in the Grand
Canyon: National Geographic Magazine, v. 26, no. 2,
p. 99–184.
Korman, J., Kaplinski, M., Hazel, J.E., III, and Melis,
T.S., 2005, Effects of the experimental fluctuating
flows from Glen Canyon Dam in 2003 and 2004
on early life history stages of rainbow trout in the
Colorado River: Final report, cooperative agreement
no. 04WRAG0006, modification no. 002.
Korman, J., Wiele, S.M., and Torizzo, M., 2004, Model-
ling effects of discharge on habitat quality and dis-
persal of juvenile humpback chub (Gila cypha) in the
Colorado River, Grand Canyon: River Research and
Applications, v. 20, p. 379–400.
Kubly, D.M., 1990, The endangered humpback chub
(Gila cypha) in Arizona: a review of past and sugges-
tions for future research, draft report: Arizona Game
and Fish Department.
Lupher, M.L., and Clarkson, R.W., 1994, Temperature
tolerance of humpback chub (Gila cypha) and Colorado
squawfish (Ptychocheilus lucius), with a description of
culture methods for humpback chub: Phoenix, Ari-
zona Game and Fish Department report, cooperative
agreement no. 9-FC-40-07940.
Maddux, H.R., Kubly, D.M., DeVos, J.C., Jr., Persons,
W.R., Staedicke, R., and Wright, R.L., 1987, Effects
of varied flow regimes on aquatic resources of Glen
and Grand Canyons: Bureau of Reclamation Glen
Canyon Environmental Studies report, National Tech-
nological Information Series PB88-1 83439/AS.
Marsh, P.C., 1985, Effect of incubation temperature on
survival of embryos of native Colorado River fishes:
Southwestern Naturalist, v. 30, p. 129–140.
Marsh, P.C., and Douglas, M.E., 1997, Predation by
introduced fishes on endangered humpback chub
and other native species in the Little Colorado River,
Arizona: Transactions of the American Fisheries
Society, v. 126, p. 343–346.
McKinney, T., Speas, D.W., Rogers, R.S., and Persons,
W.R., 1999, Rainbow trout in the Lees Ferry recre-
ational fishery below Glen Canyon Dam, Arizona,
following establishment of minimum flow require-
ments: Final report, submitted to the Grand Canyon
Monitoring and Research Center, cooperative agree-
ment no. 1425-97-FC-40-22690, 109 p.
McKinney, T., Speas, D.W., Rogers, R.S., and Persons,
W.R., 2001, Rainbow trout in a regulated river below
Glen Canyon Dam, Arizona, following increased mini-
mum flows and reduced discharge capability: North
American Journal of Fisheries Management, v. 21, p.
216–222.
Miller, R.R., 1946, Gila cypha a remarkable new species
of cyprinid fish from the Colorado River in Grand
Canyon, Arizona: Journal of the Washington Acad-
emy of Sciences, v. 36, p. 409–415.
Minckley, W.L., 1991, Native fishes of the Grand
Canyon region: an obituary?, in Marzolf, G.R., ed.,
Colorado River ecology and dam management:
Washington, D.C., National Academy Press, p. 124–
177.
Mueller, G.A., and Marsh, P.C., 2002, Lost, a desert
river and its native fishes: a historical perspective of
the lower Colorado River: Fort Collins, Colo., U.S.
Geological Survey Information and Technology
Report USGS/BRD/ITR 2002-0010, 69 p.
Niccum, J., Magnuson, Y., and Reger, S., 1998, Colorado
River-Lees Ferry status report 1994–1997: Arizona
Game and Fish Department, Statewide Fisheries
Investigation Survey of Aquatic Resources, Federal
Aid Project F-7-M, 58 p.
Robinson, A.T., and Childs, M.R., 2001, Juvenile growth
of native fishes in the Little Colorado River and in a
thermally modified portion of the Colorado River:
North American Journal of Fisheries Management, v.
21, p. 809–815.
Robinson, A.T., Clarkson, R.W., and Forrest, R.E., 1998,
Dispersal of larval fishes in a regulated river tributary:
Transactions of the American Fisheries Society, v. 127,
p. 772–786.
56 The State of the Colorado River Ecosystem in Grand Canyon
Contact Information:
Steven P. Gloss
Ecologist
U.S. Department of the Interior
U.S. Geological Survey
Southwest Biological Science Center
Tucson, AZ
sgloss@usgs.gov
Lewis G. Coggins
Fishery Biologist
U.S. Department of the Interior
U.S. Geological Survey
Southwest Biological Science Center
Flagstaff, AZ
lcoggins@usgs.gov
Stone, D., and Sponholtz, P., 2005, Translocation of
young-of-year humpback chub above Chute Falls
in the Little Colorado River, AZ: Flagstaff, Ariz.,
interim report to the Grand Canyon Monitoring and
Research Center.
Stone, D.M., 2004, Differential detection of ingested
items evacuated from genus Gila cyprinids by two non-
lethal alimentary tract lavage techniques: Journal of
Freshwater Ecology, v. 19, no. 4, p. 559–565.
Suttkus, R.D., and Clemmer, G.H., 1979, Fishes of the
Colorado River in Grand Canyon National Park, in
Proceedings of the First Annual Conference on Sci-
entific Research in the National Parks: National Park
Service Transactions and Proceedings Series, v. 5, p.
599–604.
Topping, D.J., Schmidt, J.C., and Vierra, L.E., Jr., 2003,
Computation and analysis of the instantaneous-dis-
charge record for the Colorado River at Lees Ferry,
Arizona—May 8, 1921, through September 30, 2000:
U.S. Geological Survey Professional Paper 1677, 118 p.
U.S. Department of the Interior, 1995, Operation of
Glen Canyon Dam Final Environmental Impact
Statement, Colorado River Storage Project, Coconino
County, Arizona: Salt Lake City, Utah, Bureau of
Reclamation, Upper Colorado Regional Office, 337 p.
U.S. Fish and Wildlife Service, 2002, Humpback chub
(Gila cypha) recovery goals: amendment and supple-
ment to the Humpback Chub Recovery Plan: Denver,
Colo., U.S. Fish and Wildlife Service, Mountain-
Prairie Region (6).
Valdez, R.A., and Carothers, S.W., 1998, The aquatic
ecosystem of the Colorado River in Grand Canyon.
Report to Bureau of Reclamation, Salt Lake City,
Utah: Flagstaff, Ariz., SWCA, Inc., Environmental
Consultants, 250 p.
Valdez, R.A., and Ryel, R.J., 1995, Life history and ecol-
ogy of the humpback chub (Gila cypha) in the Colorado
River, Grand Canyon, Arizona. Final report to the
Bureau of Reclamation, Salt Lake City, Utah, contract
no. 0-CS-40-09110: Logan, Utah, BIO/WEST, Inc.
Walters, C., Korman, J., Stevens, L.E., and Gold, B.,
2000, Ecosystem modeling for evaluation of adaptive
management policies in the Grand Canyon: Conserva-
tion Ecology, v. 4, no. 2, p. 1, http://www.consecol.
org/vol4/iss2/art1/, accessed July 14, 2005.
Ward, D.L., Maughan, O.E., Bonar, S.A., and Matter,
W.J., 2002, Effects of temperature, fish length, and
exercise on swimming performance of age-0 flannel-
mouth sucker: Transactions of the American Fisheries
Society, v. 131, p. 492–497.
First page photograph credit: Matt Lauretta, SWCA Environmental Consultants,
Inc., for the U.S. Geological Survey
Chapter 3
Climatic
Fluctuations,
Drought,
and Flow
in the
Colorado River
Robert H. Webb
Richard Hereford
Gregory J. McCabe
Introduction
Climate is the cumulative pattern of daily atmo-
spheric conditions in a particular geographic area, and
weather is the daily and seasonal expression of these
conditions. Climate varies over periods of years, decades,
or centuries, altering weather conditions in a region, par-
ticularly precipitation amounts and temperatures. In the
arid and semiarid Southwest, climatic fluctuations affect
many hydrologic characteristics of watersheds, includ-
ing the quantity of base flow, the occurrence of large
floods, and the timing of snowmelt runoff (Dettinger and
Cayan, 1995; Cayan and others, 1999; Stewart and oth-
ers, 2004, 2005; McCabe and Clark, in press).
Reservoirs in the Western United States, particu-
larly those in the Colorado River Basin, were built to
reduce, if not eliminate, annual variations in water
supply that occurred historically because of periods
of above- or below-average precipitation. A persistent
drought beginning in 2000 raised concern that decreases
in runoff entering Lake Powell could follow and releases
from Glen Canyon Dam could be severely reduced or
constrained. Inflows to Lake Powell on the Colorado
River were below average from 2000 through 2004,
leading to drawdown of both Lake Powell (figs. 1 and 2)
and Lake Mead, the primary flow-regulation structures
on the river. On January 27, 2005, the level of Lake
Powell was at 3,562.5 ft (1,085.9 m) (full pool capacity
is 3,700 ft (1,128 m)), and the reservoir contained 8.5
million acre-feet (maf) (10,481 million m3) of water (fig.
1), which is only 35% of the reservoir’s capacity and a
little more than 1 yr of annual flow releases. Reduction
in annual flow releases can reduce the water available for
prescribed flow releases—particularly flood releases—
designed to benefit riverine habitat within Grand
Canyon. By 2004, it was speculated that Glen Canyon
Dam would be unable to produce hydroelectric power
by 2006 or 2007 if drought conditions persisted and the
lake level continued to decline.
Conditions changed in fall and winter 2004–05 as
a series of storms led to greatly above-average precipi-
tation in the southern portion of the watershed. The
high precipitation may have been associated with the
onset of El Niño conditions in the Pacific Ocean, which
presumably could have enhanced fall and early winter
storms. On February 1, 2005, inflows to Lake Powell
were forecast to be 125% of normal, the first above-
average forecast since 1999. This reversal of conditions
58 The State of the Colorado River Ecosystem in Grand Canyon
from the previous 5 yr could suggest that the drought is
over, although some long-term records suggest that this
may not be the case since average years have occurred
within periods of extended dryness. To date, it is unclear
whether the early 21st century drought is over or not,
and the possible persistence and magnitude of the
drought are of great concern for the Glen Canyon Dam
Adaptive Management Program.
Unfortunately, the factors that caused and sustained
the early 21st century drought have not been positively
identified. Although conditions in the tropical Pacific
Ocean were considered to be ideal for drought condi-
tions in the continental United States (Hoerling and
Kumar, 2003), new studies suggest that the Atlantic
Ocean may also influence drought (Gray and others,
2004; McCabe and others, 2004). In the case of the
Colorado River, it is possible to examine the drought
in a broader historical and climatic context, which can
be developed through historical records and statistical
models. First, a historical record exists of actual obser-
vations and estimates of annual flows in the river at
various places, including Lees Ferry. Second, scientists
have gained an understanding of precipitation patterns
by using annual growth rings in trees to reconstruct the
hydrologic conditions in a basin several hundred years
before the historical record began. Third, climatologists
and other scientists have recently developed statistical
techniques and dynamical models that improve under-
standing of the relations between various ocean tempera-
ture patterns and observed precipitation patterns.
This chapter makes clear that the drought beginning
in 2000 probably had its origins in several hemispheric-
scale atmospheric and oceanic processes that affect
moisture delivery to the Colorado River Basin (fig. 3). In
this context, the chapter describes the general causes of
drought in the Southwest, the long-term perspective on
drought duration in the basin based on tree-ring recon-
structions, the use of global climate indices to explain
Colorado River flows, and scenarios of future climate
and runoff in the Colorado River Basin.
Figure 1. Fluctuations in the level of Lake Powell following
closure of Glen Canyon Dam in 1963 (from www.summittech.com/
LakePowell/LP_waterDB.php, accessed February 20, 2005).
Figure 2. Lake Powell at Glen Canyon Dam (photograph by Dale
Blank, U.S. Geological Survey).
Figure 3. Moisture sources for the Colorado River Basin
(outlined in red). Lees Ferry is the separation point between
the upper and lower Colorado River Basins as specified in the
Colorado River Compact of 1922.
Climatic Fluctuations, Drought, and Flow in the Colorado River 59
Background
Drought is caused by persistent low precipitation
over a region. As such, the severity of a drought is a
function of spatial extent, duration, and magnitude of
the precipitation deficit. Moreover, the area affected by
a drought may shift in space and time. This combina-
tion of variable factors makes drought prediction and
estimation of drought frequency extremely difficult.
The causes of persistent drought over a large drainage
basin, such as the Colorado River Basin, are particularly
difficult to determine because the basin spans a large
latitudinal range.
Sources of Moisture
The most important sources of water to the
Colorado River Basin are frontal systems that originate
in the North Pacific Ocean and occur in winter and
spring. These systems tend to carry moisture at high levels
in the atmosphere, and precipitation is orographically con-
trolled, meaning that it typically increases with elevation
in the mountains. Cold frontal systems drop substantial
amounts of snow at high elevations and rain at low eleva-
tions in the Rocky, Uinta, and Wind River Mountains,
which in turn become the headwaters of the Colorado
River and its principal tributary, the Green River (fig. 3).
The frequency and moisture content of frontal systems are
strongly affected by the strength of atmospheric circula-
tion patterns and sea-surface temperatures in the Pacific
Ocean.
There are two basic types of winter storms that
affect flow in the Colorado River. Cold winter storms
deliver moisture in the form of snow at most eleva-
tion ranges in Utah, Colorado, New Mexico, northern
Arizona, and Wyoming. These storms build snowpacks
that melt in the spring, providing runoff to the Colorado
River. Warm winter storms, originating in the tropical
Pacific Ocean, may produce rain on snowpacks, result-
ing in large runoff events and floods on major rivers,
which tend to overwhelm reservoir systems, particularly
in Arizona.
Moisture delivered to the Colorado River during
summer months typically originates from a combination
of the Gulf of Mexico, the Gulf of California, and the
eastern North Pacific Ocean. Known variously as the
“Arizona monsoon,” the “Southwestern United States
monsoon,” the “summer monsoon,” or even the “North
American monsoon,” this moisture arrives in July and
August at low atmospheric levels. The moist air rises rap-
idly over the desert landscape, spawning high-intensity
thunderstorms that produce runoff mostly at elevations
of less than 7,000 ft (2,134 m). The thunderstorms tend
to be of small spatial extent, and, although they spawn
severe flash flooding locally, few floods are generated on
larger rivers in the region.
Status and Trends
Flow at Lees Ferry
Flow in the Colorado River measured at Lees Ferry
(fig. 3), the political boundary between the upper and
lower Colorado River Basins, varied substantially during
the 20th century. Calendar-year flow volumes (fig. 4a)
were combined from three data sets that were measured
or estimated by using different techniques. From 1895
through 1922, annual flow volumes at Lees Ferry were
estimated by LaRue (1925, p. 108); from 1922 through
1962, unregulated flow was measured at the Lees Ferry
gaging station; and from 1963 through 2004, flow was
estimated as the sum of tributary flows entering Lake
Powell (Webb and others, 2004). Consumptive water
use in the basin upstream of the gage at Lees Ferry is
not accounted for in these data. As a result, flow values
measured at Lees Ferry are due to climatic fluctuations
and changes in consumptive water use in the upper basin
States of Wyoming, Colorado, Utah, and New Mexico.
The average annual flow volume shown in figure
4a was 12.3 maf (15,166 million m3) from 1895 through
2004. This volume is less than the more-commonly
quoted annual volume of 15.0 maf (18,495 million m3)
because the time series in figure 4a was not adjusted for
water consumed in the upper basin States. The period
from 1905 to 1922, which was used to estimate water
production allocated under the Colorado River Compact,
had the highest long-term annual flow volume in the
20th century, averaging 16.1 maf (19,851 million m3)
at Lees Ferry; however, the highest annual flow volume
occurred in 1984 (22.2 maf (27,373 million m3)), and the
highest 3-yr average is 20.3 maf (25,030 million m3) for
1983 through 1985. The lowest annual flow volume is
3.8 maf (4,685 million m3) in 2002, followed by 3.9 maf
(4,809 million m3) in 1934 and 4.8 maf (5,918 million
m3) in 1977. The trend in annual flow volume, which
decreased by about 0.5 maf (617 million m3) per decade
from 1895 through 2003, is due in part to upstream
water consumption.
These data show that flow in the early 21st century
is the lowest in more than a century. The current drought
60 The State of the Colorado River Ecosystem in Grand Canyon
has contributed to the lowest flow period on record,
producing an average of only 5.1 maf (6,288 million m3)
for the 3-yr period from 2002 through 2004. In contrast,
other low 3-yr averages include 6.2 maf (7,645 million
m3) for 1989 through 1991, 6.3 maf (7,768 million m3)
for 1988 through 1990, 7.3 maf (9,001 million m3) for
1954 through 1956, and 8.0 maf (9,864 million m3) for
1933 through 1935. The 5-yr average of 5.9 maf (7,275
million m3) centered on 2002 is the lowest in the 110-yr
record. By any measure, the early 21st century drought is
the most severe in the unadjusted gaging record.
The Bureau of Reclamation (BOR) adjusted the
flow record at Lees Ferry to account for consumptive
uses in the upper basin (fig. 4b). In the BOR record, flow
volumes are available by water year (October 1 through
September 30) for the period of 1905 through 2004, a
99-yr record. The adjusted average annual flow volume
at Lees Ferry is 15.0 maf (18,495 million m3), and the
decrease in flow is 350,000 acre-feet (431,550,000 m3)
per decade (fig. 4b). Using this adjusted data, the lowest
flow year was 1977 with 5.6 maf (6,905 million m3), fol-
lowed by 2002 with 6.4 maf (7,891 million m3). The 3-yr
averages for 2002 through 2004 (9.2 maf (11,344 million
m3)), 2000 through 2002 (9.45 maf (11,652 million m3)),
and 2001 through 2003 (9.51 maf (11,726 million m3))
are the lowest in the period of record. Similarly, the low-
est 5-yr average is 9.9 maf (12,207 million m3) for 2000
through 2004, which is 1 maf (1,233 million m3) less than
the average flow of the second lowest 5-yr period (1988
through 1992). Using either the actual or adjusted flow
values, the early 21st century drought produced the low-
est flows of the past century.
Figure 4. Colorado River flow volume at Lees Ferry (before 1963) and inflows to Lake Powell (after 1963). A. Actual calendar-year flow
volumes derived from three sources. From 1895 through 1922, annual flow volumes at Lees Ferry were estimated by LaRue (1925). From
1922 through 1962, flow volumes were measured at Lees Ferry, Arizona. From 1963 to 2004, inflow to Lake Powell was estimated from
gaging records on the Colorado River and its major tributaries. B. Water-year flow volumes for Lees Ferry adjusted for consumptive use
in the upper basin (Bureau of Reclamation, unpub. data, 2005).
Climatic Fluctuations, Drought, and Flow in the Colorado River 61
Tree-ring Reconstructions
of Drought
Considerable research has addressed the question
of the magnitude, frequency, and duration of droughts
affecting the Colorado River Basin, including studies
examining the effects of the most severe known droughts
on record at Lees Ferry (Tarboton, 1995). Many of these
studies are based on the seminal work of Stockton and
Jacoby (1976), who used dendrochronology to recon-
struct long-term river flows using the actual flow record
at Lees Ferry for calibration. Recent large-scale work
(e.g., Cook and others, 2004), as well as efforts within
the drainage basin (Woodhouse, 2003; Gray and oth-
ers, 2003, 2004), while suggestive, remains insufficient to
resolve the basic magnitude-frequency questions con-
cerning the early 21st century drought and its effects on
the Colorado River Basin.
What is clear from the Stockton and Jacoby (1976)
work and other studies (Salzer, 2000; Woodhouse, 2003;
Cook and others, 2004) is how unusual the high precipi-
tation of the early 20th century was in terms of runoff
in the Colorado River. The unusually wet period of the
20th century accentuates the severity of the dry condi-
tions experienced during the early 21st century drought.
The difference between extreme wet and extreme dry
conditions is accentuated because observational records
of climate and hydrologic conditions in the Colorado
River Basin generally span 100 yr or less, limiting our
ability to quantitatively understand the current drought
in a long-term context. It is possible, however, to quali-
tatively view this drought in a long-term context from
analysis of tree rings, which provide an indication of
moisture conditions going back several centuries.
Using dendrochronological reconstructions from
tree rings from the Western United States, Cook and
others (2004) analyzed long-term changes in the area
affected by drought from A.D. 800 to 2003. Although the
region they considered is far larger than the Colorado
River Basin and subject to a larger array of climatic
influences, their reconstruction provides some perspec-
tive on the 2000 through 2004 drought in the Colorado
River Basin. Cook and others (2004) concluded that
the present drought is not comparable to the so-called
“megadroughts” of A.D. 936, 1034, 1150, and 1253,
primarily because of its short duration; however, the
early 21st century drought may not yet be over. At the
very least, their drought-area reconstruction (Cook and
others, 2004) suggests that the present drought may sur-
pass other 20th century droughts in the Western United
States, including the droughts of the midcentury and the
1930s, and be comparable to droughts of the mid-19th
or late 16th centuries.
Several researchers (Tarboton, 1995; Cook and
others, 2004; Gray and others, 2004) have noted that
decadal-scale persistence of below-average precipitation
is of paramount importance when considering drought
frequency. Tarboton (1995) and Meko and others (1995)
provided data based on the Stockton and Jacoby (1976)
reconstructions that, when compared to conditions of
2001–04, suggest that the low-flow conditions of the
early 21st century may be the lowest since the drought of
A.D. 1579 to 1600.
Recent Findings
Several indices of atmospheric and oceanic processes
are used to explain climate variability in the Western
United States, including the Southern Oscillation Index
(SOI), the Pacific Decadal Oscillation (PDO), and the
Atlantic Multidecadal Oscillation (AMO). These indices
reflect short- to long-term conditions that can affect the
discharge of the Colorado River.
Southern Oscillation Index
Perhaps the most well known of the climatic indices
is the Southern Oscillation Index (SOI), which is often
used to indicate the status of the El Niño-Southern
Oscillation (ENSO) phenomenon in the Pacific Ocean.
The SOI is the measure of the strength of tropical
Pacific atmospheric circulation based on the sea-level
pressure difference between Tahiti, French Polynesia,
and Darwin, Australia (fig. 5a). Negative values, implying
weakened trade winds, are mainly the result of higher-
than-normal surface pressures at Darwin and are associ-
ated with El Niño conditions. The impacts of ENSO are
felt worldwide through disruption of the general circula-
tion of the atmosphere and associated global weather
patterns. In terms of the Colorado River Basin, ENSO
affects interannual variation of climate and precipita-
tion in Arizona (Andrade and Sellers, 1988) and helps
to explain the occurrence of floods and droughts in the
Western United States (Cayan and others, 1998, 1999).
The ENSO is a change between three basic states
of the ocean. The warm phase, called El Niño, involves
warming of the eastern Pacific Ocean off Peru and the
northward spread of warm surface water to the west
coast of the United States. Because warming of sea-sur-
face temperatures (SSTs) is a hallmark of El Niño condi-
tions (Knutson and others, 1999), several indices based
62 The State of the Colorado River Ecosystem in Grand Canyon
on SSTs have been developed, including the NINO3
index (fig. 5b). Reduced sea-level pressure over the east-
ern tropical Pacific Ocean combined with increased sea-
level pressure over Indonesia (negative SOI) leads to a
weakening in the trade winds, enabling warm water from
the central equatorial Pacific Ocean to move toward and
along the west coast of South America (positive NINO3
index). The cold phase, called La Niña, is the opposite
of the warm phase. Thus, El Niño and La Niña are the
warm and cold phases of the ENSO system, which also
includes a neutral condition that can persist for several
years between the two polar phases. ENSO phases typi-
cally last 6–18 mo and are the single most important fac-
tor affecting interannual climatic variability on a global
scale (Diaz and Markgraf, 1992).
The ENSO also affects atmospheric circulation and
SSTs in the eastern Pacific Ocean, which in turn affect
the transport of moisture across the Western United
States. During El Niño conditions, the warmer-than-
average water in the eastern tropical Pacific Ocean and
a shift in storm tracks tend to produce above-average
precipitation (Redmond and Koch, 1991), above-average
runoff (Cayan and Webb, 1992), and, potentially, floods
in the Southwest. Not all El Niño events lead to increased
runoff, however; during the 2003 El Niño, runoff was
below average.
During La Niña events, cooler-than-average SSTs
in the eastern tropical Pacific Ocean tend to cause less
moisture to flow over the continent, typically causing
below-average flow in the Colorado River and predict-
able below-average precipitation in the Southwestern
United States. This below-average precipitation occurs
despite a tendency for above-average precipitation in the
headwaters of the Colorado River Basin, although pre-
cipitation gained is negated by most of the basin having
below-average precipitation.
Figure 5. A. The Southern Oscillation Index (SOI) varies with a 4- to 7-yr periodicity between negative (El Niño) and positive (La Niña)
states. B. The NINO3 index is a standardized anomaly index of sea-surface temperatures (SSTs) in an area of the equatorial Pacific
Ocean from 150ºW to 90ºW longitude and ±5º latitude centered on the equator. Comparison of these diagrams shows that when SOI is
negative, the NINO3 index generally is positive.
Climatic Fluctuations, Drought, and Flow in the Colorado River 63
Pacific Decadal Oscillation
The Pacific Decadal Oscillation (PDO) index (fig.
6) was developed from SSTs in the Pacific Ocean north
of 20ºN latitude (Mantua and Hare, 2002). Two main
characteristics distinguish the PDO from ENSO: (1) the
PDO state (positive or negative) persists for decades,
while typical ENSO events persist for 6 to 18 mo; and
(2) the climatic signal of the PDO is most visible in the
North Pacific Ocean instead of the tropics. The PDO
index is commonly used to explain long-term periods
of above- or below-average precipitation in the West-
ern United States. When the PDO is positive, there is
colder water in the central and western Pacific Ocean
and warmer waters in the eastern Pacific Ocean; under
negative PDO, the reverse is true. Positive PDO values
are usually associated with wetter conditions in the
Southwestern United States, while negative PDO values
are suggestive of persistent drought in the Southwest.
Long-term changes in the PDO may also influence
snowmelt runoff in the Western United States, which is
occurring earlier in the year, particularly in the Pacific
Northwest and in the Sierra Nevada Range of Califor-
nia (Stewart and others, 2005).
Shifts in the phase of the PDO occurred in about
1944 and 1977 (Hereford and others, 2002; McCabe
and others, 2004); from 1999 through 2004, PDO values
have varied from negative (1999–2001) to positive (2002–
04). While this might be viewed as an inconsistency with
the persistent drought conditions during that period, the
geographic center of drought conditions shifted towards
the Pacific Northwest in a manner consistent with a posi-
tive (warm) PDO. At present, neither the causes of the
variations in PDO values nor their predictability are well
known; although, recent studies indicate that the PDO
may be associated with decadal-length periods of above-
and below-average precipitation and streamflow in the
Colorado River Basin (Hidalgo and Dracup, 2004).
Atlantic Multidecadal Oscillation
The Atlantic Multidecadal Oscillation (AMO)
(Kerr, 2000) reflects conditions in the Atlantic Ocean
that may affect climate in the continental United States
(fig. 7). The AMO is discussed only to point out that it
is an interesting and possibly significant index; much
additional research is needed to demonstrate its useful-
ness. As its name implies, AMO events have a persistence
of 20 to 35 yr. Warm conditions indicated by positive
AMO values are thought to be associated with drought
conditions (Enfield and others, 2001), such as the Dust
Bowl on the Great Plains (Schubert and others, 2004)
and other periods of drought during the last century
(McCabe and others, 2004).
Cool phases in the Atlantic Ocean occurred from
1902 to 1925 and from 1970 to 1994; these periods
coincide with generally above-average precipitation
and runoff in the Colorado River Basin. A warm phase
occurred almost continuously from 1926 to 1963, which
coincides with persistent average or below-average
rainfall and runoff in the Colorado River Basin between
the early 1930s and 1960s. More recently, the Atlantic
Ocean warmed in 1996 and remained so through 2004.
Fluctuations in the AMO combined with those of the
PDO may help explain long-term drought frequency
(Gray and others, 2003, 2004) and, therefore, fluctuation
in runoff in the Colorado River Basin.
Climate Indices and Drought
As knowledge increases about the influence of the
oceans on the climate of the United States, so too does
the awareness of the enormous complexity of the ocean-
atmosphere system, particularly its variation over time.
After intense scrutiny, scientists have learned that no
single index of the system can explain all climate varia-
tions. It is increasingly evident that the various factors
occur together in a complicated fashion. As a result,
Figure 6. The Pacific Decadal Oscillation (PDO) is typically
associated with long-term climatic variation in the Western
United States. Positive PDO values suggest wetter periods (e.g.,
1976 through 1995) for the Southwest and drier periods for the
Northwestern United States. In contrast, negative values suggest
persistent drier-than-average conditions in the Southwest (e.g.,
mid-1940s through mid-1970s).
64 The State of the Colorado River Ecosystem in Grand Canyon
researchers attempt to use a combination of indices to
explain the occurrence and spatial extent of droughts
(e.g., McCabe and others, 2004).
In terms of the Colorado River Basin, the river’s
flow is related to the indices of global climate change in
a complex way (Hidalgo and Dracup, 2004). From an
interannual perspective, large floods and high runoff
volumes typically occur during strong El Niño conditions
(e.g., 1916–17, 1983–84), whereas La Niña conditions
typically cause low-flow conditions (e.g., 1934, 1996).
Above-average precipitation during El Niño, however,
tends to occur in the southern part of the watershed
while the northern part remains dry, a situation that
tends to reverse during La Niña conditions.
Furthermore, the watershed of the Colorado River
spans more than 10º of latitude, and precipitation patterns
over that range do not necessarily respond in concert to
regional climatic fluctuations. For example, above-average
runoff in part of the watershed (e.g., the northern half)
may overcome low runoff in other parts (e.g., the south-
ern half) during some low-flow periods. As a result, much
of the variability in the annual flow record is not easily
explained by climate indices. For example, the midcen-
tury drought, which was severe on the Colorado Plateau
(Hereford and others, 2002), caused only slightly below-
average runoff in the entire basin; the average runoff vol-
ume during this period was 11.1 maf (13,686 million m3)
Figure 7. The detrended Atlantic Multidecadal Oscillation (AMO)
is related to persistent sea-surface temperature (SST) conditions
in the Atlantic Ocean. Positive values are associated with higher-
than-average drought frequencies in the United States.
for the period from 1948 to 1963. The response of Colo-
rado River flow to the interaction of these climate indices
is complicated, underscoring the concept that hydrologic
drought results from an integrated set of climatological
factors that are not easily predicted or explained.
The predicted effects of future climatic change sug-
gest overall warming conditions and decreased average
annual runoff in the basin (Christiensen and others,
2004), although a simple hydrologic response to this
complex climatic framework seems unlikely. Predicted
temperature increases suggest that snowmelt runoff may
be less prevalent and may occur earlier in spring (Stewart
and others, 2004, 2005). These analyses raise the pos-
sibility that legally mandated flow releases from Glen
Canyon Dam may be possible in only 80% of future
years owing to climatic change.
Drought Persistence and Relation
with Indices of Global Climate
Dendrochronological analyses show that since A.D.
1226, nine droughts have occurred in the Colorado
River Basin lasting 15–20 yr and four droughts have
occurred lasting more than 20 yr (Gray and others,
2003). Several of these droughts were punctuated by
above-average precipitation related to discrete El Niño
events, which could be analogous to the effect of current
El Niño conditions on the Colorado River Basin. More-
over, tree-ring records indicate that some past droughts
in the Colorado River Basin persisted for several decades
(Meko and others, 1995), leaving open the possibility that
the present drought could resume after the ongoing El
Niño ends and continue for many more years.
By using tree-ring records spanning 700 yr, Gray
and others (2003, 2004) found 30- to 70-yr multidecadal
oscillations in drought frequency in the area that includes
the headwaters of the Colorado River Basin. They also
found a strong relation between drought occurrence and
SSTs in the North Atlantic Ocean as manifested particu-
larly in the AMO index (fig. 7) but which also included
the PDO index (fig. 6). While neither index has a strong
statistical relation to annual Colorado River flow (fig. 8),
the combination may provide a context for the potential
duration of the early 21st century drought. The broad
relation between the PDO index and drought suggests
that the present drought could persist for several decades
after the end of the present El Niño period.
Climatic Fluctuations, Drought, and Flow in the Colorado River 65
Figure 8. Time series showing the complex interrelations among indices of global climate and annual flow volumes of the Colorado
River from 1895 through 2003. Colored vertical bars delineate dry (tan) and wet (light blue) climate periods. A. Southern Oscillation Index
(SOI, dimensionless). B. Pacific Decadal Oscillation (PDO, dimensionless). C. Atlantic Multidecadal Oscillation (AMO, deviation in ºC), not
detrended as in figure 7. D. Actual annual flow volume (in millions of acre-feet (maf)) passing Lees Ferry or entering Lake Powell (fig. 4).
66 The State of the Colorado River Ecosystem in Grand Canyon
Discussion and Management
Implications
From 2000 through 2004, the early 21st century
drought caused abnormally low flows in the Colorado
River and its tributaries upstream from Lake Powell.
By using either actual annual flow data or annual flow
records adjusted for consumptive uses in the upper basin,
it was found that runoff from 2000 through 2004 was the
lowest in the period of record (99–110 yr). This low flow
has caused considerable concern about the ability of the
reservoirs on the Colorado River to deliver water from
upper basin States to lower basin States. Water managers
increasingly want to know the predictability of climate
and its effects on water resources over annual, decadal,
and longer term spans.
Climate, drought, and streamflow in the Colorado
River are linked in poorly understood ways. Initial
understanding of flows in the system was based on a
relatively short historical record that is now believed to
be a period of above-average precipitation. Examination
of long-term records based on tree-ring analyses sug-
gests that drought magnitude and persistence patterns
are associated with much broader hemispheric climate
patterns; however, these correlations are imperfect and
do not provide a clear understanding of long-term pre-
cipitation patterns.
Currently, there is no reliable way to predict how long
the early 21st century drought will last in the Colorado
River Basin. Components of the climate system, such
as sea-surface temperature of the Atlantic and Pacific
Oceans, provide some context for understanding past
variations in precipitation and streamflow, but they are
insufficient for predicting the fate of the ongoing drought.
Time series of the relevant climate indices indicate a
large amount of year-to-year variability and relatively
rapid changes from one regime to another. Above-average
precipitation for winter 2004–05 and forecasts for above-
average runoff may signal the end of the drought, or the
drought conditions may resume after the present El Niño
ends. Both outcomes underscore the unpredictability of
climatic shifts affecting the Colorado River Basin.
References
Andrade, E.R., Jr., and Sellers, W.D., 1988, El Niño and
its effect on precipitation in Arizona and western New
Mexico: Journal of Climatology, v. 8, p. 403–410.
Cayan, D.R., Dettinger, M.D., Diaz, H.F., and Graham,
N.E., 1998, Decadal variability of precipitation over
western North America: Journal of Climate, v. 11, p.
3148–3166.
Cayan, D.R., Redmond, K.T., Riddle, L.G., 1999,
ENSO and hydrologic extremes in the western United
States: Journal of Climate, v. 12, p. 2881–2893.
Cayan, D.R., and Webb, R.H., 1992, El Niño/South-
ern Oscillation and streamflow in the western United
States, in Diaz, H.F., and Markgraf, V., eds., El Niño,
historical and paleoclimatic aspects of the Southern
Oscillation: Cambridge, England, Cambridge Univer-
sity Press, p. 29–68.
Christiensen, N.S., Wood, A.W., Voisin, N., Lettenmaier,
D.P., and Palmer, R.H., 2004, The effects of climate
change on the hydrology and water resources of the
Colorado River basin: Climatic Change, v. 62, p.
337–363.
Cook, E.R., Woodhouse, C.A., Eakin, C.A., Meko,
D.M., and Stahle, D.W., 2004, Long-term aridity
changes in the western United States: Science, v. 306,
p. 1015–1018.
Dettinger, M.D., and Cayan, D.R., 1995, Large-scale
atmospheric forcing of recent trends toward early
snowmelt runoff in California: Journal of Climate, v.
8, p. 606–623.
Diaz, H.F., and Markgraf, V., eds., 1992, El Niño,
historical and paleoclimatic aspects of the Southern
Oscillation: Cambridge, England, Cambridge Univer-
sity Press, 476 p.
Enfield, D.B., Mestas-Nuñez, A.M., and Trimble, P.J.,
2001, The Atlantic multidecadal oscillation and its
relation to rainfall and river flows in the continental
U.S.: Geophysical Research Letters, v. 28, p. 2077–2080.
Gray, S.T., Betancourt, J.L., Fastie, C.L., and Jackson,
S.T., 2003, Patterns and sources of multidecadal oscil-
lations in drought-sensitive tree-ring records from the
central and southern Rocky Mountains: Geophysical
Research Letters, v. 30, doi:10.1029/2002GL01154, p.
1:49-1:49-4.
Gray, S.T., Graumlich, L.J., Betancourt, J.L., and
Pederson, G.T., 2004, A tree-ring based reconstruction
of the Atlantic Multidecadal Oscillation since 1567
A.D.: Geophysical Research Letters, v. 31, L12205,
doi:10.1029/2004GL019932, 4 p.
Climatic Fluctuations, Drought, and Flow in the Colorado River 67
Hereford, R., Webb, R.H., and Graham, S., 2002,
Precipitation history of the Colorado Plateau region,
1900–2000: U.S. Geological Survey Fact Sheet 119-
02, 4 p.
Hidalgo, H.G., and Dracup, J.A., 2004, Evidence of the
signature of North Pacific multidecadal processes on
precipitation and streamflow variations in the upper
Colorado River basin, in Van Riper, C., III, and Cole,
K.L., eds., The Colorado Plateau, cultural, biological
and physical research: Tucson, University of Arizona
Press, p. 257–265.
Hoerling, M., and Kumar, A., 2003, The perfect ocean
for drought: Science, v. 299, p. 691–694.
Kerr, R.A., 2000, A North Atlantic climate pacemaker
for the centuries: Science, v. 288, p. 1984–1986.
Knutson, T.R., Kaplan, A., and Rayner, N.A., 1999, A
note on 20th century equatorial Pacific sea surface
temperatures: http://www.gfdl.noaa.gov/~tk/Note_
on_Eq_Pac_SSTs.html, accessed February 21, 2005.
LaRue, E.C., 1925, Water power and flood control of
Colorado River below Green River, Utah: U.S. Geo-
logical Survey Water Supply Paper 556, 176 p.
Mantua, N.J., and Hare, S.R., 2002, The Pacific decadal
oscillation: Journal of Oceanography, v. 58, p. 35–42.
McCabe, G.J., and Clark, M.P., in press, Trends and
variability in snowmelt runoff in the western United
States: Journal of Hydrometeorology.
McCabe, G.J., Palecki, M.A., and Betancourt, J.L., 2004,
Pacific and Atlantic Ocean influences on multidecadal
drought frequency in the United States: Proceed-
ings of the National Academy of Science, v. 101, p.
4136–4141.
Meko, D.M., Stockton, C.W., and Boggess, W.R., 1995,
The tree-ring record of severe sustained drought:
American Water Resources Association, Water
Resources Bulletin, v. 31, p. 789–801.
Redmond, K.T., and Koch, R.W., 1991, Surface climate
and streamflow variability in the western United States
and their relationship to large-scale circulation indices:
Water Resources Research, v. 27, p. 2381–2399.
Salzer, M.W., 2000, Dendroclimatology in the San Fran-
cisco Peaks region of northern Arizona, USA: Tucson,
University of Arizona, unpublished Ph.D. disserta-
tion, 211 p.
Schubert, S.D., Suzrez, M.J., Pegion, P.J., Koster, R.D.,
and Bacmeister, J.T., 2004, On the cause of the 1930s
Dust Bowl: Science, v. 303, p. 1855–1859.
Stewart, I.T., Cayan, D.R., and Dettinger, M.D., 2004,
Changes in snowmelt runoff timing in western North
America under a ‘business as usual’ climate change
scenario: Climatic Change, v. 62, p. 217–232.
Stewart, I.T., Cayan, D.R., and Dettinger, M.D., 2005,
Changes toward earlier streamflow timing across
western North America: Journal of Climate, v. 18, p.
1136–1155.
Stockton, C.W., and Jacoby, G.C., Jr., 1976, Long-term
surface-water supply and streamflow trends in the
upper Colorado River basin based on tree-ring analy-
ses: Lake Powell Research Project Bulletin No. 18, 70 p.
Tarboton, D.G., 1995, Hydrologic scenarios for severe
sustained drought in the southwestern United States:
American Water Resources Association, Water
Resources Bulletin, v. 31, p. 803–813.
Webb, R.H., McCabe, G.J., Hereford, R., and
Wilkowske, C., 2004, Climatic fluctuations, drought,
and flow of the Colorado River: U.S. Geological Sur-
vey Fact Sheet 2004-3062, 4 p.
Woodhouse, C.A., 2003, A 431-yr reconstruction of
western Colorado snowpack from tree rings: Journal
of Climate, v. 16, p. 1551–1561.
68 The State of the Colorado River Ecosystem in Grand Canyon
Contact Information:
Robert H. Webb
Research Hydrologist
U.S. Department of the Interior
U.S. Geological Survey
Water Resources Discipline
Tucson, AZ
rhwebb@usgs.gov
Richard Hereford
Research Geologist (emeritus)
U.S. Department of the Interior
U.S. Geological Survey
Western Earth Surface Processes Team
Flagstaff, AZ
rhereford@usgs.gov
Gregory J. McCabe
Physical Scientist
U.S. Department of the Interior
U.S. Geological Survey
Water Resources Discipline
Denver, CO
gmccabe@usgs.gov
First page photograph credit: Andrew Pernick, Bureau of Reclamation
A
n
d
r
e
w
P
e
r
n
i
c
k
,
B
u
r
e
a
u
o
f
R
e
c
l
a
m
a
t
i
o
n
Chapter 4
Water Quality
in Lake Powell
and the
Colorado River
William S. Vernieu
Susan J. Hueftle
Steven P. Gloss
Introduction
Water temperature, nutrient concentrations, turbid-
ity, and other water-quality parameters are of interest
to managers and scientists because these parameters
influence a range of ecosystem components, from sup-
port of aquatic microorganisms and invertebrates to the
behavior of native and nonnative fishes. For example,
declines of Colorado River Basin native fishes and
changes in their condition have been attributed, in part,
to low water temperatures downstream from dams, such
as Glen Canyon Dam, that release water from deeper
portions of the reservoir (Clarkson and Childs, 2000).
Similarly, water quality is an important determinant of
food-web structure in aquatic habitats and abundance of
consumers like fish in those food webs (Carpenter and
Kitchell, 1996; Wetzel, 2001).
Any investigation of the dynamics of the Colorado
River ecosystem in Grand Canyon must not only
document and understand the water quality in Grand
Canyon itself but also the water quality in Lake Powell,
the reservoir created by Glen Canyon Dam. The
impoundment of a river system in a reservoir alters
downstream water quality in many ways (Nilsson and
others, 2005). The formation of Lake Powell in 1963 was
accompanied by reductions in suspended-sediment and
nutrient transport and by changes in seasonal tempera-
tures, discharge levels, and benthic community struc-
ture of the Colorado River (Paulson and Baker, 1981;
Stevens and others, 1997; Topping and others, 2000
a, b). More recently, reservoir and downstream water
quality has been affected by reservoir drawdown from
a 5-yr basinwide drought in the Western United States.
Water released from Glen Canyon Dam in 2003 and
2004 was the warmest recorded since August 1971, when
Lake Powell was in its initial filling period (initial filling
of the reservoir began in 1963 with the closure of Glen
Canyon Dam, and it reached full pool of 3,700 ft for the
first time in 1980). Changes in stratification and the fate
of inflow currents in Lake Powell under various storage
conditions, as well as various operational scenarios such
as experimental releases and a proposed temperature
control device, could have significant effects on the qual-
ity of water released from Glen Canyon Dam.
This chapter provides an overview of water-quality
trends and conditions in Lake Powell and the Grand
Canyon ecosystem. Because Lake Powell and Glen
Canyon Dam operations have a strong influence on
70 The State of the Colorado River Ecosystem in Grand Canyon
downstream water quality, the water quality of the
reservoir is discussed in some detail. The chapter also
addresses recent drought-induced changes and the effects
of the modified low fluctuating flow (MLFF) alterna-
tive. The monitoring of water quality in Lake Powell
is conducted by the U.S. Geological Survey’s (USGS)
Grand Canyon Monitoring and Research Center under
separate funding from the Bureau of Reclamation
and is not funded by the Glen Canyon Dam Adaptive
Management Program.
Background
Glen Canyon Dam has a structural height of 710 ft
(216 m). This high, concrete-arch dam backs up water
for 186 mi (299 km) to form Lake Powell, the second
largest reservoir in the United States. Lake Powell had
an original capacity of 27.1 million acre-feet (maf)
(33,414 million m3) and a surface area of 161,390 acres
(65,315 ha) at full pool elevation of 3,700 ft (1,128 m).
By 1986, this capacity had been reduced to 26.2 maf
(32,305 million m3) because of an estimated loss of
capacity of 30,000 acre-feet (af) (36,990,000 m3) per
year resulting from sedimentation (Ferrari, 1988). Water
can be released from Glen Canyon Dam through three
separate structures (spillways, penstocks, and river outlet
works). The majority of water is routed through eight
penstocks, which feed the powerplant turbines. The
penstock inlets are at an elevation of 3,470 ft (1,058 m)
and have a maximum combined discharge capacity of
approximately 33,200 cubic feet per second (cfs) when
the reservoir is full. Water can also be released from (1)
the river outlet works at an elevation of 3,374 ft (1,028
m) and (2) two spillways at an elevation of 3,648 ft (1,112
m), both of which bypass the powerplant turbines and
have discharge capacities of 15,000 cfs and 208,000 cfs,
respectively (Bureau of Reclamation, 1981) (fig. 1).
Lake Powell
Glen Canyon Dam began storing water on March
13, 1963, and full pool elevation was reached on June
22, 1980. Ninety-six percent of the reservoir’s inflow is
received from the Colorado and San Juan Rivers; the
majority of this inflow is received from May to July as
the result of snowmelt in the Rocky Mountains (Stanford
and Ward, 1991). The impoundment of the Colorado
River by Glen Canyon Dam altered the quality, seasonal
release volumes, and the amount of daily fluctuations
for the Colorado River ecosystem downstream of the
dam. Colorado River water is now transformed by an
approximate 2-yr residence time in Lake Powell and
by the structure and operation of Glen Canyon Dam.
These factors influence the temperature, suspended and
dissolved solids, nutrients, and organisms that pass down-
stream as well as the volume of water released and the
magnitude of fluctuations.
Lake Powell has a maximum depth immediately
upstream of Glen Canyon Dam of approximately 515 ft
(157 m) at full pool elevation; the lake is vertically strati-
fied into density layers and differs longitudinally as the
currents move through the reservoir. Vertical stratifica-
tion varies seasonally and is determined by the relative
density of the different layers of the reservoir. Density
is determined by water temperature and the amount of
dissolved minerals and suspended solids. The surface
layer of the reservoir, or epilimnion, warms through
summer and is eventually mixed with deeper water by
the wind and convective currents during the winter cool-
ing period, which extends from October to early March.
The epilimnion exhibits the highest level of biological
activity because of warm temperatures and light avail-
ability. Water temperature decreases with depth in the
metalimnion, the layer that separates the epilimnion
from the bottom layer of the reservoir, or hypolimnion.
The hypolimnion consistently exhibits lower tempera-
tures, lower dissolved oxygen levels, and higher salinity
concentrations than the other layers of the reservoir.
Because of the subsurface position of the penstocks,
water may be withdrawn from the epilimnion, metalim-
nion, or hypolimnion depending on reservoir level, reser-
voir hydrodynamics, timing and strength of stratification,
and magnitude of withdrawals.
Longitudinal variation in water quality is the result
of currents moving through the reservoir. The por-
tions of the reservoir farthest from the dam exhibit
characteristics similar to those of the river entering the
reservoir, with more variable temperature and salinity
patterns and higher sediment and nutrient concentra-
tions. Primary productivity from photosynthesis is limited
by light availability in this more turbid riverine zone.
The deeper portions of the reservoir closest to the dam,
or the lacustrine zone, exhibit characteristics similar to
those of a lake system, with more stable temperature and
salinity patterns, low suspended-sediment concentration,
and lower nutrient concentrations. Primary productivity
in this zone is limited by nutrient availability. A transition
zone of intermediate characteristics separates the river-
ine and lacustrine zones (Kimmel and Groeger, 1984;
Department of the Army, Corps of Engineers, 1987;
Ford, 1990). The relative location of these zones depends
on reservoir levels and the magnitude of inflows. In the
Water Quality in Lake Powell and the Colorado River 71
main channel of Lake Powell, the riverine zone extends
from the Colorado River inflow to Hite Bay, the transi-
tion zone extends from Hite Bay to the Bullfrog Bay
area, and the lacustrine zone extends from Bullfrog Bay
to Glen Canyon Dam (figs. 1 and 2).
The depth at which river water enters the reservoir
is dictated by its density relative to the density of the
water already in the reservoir. Spring and early summer
snowmelt runoff entering the reservoir tends to be dilute,
has warmed during its passage through the canyonlands,
and represents the lowest density water entering the
reservoir during the year. Consequently, this water travels
through the reservoir as an overflow density current.
During the winter months, inflows are colder and more
saline and represent the highest density water entering
the reservoir. Depending on the relative density of the
hypolimnion, winter inflows will either flow along the
bottom of the reservoir, routing fresh water to the hypo-
limnion and displacing older water upward, or flow into
intermediate layers, leaving deeper waters stagnant.
Convective mixing takes place in the epilimnion as
the reservoir cools during the fall and winter months.
By the end of the calendar year, convective mixing in
the upper layers progresses to the point that penstock
withdrawals begin to exhibit characteristics of the
epilimnion, which contains the warmest water in the
reservoir at that time of year, despite the cooler weather
conditions. This convective mixing results in the warmest
release temperatures of the year occurring in late fall or
early winter.
Downstream of Glen Canyon Dam
Changes to the chemical and physical quality of the
water of the Colorado River after its release from Glen
Canyon Dam are affected by ambient meteorological
conditions, primary production and respiration from the
aquatic environment, aeration from rapids, inputs from
other tributary sources and overland flow, and various
aspects of the operation of Glen Canyon Dam.
Figure 1. Profile of Lake Powell from Glen Canyon Dam to the inflow of the Colorado River, illustrating the vertical stratification and
horizontal zonation of the reservoir at or near full pool elevation, September 1999. Also shows the elevations of each of the three
release structures and their capacities as well as an approximation of the wedge of deltaic sediments. Y axis on left is measurement of
elevation above mean sea level and on right is actual depth.
72 The State of the Colorado River Ecosystem in Grand Canyon
Water released from Glen Canyon Dam is usually
colder than the surrounding environment and warms as
it flows downstream with exposure to solar radiation and
warmer ambient air temperatures. The exception to this
pattern is during portions of the winter months when
dam releases are slightly warmer than the surrounding
environment and cool as they flow downstream before
warming again in lower elevation reaches.
The aquatic environment affects dissolved oxygen
concentrations and pH in the tailwater (referred to as the
Lees Ferry reach elsewhere in this report), which is the
15 mi (24 km) of the river that extends downstream from
Glen Canyon Dam to Lees Ferry. This area is free of
significant tributary sediment inputs that limit light avail-
ability for primary production (Yard and others, 2005).
As a result of photosynthetic activity, therefore, dissolved
oxygen concentrations and pH in the tailwater display
daily oscillations at Lees Ferry. During daylight hours dis-
solved oxygen concentrations and pH increase because
of the addition of oxygen and removal of carbon dioxide
during photosynthesis. The opposite occurs at night
when respiratory processes become dominant (Marzolf
and others, 1999; U.S. Geological Survey, 2001).
Under normal powerplant discharges, limited aera-
tion of the river occurs in the tailwater reach of the river
compared to downstream reaches. Generally, released
water that may be lower in dissolved oxygen does not
reach full saturation until the first rapids in Marble
Canyon, where the water is aerated by turbulence;
however, during periods when the river outlet works are
Figure 2. Lake Powell water-quality sampling sites.
Water Quality in Lake Powell and the Colorado River 73
operated, such as during the 1996 beach/habitat-build-
ing flow or the 2004 experimental high flow, turbulence
immediately below the dam is sufficient to bring release
water up to full oxygen saturation (Hueftle and Stevens,
2001).
Various tributaries that enter Grand Canyon
can significantly affect water quality of the Colorado
River below Glen Canyon Dam. The Paria and Little
Colorado Rivers can carry large amounts of fine sedi-
ment that limit light availability for primary production
and may enhance conditions for native fish that use tur-
bid water for cover from predation (Shannon and others,
1994; Topping and others, 2000 a, b). Some tributaries,
such as the Little Colorado River, are significant sources
of salinity to the mainstem Colorado River (Cole and
Kubly, 1976).
Water-quality Monitoring
Lake Powell
The purpose of water-quality monitoring in Lake
Powell is to document and understand the water-quality
changes that occur during the residence time of the
water in the reservoir and how those changes may affect
the quality of water being released from Glen Canyon
Dam under various conditions.
Water-quality monitoring of Lake Powell currently
has two main components. Monthly surveys of the fore-
bay, the pool of water in front of the dam, take place at
the mouth of Wahweap Bay, approximately 1.5 mi (2.4
km) upstream of Glen Canyon Dam, to document the
quality of water in dam releases. Reservoir-wide surveys
are conducted quarterly to describe seasonal changes in
the stratification and hydrodynamics of the reservoir and
to better understand the reason for observed changes in
downstream releases.
Water-quality sampling in Lake Powell was initiated
by the Bureau of Reclamation in 1964 and continued
through 1990, including several phases of differing
sampling frequencies for the reservoir and forebay. Glen
Canyon Environmental Studies conducted the monitor-
ing from 1990 to 1996. The USGS Water Resources
Discipline conducted monitoring in Lake Powell on
several dates in 1992, 1994, and 1995 (U.S. Geological
Survey, 1998). Since 1997, monitoring has been con-
ducted by the USGS Grand Canyon Monitoring and
Research Center.
Monthly monitoring of the forebay allows for the
observation of conditions immediately upstream of Glen
Canyon Dam and for the description of the dynamics
of the water column that is the immediate source for
downstream releases. Quarterly reservoir-wide sampling
describes seasonal conditions at 20–25 stations through-
out the reservoir during the maximum extent of winter
convective mixing, spring runoff, post runoff/late sum-
mer stratification, and early winter conditions during the
early phases of convective mixing (fig. 2, table 1).
At each station, data on basic water-quality param-
eters—temperature, specific conductance, dissolved
oxygen, pH, oxidation-reduction potential, and turbid-
ity—are collected through the water column. At selected
depths, chemical (major ions and nutrients) and biologi-
cal (chlorophyll and plankton) sampling is performed to
characterize the major strata in the water column. Major
ions are the common negative (e.g., chloride) and positive
(e.g., calcium) ions that constitute the majority of miner-
als dissolved in water. Nutrients represent the total and
dissolved fractions of compounds of phosphorus and
nitrogen, which are essential for the production of plant
life (algae or phytoplankton).
Glen Canyon Dam Tailwater
Water-quality monitoring activities in the dam’s
tailwater assess the initial quality of water leaving the res-
ervoir and entering Grand Canyon. These baseline mea-
surements are important for detecting changes occurring
in Grand Canyon and for understanding the relationship
between the quality of water leaving the reservoir and its
relationship to the downstream aquatic ecosystem (fig. 3).
The USGS recorded daily instantaneous water
temperatures at Lees Ferry from 1949 to 1977 (U.S.
Geological Survey, 2004). Since then, temperatures
recorded at Lees Ferry reflect mean daily values of mul-
tiple observations (U.S. Geological Survey, 1985–2004).
Glen Canyon Environmental Studies began monitoring
the temperature and conductivity of dam releases in
1988 by using remotely deployed, continuously log-
ging monitors. In 1991, this program was expanded to
include continuous monitoring at Lees Ferry. Dissolved
oxygen and pH measurements were added to the moni-
toring protocol shortly afterwards.
Tailwater monitoring activities currently include the
continuous measurement of temperature, salinity, dis-
solved oxygen, and pH and monthly sampling for phos-
phorus, nitrogen, major-ion chemistry composition, and
biological indicators such as chlorophyll and plankton.
74 The State of the Colorado River Ecosystem in Grand Canyon
Table 1. Lake Powell and tailwater sampling sites.
Site name Distance in Chemical and
miles (kilometers) biological sampling
from Glen Canyon Dam
Tailwater
Colorado River below Glen Canyon Dam 0 X
Colorado River at Lees Ferry -15.5 (-24.9) X
Colorado River main channel
Wahweap 1.5 (2.4) X
Crossing of the Fathers 28.1 (45.2) X
Oak Canyon 56.2 (90.5) X
San Juan River confluence 62.2 (100.1)
Escalante 72.6 (116.9) X
Iceberg 86.7 (139.5)
Lake Canyon 98.6 (158.7)
Bullfrog Bay 104.3 (167.9) X
Moki Canyon 111.8 (179.9)
Knowles Canyon 120.1 (193.3)
Lower Good Hope Bay 129.6 (208.5) X
Scorup Canyon 140.1 (225.5)
Hite Basin 148.3 (238.7) X
Colorado River inflow 149.1–185.8 (240.0–299.0) X
San Juan River arm
Cha Canyon 12.0 (19.3) X
Lower Piute Bay 20.4 (32.9)
Upper Piute Bay 26.8 (43.1) X
Lower Zahn Bay 38.8 (62.5)
Mid Zahn Bay 42.6 (68.6)
San Juan inflow 32.3–54.1 (52.0–87.0) X
Escalante River arm
Escalante at Clear Creek 4.5 (7.2)
Escalante at Davis Gulch 7.4 (11.9) X
Escalante at Willow Creek 12.4 (20.0)
Escalante inflow 13.7–24.8 (22.0–40.0) X
Downstream Thermal Monitoring
in Grand Canyon
Downstream thermal monitoring provides an
indication of status and trends in water temperature and
how warming is affected by river reach, seasonality, and
dam operations. Concerns about the effects of the ther-
mal regime on both native and nonnative fish resulted
in the development of a continuous thermal monitoring
program in Grand Canyon beginning in 1990. Thermal
monitoring was conducted at 10 mainstem stations at
intervals of roughly 30 mi (48 km) and at 8 additional
sites on major tributaries. Tributary sites have been
monitored since 1994, providing thermal baseline data
for streams that may act as warmwater refugia for many
aquatic species, particularly native fish. In 2005, thermal
monitoring in tributaries was reduced to four sites, the
Paria River, the Little Colorado River, Kanab Creek,
and Havasu Creek.
In 2002, the thermal monitoring program in the
mainstem Colorado River was expanded to include
multiparameter monitoring stations throughout Grand
Canyon to collect time-series measurements of water
temperature, specific conductance, dissolved oxygen,
and pH at five sites where suspended-sediment transport
is also monitored. In 2005, mainstem monitoring was
reduced to temperature and specific conductance mea-
surements (fig. 3).
Water Quality in Lake Powell and the Colorado River 75
Trends and Current
Conditions
Hydrology
Because of a prolonged drought between 2000 and
2005, Lake Powell water storage was reduced by approx-
imately 60%. Water year (WY) 2004, which ended on
September 30, 2004, was the fifth consecutive year of
below-normal inflows to Lake Powell; inflows were at
51% of average in WY 2004 (table 2). Inflow in WY
2002 was the lowest observed since the completion of
Glen Canyon Dam in 1963. This drought period resulted
in a 130 ft (40 m) drop in reservoir elevation and a 13
maf (16,029 million m3) decline in storage in Lake Powell
by the end of WY 2004 (fig. 4). While precipitation in
the upper Colorado River Basin increased substantially
during the first part of WY 2005, storage in Lake Powell
continued to decline until the reservoir reached an eleva-
tion of 3,555 ft (1,084 m) on April 8, 2005, after which
snowmelt runoff and reduced dam releases increased the
reservoir elevation. Average unregulated inflow to Lake
Powell is 12.056 maf (14,865 million m3), as determined
from the 30-yr record that spans WY 1971 through WY
2000 (Tom Ryan, Bureau of Reclamation, oral com-
mun., 2005); however, the average inflow for the water
years from 2000 to 2004 was 5.962 maf (7,351 million
m3) (table 2).
Figure 3. Grand Canyon water-quality sampling sites.
76 The State of the Colorado River Ecosystem in Grand Canyon
Salinity
Salinity levels are of concern throughout the entire
Colorado River Basin because high salinity can be
damaging to soils and crops. Furthermore, treaty obliga-
tions with Mexico limit the salinity of water that can be
delivered to that country. As the Colorado River flows
to the Gulf of California, it leaches salts from soils and
other geologic substrates through and over which it flows.
Salinity levels are also increased by irrigation returns,
by evaporation in storage facilities, and by rate of flow
(slow-flowing water picks up higher levels of dissolved
solids than do high flows during runoff).
Table 2. Recent inflows and releases at Glen Canyon Dam (maf = million acre-feet).
Water April–July Percent WY Percent Glen End of End of
year unregulated of unregulated of Canyon year year
(WY) inflow average inflow average Dam release storage elevation
(maf) (maf) (maf) (maf) (ft)
1998 8.625 112 13.661 116 13.511 22.403 3687.7
1999 7.621 99 12.71 108 11.202 22.997 3691.6
2000 4.352 56 7.310 62 9.380 20.939 3677.8
2001 4.301 56 6.955 59 8.238 19.135 3664.8
2002 1.115 14 3.058 25 8.230 14.468 3626.5
2003 3.918 51 6.358 53 8.228 12.110 3603.8
2004 3.640 46 6.128 51 8.231 9.169 3570.8
Figure 4. Lake Powell surface elevation, 1963–2005.
Periodically the salinity of water released from the
dam increases as a result of drought. This increase is
due to a combination of factors, including increases in
the salinity of base flows into the reservoir, lack of large
volumes of dilute snowmelt runoff, and reduced reservoir
volume to dilute the effects of reservoir inflows. At the
end of WY 2004, releases from Glen Canyon Dam had a
specific conductance of approximately 850 microsiemens
per centimeter (S/cm) at 25°C, corresponding to a total
dissolved solids concentration of 575 mg/L (fig. 5).
Water Temperature
Impounding water in Lake Powell significantly
affected the water temperature of dam releases to the
Colorado River ecosystem because of reservoir stratifica-
tion and the location of the penstock release structures
(fig. 1). During the summer months, the epilimnion of
Lake Powell warms considerably from inflows, ambient
air temperature, and solar radiation, reaching tempera-
tures as high as 86°F (30°C); however, the hypolimnion is
isolated from these processes, maintaining temperatures
between 43°F and 48°F (6°C and 9°C).
Before closure of the dam, mean water tempera-
ture for what is now the tailwater was approximately
57°F (14°C), ranging from 32°F to 80°F (0°C to 27°C)
over the course of a year (U.S. Geological Survey, 2004).
Before 1973, during the reservoir’s initial filling stage,
release temperatures were affected by surface or epi-
limnetic withdrawals because of the proximity of the
reservoir’s surface to the penstock withdrawal zone. Max-
Water Quality in Lake Powell and the Colorado River 77
imum release temperature during that period occurred
during the months of August and September, reflecting
the surface warming of the reservoir.
Trends in water temperature of the tailwater sta-
bilized from 1973 to 2003, when the reservoir surface
elevations were above 3,600 ft (1,097 m) and the epilim-
nion was situated above the penstock withdrawal zone.
During this period, release temperatures as measured at
Lees Ferry averaged 48.7ºF (9.3°C). Temperatures fluctu-
ated between 44°F and 54°F (7°C and 12°C), with minor
excursions beyond this range during periods of spillway
releases (fig. 6). Under these conditions, there was some
seasonality to Glen Canyon Dam release temperatures,
with slight warming beginning in May and June and
increasing through the year. The highest temperatures
occurred at the end of December as a result of the influ-
ence of the relatively warm, convectively mixed epi-
limnion on penstock releases. Peak temperatures under
these conditions appeared to be affected by the volume
of the previous year’s snowmelt runoff, which affects the
thickness of the warm epilimnion near the dam dur-
ing the latter months of the year. Although seasonality
in temperature patterns exists in the postdam era, the
annual variation has been reduced to approximately 9ºF
(5°C) from approximately 48ºF (27°C) in the predam
era. Also, the highest river temperatures immediately
below the dam now occur in late fall or winter instead of
in summer, which is when they occurred in the predam,
unregulated river.
The water level of the reservoir dropped more than
140 ft (42 m) between 1999 and 2005 as a result of a
basinwide drought that began in 2000 (fig. 4). This drop
placed the warmer epilimnetic water much closer to
the penstock withdrawal zone and resulted in reservoir
releases being drawn from this epilimnetic layer. Substan-
tially warmer release temperatures have occurred in the
fall and early winter months since 2003. An annual max-
imum mean daily release temperature of 55°F (12.9°C)
was observed on November 14, 2003; on November 6,
2004, the annual maximum mean daily temperature
reached 59°F (15°C) (fig. 6). These values represent the
highest release temperatures from Glen Canyon Dam
since August 1971, when the reservoir was filling. As of
July 11, 2005, the mean daily release temperatures had
reached 56.4°F (13.6°C), showing earlier warming and
higher temperatures than had occurred in the past 2 yr.
Seasonal and longitudinal water temperature pat-
terns in Grand Canyon have been measured from 1994
to 2005 from Glen Canyon Dam to Diamond Creek,
241 mi (388 km) below Glen Canyon Dam (fig. 7). Dur-
ing summer months, gradual downstream warming
occurs because of the transfer of heat from the warmer
surrounding air mass, heat stored in the canyon walls
adjacent to the river, and solar radiation.
A comparison of weekly average increase in water
temperature between Glen Canyon Dam and Diamond
Creek to average weekly discharge during mid-June from
1994 to 2004 demonstrates the effect of Glen Canyon
Dam releases on warming patterns in the Colorado
River in Grand Canyon (fig. 8). High steady flows of
approximately 26,000 cfs in 1997 resulted in 9°F (5°C)
warming at Diamond Creek, while low steady flows of
Figure 5. Mean daily specific conductance (in microsiemens per
centimeter (μS/cm) at 25°C), an indicator of salinity, below Glen
Canyon Dam, 1995–2005.
Figure 6. Daily water temperature (red line) at Lees Ferry as
affected by changes in Lake Powell’s elevation (green line).
78 The State of the Colorado River Ecosystem in Grand Canyon
8,000 cfs in 2000 exhibited 18°F (10°C) warming. This
difference is because large volumes of water have greater
mass and a lower surface area to volume ratio as well as
less exposure time for atmospheric heat exchange that is
due to higher velocity, reducing the amount of warming
from ambient temperatures and solar radiation. The
warming occurring at low discharges affects water tem-
peratures in lower Grand Canyon to a greater degree
than the elevated release temperatures observed in the
past 2 yr.
Lateral variation in river temperature also occurs
throughout Grand Canyon. Substantial warming occurs
in various nearshore environments, ranging from shallow,
open-water areas to enclosed backwaters. Water in cer-
tain nearshore environments becomes isolated from mix-
ing with the main channel current and warms with solar
radiation and equilibration with ambient temperatures.
These environments may be important to the survival,
growth, and eventual recruitment of the larval life stages
of native fish (see chapter 2, this report).
Figure 7. Water temperatures along the Colorado River from Glen Canyon Dam to Diamond Creek, 1994–2005. Black dots represent
monitoring locations.
Figure 8. Mid-June warming above release temperatures
at Diamond Creek, 1994–2004, as a function of mean weekly
discharge (in cubic feet per second). Warming at Diamond Creek
= 0.000532 * Q + 21.01.
Water Quality in Lake Powell and the Colorado River 79
Turbidity and
Suspended Sediment
Construction of Glen Canyon Dam dramatically
altered the sediment-transport processes of the Colorado
River. Before the completion of Glen Canyon Dam, the
total sand supply to Grand Canyon, from the Colorado
River upstream from Lees Ferry, with the Paria and Little
Colorado Rivers combined, was approximately 29 mil-
lion tons (26 million Mg). Today, because Lake Powell
traps all of the sediment upstream from Glen Canyon
Dam, the Paria River is the primary source of sand to
Marble Canyon, supplying approximately 6% of predam
sand levels (see chapter 1, this report). Only a small por-
tion of the suspended sediment entering Lake Powell is
transported for any distance because most of it is depos-
ited near the inflows of major tributaries.
Turbidity and suspended-sediment concentrations
are of interest in the downstream environment because
water clarity affects the amount of light available for
photosynthesis for downstream algal communities, which
are an important part of the overall food base for native
and nonnative fishes. Turbidity also affects the behavior
and distribution of various native and nonnative fishes
in providing cover from various predators or by affect-
ing sight-feeding abilities. Turbidity is measured in both
Lake Powell and downstream. Turbidity measurements
in Lake Powell indicate the location of advective tribu-
tary inflows and also can be used as an indicator of
primary productivity in the reservoir because increased
turbidity indicates the presence of phytoplankton. The
rather abrupt decrease in filamentous green alga below
Lees Ferry most probably results from inputs of sediment
from major tributaries, including the Paria and Little
Colorado Rivers, which reduce light penetration (Cole
and Kubly, 1976; Stevens and others, 1997).
Nutrients
Nutrients such as phosphorus, nitrogen, and silica
are essential for microbial production and algal growth.
Most phosphorus entering Lake Powell is associated
with suspended clays in the inflows of river water.
The reservoir acts as a nutrient sink, especially for
phosphorus. More than 95% of phosphorus reaching
Lake Powell is in particulate form or is associated with
suspended sediment particles. A large fraction of this
phosphorus load is deposited within the reservoir by
sedimentation (Gloss, 1977). Most of the remaining dis-
solved phosphorus is removed from the water by uptake
from biological activity.
Bioproduction in Lake Powell apparently is directly
related to the intensity and duration of enriched spring
inflow events that are responsible for delivering the bulk
of nutrient capital to the reservoir (Gloss and others,
1980). Surface concentrations of dissolved phosphorus
generally decline from the upper end of the reservoir to
the dam because of the uptake from primary production,
to the point that dissolved phosphorus is usually below
detection limits within 30–60 mi (48–97 km) upstream
from the dam in the upper water column. Phosphorus
is the limiting factor for primary production near the
dam, while low light availability is the limiting factor to
productivity in the upper portion of Lake Powell because
of turbidity from inflow currents, especially during early
summer months (Gloss and others, 1980) (fig. 9a).
Nitrate-nitrogen concentrations from the surface of
the reservoir forebay fluctuate in a manner that reflects
the utilization of the nitrogen by algae and begin to
increase in fall as primary production slows. Nitrate-
nitrogen concentrations show a peak in winter when
temperatures are coolest and productivity is relatively
low. During the summer months, when primary pro-
ductivity is at a maximum, nitrate-nitrogen concentra-
tions reach a minimum because of uptake by primary
producers. From 1999 to 2004, surface nitrate-nitrogen
concentrations in the forebay above Glen Canyon Dam
averaged 0.09 parts per million (ppm). Nitrate-nitrogen
concentrations in the deepest part of the hypolimnion
averaged 0.39 ppm, about four times higher than surface
concentrations. For the same period, nitrate-nitrogen
concentrations in Glen Canyon Dam releases averaged
0.29 ppm (fig. 9b).
The highest productivity in Lake Powell is seen in
surface waters of the reservoir and results from a com-
bination of temperature, light availability, and nutrient
concentrations. Because primary productivity processes
consume nutrients, nutrient concentrations are eventu-
ally depleted in the surface waters of Lake Powell and
remain at elevated concentrations in the hypolimnion,
where there is little primary production taking place.
Consequently, hypolimnetic releases from Glen Canyon
Dam are relatively nutrient rich. Periods of epilimnetic
releases from Glen Canyon Dam may cause a reduc-
tion in the amount of nutrients available to the down-
stream ecosystem.
Dissolved Oxygen
Dissolved oxygen concentrations in Lake Powell are
affected by inflow, seasonal water-circulation patterns,
and biological processes (Johnson and Merritt, 1979).
80 The State of the Colorado River Ecosystem in Grand Canyon
The spring snowmelt runoff enters Lake Powell as an
overflow density current, representing the lowest density
water entering the reservoir during the year. This inflow
current then travels through the reservoir slightly below
the surface, eventually reaching Glen Canyon Dam by
late summer or early fall. A large amount of suspended
sediment, nutrients, and organic material may be associ-
ated with this inflow current. As the inflow current trav-
els through the reservoir, the organic material undergoes
bacterial decomposition, removing large amounts of
oxygen from this water. This situation, combined with
decomposition of plankton from the epilimnion, results
in a marked reduction of dissolved oxygen in the met-
alimnion of the reservoir by late summer.
During periods of reservoir drawdown, as in the
past several years, tributary inflows cause the resus-
pension of exposed deltaic sediments in the upstream
portions of the reservoir. This resuspension entrains
large amounts of suspended sediment and decompos-
ing organic material in the advective inflow currents
that move through the reservoir. In September 2003, the
inflow plume extended downstream to Padre Bay, about
28 mi (45 km) above Glen Canyon Dam, at depths of
50–80 ft (15–25 m) (fig. 10). Dissolved oxygen concentra-
tions at or near zero were observed throughout its extent.
This plume reached Glen Canyon Dam by October and
was eventually dissipated by convective mixing in the
following months. The pattern was observed to a lesser
extent in 2004 because runoff volumes were smaller.
The hypolimnion of Lake Powell is isolated from
photosynthetic oxygenation and reaeration caused
by surface mixing and acts as a sink for organic mat-
ter falling through the water column. As a result, the
water at the bottom of Lake Powell is often oxygen poor
throughout the year. Depending on the density of cold
winter inflows relative to that of the receiving reservoir,
winter inflows will either form an interflow between
layers of higher and lower density or may underflow the
entire reservoir, displacing hypolimnetic water upwards
for withdrawal by dam releases. In the former case, the
density of the hypolimnetic water is too great to allow
an underflow current, and as a result, the hypolimnion
remains stagnant and oxygen concentrations continue to
decline. In the latter case, the hypolimnion is refreshed
with oxygenated water.
There have been three distinct periods of hypolim-
netic stagnation during Lake Powell’s history, usually
lasting several years and characterized by a buildup of
relatively saline water, followed by decreasing dissolved
oxygen concentrations. These periods of hypolimnetic
stagnation appear to have been dissipated by a series of
above-average inflows that flushed the reservoir. Above-
normal inflows to Lake Powell during the mid-1990s left
the reservoir fairly dilute, and distinct winter underflows
have occurred since 1999 (fig. 11).
Dissolved oxygen concentrations in the tailwater are
usually slightly below saturation but have not dropped to
concentrations low enough to affect the aquatic ecosys-
tem in Grand Canyon. As the reservoir ages or there are
periods of extended drought, however, it is likely that the
chances of water low in dissolved oxygen being released
from Glen Canyon Dam will increase.
Plankton
The epilimnion of Lake Powell is fairly warm and
receives abundant sunlight for photosynthesis through
much of the year. As a result, the majority of phyto-
Figure 9. A. Total phosphorus concentrations measured (in parts
per million) in the forebay of Lake Powell and Glen Canyon Dam
releases, 1999–2004. B. Dissolved nitrate-nitrogen concentrations
measured (in parts per million) in the forebay of Lake Powell and
Glen Canyon Dam releases, 1999–2004.
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.10
1/1/1999 1/1/2000 1/1/2001 1/1/2002 1/1/2003 1/1/2004 1/1/2005
P (ppm)
Forebay surface
Forebay bottom
Below dam
0
0.1
0.2
0.3
0.4
0.5
0.6
1/1/1999 1/1/2000 1/1/2001 1/1/2002 1/1/2003 1/1/2004 1/1/2005
N (ppm)
Forebay surface
Forebay bottom
Below dam
A.
B.
Water Quality in Lake Powell and the Colorado River 81
plankton and zooplankton reside in this layer, support-
ing a recreational fishery in Lake Powell. Under peri-
ods of epilimnetic withdrawals—for example during
winter mixing or during the recent late-summer events
of 2003, 2004, and 2005, when warmer epilimnetic
waters are released—plankton from the epilimnion
can be released downstream, potentially providing an
alternate food source for the downstream ecosystem (fig.
12). Little is known about the downstream importance
of plankton and other organic matter released from
Glen Canyon Dam.
Effects of Dam Operations
Operation of Glen Canyon Dam affects the water
quality of Lake Powell and downstream releases. High
sustained penstock releases during 1973, through the
mid-1980s, and under modified low fluctuating flow
(MLFF) operations in 1997 acted to route increased
volumes of water through the reservoir. The operation
of different release structures can affect the downstream
environment by withdrawing water of different quality
than that at the penstock elevation. Since the implemen-
tation of the 1996 Record of Decision and the MLFF
alternative, the river outlet works have been used dur-
ing the 1996 beach/habitat-building flow (Hueftle and
Stevens, 2001) and the 2004 experimental high flow.
During these events, water of cooler temperature and
higher salinity and nutrient content was released, and
the turbulence created downstream of the dam increased
oxygen concentrations.
During periods of normal operations under the
MLFF, daily fluctuations in temperature, specific con-
ductance, and dissolved oxygen can be observed in
Glen Canyon Dam releases because of the effects of
various discharge volumes and fluctuation patterns
on the dimension of the withdrawal zone in the reser-
voir. Large or fluctuating releases draw water from a
thicker withdrawal zone than do small or steady releases
(Monismith and others, 1988; Ford, 1990; Casamitjana
and others, 2003).
Figure 10. Dissolved oxygen concentrations (in parts per million) in the main channel of Lake Powell, from Glen Canyon Dam to the
Colorado River inflow, September 2003.
82 The State of the Colorado River Ecosystem in Grand Canyon
Figure 11. Forty-year temperature (degrees Fahrenheit), specific conductance (microsiemens per centimeter (μS/cm) at 25°C), and dissolved oxygen (parts per million) patterns
in the forebay of Lake Powell, December 1964 to March 2005. Elevations of spillways, penstocks, and river outlet works are indicated by horizontal lines in each plot. Warmer
colors represent higher values for each parameter. Shaded areas represent areas of low temporal resolution or poor data quality. Note three periods of hypolimnetic stagnation
(red areas in conductance section), followed by periods of hypolimnetic hypoxia (blue areas in dissolved oxygen section). Dissolved oxygen section shows recent pattern of
winter underflows since 1999.
Water Quality in Lake Powell and the Colorado River 83
In terms of downstream water quality, the mag-
nitude of discharge and the volume of water in the
river determine how much water is exposed to ambient
air temperatures and solar radiation, which, in turn,
determines the amount of in-stream warming in Grand
Canyon. The magnitude of dam discharges also influ-
ences the amount of sediment in suspension, which
can limit light availability to the downstream aquatic
environment.
There are many interactions at work, and the use
of a hydrodynamic water-quality model will help under-
stand the changes in Lake Powell and their effects on
downstream water quality.
Conclusion and
Recommendations
The current drought demonstrates how long-term
climatic trends influence inflows to Lake Powell and
affect the quality of dam releases. As the elevation of
Lake Powell drops, the warmer surface water is drawn
downstream through the penstocks, seasonally increasing
downstream temperatures. Furthermore, salinity con-
centrations increase, more epilimnetic biota are exported
downstream, and dissolved oxygen levels may decrease
seasonally. Current dissolved-oxygen minimums are
within the tolerance limits of tailwater organisms, includ-
ing rainbow trout (Oncorhynchus mykiss). Entrainment of
an oxygen-depleted inflow plume or hypolimnion, along
with the aging of the reservoir, however, could affect the
downstream ecosystem.
If the drought continues, these trends will be exac-
erbated as summer and fall releases draw increasingly
from the warmer surface waters of the reservoir. Release
patterns during an extended drought may replicate those
of the 1960s when the reservoir began filling, including
wider fluctuations in water quality that may begin to
mimic predam conditions. These conditions create an
opportunity to observe some of the possible effects in
the Colorado River ecosystem of a temperature con-
trol device, which would route warmer surface waters
through the Glen Canyon Dam powerplant.
Future Monitoring and Modeling
Current understanding of water quality downstream
of Glen Canyon Dam in Grand Canyon is limited
because more extensive sampling has only recently been
initiated and because there has been a lack of modeling
Figure 12. Plankton concentrations and biomass in Lake
Powell forebay.
for river-water quality. Exceptions are water tempera-
ture and sediment concentrations (see chapter 1, this
report), which have been more extensively monitored
and analyzed. Future water-quality programs should
emphasize a model-based approach and close linkages
with other sampling programs like aquatic ecology and
fine-sediment monitoring.
The extensive 40-yr database available for Lake
Powell presents a very clear opportunity to model the
hydrodynamic properties of the reservoir under a variety
84 The State of the Colorado River Ecosystem in Grand Canyon
of inflow and operational conditions. Such modeling not
only would increase the understanding and characteriza-
tion of reservoir limnology but also would serve as an
important predictive tool to anticipate what the qual-
ity of water released to the Colorado River ecosystem
in Grand Canyon will be. A number of river-flow and
water-quality models are available and could be used or
modified to better understand and predict downstream
changes in water quality and their effects on various
components of the ecosystem.
References
Bureau of Reclamation, 1981, Water and power
resources: project data: Denver, Colo., Bureau of
Reclamation.
Carpenter, S.R., and Kitchell, J.F., eds. 1996, The
trophic cascade in lakes: Cambridge, England,
Cambridge University Press, 385 p.
Casamitjana, X., Serra, T., Colomer, J., Baserba, C., and
Pérez-Losada, J., 2003, Effects of the water withdrawal
in the stratification patterns of a reservoir: Hydrobio-
logia, v. 504, p. 21–28.
Clarkson R.W., and Childs, M.R., 2000, Temperature
effects of hypolimnial-release dams on early life stages
of Colorado River basin big-river fishes: Copeia, v. 2,
p. 402–412.
Cole, G., and Kubly, D.M., 1976, Limnologic studies
in the Colorado River from Lees Ferry to Diamond
Creek: Colorado River Research Program Technical
Report No. 8, Colorado River Research Series Contri-
bution No. 37, Grand Canyon National Park.
Department of the Army, Corps of Engineers, 1987,
EM 1110-2-1201, Engineering and design, reservoir
water quality analysis: Washington, D.C., p. 2–11.
Ferrari, R.L., 1988, 1986 Lake Powell survey: Denver,
Colo., Bureau of Reclamation.
Ford, D.E., 1990, Reservoir transport processes, in
Thornton, K.W., Kimmel, B.L., and Payne, F.E., eds.,
Reservoir limnology: ecological perspectives: New
York, Wiley, p. 15–41.
Gloss, S.P., 1977, Application of the nutrient load-
ing concept to Lake Powell, the effects of nutrient
perturbations on phytoplankton productivity, and
levels of nitrogen and phosphorus in the reservoir:
Albuquerque, University of New Mexico, Ph.D. thesis,
225 p.
Gloss, S.P., Mayer, L.M., and Kidd, D.E., 1980, Advec-
tive control of nutrient dynamics in the epilimnion of
a large reservoir: Limnology and Oceanography, v. 25,
p. 219–228.
Hueftle, S.J., and Stevens, L.E., 2001, Experimental
flood effects on the limnology of Lake Powell reservoir,
southwestern USA: Ecological Applications, v. 11, p.
644–656.
Johnson, N.M., and Merritt, D.H., 1979, Convec-
tive and advective circulation of Lake Powell, Utah
and Arizona, during 1972–1975: Water Resources
Research, v. 15, p. 873–884.
Kimmel, B.L., and Groeger, A.W., 1984, Factors control-
ling phytoplankton production in lakes and reservoirs:
a perspective, in Lake and reservoir management:
Washington, D.C., U.S. Environmental Protection
Agency, EPA 440/5/84-001, p. 277–281.
Marzolf, G.R., Bowser, C.J., Hart, R.J., Stephens, D.W.,
and Vernieu, W.S., 1999, Photosynthetic and respira-
tory processes: an open-stream approach, in Webb,
R.H., Schmidt, J.C., Marzolf, G.R., and Valdez, R.A.,
eds., The 1996 controlled flood in Grand Canyon:
Washington, D.C., American Geophysical Union,
Geophysical Monograph Series, v. 110, p. 205–216.
Monismith, S., Imberger, J., and Billi, G., 1988, Shear
waves and unsteady selective withdrawal: Journal of
Hydraulic Engineering, v. 114, no. 9, p. 1134–1152.
Nilsson, C., Reidy, C.A., Dynesius, M., Revenga, C.,
2005, Fragmentation and flow regulation of the
world’s large river systems: Science, v. 308, p. 405–408.
Paulson, L.J., and Baker, J.R., 1981, Nutrient interactions
among reservoirs on the Colorado River, in Stephan,
H.G., ed., Proceedings of the Symposium on Surface
Water Impoundments: New York, American Society
of Civil Engineers, p. 1648–1656.
Shannon, J.P., Blinn, D.W., and Stevens, L.E., 1994,
Trophic interactions and benthic animal community
structure in the Colorado River, Arizona, U.S.A.:
Freshwater Biology, v. 31, p. 213–220.
Water Quality in Lake Powell and the Colorado River 85
Stanford, J.A., and Ward, J.V., 1991, Limnology of
Lake Powell and the chemistry of the Colorado
River, in Colorado River ecology and dam manage-
ment: Washington, D.C., National Academy Press, p.
75–101.
Stevens, L.E., Shannon, J.P., Blinn, D.W., 1997,
Colorado River benthic ecology in Grand Canyon,
Arizona, USA: dam, tributary, and geomorphological
influences: Regulated Rivers: Research & Manage-
ment, v. 13, p. 129–149.
Topping, D.J., Rubin, D.M., and Vierra, L.E., Jr., 2000a,
Colorado River sediment transport: pt. 1: natural
sediment supply limitations and the influence of the
Glen Canyon Dam: Water Resources Research, v. 36,
p. 515–542.
Topping, D.J., Rubin, D.M., and Vierra, L.E., Jr., 2000b,
Colorado River sediment transport: pt. 2: systematic
bed-elevation and grain-size effect of sand supply limi-
tation: Water Resources Research, v. 36, p. 543–570.
U.S. Geological Survey, 1985–2004, Water resource data
for Arizona, water year 1985–1996: U.S. Geological
Survey Water Data Reports AZ-85-1 through 04-1.
U.S. Geological Survey, 1998, Depth profiles of tem-
perature, specific conductance, and oxygen concentra-
tions in Lake Powell, Arizona-Utah, 1992–1995: U.S.
Geological Survey Open-File Report 97-835.
U.S. Geological Survey, 2001, Daily and seasonal vari-
ability of pH, dissolved oxygen, temperature, and
specific conductance in the Colorado River between
the forebay of Glen Canyon Dam and Lees Ferry,
northeastern Arizona, 1998–1999: U.S. Geological
Survey Open-File Report 01-222.
U.S. Geological Survey, 2004, National water informa-
tion system, water resource data for Arizona online
data retrieval, water years 1949–2004: U.S. Geological
Survey Water Data Report, http://waterdata.usgs.
gov/nwis/uv?09380000, accessed December 2, 2004.
Wetzel, R.G., 2001, Limnology, lake and river ecosys-
tems (3d ed.): San Diego, Calif., Academic Press.
Yard, M.D., Bennett, G.E., Mietz, S.N., Coggins, L.G.,
Stevens, L.E., Hueftle, S., and Blinn, D.W., 2005,
Influence of topographic complexity on solar insola-
tion estimates for the Colorado River, Grand Canyon,
AZ: Ecological Modelling, v. 183, p. 157–172.
Contact Information:
William S. Vernieu
Hydrologist
U.S. Department of the Interior
U.S. Geological Survey
Southwest Biological Science Center
Flagstaff, AZ
bvernieu@usgs.gov
Susan J. Hueftle
Limnologist
U.S. Department of the Interior
U.S. Geological Survey
Southwest Biological Science Center
Flagstaff, AZ
shueftle@usgs.gov
Steven P. Gloss
Ecologist
U.S. Department of the Interior
U.S. Geological Survey
Southwest Biological Science Center
Tucson, AZ
sgloss@usgs.gov
First page photograph credit: Andrew Pernick, Bureau of Reclamation
Chapter 5
Aquatic
Ecology:
the Role
of Organic
Matter and
Invertebrates
Theodore A. Kennedy
Steven P. Gloss
Introduction
Closure of Glen Canyon Dam and the beginning of
flow regulation of the Colorado River through Grand
Canyon in 1963 changed the river through the canyon
considerably. The river changed from having highly
variable discharge rates and temperatures and high
suspended-sediment loads to having a relatively constant
flow regime (Topping and others, 2003), cold and con-
stant water temperatures, and suspended-sediment loads
that are dramatically reduced relative to predam levels
(see chapter 1, this report). These changes in the physical
environment, coupled with changes in the quantity and
types of organic matter present in the Colorado River
and intentional introductions of aquatic invertebrates
that occurred shortly after Glen Canyon Dam was closed
(Blinn and Cole, 1991), have led to substantial changes in
the kinds of aquatic invertebrates present in the Grand
Canyon ecosystem. Since the closure of the dam, con-
siderable effort has been directed toward understanding
the aquatic ecology of this altered ecosystem (Blinn and
Cole, 1991).
This chapter describes the results of the research
and monitoring activities that have investigated the
kinds of organic matter and invertebrate communities
in the Colorado River below Glen Canyon Dam. Col-
lectively, organic matter and the aquatic invertebrates
that consume it largely constitute the food base for fish in
the Colorado River ecosystem. This chapter focuses on
patterns, trends, and important controls on the amount
and sources of organic matter and invertebrates that are
primary food resources for humpback chub (Gila cypha)
and rainbow trout (Oncorhynchus mykiss) in an effort to
understand the role that food plays in determining the
distribution, population density, and growth of these fish
in this ecosystem. Furthermore, most of the research and
monitoring that have been conducted on organic mat-
ter and invertebrates in this ecosystem have centered on
the food items that are important for these two species.
This chapter also addresses how organic matter and
invertebrates are affected by the timing and magnitude
of water releases from Glen Canyon Dam, including the
modified low fluctuating flow (MLFF) alternative, which
was implemented in 1996 and continues as the operating
regime for Glen Canyon Dam today. Finally, this chap-
ter concludes with a brief discussion of recommended
research directions and management actions.
88 The State of the Colorado River Ecosystem in Grand Canyon
Background
Virtually all food webs, including those in rivers,
are fueled by energy that comes from autotrophs (also
known as primary producers), which are organisms that
can convert sunlight into chemical energy. Examples of
autotrophs include vascular plants and algae. Without
autotrophs there would be no food energy available to
other organisms that lack the capability to fix light energy.
In rivers, this autotrophic material can come from two
places: the terrestrial environment, such as leaves fall-
ing into a river from trees lining the river’s banks, or the
aquatic environment, such as algae growing on river
rocks. Terrestrially derived material is an extremely abun-
dant source of energy in many streams and rivers (Bayley,
1989; Meyer and Edwards, 1990). Although algae often
represent just a small fraction of the available energy in
river ecosystems, they are frequently an important energy
source that contributes to secondary production (Bilby
and Bisson, 1992; Lewis and others, 2001; Thorp and
Delong, 2002) because they are far more nutritious than
terrestrial material (Anderson and Sedell, 1979). Aquatic
or terrestrial autotrophic material is also called “organic
matter” and provides the energy that supports consumers
at higher trophic levels. Trophic levels are groups of organ-
isms that occupy the same position in a food web (fig. 1).
The importance of understanding patterns, trends,
and controls of organic matter and invertebrates is
reflected in the goals of the Glen Canyon Dam Adaptive
Management Program. For example, the program’s first
goal is to “protect or improve the aquatic foodbase so
that it will support viable populations of desired species
at higher trophic levels” (Glen Canyon Dam Adaptive
Management Program, 2001, p. 11). Two additional
goals are to maintain populations of rainbow trout in the
Lees Ferry reach and maintain, and ultimately increase,
populations of native fish, particularly endangered
humpback chub, in sections of the river downstream of
the Lees Ferry reach. Recent trends for important fish
in the Grand Canyon ecosystem may be partly due to
changes in food resources or to an increase in the severity
of competition between humpback chub and other fish
and highlight the need for continued research on organic
matter and invertebrates. The number of rainbow trout
in the Lees Ferry reach has generally been high since
intensive population measurements began in 1991, but
the condition of fish as determined by weight relative to
length declined during the late 1990s (McKinney and
others, 2001). Also, the average condition (Meretsky and
others, 2000) and size of humpback chub populations
have declined considerably since intensive measurements
began in 1986 (see chapter 2, this report).
Figure 1. Idealized and simplified food-web diagrams for two different sections of the Colorado River ecosystem, the Lees Ferry reach
and the Colorado River at the Little Colorado River (LCR) confluence. The sizes of the boxes reflect qualitative differences in the standing
mass within trophic levels, and the size of the arrows reflects qualitative differences in the amount of food energy moving between
trophic levels. Green arrows linking algae to fish are used to highlight the fact that fish consume algae but do not actually derive
significant nutrients or energy from them. Rainbow trout also consume New Zealand mudsnails, but the snails often survive passage
through the gut alive and intact, which is noted with the
X
. Humpback chub may be capable of actually digesting mudsnails because
they are capable of crushing their shells; however, it is unclear whether humpback chub ingest snails. Areas of greatest uncertainty are
noted with dashed lines.
Aquatic Ecology 89
Several of the hypothesized causes of humpback
chub decline are, in part, based on the assumption that
there is a limited amount of food available in the river
to support populations of humpback chub and other
fish. For example, Gloss and Coggins (see chapter 2,
this report) list competition with nonnative and native
fish as one of several possible reasons for the decline
of the endangered humpback chub populations. Thus,
documenting the food resources that humpback chub
and rainbow trout are dependent on and whether these
resources vary over space and time will help determine
the validity of the food-limitation hypothesis. This
information will also clarify the role that food availability
plays in determining the population density and condi-
tion of both native and nonnative fishes and may prove
useful as an indicator of ecosystem health.
Status and Trends
Identifying the Food Items of Fish
A great deal is known about the types of food items
consumed by humpback chub and rainbow trout. In gen-
eral, both fish appear to consume mostly the introduced
invertebrate Gammarus lacustris (small crustaceans, also
called scuds or side-swimmers, hereafter Gammarus) (fig.
2), larval chironomids (midges, also called bloodworms),
larval simuliids (black flies), terrestrial invertebrates,
and the filamentous algae Cladophora glomerata (hereaf-
ter Cladophora) (fig. 3). The small aquatic invertebrates
mentioned range from about 0.25 to 1 inch (6 to 25 mm)
in length, and individual Cladophora filaments can attain
nearly 20 ft (about 6 m) in length.
Gut pumping was used to nondestructively inves-
tigate the food items consumed by humpback chub
collected from two known chub aggregations, the Little
Colorado River confluence aggregation (at about RM
61) and the Middle Granite Gorge aggregation (at about
RM 127) (Valdez and Ryel, 1995). Cladophora represented
24% of gut contents by volume for chub at the Little
Colorado River confluence site, with invertebrates repre-
senting the remaining 76%. In contrast, humpback chub
at the Middle Granite Gorge site consumed exclusively
invertebrates. Of the invertebrates, simuliids, Gammarus,
and terrestrial invertebrates were the most common
items consumed. Valdez and Ryel (1995) also quantified
the density of algae and specific invertebrates in the drift
(fig. 4) and then compared chub diets with the avail-
ability of food items in the drift to determine whether
chub were “selectively” feeding. These analyses indicated
that chub selectively avoided consuming Cladophora, even
though it represented 88%–93% of potential food in the
drift, and chub generally consumed simuliids, chirono-
mids, and Gammarus in “approximate proportion to their
availability in the drift” (Valdez and Ryel, 1995, p. 9–13).
The feeding habitats of rainbow trout in the Lees
Ferry reach were determined by McKinney and Speas
(2001). They analyzed the stomach contents of 658
rainbow trout caught in the Lees Ferry tailwater from
1991–97 and found that Cladophora, Gammarus, and chi-
ronomids accounted for more than 90% of the stomach
contents by volume. Of the invertebrates consumed by
these fish, Gammarus and chironomids together accounted
for more than 90% of the total by volume.
For a complete understanding of the energy sources
that are driving a food web, it is important to know
not only what the fish are consuming but also what the
invertebrates themselves are consuming. To this end, the
diets of some aquatic invertebrates that are commonly
consumed by fish have also been investigated by using
gut-content analysis and habitat-choice experiments.
Pinney (1991) found that diatoms, a class (Bacillariophy-
ceae) of microscopic algae common in aquatic envi-
ronments, made up more than 93% by volume of gut
contents for Gammarus in the Lees Ferry reach. Shannon
and others (1994) used field and lab-based habitat-choice
Figure 2. A preserved Gammarus lacustris (also known as a
scud or side-swimmer), which is consumed by humpback chub
and rainbow trout. Live animals are more translucent (photograph
by Michael Booth).
90 The State of the Colorado River Ecosystem in Grand Canyon
experiments to determine that Gammarus preferred
Cladophora as habitat over all other choices, including
Oscillatoria spp. (another species of filamentous algae),
gravel, and detritus. These same researchers determined
that Gammarus were only relying on Cladophora as habi-
tat and were actually eating diatoms that were attached
to Cladophora. Stevens and others (1997) determined the
diet composition of chironomids by using gut-content
analysis and found that the relative importance of algae
in chironomid diets declined with distance downstream
from the dam: algae represented 61.4% of chironomid
diets at Lees Ferry, 30.7% at RM 32, and only 7.5% at
RM 224. This trend is consistent with observed down-
stream declines in algae biomass (discussed below) and
indicates that algae may not be the most common form
of organic matter consumed by invertebrates at down-
stream locations.
Identifying the food items consumed by fish and
invertebrates by using gut-content analysis provides an
indication of the food resources that are most impor-
tant to fish; however, relying solely on this approach
also has weaknesses. First, gut contents only reflect the
items consumed by fish or invertebrates within about an
hour of their capture, providing only a “snapshot” of
the food items consumed. Even in this short timeframe,
however, labile food items may be more readily digested
than others, leaving behind the more resistant items and
the appearance that these items are the most important
food sources. Furthermore, if, for example, rainbow
trout consume other fish very infrequently, the snapshot
taken through gut-content analysis is unlikely to detect
this relatively rare event. Yet these infrequent events of
predation may be a significant source of calories and
nutrients for the fish and may represent an important
type of food that might be overlooked when using only
gut-content analysis. Second, just because an item is
consumed by a fish does not mean that it is actually an
important source of energy or nutrients. For example,
both humpback chub and rainbow trout regularly
consume the filamentous algae Cladophora, but energetic
and stable-isotope analyses (discussed below) indicate
that this material is not actually assimilated because it
is difficult to digest and is low in essential nutrients such
as fatty acids, nitrogen, and phosphorus (Angradi, 1994;
McKinney and Speas, 2001). It has been suggested that
the humpback chub and rainbow trout that consume
Cladophora may actually be after the more nutritious
invertebrates that are imbedded in Cladophora (Valdez and
Ryel, 1995). Finally, the relative contribution of terres-
trial and aquatic organic matter to invertebrate growth
Permanently submerged zone
Varial zone
Figure 3. Cobble bar in Glen Canyon
showing the varial zone (shoreline
habitat that is both inundated and
exposed to air for long periods
each day) and the presence of the
filamentous algae Cladophora in
the permanently submerged zone.
Cladophora is unable to grow in the
varial zone because dam operations
result in discharge rates that
regularly expose the varial zone to air
(photograph by Theodore Kennedy,
U.S. Geological Survey).
Aquatic Ecology 91
and production, and hence fish growth and production,
is often unclear when only gut-content analysis is used.
To determine what is being consumed and actually
assimilated by fish and small invertebrates and to quan-
tify the relative importance of terrestrial and aquatic
organic matter in fueling fish growth require a combina-
tion of gut-content analysis and a specialized technique
known as stable-isotope analysis. Terrestrial and aquatic
organic matter often has distinct stable-isotope signa-
tures (Fry and Sherr, 1984) that are largely conserved up
the food chain, and these signatures provide informa-
tion about the source of energy at the base of the food
web (Peterson and Fry, 1987). For example, if Gammarus
consume exclusively algae, then they will have a carbon
stable isotope signature identical to that of algae, as will
a fish that consumes exclusively Gammarus. In contrast,
nitrogen stable isotope values change predictably with
each link in a food chain, increasing at a rate of 3.4 parts
per thousand (‰) with each link, and therefore provide
an indication of trophic position (Minagawa and Wada,
1984). Thus, herbivores typically have nitrogen stable
isotope signatures that are 3.4‰ higher than plants, and
primary carnivores in turn have nitrogen stable isotope
signatures that are 3.4‰ higher than herbivores and
6.8‰ higher than plants. Also, stable-isotope analyses
provide time-integrated measures of diet. That is, the
carbon and nitrogen stable isotope signatures for a large
fish will usually reflect the food it has consumed and
assimilated over the past several months, which provides
a contrast to the snapshot picture of diets obtained with
gut-content analysis.
A combination of gut-content analysis of rainbow
trout and stable-isotope analysis of the entire food web
provided a clear picture of the aquatic food web in the
Lees Ferry reach of the Colorado River (Angradi, 1994).
Rainbow trout gut-content data collected by Angradi
(1994) were remarkably similar to those reported by
McKinney and Speas (2001): both studies showed
Gammarus, chironomids, and Cladophora to be the domi-
nant food items. Analysis of stable-isotope ratios showed
that rainbow trout were assimilating nutrients from
only the Gammarus and chironomids and not from the
Cladophora they consumed (Angradi, 1994). Furthermore,
Gammarus and chironomids were feeding almost exclu-
sively on benthic algae, which are algae attached to the
bottom of the riverbed. Thus, the entire food web of the
Colorado River in the Lees Ferry reach, from aquatic
invertebrates to fish, was based on algae.
Our understanding of the aquatic food web at
sites downstream of Lees Ferry is much more limited.
Figure 4. Volume of specific invertebrates in humpback chub stomachs from the Little Colorado River confluence aggregation (about
RM 61) and composition of these same items in the drift. Data were collected during 1992–93 and exclude algae (Cladophora) that were
consumed by chub. Modified from Valdez and Ryel (1995).
92 The State of the Colorado River Ecosystem in Grand Canyon
Although gut-content analysis for humpback chub
and rainbow trout collected at downstream locations
indicated that both species consume mostly aquatic
invertebrates and algae, the relative importance of
terrestrial and aquatic energy sources remains unclear.
Using stable-isotope analysis, Haden and others (1999)
investigated the diet of humpback chub and other fish
in the Little Colorado River, the largest tributary of the
Colorado River in Grand Canyon and the single most
important breeding habitat for humpback chub. They
found a complex food web in the Little Colorado River
with small chub, less than 6 inches (<150 mm) in length,
relying heavily on invertebrates, especially chironomids,
mayflies (Ephemeroptera), and caddisflies (Trichoptera),
while large chub, greater than 6 inches (>150 mm) in
length, were found to be relying on invertebrates and
small fish. They also found evidence that terrestrial
and aquatic organic matter was fueling the food web,
although they were unable to determine the proportional
contribution of each energy source. Angradi (1994)
investigated food-web structure in tributary streams of
the Colorado River and found that the food web in some
tributaries was supported by leaf litter from streamside
vegetation while others were supported by leaf litter
from upland plants. Shannon and others (2001a) col-
lected samples of algae, aquatic invertebrates, and eight
species of fish (rainbow trout and humpback chub
were collected, but the identity of the other six species
was not specified) from seven sites that span the entire
Grand Canyon ecosystem and found that carbon isotope
values for algae increased consistently with downstream
distance. Further, they found that invertebrate and fish
isotope values roughly tracked the downstream shift in
algae isotope values, providing evidence that algae is
contributing to invertebrate and fish growth along the
entire length of the Grand Canyon ecosystem.
Spatial and Temporal Patterns in
Organic Matter and Invertebrates
There are very few data on the relative abundance
of terrestrial or aquatic organic matter, or the density
or kinds of invertebrates present, in the Grand Can-
yon ecosystem before the construction of Glen Can-
yon Dam (Blinn and Cole, 1991). In general, aquatic
invertebrate diversity (the number of different species)
has declined following closure of Glen Canyon Dam,
while invertebrate density and biomass have probably,
perhaps even dramatically, increased (Blinn and Cole,
1991, and references therein). Comparison of inverte-
brate diversity in tributaries relative to the mainstem
provides an indication of changes in the invertebrate
fauna that have occurred following closure of Glen
Canyon Dam; Hofknecht (1981) found 52 insect families
in tributaries of the Grand Canyon ecosystem, com-
pared to just 5 insect families for the mainstem Colorado
River. Haden and others (2003) studied relatively pris-
tine and free-flowing sections of the lower Green River
and the Colorado River in Canyonlands National Park
in Utah to determine what the food web in the predam
Colorado River in Grand Canyon might have looked
like. They found that terrestrial organic matter was the
primary energy source for aquatic invertebrates in this
free-flowing reach because high levels of suspended sedi-
ment prohibited algae growth. They also found an inver-
tebrate community that was markedly different from
that in the Grand Canyon ecosystem; the invertebrate
community in the free-flowing reach was dominated by
filter feeders (simuliids and caddisflies) and collectors
(mayflies and chironomids), reflecting the importance
of terrestrial organic matter to this system. Prior to the
closure of Glen Canyon Dam, the Colorado River con-
tained large quantities of coarse woody debris (i.e., whole
trees and branches) and other terrestrial plant material
that were transported from upstream sources (Valdez and
Carothers, 1998). This material accumulated along river
banks and in eddies and supported a high diversity and
abundance of terrestrial invertebrates. When this mate-
rial was entrained by the river during spring floods, the
terrestrial invertebrates probably served as an important
food resource for fish in the Colorado River (Valdez and
Carothers, 1998).
Because Glen Canyon Dam has created clear water
conditions that allow sunlight to reach the river bottom,
algal standing mass is extremely high in the Lees Ferry
reach of the Grand Canyon ecosystem. Much of the
terrestrial organic matter that formerly moved through
the Colorado River system is now trapped behind Glen
Canyon Dam. Stevens and others (1997) quantified
river-bottom algae and invertebrate standing mass at 11
stations between Lees Ferry and Diamond Creek on a
bimonthly basis during 1991. They found that Cladophora
was the dominant algae throughout the Lees Ferry reach,
exhibiting an average of 0.5 oz carbon (C)/yd² (15.5
g C/m2). Downstream of the Paria River confluence,
Cladophora standing mass abruptly decreased to 0.01 oz
C/yd² (0.5 g C/m2), and it remained low at the remain-
ing downstream sampling stations (fig. 5). Oscillatoria spp.,
mat-forming algae, tended to dominate aquatic habitats
at sites downstream of Lees Ferry with average biomass
of 0.02 oz C/yd² (0.6 g C/m2) at the site immediately
downstream of the Paria River.
Aquatic Ecology 93
Figure 5. Downstream patterns of sediment concentration and biomass of Cladophora and macroinvertebrates along the Colorado
River ecosystem. Modified from The Colorado River Through Grand Canyon by Steven W. Carothers and Bryan T. Brown. © 1991 The
Arizona Board of Regents. Reprint by permission of the University of Arizona Press.
The species composition and biomass of aquatic
invertebrates also vary with distance downstream.
Stevens and others (1997) reported that Gammarus and
chironomids were the dominant aquatic invertebrates
in the Lees Ferry reach, while simuliids were the domi-
nant invertebrates at sites downstream from the Paria
River. Sublette and others (1998) identified 38 species
of chironomids in the Grand Canyon ecosystem, and
Stevens and others (1998) studied the factors that influ-
ence chironomid distribution in the Grand Canyon
ecosystem. Stevens and others (1998) found that turbidity
strongly influenced chironomid diversity, with 11 species
present in the clear water of the Lees Ferry reach, 18
species present in what they termed the “variably turbid”
segment of the Colorado River (Lees Ferry to Little
Colorado River confluence), and 24 species in the “usu-
ally turbid” segment (Little Colorado River confluence
to Diamond Creek). In contrast, the biomass of inverte-
brates declined downstream (Stevens and others, 1997),
with mean biomass of 0.09 oz C/yd² (2.9 g C/m2) in the
Lees Ferry reach and much lower values, less than 0.003
oz C/yd² (<0.1 g C/m2), at downstream locations.
Even though the Lees Ferry reach accounted for
only 6.9% of the aquatic habitat in the 242 mi (390 km)
of river studied by Stevens and others (1997), it sup-
ported 63.5% of the primary producer biomass and
94 The State of the Colorado River Ecosystem in Grand Canyon
87% of the invertebrate biomass in the entire study area.
Stevens and others (1997) attributed the downstream
decline in Cladophora biomass, and hence the invertebrates
which are dependent on Cladophora and its attached
diatoms (i.e., especially Gammarus and chironomids), to
episodic inputs of suspended sediments from tributar-
ies such as the Paria and the Little Colorado Rivers that
reduce water clarity and light penetration enough to limit
algal production (fig. 5).
In contrast to the patterns described above, river-
bottom detritus (nonliving organic matter), which can
be derived from both terrestrial and aquatic sources,
peaked at approximately RM 124. This peak could
have occurred here because detritus is transported from
upstream locations and accumulates in this region of the
river. Detritus is an important component of the aquatic
food web because fish at downstream locations regularly
consume simuliids that feed on detritus via filter feeding.
In general, the overall quantity of drifting organic matter
increases and the composition changes from predomi-
nantly aquatic to terrestrial material with distance down-
stream (Shannon and others, 1996; Benenati and others,
2001). Shannon and others (1996) noted that tributary
inputs of organic matter constituted less than 0.1% of
the total organic drift of the Colorado River. We suspect,
however, that 0.1% is a gross underestimate of tributary
organic inputs because it does not appear that sampling
of tributary organic inputs was carried out during peri-
ods of flooding. Determining whether tributary inputs of
organic matter and energy are important to the food web
within the Colorado River will be a major focus of future
aquatic ecology efforts, as outlined in the Discussion and
Future Research Needs section of this chapter.
The density of many important components of the
aquatic food web appears to vary with season. The den-
sities of Cladophora, the dominant algae in the Lees Ferry
reach, and Oscillatoria, the dominant algae at downstream
locations, vary over time with peak density occurring
during summer (Stevens and others, 1997). The high-
est density of aquatic invertebrates also occurred during
the summer (Stevens and others, 1997). In contrast, the
concentration of river-bottom detritus was greatest dur-
ing autumn (Stevens and others, 1997), perhaps because
this is when trees adjacent to the river were dropping
their leaves or because lower river flows during this time
of year allowed the material to settle out of the water
column and accumulate on the river bottom.
It should be noted that because of logistical chal-
lenges, samples from only about 10 locations along the
entire 241 mi (386 km) of the Grand Canyon ecosystem
have been used to characterize spatial and temporal
variability of organic matter and invertebrates. Further,
much of the sampling (e.g., Shannon and others, 1994,
2001b; Blinn and others, 1995, 1998; Stevens and others,
1997; Benenati and others, 1998; McKinney and others,
1999) focused heavily on quantifying organic matter and
invertebrate dynamics at cobble bar habitats; yet, cobble
bars make up less than 10% of the aquatic habitat down-
stream from Glen Canyon Dam (Mietz, 2003). Thus,
these results may not accurately characterize spatial and
temporal variability of organic matter and invertebrates
within the Grand Canyon ecosystem.
Mathematical modeling is a powerful tool that can
be used to predict or estimate variables of interest (e.g.,
algae productivity) across large areas, such as the Grand
Canyon ecosystem, where logistics prevent intensive
field sampling. Yard and others (2005) measured and
modeled the influence of canyon orientation and topo-
graphic complexity on solar inputs to the Grand Canyon
ecosystem. Light is the resource that most often limits
the growth of algae and plants in aquatic environments
(Wetzel, 2001), so these results provide an indication of
potential algae growth across the entire Grand Canyon
ecosystem. One of the most striking results of these
modeling efforts is that river reaches that are oriented
east-west receive far less solar radiation during the winter
months relative to north-south reaches because the sun
is lower on the horizon at this time of year. That is, the
river has a clear view of the sun as it traces a path across
the horizon during the winter months along north-south
reaches, but the sun never gets high enough on the hori-
zon to shine on the river along east-west reaches. Thus,
Yard and others (2005) forecast that algae production
should vary predictably with canyon orientation and sea-
son because of differences in solar radiation and because
of the general downstream decline in algae production
that is associated with tributary sediment inputs that
reduce water clarity and light penetration.
The Influence of Dam Releases on
Organic Matter and Invertebrates
Understanding how Glen Canyon Dam discharge
regimes influence Cladophora and associated inverte-
brates has been a major focus of recent research efforts
(Angradi, 1994; Shannon and others, 1994, 1996, 2001b;
Valdez and Ryel, 1995; McKinney and Speas, 2001).
With discharge from Glen Canyon Dam fluctuating as
much as 8,000 cubic feet per second (cfs) daily, there
exists a large varial zone of shoreline habitat that is both
inundated and exposed to air for long periods each day
(fig. 3). Several studies have determined that the varial
zone supports a relatively low density of algae, which is
Aquatic Ecology 95
often dominated by Oscillatoria spp. because Cladophora
and associated invertebrates cannot thrive in the varial
zone (Blinn and others, 1995; Shaver and others, 1997;
Benenati and others, 1998). Specifically, Blinn and others
(1995) found fourfold higher invertebrate mass in per-
manently submerged zones compared to the varial zone.
Using a series of in situ experiments, they determined
that snails readily recolonized cobbles that were resub-
merged after initially being subjected to long-term desic-
cation; however, the density of Cladophora,Gammarus, and
chironomids on resubmerged cobbles was still less than
30% of control sites after 4 mo.
Fluctuations in river flows also have an impact on
drifting organic matter because periodic desiccation can
weaken algae and invertebrates, making them more sus-
ceptible to fragmentation and entrainment by the river.
Moreover, higher river flows lead to more turbulent and
faster water that is more likely to entrain organic matter
and invertebrates. Shannon and others (1996) found that
the quantity of drifting organic matter increased with
discharge.
Historical data are insufficient to quantitatively
determine what impact the MLFF alternative has had
on organic matter and invertebrates; however, it is pos-
sible to qualitatively describe the likely impacts of this
flow regime based on the research described above. By
restricting daily fluctuations in discharge to less than
8,000 cfs and limiting minimum discharge to 5,000
cfs, MLFF flows have reduced the size of the varial
zone and increased the amount of river bottom that is
permanently submerged. Both of these changes prob-
ably increased the productivity and standing mass of
important components of the aquatic food web, includ-
ing Cladophora and Gammarus; however, abrupt changes
in monthly release volumes that are permitted under
the Record of Decision and MLFF may be detrimental
to algae and aquatic invertebrates. For example, when
monthly release volumes are decreased, the amount of
river-bottom habitat that is permanently submerged
also decreases. It seems likely that there is a subsequent,
abrupt decrease in the quantity of food available for fish
in the Grand Canyon ecosystem. Further, when monthly
release volumes are abruptly increased, algae and inver-
tebrates that were in shallow, nearshore habitats may
find themselves under several yards of water. While this
increase in monthly volume may eventually lead to an
increase in the quantity of food available (by increasing
available habitat), it seems likely that there is a short-
term decrease in food as algae and invertebrates adjust to
the new conditions.
Experimental high flows, or controlled floods, have
been used principally as a tool to restore sandbars in the
Grand Canyon ecosystem, but these floods also impact
organic matter and invertebrates. Blinn and others (1999)
found that the 1996 beach/habitat-building flow scoured
more than 90% of the primary producer biomass (i.e.,
algae and submerged aquatic plants) and about 50%
of the river-bottom invertebrates from a site at Lees
Ferry; primary producers (1 mo) and invertebrates (2
mo) quickly recovered to preflood levels. In contrast,
McKinney and others (1999) found that the 1996 beach/
habitat-building flow caused short-term reductions in the
standing mass of primary producers and invertebrates
only in depositional habitats (i.e., areas of sand/silt)
and not in more resistant habitats like cobble bars (fig.
6). Brock and others (1999) found that the 1996 beach/
habitat-building flow actually led to significant increases
in algae production rates (the rate at which photosynthe-
sis is occurring within the algae); they hypothesized that
algae production increased because the flood removed
senescent, or old, material and detritus from the algae.
Marzolf and others (1999) measured oxygen production,
a byproduct of photosynthesis, along several river seg-
ments within the Lees Ferry reach and found that those
segments produced less oxygen after the flood relative to
preflood values. These data suggest that the 1996 beach/
habitat-building flow did in fact scour large quantities
of algae and aquatic macrophytes from the Lees Ferry
reach, resulting in a systemwide reduction in primary
production. Even though the flood may have reduced the
standing mass of invertebrates from some areas in the
Lees Ferry reach, the quantity of food items in rainbow
trout stomachs was actually greater immediately after the
flood relative to before the flood (McKinney and others,
1999). Blinn and others (1999) used stable-isotope analy-
ses to determine that riparian vegetation and upland
vegetation were the dominant types of drifting organic
matter during the flood, while river-bottom algae were
the dominant drifting organic matter during normal dam
operations. Thus, experimental high flows can scour ben-
thic algae and invertebrates and capture large quantities
of terrestrial organic matter, which may temporarily
increase the amount of food available for fish.
Recent Findings
New Zealand Mudsnail Invasion
Biological invasions represent a significant threat to
the persistence of resident species because invaders are
capable of altering food-web structure, rates of disease
96 The State of the Colorado River Ecosystem in Grand Canyon
or parasitism, and the amount or type of energy at the
base of a food web (Vitousek, 1990; Wilcove and others,
1998; Kennedy and Hobbie, 2004). Although the food
web in the Colorado River below Glen Canyon Dam
has already been dramatically changed because of the
installation of the dam and intentional introductions of
nonnative sport fish, it is still susceptible to the impacts
of biological invasions.
The New Zealand mudsnail (Potamopyrgus antipo-
darum), a species that is rapidly spreading throughout
North American rivers and lakes (Hall and others, 2003,
and references therein), has recently invaded the Grand
Canyon ecosystem. The presence of the mudsnail in
Grand Canyon was first identified from samples collected
in March 2002; however, the analysis of archived col-
lections revealed that mudsnails were actually present as
early as May 1995 (Benenati and others, 2002). Presently,
this small snail (fig. 7), which measures approximately 0.2
inch (about 5 mm), occurs in high densities throughout
the Grand Canyon ecosystem, particularly in the Lees
Ferry reach where densities of more than 32,800/yd2
(40,000 snails/m2) have been found (Benenati and others,
2002). Before the New Zealand mudsnail invasion, snails
represented a minor component of total invertebrate
mass in the Lees Ferry reach (less than 6% of the total
in 1992). Snails are now the dominant category of
invertebrate in the Lees Ferry reach, representing more
than 66% of invertebrate mass in 2001 (Benenati and
others, 2002).
The potential impacts of the New Zealand mudsnail
invasion on the aquatic food web of the Grand Canyon
ecosystem are significant because nonnative rainbow
trout apparently cannot easily digest the snails (fig. 1);
snails eaten by trout often survive intact after passage
through the gut (Mike Yard, oral commun., 2004),
perhaps because of the snails’ protective operculum.
Humpback chub have pharyngeal gills that are capable
of crushing snail shells and therefore might be able to
more completely digest them; however, the only detailed
diet analysis for chub was conducted before the mud-
snail invasion of the Colorado River ecosystem. Because
Figure 6. Algae (A) and invertebrate (B) response to the
1996 beach/habitat-building flow. The flood occurred between
the March and April sampling dates. Bars represent average
standing mass (algae) or density (Gammarus) at cobble bars
in the Lees Ferry reach, while the thin lines on top of each bar
represent 1 standard error (an indication of the uncertainty
associated with each estimate). Data from McKinney and others
(1999), table 1, p. 251.
Figure 7. New Zealand mudsnails on a concrete anchor in Glen
Canyon. The density of snails shown here is typical for the Glen
Canyon ecosystem (photograph by Michael Booth).
A.
B.
Aquatic Ecology 97
rainbow trout cannot easily digest the mudsnail, the
mudsnail may have a competitive advantage over other
herbivores such as Gammarus, potentially allowing it to
displace these other organisms as it continues to spread.
If the density of Gammarus and other invertebrates that
are regularly consumed by fish declines because of com-
petition with the New Zealand mudsnail, it seems likely
that these changes will affect fish density and condition.
Even if mudsnails do not compete with or displace herbi-
vores such as Gammarus, which seems unlikely given their
extremely high density, they dramatically alter important
ecosystem functions, including rates of nutrient cycling
and primary production (Hall and others, 2003). Unfor-
tunately, historical data are inadequate to determine
whether the density or biomass of Gammarus and other
invertebrates that are important food items for fish in the
Colorado River ecosystem has been affected by the New
Zealand mudsnail invasion.
Discussion and Future
Research Needs
With recent declines in the size of native fish popu-
lations and the condition of rainbow trout (see chapter 2,
this report), understanding what food resources drive fish
growth and production, as well as what sources of energy
drive production of these food resources, becomes
increasingly important. Previous research on the aquatic
food web has produced a relatively clear picture of the
food habits of fish in the Grand Canyon ecosystem.
Aquatic invertebrates—particularly Gammarus, simuliids,
and chironomids—appear to be the most important food
items for both rainbow trout and humpback chub. In the
upper reaches of the Colorado River near Lees Ferry,
it appears that invertebrates, and therefore the fish that
consume them, are fueled almost exclusively by algae,
particularly diatoms attached to Cladophora. Although
there is evidence that the invertebrates consumed by fish
at downstream locations are relying on both aquatic and
terrestrial organic matter, the relative importance and
sources of this material remain unclear. This uncertainty
prevents a complete understanding of the role that food
availability plays in determining the condition and popu-
lation size of native and nonnative fish.
Food-web analysis provides a framework for quan-
tifying the movement of terrestrial and aquatic mate-
rial into higher trophic levels, the trophic positions of
consumers, and the importance of interactions such as
competition and predation. Applied research efforts have
repeatedly benefited from studying an ecosystem from
a food-web perspective (Winemiller and Polis, 1996).
Fisheries management, in particular, can benefit from a
food-web perspective because it is critical for accurately
predicting the responses of both predators and prey to
management actions (Parsons, 1992). Although many
food-web studies of terrestrial and aquatic systems have
focused on trophic pathways based exclusively on aquatic
production, it is increasingly recognized that leaf litter
and other types of terrestrial organic matter play a major
role in determining ecosystem structure and function
(Winemiller and Polis, 1996).
In the case of the Grand Canyon ecosystem, previ-
ous research efforts on the aquatic food web have focused
almost exclusively on trophic pathways associated with
aquatic organic matter, namely the filamentous algae
Cladophora, the diatoms attached to this algae, and the
invertebrate consumers of these attached diatoms (Usher
and Blinn, 1990; Hardwick and others, 1992; Shannon
and others, 1994). Focusing on these aquatic sources
of organic matter seems appropriate for the tailwater
section of the Colorado River, that is, from the dam
to Lees Ferry. Yet, the tailwater is not representative of
downstream portions of the river and accounts for less
than 7% of the total wetted area of the Grand Canyon
ecosystem. A thorough analysis of the trophic signifi-
cance of terrestrially derived material has never been
conducted, although there are some correlative data
that support the contention that the downstream decline
in algae production limits secondary production at
downstream sites (Shaver and others, 1997; Stevens and
others, 1997). Importantly, the one detailed food-web
analysis that has been conducted in the Grand Canyon
ecosystem indicates that terrestrially derived carbon is
contributing to invertebrate and fish production at down-
stream tributaries (Angradi, 1994).
Given these considerations, it seems clear that future
research and monitoring efforts should take a broader
view of the food web and attempt to document the
relative importance of aquatic and terrestrial organic
matter to invertebrate and fish production. For example,
if invertebrates, and by extension humpback chub and
rainbow trout, are dependent on algae throughout the
ecosystem, a systemwide reduction in algae production
would likely have strong negative consequences for fish,
especially because algae biomass is already very limited
at downstream sites. Alternatively, if terrestrial organic
matter is fueling production of invertebrates at down-
stream sites, then findings from the proposed food-web
research would provide managers with some of the
information necessary to assess the effects of proposed
management actions, including sediment augmentation
and thermal modifications, on fish production.
98 The State of the Colorado River Ecosystem in Grand Canyon
Food-web analysis should continue to focus on the
two most ecologically and economically important fish
species, humpback chub and rainbow trout, and use a
combination of gut-content and stable-isotope analysis.
To determine whether the resource base of the food web
shifts downstream and is affected by tributary inputs of
organic matter, a food-web analysis should encompass
the entire study area including the major tributaries
of the Colorado River, including the Paria and Little
Colorado Rivers. Shannon and others (2001a) found
downstream shifts in the isotopic composition of algae,
invertebrates, and fish. This downstream shift may com-
plicate interpretation of stable-isotope data, but another
technique for determining trophic linkages, known
as quantitative fatty acid analysis (Iverson and others,
2004), may allow investigators to resolve any uncertain-
ties associated with stable-isotope analysis. Manipula-
tive experiments involving New Zealand mudsnails also
represent an important research direction because these
experiments could help scientists determine whether or
not this invasive species is having a negative impact on
important food items for fish and identify the factors that
control New Zealand mudsnail density. Collectively, these
activities will determine the short- and long-term feeding
habits of humpback chub and rainbow trout, the energy
resources at the base of the food web, whether hump-
back chub are consuming and digesting New Zealand
mudsnails, and the impact of the mudsnail invasion on
the aquatic food web.
References
Anderson, N.H., and Sedell, J.R., 1979, Detritus pro-
cessing by macroinvertebrates in stream ecosystems:
Annual Review of Entomology, v. 24, p. 351–377.
Angradi, T.R., 1994, Trophic linkages in the lower
Colorado River—multiple stable isotope evidence:
Journal of the North American Benthological Society,
v. 13, p. 479–495.
Bayley, P.B., 1989, Aquatic environments in the Amazon
Basin, with an analysis of carbon sources, fish produc-
tion, and yield, in Proceedings of the International
Large River Symposium: Canadian Special Publication
of Fisheries and Aquatic Sciences, v. 106, p. 399–408.
Benenati, E.P., Shannon, J.P., Hagan, J.S., and Blinn, D.W.,
2001, Drifting fine particulate organic matter below Glen
Canyon Dam in the Colorado River, Arizona: Journal of
Freshwater Ecology, v. 16, p. 235–248.
Benenati, P.L., Shannon, J.P., and Blinn, D.W., 1998,
Desiccation and recolonization of phytobenthos in a
regulated desert river: Colorado River at Lees Ferry,
Arizona, USA: Regulated Rivers: Research & Man-
agement, v. 14, p. 519–532.
Benenati, P.L., Shannon, J.P., Haden, A., Straka, K.,
and Blinn, D.W., 2002, Monitoring and research: the
aquatic food base in the Colorado River, Arizona
during 1991–2001: Flagstaff, Northern Arizona
University.
Bilby, R.E., and Bisson, P.A., 1992, Allochthonous versus
autochthonous organic matter contributions to the
trophic support of fish populations in clear-cut and old
growth forested streams: Canadian Journal of Fisher-
ies and Aquatic Sciences, v. 49, p. 540–551.
Blinn, D.W., and Cole, G.A., 1991, Algal and inverte-
brate biota in the Colorado River: comparison of
pre- and post-dam conditions, Colorado River ecology
and dam management: Washington, D.C., National
Academy Press, p. 102–123.
Blinn, D.W., Shannon, J.P., Benenati, P.L., and Wilson,
K.P., 1998, Algal ecology in tailwater stream commu-
nities: the Colorado River below Glen Canyon Dam,
Arizona: Journal of Phycology, v. 34, p. 734–740.
Blinn, D.W., Shannon, J.P., Stevens, L.E., and Carder,
J.P., 1995, Consequences of fluctuating discharge for
lotic communities: Journal of the North American
Benthological Society, v. 14, p. 233–248.
Blinn, D.W., Shannon, J.P., Wilson, K.P., O’Brien, C.,
and Benenati, P.L., 1999, Response of benthos and
organic drift to a controlled flood, in Webb, R.H.,
Schmidt, J.C., Marzolf, G.R., and Valdez, R.A., eds.,
The controlled flood in Grand Canyon: Washington,
D.C., American Geophysical Union, Geophysical
Monograph Series, v. 110, p. 259–272.
Brock, J.T., Royer, T.V., Snyder, E.B., and Thomas, S.A.,
1999, Periphyton metabolism: a chamber approach,
in Webb, R.H., Schmidt, J.C., Marzolf, G.R., and
Valdez, R.A., eds., The controlled flood in Grand
Canyon: Washington, D.C., American Geophysical
Union, Geophysical Monograph Series, v. 110, p.
217–223.
Carothers, S.W., and Brown, B.T., 1991, The
Colorado River through Grand Canyon: natural
history and human change: Tucson, University of
Arizona Press, 235 p.
Aquatic Ecology 99
Fry, B., and Sherr, E.B., 1984, δ13C measurements as indi-
cators of carbon flow in marine and freshwater ecosys-
tems: Contributions to Marine Science, v. 27, p. 13–47.
Glen Canyon Dam Adaptive Management Program,
2001, Strategic plan.
Haden, A., Blinn, D.W., and Shannon, J.P., 1999, Food
base studies and stable isotope analysis of the diet of
humpback chub (Gila cypha) in the Little Colorado
River, Coconino County, AZ: Flagstaff, Northern
Arizona University, Department of Biological Sciences.
Haden, G.A., Shannon, J.P., Wilson, K.P., and Blinn,
D.W., 2003, Benthic community structure of the
Green and Colorado Rivers through Canyonlands
National Park, Utah, USA: Southwestern Naturalist, v.
48, p. 23–35.
Hall, R., Tank, J.L., and Dybdahl, M.F., 2003, Exotic
snails dominate nitrogen and carbon cycling in a
highly productive stream: Frontiers in Ecology and the
Environment, v. 1, p. 407–411.
Hardwick, G.G., Blinn, D.W., and Usher, H.D., 1992,
Epiphytic diatoms on Cladophora glomerata in the
Colorado River, Arizona: longitudinal and vertical dis-
tribution in a regulated river: Southwestern Naturalist,
v. 37, p. 148–156.
Hofknecht, G.W., 1981, Seasonal community dynamics
of aquatic invertebrates in the Colorado River and its
tributaries within Grand Canyon, Arizona: Flagstaff,
Northern Arizona University, M.S. thesis, 105 p.
Iverson, S.J., Field, C., Bowen, W.D., and Blanchard,
W., 2004, Quantitative fatty acid signature analysis: a
new method of estimating predator diets: Ecological
Monographs, v. 74, p. 211–235.
Kennedy, T.A., and Hobbie, S.E., 2004, Saltcedar
(Tamarix ramosissima) invasion alters organic matter
dynamics in a desert stream: Freshwater Biology, v. 49,
p. 65–76.
Lewis, W.M., Hamilton, S.K., Rodriguez, M.A.,
Saunders, J.F., and Lasi, M.A., 2001, Foodweb analy-
sis of the Orinoco floodplain based on production
estimates and stable isotope data: Journal of the North
American Benthological Society, v. 20, p. 241–254.
Marzolf, G.R., Bowser, C.J., Hart, R., Stephens, D.W.,
and Vernieu, W.S., 1999, Photosynthetic and respira-
tory processes: an open stream approach, in Webb,
R.H., Schmidt, J.C., Marzolf, G.R., and Valdez,
R.A., eds., The controlled flood in Grand Canyon:
Washington, D.C., American Geophysical Union,
Geophysical Monograph Series, v. 110, p. 205–215.
McKinney, T., Rogers, R.S., Ayers, A.D., and Persons,
W.R., 1999, Lotic community responses in the Lees
Ferry reach, in Webb, R.H., Schmidt, J.C., Marzolf,
G.R., and Valdez, R.A., eds., The controlled flood in
Grand Canyon: Washington, D.C., American Geo-
physical Union, Geophysical Monograph Series, v.
110, p. 249–258.
McKinney, T., and Speas, D.W., 2001, Observations of
size-related asymmetries in diet and energy intake of
rainbow trout in a regulated river: Environmental Biol-
ogy of Fishes, v. 61, p. 435–444.
McKinney, T., Speas, D.W., Rogers, R.S., and Persons,
W.R., 2001, Rainbow trout in a regulated river below
Glen Canyon Dam, Arizona, following increased mini-
mum flows and reduced discharge variability: North
American Journal of Fisheries Management, v. 21, p.
216–222.
Meretsky, V.J., Valdez, R.A., Douglas, M.E., Brouder,
M.J., Gorman, O.T., and Marsh, P.C., 2000, Spatio-
temporal variation in length-weight relationships of
endangered humpback chub: implications for conser-
vation and management: Transactions of the Ameri-
can Fisheries Society, v. 129, p. 419–428.
Meyer, J.L., and Edwards, R.T., 1990, Ecosystem metab-
olism and turnover of organic carbon along a black-
water river continuum: Ecology, v. 71, p. 668–677.
Mietz, S.W., 2003, Evaluating historical electrofishing
distribution in the Colorado River, Arizona, based
on shoreline substrate: Flagstaff, Northern Arizona
University, M.S. thesis, 74 p.
Minagawa, M., and Wada, E., 1984, Stepwise enrich-
ment of 15N along food chains: further evidence and
the relation between 15N and animal age: Geochimica
et Cosmochimica Acta, v. 48, p. 1135–1140.
Parsons, T.R., 1992, The removal of marine preda-
tors by fisheries and the impact of trophic structure:
Marine Pollution Bulletin, v. 26, p. 51–53.
100 The State of the Colorado River Ecosystem in Grand Canyon
Peterson, B.J., and Fry, B., 1987, Stable isotopes in
ecosystem studies: Annual Review of Ecology and
Systematics, v. 18, p. 293–320.
Pinney, C.A., 1991, The response of Cladophora glomerata
and associated epiphytic diatoms to regulated flow, and
the diet of Gammarus lacustris, in the tailwaters of Glen
Canyon Dam: Flagstaff, Northern Arizona University,
M.S. thesis, 81 p.
Shannon, J.P., Blinn, D.W., Benenati, P.L., and Wilson,
K.P., 1996, Organic drift in a regulated desert river:
Canadian Journal of Fisheries and Aquatic Sciences, v.
53, p. 1360–1369.
Shannon, J.P., Blinn, D.W., Haden, G.A., Benenati, E.P.,
and Wilson, K.P., 2001a, Food web implications of
delta C-13 and delta N-15 variability over 370 km of
the regulated Colorado River, USA: Isotopes in Envi-
ronmental and Health Studies, v. 37, p. 179–191.
Shannon, J.P., Blinn, D.W., McKinney, T., Benenati,
E.P., Wilson, K.P., and O’Brien, C., 2001b, Aquatic
food base response to the 1996 test flood below Glen
Canyon Dam, Colorado River, Arizona: Ecological
Applications, v. 11, p. 672–685.
Shannon, J.P., Blinn, D.W., and Stevens, L.E., 1994,
Trophic interactions and benthic animal community
structure in the Colorado River, Arizona, USA: Fresh-
water Biology, v. 31, p. 213–220.
Shaver, M.L., Shannon, J.P., Wilson, K.P., Benenati, P.L.,
and Blinn, D.W., 1997, Effects of suspended sedi-
ment and desiccation on the benthic tailwater com-
munity in the Colorado River, USA: Hydrobiologia,
v. 357, p. 63–72.
Stevens, L.E., Shannon, J.P., and Blinn, D.W., 1997,
Colorado River benthic ecology in Grand Canyon,
Arizona, USA: dam, tributary and geomorphological
influences: Regulated Rivers: Research & Manage-
ment, v. 13, p. 129–149.
Stevens, L.E., Sublette, J.E., and Shannon, J.P., 1998,
Chironomidae (Diptera) of the Colorado River, Grand
Canyon, Arizona, USA, II: factors influencing distri-
bution: Great Basin Naturalist, v. 58, p. 147–155.
Sublette, J.E., Stevens, L.E., and Shannon, J.P., 1998,
Chironomidae (Diptera) of the Colorado River, Grand
Canyon, Arizona, USA, I: systematics and ecology:
Great Basin Naturalist, v. 58, p. 97–146.
Thorp, J.H., and Delong, M.D., 2002, Dominance of
autochthonous autotrophic carbon in food webs of
heterotrophic rivers: Oikos, v. 96, p. 543–550.
Topping, D.J., Schmidt, J.C., and Vierra, L.E., Jr., 2003,
Computation and analysis of the instantaneous-dis-
charge record for the Colorado River at Lees Ferry,
Arizona—May 8, 1921, through September 30, 2000:
U.S. Geological Survey Professional Paper 1677.
Usher, H.D., and Blinn, D.W., 1990, Influence of various
exposure periods on the biomass and chlorophyll a of
Cladophora glomerata (Chlorophyta): Journal of Phycol-
ogy, v. 26, p. 244–249.
Valdez, R.A., and Carothers, S.W., 1998, The aquatic
ecosystem of the Colorado River in Grand Canyon.
Report to Bureau of Reclamation, Salt Lake City,
Utah: Flagstaff, Ariz., SWCA, Inc., Environmental
Consultants, 250 p.
Valdez, R.A., and Ryel, R.J., 1995, Life history and ecol-
ogy of the humpback chub (Gila cypha) in the Colorado
River, Grand Canyon, Arizona: Final report to Bureau
of Reclamation, contract no. 0-CS-40-09110.
Vitousek, P.M., 1990, Biological invasions and ecosystem
processes: towards an integration of population biol-
ogy and ecosystem studies: Oikos, v. 57, p. 7–13.
Wetzel, R.G., 2001, Limnology (3d ed.): San Diego,
Calif., Academic Press, 1006 p.
Wilcove, D.S., Rothstein, D., Dubow, J., Phillips, A.,
and Losos, E., 1998, Quantifying threats to imper-
iled species in the United States: BioScience, v. 48, p.
607–615.
Winemiller, K.O., and Polis, G.A., 1996, Food webs:
what can they tell us about the world?, in Polis, G.A.,
and Winemiller, K.O., eds., Food webs: integration of
patterns and dynamics: New York, Chapman & Hall,
p. 1–22.
Yard, M.D., Bennett, G.E., Mietz, S.N., Coggins, L.G.,
Stevens, L.E., Hueftle, S., and Blinn, D.W., 2005,
Influence of topographic complexity on solar insola-
tion estimates for the Colorado River, Grand Canyon,
AZ: Ecological Modelling, v. 183, p. 157–172.
Aquatic Ecology 101
Contact Information:
Theodore A. Kennedy
Aquatic Ecologist
U.S. Department of the Interior
U.S. Geological Survey
Southwest Biological Science Center
Flagstaff, AZ
tkennedy@usgs.gov
Steven P. Gloss
Ecologist
U.S. Department of the Interior
U.S. Geological Survey
Southwest Biological Science Center
Tucson, AZ
sgloss@usgs.gov
J
e
f
f
S
o
r
e
n
s
e
n
,
A
r
i
z
o
n
a
G
a
m
e
a
n
d
F
i
s
h
D
e
p
a
r
t
m
e
n
t
First page photograph credit: Jeff Sorensen, Arizona Game and Fish Department
Chapter 6
Riparian
Vegetation
and Associated
Wildlife
Barbara E. Ralston
Introduction
In the Western United States, riparian areas are
conspicuous as narrow belts of dense, green vegetation
along streams and rivers. Fluvial marshes—areas of
wetland characterized by emergent herbaceous plants
like sedges (Carex sp.), grasses (Poaceae family), and cat-
tails (Typha sp.)—can be part of the riparian areas that
are associated with erosion and sediment deposition
patterns of the adjoining stream or river. The riparian
and wetland plant community is dependent on surface
water and groundwater flows (Busch and Smith, 1995;
Stromberg and others, 1996; Stromberg, 2001) and is
transitional between aquatic and upland systems. In
Grand Canyon, the upland system is characterized by
limited moisture and includes Great Basin desertscrub,
Mohave desertscrub, and Sonoran desertscrub plant
constituents (classifications per Brown, 1982), such as
mormon tea (Ephedra nevadensis), sagebrush (Artemisia
spp.), white brittle bush (Encelia farinosa), and barrel
cactus (Ferocactus cylindraceus). The dry uplands are a stark
contrast to the lush plant community along the Colorado
River in Grand Canyon (figs. 1a–d).
Riparian areas are a junction between aquatic and
terrestrial habitat types. In the West, they tend to exhibit
higher levels of species diversity, richness, and population
densities than either adjacent habitat. Because of these
characteristics, riparian areas are of high value to man-
agers, scientists, and the public, particularly to Native
American communities (see chapter 11, this report). The
importance of riparian areas in the maintenance of bio-
diversity is well documented (Nilsson and others, 1989;
Naiman, 1992; Nilsson, 1992; Decamps, 1993; Lock and
Naiman, 1998; Saab, 1999; National Research Council,
2002). Riparian areas are especially important in the
Southwestern United States, where more than 50% of
166 species of breeding birds in the lowlands are com-
pletely dependent on water-related habitat (Johnson and
others, 1977; Farley and others, 1994). Riparian areas
also buffer the movement of materials, such as nitrogen
and carbon, between aquatic and terrestrial environ-
ments and help retain nutrients along the river or stream
channel (Schlosser and Karr, 1981; Jacobs and Gilliam,
1985; Naiman and Decamps, 1990). Vegetation along a
water course also provides cover and food such as insects
and seeds for animal life. In Grand Canyon, whitewater
recreationists and hikers also use the shade of riparian
shrubs and trees in the hot summer months. Understand-
104 The State of the Colorado River Ecosystem in Grand Canyon
Figure 1. A. Riparian zone along the Colorado River in Grand Canyon. The predam high-water zone is noticeable as a line of vegetation
well above the shoreline. Postdam riparian vegetation has progressed downslope and become thicker (photograph © 2005 Geoff
Gourley; used with permission). B. Tamarisk (Tamarix ramosissima), which is one of the species that was found along the predam high-
water zone. Tamarisk was introduced to the Colorado River Basin in the 1880s and was present in Grand Canyon in 1938 (photograph
by Jeff Sorensen, Arizona Game and Fish Department). C. Seep willow (Baccharis sp.), a new high-water zone constituent. This
species was also present before the dam, but in lower densities (photograph by Jeff Sorensen, Arizona Game and Fish Department).
D. Examples of fluvial marsh and postdam species. The foreground is composed of common reed (Phragmites australis) and water
sedge (Carex aquatilis), but the background shows tamarisk and arrowweed (Pluchea sericea) (photograph by Jeff Sorensen, Arizona
Game and Fish Department).
A. B.
C. D.
Riparian Vegetation and Associated Wildlife 105
ing how this community has changed over time is key to
developing appropriate management strategies for this
important resource.
This chapter describes changes in the riparian and
fluvial marsh communities along the Colorado River in
Grand Canyon from the closure of the Glen Canyon
Dam and the beginning of the regulation of the river in
1963 to the present. To provide a better understanding
of how dam operations have affected riparian vegetation,
we examine changes in Grand Canyon riparian vegeta-
tion during three periods of time (1963–80; 1981–91;
1991–present) that correspond to major operational
changes at Glen Canyon Dam. The effects on riparian
vegetation of both the modified low fluctuating flow
(MLFF) alternative, which was implemented begin-
ning in 1996, and the recent drought are discussed. The
chapter concludes with a summary of the findings with
respect to riparian vegetation as habitat and its relation-
ship to other resources and with a discussion of moni-
toring priorities within the context of the Glen Canyon
Dam Adaptive Management Program.
Background
Predam Vegetation
Much of what is known about predam vegetation
comes from the 1938 investigation of the Colorado River
corridor by Clover and Jotter (1944). Predam vegeta-
tion in the high-elevation benches (fig. 2), the areas well
above the river that are less frequently scoured by floods,
was dominated by mesquite (Prosopis glandulosa), catclaw
acacia (Acacia greggii), Apache plume (Fallugia paradoxa),
and tamarisk (Tamarix ramosissima). Tamarisk was intro-
duced to the Colorado River Basin in the 1800s and was
present in Grand Canyon in 1938 (Clover and Jotter,
1944). These same investigators noted the presence of
coyote willow (Salix exigua), rabbitbrush (Chrysothamnus
nauseosus), four-wing salt bush (Atriplex canescens), and
Goodding’s willow (Salix gooddingii) along or close to the
shoreline (moist sand) at Lees Ferry. On higher benches,
they found arrowweed (Pluchea sericea) and four-wing salt
bush. Mormon tea and rabbitbrush were found in the
talus above the flood plain. The presence of plant cover
from the river up to the talus at Lees Ferry (RM 0) in
1938 illustrated the degree of community development.
Greater vegetation cover may have occurred previously
because the 1930s was the period of the lowest dis-
charges in the record (Topping and others, 2003).
Elsewhere in the river corridor, Clover and Jotter
(1944) documented patchy riparian development, includ-
ing the absence of vegetation in the moist sand zone
at President Harding Rapids (approximately RM 43.8)
(Stevens, 1990) associated with a recent sand deposit.
They also noted the existence of marsh emergent spe-
cies, including cattails and rushes (Juncaceae family).
The predam riparian zone of Grand Canyon was
periodically disturbed with floods of variable frequency
and magnitude that redistributed sediment of vary-
ing grain size and influenced what plants colonized the
bare area. Plants like acacia and mesquite located in the
high-water zone were disturbed less frequently than were
herbaceous and marsh plants located near the lower
benches. The riparian assemblage likely represented
different stages of maturity and succession as it pro-
gressed either landward or downstream from a tributary
source of disturbance. At the higher benches, the vegeta-
tion would be the most mature and stable, while at the
shoreline, the vegetation would be composed of more
flood-tolerant pioneering species. Areas from the channel
upslope to the former high-water zone (also known as the
old high-water zone) were composed of a mix of peren-
nial and annual plants that corresponded with tolerances
to moisture and disturbance. (Hereafter, refer to fig. 2 for
references that relate discharge to riparian elevations.)
Variables Affecting
Vegetation Change
The abundance, distribution, and composition of
riparian and fluvial marsh vegetation along the river
corridor in Grand Canyon are influenced by many
variables including yearly discharge, soils, sedimentation,
time since disturbance, and temperature (Turner and
Karpiscak, 1980; Baker, 1989; Stromberg and Patten,
1991; Busch and Smith, 1995; Stevens and others, 1995;
Stromberg, 2001). A conceptual model illustrates some
of the linkages between physical processes and riparian
habitat (fig. 3); however, the model shown in figure 3 is
not comprehensive with respect to all variables that affect
riparian habitat.
Following closure of Glen Canyon Dam in 1963,
changes in the hydrologic and sediment regime occurred
that affected vegetation in Grand Canyon. The opera-
tions reduced annual peak discharge and duration and
increased the yearly base flow (Topping and others,
2003). The yearly hydrograph was replaced by monthly
volume releases that followed energy demands (high
releases in December–February and in July–September)
rather than seasonal patterns (high flows occurring
106 The State of the Colorado River Ecosystem in Grand Canyon
in June and July). Postdam median daily discharge
increased to 12,600 cubic feet per second (cfs), which
was 58% greater than the predam volumes of 7,980 cfs
(Topping and others, 2003). Larger average discharges
increased sediment-export rates (Topping and others,
2000) and reduced sand storage along channel margins
and reattachment bars associated with debris fan-eddy
complexes (Schmidt and Rubin, 1995). Both channel
margins and reattachment bars are substrates for plant
colonization. Higher sediment-export rates were most
noticeable in Marble Canyon (Topping and others,
2000). Also, the dam reduced the amount of upstream
sediment coming into the river by 99.9% (Topping and
others, 2000); tributaries below the dam, including the
Paria and Little Colorado Rivers, now provide the bulk
of sand and smaller sediment fractions (see chapter 1,
this report).
Disturbance frequency along a river or stream is
one of several variables that affect riparian community
development (Stromberg and Patten, 1991; Bendix,
1994; Toner and Keddy, 1997). Scour and sediment
reworking within depositional environments like debris
fans, channel margins, and return channels (Schmidt,
1990; Schmidt and Rubin, 1995) provided sites for
colonization by marsh and riparian plant species (Clover
and Jotter, 1944; Turner and Karpiscak, 1980; Gecy and
Wilson, 1990; Decamps, 1993). The pioneer assemblage
may be from an introduced seed source or from vegeta-
tive regrowth following scour (Gecy and Wilson, 1990).
Before regulation of the Colorado River through Glen
and Grand Canyons, large portions of the river’s flood
plain were periodically scoured. Predam 1-yr return
flood discharge reached approximately 50,000 cfs, with
larger discharges of 120,000 cfs occurring every 6 yr on
average (fig. 2) (Topping and others, 2003). Yearly flood-
ing reduced vegetation below the 50,000-cfs water-sur-
face elevation, while larger, less frequent floods affected
vegetation communities on higher benches. Lower peak
flows caused by Glen Canyon Dam allowed species,
including nonnative plants, to occupy lower flood-plain
Figure 2. Habitat zones according to inundation frequency and flow magnitude (cubic feet per second). Regulation of the Colorado
River by Glen Canyon Dam has resulted in reduced flood frequencies and magnitudes. In the postdam setting, the area below 50,000-
cfs water-surface elevation is the active riparian zone. (In the postdam period, the active riparian zone remained at 50,000-cfs water-
surface elevation during the 1980s, dropping to 45,000 cfs after 1991, when new restrictions constrained dam operations to minimize
resource impacts.) Above the active riparian zone are the higher elevation benches, areas well above the river that are less frequently
scoured by floods and were subjected to a predam 6-yr return flood frequency. Within the active riparian zone are bands of vegetation
that follow a moisture gradient from water-tolerant plants located near shoreline to species that tolerate drier upslope conditions.
Figure modified from Carothers and Aitchison (1976) with data from Topping and others (2003).
Riparian Vegetation and Associated Wildlife 107
benches. Riparian areas are particularly prone to inva-
sion by nonnative plants because they are frequently dis-
turbed by flood events, which create favorable conditions
for the seeds of nonnative plants that can be dispersed
long distances by a variety of processes (Decamps, 1993).
Status and Trends
Riparian community changes following the closure of
Glen Canyon Dam in 1963 occurred over three periods.
Period I: Initial Vegetation
Expansion (1963–80)
Immediately following the closure of Glen Canyon
Dam, operations focused on filling Lake Powell, deliver-
ing water to Lake Mead, and producing peak power.
Discharges were reduced to between 1,000 and 20,000
cfs during this period (median discharge was 9,490 cfs in
the 1960s). A series of discharges of 50,000 cfs con-
ducted in 1965 cleaned the channel below the dam and
raised the elevation of Lake Mead (Topping and others,
2003). Daily fluctuations in the 1970s were large, varying
between 4,000 and 25,000 cfs, with a median discharge
of 11,600 cfs (Topping and others, 2003). The result
of dam operations during this period was to encourage
plant colonization along the channel in the low-elevation
benches.
Vegetation expansion below the 50,000-cfs water-
surface elevation was documented by Turner and
Karpiscak (1980), who used repeat photography from
historical expeditionary trips through Grand Canyon
such as J.W. Powell’s second trip in 1872 (Darrah, 1948),
the Robert Stanton expedition in 1889 (Stanton, 1965),
and the U.S. Geological Survey expedition in 1921
(LaRue, 1925). Although Turner and Karpiscak (1980)
did not quantify vegetation change, they did qualitatively
demonstrate an increase in vegetation in the postdam
Figure 3. A conceptual model of physical factors that affect riparian vegetation development along the river corridor. Thicker arrows
indicate a greater degree of effect on riparian vegetation. The closure of Glen Canyon Dam and the beginning of flow regulation of the
Colorado River through Grand Canyon in 1963 all but eliminated the mainstem sand supply to Grand Canyon. Also, dam operations are
now independent of tributary inputs of sediment. Taken together, dam-induced changes in both sand supply and flow have altered the
sedimentary processes that provide substrate for riparian vegetation.
108 The State of the Colorado River Ecosystem in Grand Canyon
fluvial sediment zone (up to 30,000 cfs) and in the
predam fluvial sediment zone (30,000–85,000 cfs; 2-yr
return period) (Topping and others, 2003). They noted
dense stands of tamarisk, coyote willow, and arrowweed
throughout the corridor, with desert broom (Baccharis
sarothroides), Bermudagrass (Cynodon dactylon), Emory seep
willow (Baccharis emoryi), and some cottonwoods (Populus
fremontii) along the postdam fluvial zone. The expansion
of coyote willow, tamarisk, and arrowweed within the
predam and postdam fluvial zones was predictable given
the creation of a stable water source and exposed land
area (figs. 4a and b). The expansion of emergents such
as cattails along the channel was also noted by Turner
and Karpiscak (1980). The higher bench (land above
>50,000 cfs) was found to be changing at a slower rate
and composed of predam high-bench species like acacia,
mesquite, sand dropseed (Sporobolus cryptandrus), and
Apache plume.
The qualitative findings of Turner and Karpiscak
(1980) are consistent with a land area change study
of selected sites in Marble Canyon and upper Grand
Canyon through the use of geographic information
systems (Waring, 1995). Waring estimated a 100%
increase in vegetation in the postdam and predam fluvial
sediment zones between 1965 and 1973 (56.5 acres vs.
108.6 acres (228,503 m2 vs. 439,420 m2)). Anderson and
Ruffner (1988) examined the predam high-bench terrace
vegetation and determined that this zone showed little
recruitment of new acacia or mesquite individuals. In
other words, this zone was not showing signs of replace-
ment of similar species. They hypothesized that the
vegetation was becoming more mature, with individuals
becoming larger and more closely spaced. Over time,
the vegetation in this zone would become less dense as
mature individuals died and were not replaced. Species
found in this predam bench would “move shoreward”
over time.
Period II: Inundation and
Habitat Reworking (1981–90)
In 1980, Lake Powell reached full pool elevation,
and operations over the next decade focused primarily
on water delivery and power generation. Because the
early 1980s was a wet period, causing a high-release
spill of 97,000 cfs in June 1983 (Martin, 1989), however,
Glen Canyon Dam was also operated to manage spring
inflows and protect the integrity of the dam. As a result,
the 1980s produced several years of releases greater
than 20,000 cfs for portions of the year (U.S. Geological
Survey, 2005), with higher releases occurring in spring
to reduce the frequency of spills. Median discharge for
the decade was 15,900 cfs (Topping and others, 2003),
approximately 32% greater than the median releases
of the 1970s. Fine-sediment erosion and export from
the Marble Canyon and upper Grand Canyon reaches
occurred in this decade (Topping and others, 2000;
Schmidt and others, 2004). Sediment export exceeded
inputs during these events, resulting in an overall loss of
sediment in the system.
The higher peak and median discharge presented
situations of sustained inundation of riparian vegetation
along the channel and increased water-table elevations
that promoted expansion of woody vegetation in this
decade (figs. 4b and c). The peak flow likely redistributed
seeds from the predam flood-plain surfaces, which may
have promoted establishment and growth of acacia and
mesquite within the lower elevation benches. Waring
(1995) detected a 13% decrease in vegetation occupy-
ing area below 50,000-cfs discharge. Waring showed
an increase in vegetated area in the higher elevation
benches for 1984 compared with 1975; measurements of
mesquite in the predam flood plain by other researchers,
however, did not indicate a growth response to the flood
events of the 1980s (Anderson and Ruffner, 1988).
Period III: Low Fluctuating
Flows and Experimentation
(1991–present)
River Flows
The operation of Glen Canyon Dam since 1991 has
focused on meeting water allocation requirements, pro-
ducing power, and complying with environmental con-
straints designed to minimize the effects of Glen Canyon
Dam on the erosion of recreational and archaeological
sites and on the deterioration of habitats for endangered
species, particularly humpback chub (Gila cypha) (U.S.
Department of the Interior, 1995). During this period,
flows have been further stabilized, not varying more
than 8,000 cfs daily, though median annual flows have
decreased only 15% when compared with those of the
1980s (13,500 cfs vs. 15,900 cfs) (Topping and others,
2003). The frequency of high peak flows has diminished
with two experimental high-flow events of 41,000 cfs
and 45,000 cfs occurring in November 2004 for 2.5 d
and in March 1996 for 7 d (Webb and others, 1999). As
a result, since 1991 the active riparian zone has been
reduced to 45,000-cfs water-surface elevation. Smaller,
shorter duration spikes of up to 31,000 cfs occurred in
Riparian Vegetation and Associated Wildlife 109
Figure 4. A. Aerial photograph from 1965 at RM 55.4 showing sparse vegetation occupying the sandbar and beginning to expand into
the postdam fluvial zone. B. Aerial photograph from 1984 of RM 55.4 showing effects of flooding, which scoured low-lying riparian
habitats. Vegetation occupies larger expanses along the shoreline and upslope from the river. C. Aerial photograph from 1994 of RM
55.4 showing expansion of riparian and marsh vegetation since 1984 on the sandbar and along the shoreline (source: U.S. Geological
Survey file photographs).
A.
B.
C.
110 The State of the Colorado River Ecosystem in Grand Canyon
1997 and 2000 (Schmidt and others, 2004). Peak flows
(flows >120,000 cfs) have also been reduced by more
than 50% of mean long-term high flows experienced
before 1963 (Topping and others, 2003). As a result, the
active riparian zone in Grand Canyon has contracted in
width to that area below approximately 50,000-cfs water-
elevation stage, with greatest change occurring below
30,000 cfs. The higher elevation bench (>50,000 cfs)
persists through yearly rainfall events and is little affected
by operations compared to predam hydrology.
Response in the Plant Community
This 14-yr period of stable but fluctuating flows
and experimentation resulted in the expansion of vegeta-
tion into open areas, including into campsites and chan-
nel margins within the active riparian zone (Kearsley
and others, 1994; Webb and others, 2002); the reduction
of marsh habitat associated with eddy return channels
(i.e., backwaters) (Stevens and others, 1995; Goeking and
others, 2003); and most recently (since 2002), a reduction
of vegetative cover in low channel positions (Kearsley,
2004b). The community as a whole has become more
mature, providing complex habitat for riparian breeding
birds. With the onset of the drought in 2000 there has
also been a reduction in cover of annual and perennial
grasses in areas located above flows of 35,000-cfs eleva-
tion (Kearsley, 2004b).
Forty-one percent of camping sites surveyed
between 1983 and 1991 were determined to be unusable
because of vegetation overgrowth (Kearsley and others,
1994). Vegetation expansion into campsites occurred
in reaches that are classified as wide reaches (Schmidt,
1990) or in areas where more sediment is available for
plant establishment. Kaplinski and others (2005) dis-
cussed trends in vegetation expansion into campsites
since 1991; their findings are summarized in chapter
12 of this report. Vegetation expansion between the
high-water periods of 1984 and 1992 was supported
by Waring (1995), who showed expansion during this
period at a percentage of change similar to that which
occurred with initial dam closure through 1973. Waring
(1995) speculated that the rate of vegetation expansion
increased during the early 1990s with the implementa-
tion of interim operating criteria. Expansion within
the zone between shoreline and up to the 50,000-cfs
water-surface elevation included the establishment by
nonnative plants and pioneer species like camel thorn
(Alhagi maurorum) and clonal growth by woody vegetation,
including arrowweed and coyote willow. This vegetation
expansion resulted in an increase in riparian bird habitat
(see chapter 7, this report). Vegetation expansion was
greatest in channel margin habitats used primarily by
wildlife and was least in sites adjacent to rapids associ-
ated with debris fans where disturbance was more likely
to occur (Melis and others, 1995; Waring, 1995; Webb
and others, 2002).
Debris fan-eddy complexes (Schmidt and Rubin,
1995) are geomorphic features that support fluvial marsh
habitat, primarily because they are low-velocity habitats
that accumulate silt and clay fractions (Schmidt and
Rubin, 1995; Stevens and others, 1995). Daily inunda-
tion frequency, soil texture, and distance from the dam
influence marsh locations and assemblages (Stevens
and others, 1995). Wet-marsh constituents like cattails,
sedges, and common reed (Phragmites australis) are found
in sites with increased inundation frequency, while drier
marsh-associated species like tamarisk, arrowweed,
horsetails (Equisetum sp.), and willows (Salix sp.) are associ-
ated with lower inundation frequencies. Interim operat-
ing criteria, initiated in 1991 and followed by the MLFF
alternative in 1996, reduced inundation frequency. This
change is coincident with a reduction in wet-marsh habi-
tat since 1991 (Stevens and others, 1995; Kearsley and
Ayers, 1996).
Geomorphic Effects
The effect of geomorphology on plant assemblages
is illustrated in marsh plots that were surveyed in the
mid-1990s. Narrow reaches of the river such as that of
Marble Canyon experienced losses of marsh patches,
which correspond to a reported loss of sediment in this
reach during the same time (Kearsley and Ayers, 1996;
Schmidt and others, 2004). Wider reaches found near
the Little Colorado River and in western Grand Canyon
that have greater sediment-storage capacities showed
gains and losses of marsh patches during these same
years in the mid-1990s. The variability in the number
of patches within these reaches may reflect local sedi-
ment inputs from ungaged tributaries as well as inputs
from the Little Colorado River. For these same years, a
drying trend (i.e., plants encountered were associated
with lower moisture gradients) was noted for riparian
plants (Kearsley and Ayers, 1996). Reduced numbers of
marshes support a hypothesis that interim flows reduced
inundation frequency and that species encountered were
more likely to be associated with a lower moisture gradi-
ent. Alternatively, the change may reflect infilling and
riparian community succession. Marsh census numbers
since 1995 are not available, but geomorphic studies of
debris fan-eddy complexes detected reduction in back-
waters from 1984 to 2000 (Goeking and others, 2003). A
reduction in these sites may be an indication that fine-
Riparian Vegetation and Associated Wildlife 111
sediment habitats were also declining during this period.
Sediment export, particularly silt and clay, and reduced
fluctuations likely reduced the area of cover represented
by wet-marsh species in the 1990s through 2001.
Hydrology and Climatic Influences
Reservoir levels, yearly operations, and local pre-
cipitation affect riparian vegetation growth and develop-
ment within all vegetation zones along the river corridor.
A persistent, basinwide drought was identified in July
2000 by the National Drought Mitigation Center (www.
drought.unl.edu/dm/archive/2000/drmon0725.htm,
accessed February 8, 2005). Since then, inflows to Lake
Powell have been below average, leading to drawdown
of both Lake Powell and Lake Mead (see chapter 4, this
report). As a result, Glen Canyon Dam has released the
minimum amount of water needed to meet delivery
requirements, which is 8.23 million acre-feet (10,148
million m3). While overall volume of delivery has been
reduced, monthly median flows have not changed appre-
ciably. Beginning in 2002, the months from January
to March and from June to August are dominated by
power-generation flows in summer months and fish sup-
pression measures in winter months (U.S. Department
of the Interior, 2004a). Fluctuations vary from 5,000 to
20,000 cfs daily in the winter months and by 8,000 cfs
daily in summer, with base flow being approximately
10,000–12,000 cfs. Other months have lower volumes
allocated with corresponding lower base flow and
reduced daily range (e.g., 5,000–10,000 cfs in April). The
abrupt shifts in monthly volumes in April and Septem-
ber expose areas in spring for plant establishment but
are soon followed by high summer fluctuations in June.
By September, the area inundated to 17,000-cfs water-
surface elevation is often sparsely vegetated and reduced
in sediment (Kearsley, 2004b). Precipitation variability
and operational shifts of Glen Canyon Dam resulted in
continued changes in the vegetation cover, abundance,
and density along the river corridor.
Recent Monitoring Results
Between 2001 and 2003, riparian vegetation was
affected both by changes in dam operations and by a
persistent drought. The summers of 2002 and 2003 had
higher daily minimum flows than 2001, and these years
also had winter (January to March) discharges that varied
from 5,000 cfs to 20,000 cfs. Vegetation volume (a sur-
rogate for structure) in the active riparian zone (5,000 cfs
to 45,000 cfs) responded markedly each year. Between
2001 and 2002 volume decreased by 15% but had
recovered by approximately the same amount between
2002 and 2003 (fig. 5) (Kearsley, 2004a). The recovery
was attributable to the operational change that took
place in January 2003. How these volumes may influence
riparian bird density or abundances between years is not
known. In contrast, vegetation at higher water-surface
elevations (>45,000 cfs) changed little between years (fig.
5) (Kearsley, 2004a). Vegetation at higher water-surface
elevations may respond more to localized precipitation
events than to dam operations (Kearsley, 2004a). In
general, operations had the greatest effect on vegetation
located below the 35,000-cfs water-surface elevation.
Measures of plant abundance, species richness,
diversity, and distribution all showed a decline since
2001 (Kearsley, 2004b). Operations and local precipita-
tion differentially affected plants along the elevational
gradient. Herbaceous annuals and perennials like cheat
grass (Bromus tectorum), sand dropseed, and spiny aster
(Chloracantha spinosa) located above 35,000-cfs water-
surface elevation were affected by yearly precipitation
and showed the greatest decline in cover (fig. 6a). The
effect of the drought on higher elevation plants was also
evident when species composition was examined. Spe-
cies richness changed significantly at sites at 45,000 cfs
and 60,000 cfs (fig. 6b) (Kearsley, 2004b); the change
was associated mostly with a loss of annual and rarely
encountered plant species. Increased summer precipi-
tation in 2003 was responsible for increases in species
richness in both of these elevations (fig. 6b). Annuals
appearing in wetter years likely contributed to these
increases (Kearsley, 2004b). Compositional shifts did not
Figure 5. Change in vegetation volume (a surrogate measure
of structure) from 2001 to 2003 in the riparian zone along the
Colorado River in Grand Canyon at surface-water elevations
above and below 50,000-cfs surface elevation. Figure modified
from Kearsley (2004a).
112 The State of the Colorado River Ecosystem in Grand Canyon
occur for zones below 35,000 cfs, and these data, in com-
bination with vegetation volume measurements, suggest
that changes in this zone were associated with increased
growth of established woody species rather than with
recruitment or mortality. Moreover, reductions in cover,
richness, and diversity in lower water-surface elevation
locations (<25,000 cfs) were caused by sediment loss,
inundation, or scour that coincided with operational
changes in January 2003. Dam operations influenced
changes in vegetation beyond recorded discharge lev-
els, possibly up to approximately 15,000 cfs, while local
precipitation appeared to have a greater influence, in the
short term, on vegetation above the 35,000-cfs water-sur-
face elevation.
Since the 1990s, reservoir levels, yearly operations,
and local precipitation have affected riparian vegeta-
tion growth and development along the river corridor in
Grand Canyon (figs. 3, 5, and 6). Though other factors
do affect riparian vegetation dynamics, these variables
appear to be significant drivers in riparian vegetation
development and change. The riparian zone in Grand
Canyon has contracted shoreward as flows have stabi-
lized. Riparian vegetation at water-surface elevations
up to about 15,000 cfs above daily maximum discharge
responds to operational changes. Vegetation below a
water-surface elevation of 45,000 cfs has become denser
and has expanded into open sites, including campsites.
Vegetation cover and richness at low water-surface
elevation locations (below 20,000 cfs) are most directly
affected by dam operations. The fluvial marsh com-
munity is the most responsive of vegetative communi-
ties within Grand Canyon to changes in hydrology and
sediment supply. Not surprisingly, cover and richness
decline during flow fluctuations that promote scour and
sediment export (Kearsley, 2004b). At the same time
these flows can increase inundation frequency at higher
elevations and can shift plant composition to more water-
tolerant species like cattails and rushes as a result of
annual changes in operations (Stevens and others, 1995).
Essentially, the riparian area has declined quantitatively in
some aspects (less spatial coverage, fewer numbers of spe-
cies) and has changed qualitatively (denser, more mature).
Riparian Vegetation as
Terrestrial Habitat
As stated in the Introduction of this chapter, ripar-
ian communities in the Southwestern United States play
an important role as wildlife habitat (Carothers and
Brown, 1991; Farley and others, 1994; Skagen and oth-
Figure 6. A. Percent change in vegetation covers at five water-
surface elevations (cubic feet per second) between 2001 and 2003
in the riparian zone along the Colorado River in Grand Canyon.
Cover has declined in all zones since 2001. Discharge had the
greatest effect on species richness at water-surface elevations
from 35,000 cfs to 15,000 cfs. Overall vegetation cover within the
riparian zone is not dense although cover is greatest at the 25,000-
cfs water-surface elevation, which corresponds with areas used by
campers and breeding birds. Figure modified from Kearsley (2004b).
B. Species richness in five water-surface elevations (cubic feet per
second) from 2001 to 2003 and change between years in the riparian
zone along the Colorado River in Grand Canyon. Species richness
is a measure of the total number of species found at each water-
surface level and is an indication of biodiversity. Species richness
in this case is measured from a scale of 0 to 10 with 10 being the
highest amount of richness. Overall species richness within the
riparian zone is greatest in those zones that are above the 25,000-
cfs water-surface elevation where vegetation is not directly
scoured by flow. Species richness in the upper water-surface
elevations is more affected by yearly precipitation. Species at
upper surface water elevations have declined since 2001 because
of drought conditions and include annual cheat grass (Bromus
tectorum) and perennial sand dropseed (Sporobolus cryptandrus).
Figure modified from Kearsley (2004b).
A.
B.
Riparian Vegetation and Associated Wildlife 113
ers, 1998; Stevens and Ayers, 2002). Along the Colorado
River, riparian birds have had a greater emphasis placed
on them than other types of wildlife in terms of monitor-
ing and are treated in a separate chapter of this report
(chapter 7). Wildlife other than endangered species
and birds has not been emphasized to date in the Glen
Canyon Dam Adaptive Management Program. For these
reasons, the following section addresses threatened and
endangered species that occupy riparian habitat marshes
and springs within the Colorado River ecosystem. Other
faunal constituents are briefly mentioned.
Threatened and
Endangered Species
Kanab Ambersnail
The Kanab ambersnail (Oxyloma haydeni ssp. kanaben-
sis) is a terrestrial succineid snail (fig. 7) associated with
wetland and spring vegetation on the Colorado Plateau.
The snail was listed as endangered in 1992 (England,
1992). Presently, the species is found at three locations:
Vaseys Paradise and Elves Chasm (a translocated popula-
tion), in Grand Canyon National Park, and private land
in southern Utah. Data presented here pertain to the
snails located at Vaseys Paradise.
Vaseys Paradise (figs. 8a–c) is a small patch of
spring-fed riparian vegetation at RM 31.8 (Stevens,
1990). Ambersnails are found in the vegetation, usually
associated with cardinal monkeyflower (Mimulus cardina-
lis) (fig. 8b), watercress (Rorippa nasturtium-aquaticum) (fig.
8c), and water sedge (Carex aquatilis). Ambersnail adults
overwinter and reproduce in spring. Recruitment into
the adult population takes place during summer and fall
(Stevens and others, 1998; Nelson, 2001).
Habitat
The greatest gains in habitat area, measured by
traditional land-survey methods, occurred between fall
2001 and fall 2002 when snail habitat at Vaseys Paradise
increased 23% in area (~2,374 ft² vs. ~3,103 ft² (220.6
m2vs. 288.4 m2)) (fig. 9a) (U.S. Geological Survey, unpub.
data, 1998–2004). Increases in measured habitat may
be attributable to low minimum flows in 2001 that
increased area for colonization by watercress, monkey-
flower, or other plants; however, watercress, which is a
species adapted to disturbance and that requires sus-
tained moisture, has decreased since 1998 because spring
discharges declined in association with the drought. At
the same time, monkeyflower increased (U.S. Geological
Survey, unpub. data, 1998–2004) and accounted for most
of the habitat increase measured between fall 2001 and
fall 2002. Monkeyflower, while still requiring moisture,
appears to be more tolerant of drier habitats. Water
sedge is patchily distributed in Kanab ambersnail habitat
and is a source of forage for bighorn sheep (Ovis canaden-
sis). As a site that provides a reliable source of vegetation
in a drought, the springs are now habitually visited by
bighorn sheep, resulting in vegetation used by the snails
being regularly trampled.
Snail Abundances
The number of snails has not changed significantly
since 1998. Fall numbers generally exceed spring num-
bers as would be expected with seasonal recruitment (fig.
9b; U.S. Geological Survey, unpub. data, 1998–2004).
Curiously, while habitat has increased, snail numbers
have not had a correlated increase. The lack of increase
in snail numbers may be associated with soil moisture,
shifts in plant-species composition, and mortality associ-
ated with trampling by bighorn sheep rather than with
the amount of habitat available.
Figure 7. Kanab ambersnail (Oxyloma haydeni ssp.
kanabensis), which is monitored at Vaseys Paradise
(photograph by Roy Averill-Murray, Arizona Game and
Fish Department).
114 The State of the Colorado River Ecosystem in Grand Canyon
Figure 8. A. Vaseys Paradise along the Colorado River, which is one of three locations known to support the endangered
Kanab ambersnail (Oxyloma haydeni ssp. kanabensis). Cardinal monkeyflower (Mimulus cardinalis) (B) and watercress (Rorippa
nasturtium-aquaticum) (C) are primary plant species associated with Kanab ambersnail (photographs by Jeff Sorensen, Arizona Game
and Fish Department).
Other Wildlife
Invertebrates
Invertebrate species, as in other ecosystems, account
for the greatest number of species found along the river.
There are several thousand invertebrate species from
over 200 families (Stevens and Ayers, 2002; Lightfoot
and others, 2004), including scorpions, spiders, flies,
ants, moths, and butterflies. Surveys for invertebrates
conducted over the past several years identified either
range extensions for species (e.g., butterflies: Arizona
powdered-skipper (Systacea xampa), piute agave skipper
(Agathymus alliea piute), desert marble (Euchloe lotta), and
desert elfin (Callophrys fotis)) or species not previously
known to exist (e.g., moth, Schinia immaculate) (Stevens
and Ayers, 2002; Pogue, 2004). Invertebrate composition
associated with higher elevation riparian vegetation con-
sists largely of native invertebrate taxa, while the lower
elevation riparian vegetation includes a mix of native
and nonnative invertebrate species (Lightfoot and oth-
ers, 2004). The mix of species is not unexpected because
vegetation in this area consists of a more pronounced
mix of native and nonnative plants.
Amphibians, Mammals,
and Reptiles
Surveys for mammals, reptiles, and amphibians
have been sporadic (Carothers and Aitchison, 1976;
Warren and Schwalbe, 1986; Frey, 2003). Past surveys
found 14 mammal species, 16 reptile species, and 4
amphibian species along the corridor. Amphibians
of special concern are detailed in the accompanying
text box. Only the deer mouse (Peromyscus manicula-
tus) is restricted to the riparian zone (Frey, 2003; U.S.
Department of the Interior, 2004b). Larger mammals
include beavers (Castor canadensis), coyotes (Canis latrans),
bighorn sheep, mule deer (Odocoileus rafinesque), moun-
tain lions (Puma concolor), and bobcats (Lynx rufus) (U.S.
Department of the Interior, 2004b). Mountain lions
and bobcats are seen infrequently. Of these mammals,
beavers appear to have expanded their numbers since
the 1960s (Carothers and Brown, 1991) in association
with riparian vegetation expansion. Beavers appear to
be relatively evenly distributed throughout the river cor-
ridor (U.S. Department of the Interior, 2004b).
Discussion and Future
Research Needs
A long-term data set associated with changes in
riparian vegetation is lacking for the Colorado River
ecosystem. The data that are provided here are results
associated with specific research questions of 2- to 3-yr
duration rather than monitoring to detect trends. Trend
detection associated with riparian vegetation requires
local and regional scale monitoring because local and
regional hydrology and geomorphology affect the ripar-
ian zone. The long-term goal for monitoring in the
A. B. C.
Riparian Vegetation and Associated Wildlife 115
Figure 9. A. Change in the area of Kanab ambersnail (Oxyloma haydeni ssp. kanabensis) habitat (ft2) at Vaseys Paradise from 1998
through 2004 from spring and fall surveys. B. Estimated snail abundances at Vaseys Paradise from 1998 through spring 2004 from spring
and fall surveys at Vaseys Paradise (U.S. Geological Survey, unpub. data).
A. B.
Colorado River should be to use yearly data on cover,
species richness, and diversity in concert with semi-
decadal vegetation mapping data to discern operational
versus climatic effects on the riparian vegetation. Large-
scale trend detection at the reach or regional scale will be
able to provide information about changes in vegetated
area, increases and decreases in vegetation patches, and
the vegetation classes that have changed the most. The
mapping data can also be integrated with bird monitor-
ing, and other faunal resource monitoring that may be
developed, to produce information about habitat quality.
Habitat quality affects invertebrate communities that are
food resources for both riparian breeding birds and fish
communities along the river corridor.
Linkages among vegetation, faunal assemblages,
and habitat quality are needed before comprehensive
assessments of the riparian zone for the Colorado River
can be made. One step toward a comprehensive assess-
ment has been initiated by an inventory of invertebrates
along the corridor. The intent of the inventory is to
determine what types of invertebrates exist along the
corridor, whether or not any species can be used to indi-
cate environmental conditions, and in what quantities
these potential indicators occur. Both quantity and type
of invertebrates encountered are affected by vegetation
assemblages. The challenge associated with riparian veg-
etation is to determine how dam operations affect plant
species assemblages and densities, which, in turn, influ-
ence habitat quality and food resources for vertebrates
found along the river corridor.
In the long term, continued loss of sediment along
the shoreline and changes in the size fraction of the
substrate will reduce available colonizing substrate and
affect subsequent species establishment (i.e., marsh com-
munities may shift to constituents that persist in coarser
substrates) (Stevens and others, 1995). At elevations
above the 20,000-cfs flow but still in the active riparian
zone, woody vegetation is becoming more mature and
less diverse in association with the reduction in high
flow frequency. Under current operations (modified low
fluctuating flows), precipitation affects vegetation above
the 35,000-cfs water-surface elevation more than opera-
tions do (Kearsley, 2004 a, b). Remnants of the predam
high-water riparian zone have remained relatively
unchanged, depending on seasonal precipitation rather
than on yearly hydrology for its maintenance. Implica-
tions for these changes within the river corridor include
the potential reduction in numbers of some riparian
bird species that depend on wetter marsh-plant species
for nesting or food resources, though other riparian bird
species may benefit from the more mature, dense habitat
(Anderson and Ohmart, 1984; Farley and others, 1994).
Furthermore, continued loss of campsite area associated
with vegetation expansion may occur. The trade-offs
between recreation and wildlife habitats are value-based
management decisions that the adaptive management
program will have to address. Some of these trends could
be ameliorated through higher frequency disturbances
up to and above 45,000-cfs discharges, but these events
should be timed to coincide with sufficient sediment
inputs and existing system supply.
116 The State of the Colorado River Ecosystem in Grand Canyon
Declining Riparian Species:
Leopard Frogs in Grand
Canyon and Glen Canyon
Charles Drost
Amphibians have been relatively neglected in stud-
ies of plants and animals and of the effects of dams in
Grand and Glen Canyons. Amphibians were surveyed
along the Colorado River in Glen Canyon before the
construction of Glen Canyon Dam (Woodbury, 1959),
but extensive surveys were not conducted in Grand
Canyon until well after the completion of the dam (e.g.,
Aitchison and others, 1974; Suttkus and others, 1976).
Over the last 15 yr there has been increasing recognition
and concern about declines in amphibian populations
in areas throughout the world (Wyman, 1990; Wake,
1991; Vial and Saylor, 1993). The cause of many of
these declines is unknown, but they have even occurred
in national parks and other protected areas, suggest-
ing that causes are widespread regional ones (Blaustein
and Wake, 1990). We describe here the current status of
amphibians in the Colorado River corridor of Grand
Canyon National Park and Glen Canyon National
Recreation Area based on extensive surveys conducted
over the last 10 yr throughout Glen and Grand Canyons,
from the uppermost end of Lake Powell to the upper end
of Lake Mead.
Surveys found healthy, widespread populations
of two species of toads (Woodhouse’s (Bufo woodhou-
sii) and red-spotted (B. punctatus)); the canyon treefrog
(Hyla arenicolor); and the tiger salamander (Ambystoma
tigrinum). Northern leopard frogs (Rana pipiens) (fig. 1),
on the other hand, have declined substantially, paral-
leling losses reported in other areas of western North
America. Leopard frog populations have disappeared
from 70% of sites where they were formerly found and
have declined in numbers at some sites where they still
occur (fig. 2). Some of the losses are of riverside popula-
tions in Glen Canyon, which were inundated by Lake
Powell (Drost and Sogge, 1993); however, other popula-
tions have been lost from side canyons off of the lake
that are not impacted directly by inundation. Currently,
seven leopard frog populations are known to occur in
side canyons of Lake Powell, concentrated around the
Escalante River area of the lake. The status of the spe-
cies at some sites is uncertain. Earlier surveys found frogs
as far upstream as Dark Canyon, near Hite, but none
have been seen there recently. A small population in
Wilson Creek off the San Juan River area of the lake has
not been seen since 1994.
The predam distribution of northern leopard frogs
in the Grand Canyon reach of the Colorado River is
unknown because of the lack of early surveys. Compi-
lation of reports from more recent surveys shows that
northern leopard frogs occurred at least as far down-
stream as Cardenas Creek (RM 71) along the river corri-
dor and in side canyons as far as Bright Angel Creek and
Kanab Creek. In spite of intensive searches of potential
habitat along the river and in side canyons with peren-
nial streams, the only known remaining population is at
a spring-fed marsh between Glen Canyon Dam and Lees
Ferry. This population was discovered in 1992 (Drost and
Sogge, 1993) and has experienced wide year-to-year fluc-
tuations in numbers. The most recent surveys indicate
a sharp decline in population size, with only two adult
individuals found in 2004. Marsh vegetation at the site
has become very dense, reducing areas of open water,
and this reduction may be an important factor in the
decline of this population.
Although survey work is continuing, it is clear, based
on the historical record, that there has been a severe
contraction of the northern leopard frog’s range in both
Glen Canyon and Grand Canyon. A major concern
for remaining populations of frogs is that most or all of
them are now effectively isolated from each other. No
other extant populations have been found along the river
below Glen Canyon Dam, so the population below Glen
Figure 1. Northern leopard frog (Rana pipiens) (photograph by
Charles Drost, U.S. Geological Survey).
Riparian Vegetation and Associated Wildlife 117
Figure 2. Northern leopard frog (Rana pipiens) sites in Glen Canyon above Glen Canyon Dam (A) and the corridor of the
Colorado River and its tributaries below Glen Canyon Dam (B). Northern leopard frog numbers have declined substantially,
paralleling losses reported in other areas of Western North America. Leopard frog populations have disappeared from
70% of sites where they were formerly found and have declined in numbers at some sites where they still occur.
A.
B.
118 The State of the Colorado River Ecosystem in Grand Canyon
Canyon Dam is completely isolated. In addition, no
frogs have been found along the shores of Lake Powell.
The combination of deep lake waters, lack of vegetation
cover, and large numbers of predatory fish in the lake prob-
ably prevents any movement of frogs among side canyons.
One unexpected, positive finding of the surveys was
the discovery of a previously unknown population of a
second leopard frog species in western Grand Canyon.
In spring 2004, small numbers of leopard frogs were
found in a pool in Surprise Canyon (RM 248) (Gelczis
and Drost, 2004). The frogs are clearly different from
the northern leopard frogs found farther upstream.
Genetic studies of the population are still in progress, but
the frogs are apparently the lowland leopard frog (Rana
yavapaiensis). This location represents a significant north-
ward extension in range for this relatively rare species.
There are potential threats at the site in the form of non-
native predatory fish and crayfish, but this new popula-
tion appears to be healthy and thriving.
References
Aitchison, S., Carothers, S.W., Karpiscak, M.M.,
Theroux, M.E., and Tomko, D.S., 1974, An ecologi-
cal survey of the Colorado River and its tributar-
ies between Lees Ferry and the Grand Wash Cliffs:
Flagstaff, Museum of Northern Arizona, Department
of Biology, contract report nos. CX821040079 (final
report) and CX821500007 (progress report), 235 p.
Blaustein, A.R., and Wake, D.B., 1990, Declining
amphibian populations: a global phenomenon?:
Trends in Ecology and Evolution, v. 5, p. 203–204.
Drost, C.A., and Sogge, M.K., 1993, Survey of an
isolated northern leopard frog population along the
Colorado River in Glen Canyon National Recreation
Area: Flagstaff, Ariz., National Park Service, Cooper-
ative Park Studies Unit, Northern Arizona University,
Northern Arizona University Report, 34 p.
Gelczis, L., and Drost, C., 2004, A surprise in Surprise
Canyon: a new amphibian species for Grand Canyon
National Park: Nature Notes, Fall 2004, p. i–iii.
Suttkus, R.D., Clemmer, G.H., Jones, C., and Shoop,
C.R., 1976, Survey of fishes, mammals and herpe-
tofauna of the Colorado River in Grand Canyon:
Colorado River Research Program final report, tech-
nical report no. 5, U.S. Department of the Interior,
National Park Service.
Vial, J.L., and Saylor, L., 1993, The status of amphib-
ian populations. Working document no. 1, Declining
Amphibian Populations Task Force: Gland, Switzer-
land, International Union for the Conservation of
Nature and Natural Resources, 98 p.
Wake, D.B., 1991, Declining amphibian populations:
Science, v. 253, p. 860.
Woodbury, A.M., 1959, Amphibians and reptiles of Glen
Canyon. Ecological studies of flora and fauna in Glen
Canyon: University of Utah Anthropological Papers
No. 40, University of Utah (Glen Canyon Series 7).
Wyman, R., 1990, What’s happening to the amphib-
ians?: Conservation Biology, v. 4, p. 350–352.
Contact Information:
Charles A. Drost
Research Zoologist
U.S. Department of the Interior
U.S. Geological Survey
Southwest Biological Science Center
Flagstaff, AZ
charles_drost@usgs.gov
Riparian Vegetation and Associated Wildlife 119
References
Anderson, B.W., and Ohmart, R.D., 1984, Vegetation
management study for the enhancement of wild-
life along the lower Colorado River: Boulder City,
Nev., report for U.S. Bureau of Reclamation, Lower
Colorado Region, 529 p.
Anderson, L.S., and Ruffner, G.A., 1988, Effects of the
post-Glen Canyon Dam flow regime on the old high
water zone plant community along the Colorado River
in Grand Canyon: Glen Canyon Environmental Stud-
ies, executive summaries of technical reports, NTIS
no. PB-183504/AS.
Baker, W.L., 1989, Macro- and micro-scale influences on
riparian vegetation in western Colorado: Annals of the
Association of American Geographers, v. 79, p. 65–78.
Bendix, J., 1994, Among-site variation in riparian veg-
etation of the southern California transverse ranges:
American Midland Naturalist, v. 132, p. 136–151.
Brown, D.E., 1982, Biotic communities of the American
Southwest-United States and Mexico: Desert Plants, v.
4, nos. 1–4.
Busch, D.E., and Smith, S.D., 1995, Mechanisms associ-
ated with decline of woody species in riparian ecosys-
tems of the southwestern U.S.: Ecological Monographs,
v. 65, p. 347–370.
Carothers, S.W., and Aitchison, S.W., eds., 1976, An
ecological survey of the riparian zones of the Colorado
River between Lees Ferry and the Grand Cliffs: final
report. Grand Canyon National Park Technical Report
Series No. 10: Washington, D.C., U.S. Department of
the Interior, National Park Service.
Carothers, S.W., and Brown, B.T., 1991, The Colorado
River through Grand Canyon: natural history and
human change: Tucson, University of Arizona Press.
Clover, E.U., and Jotter, L., 1944, Floristic studies in the
canyon of the Colorado and tributaries: American
Midland Naturalist, v. 32, p. 591–642.
Darrah, W.C., 1948, Beaman, Fennemore, Hillers,
Dellenbaugh, Johnson and Hattan: Utah Historical
Quarterly, v. 16, p. 491–503.
Decamps, H., 1993, River margins and environmental
change: Ecological Applications, v. 3, p. 441–445.
England, J.L., 1992, Endangered and threatened wildlife
and plants; final rule to list the Kanab ambersnail as
endangered: Federal Register, v. 57, no. 75, p. 13657–
13661.
Farley, G.H., Ellis, L.M., Stuart, J.N., and Scott, N.J., Jr.,
1994, Avian species richness in different-aged stands
of riparian forest along the middle Rio Grande, New
Mexico: Conservation Biology, v. 8, p. 1098–1108.
Frey, J., 2003, Grand Canyon inventory and monitoring
2003 mammal report, in Kearsley, M.J.C., ed., Inven-
tory and monitoring of terrestrial riparian resources
in the Colorado River corridor of Grand Canyon: an
integrative approach: Annual report from Northern
Arizona University submitted to Grand Canyon Moni-
toring and Research Center, U.S. Geological Survey,
Flagstaff, Ariz., 198 p.
Gecy, J.L., and Wilson, M.V., 1990, Initial establishment
of riparian vegetation after disturbance by debris flows
in Oregon: American Midland Naturalist, v. 123, p.
282–291.
Goeking, S.A., Schmidt, J.C., and Webb, M.K., 2003,
Spatial and temporal trends in the size and number
of backwaters between 1935 and 2000, Marble and
Grand Canyons, Arizona: Final report from Utah State
University to Grand Canyon Monitoring and Research
Center, U.S. Geological Survey, Flagstaff, Ariz.
Jacobs, T.C., and Gilliam, J.W., 1985, Riparian losses of
nitrate from agricultural drainage waters: Journal of
Environmental Quality, v. 14, p. 540–551.
Johnson, R.R., Haight, L.T., and Simpson, J.M., 1977,
Endangered species vs. endangered habitats—a
concept, in Johnson, R.R., and Jones, D.A., Jr., tech.
coords., Importance, preservation and management of
riparian habitat: a symposium: USDA Forest Service
General Technical Report RM-43, p. 68–79.
Kaplinski, M., Hazel, J.E., and Parnell, R., 2005, Camp-
site area monitoring in the Colorado River ecosystem:
1998 to 2003: Final report from Northern Arizona
University submitted to Grand Canyon Monitoring
and Research Center, Flagstaff, Ariz., 39 p.
Kearsley, L.H., Schmidt, J.C., and Warren, K.D., 1994,
Effects of Glen Canyon Dam on Colorado River sand
deposits used as campsites in Grand Canyon National
Park, USA: Regulated Rivers: Research & Manage-
ment, v. 9, p. 137–149.
120 The State of the Colorado River Ecosystem in Grand Canyon
Kearsley, M.J.C., 2004a, Vegetation structure, in Kearsley,
M.J.C., ed., Inventory and monitoring of terrestrial
riparian resources in the Colorado River corridor
of Grand Canyon: an integrative approach: Report
from Northern Arizona University submitted to
Grand Canyon Monitoring and Research Center, U.S.
Geological Survey, Flagstaff, Ariz., 218 p.
Kearsley, M.J.C., 2004b, Vegetation dynamics, in
Kearsley, M.J.C., ed., Inventory and monitoring of
terrestrial riparian resources in the Colorado River
corridor of Grand Canyon: an integrative approach:
Report from Northern Arizona University submitted to
Grand Canyon Monitoring and Research Center, U.S.
Geological Survey, Flagstaff, Ariz., 218 p.
Kearsley, M.J.C., and Ayers, T.J., 1996, The effects of
interim flows from Glen Canyon Dam on riparian veg-
etation in the Colorado River corridor, Grand Canyon
National Park, Arizona: Final report from Northern
Arizona University submitted to the Grand Canyon
Science Center, Grand Canyon National Park, Grand
Canyon, Ariz., 40 p., appendices.
LaRue, E.C., 1925, Water power and flood control of
the Colorado River below Green River, Utah: U.S.
Geological Survey Water Supply Paper 556, 176 p.
Lightfoot, D., Brantley, S., Cobb, N., and Delph, R.,
2004, Arthropod communities across three hydrologic
riparian zones along the Colorado River in Grand
Canyon, in Kearsley, M.J.C., ed., Inventory and moni-
toring of terrestrial riparian resources in the Colorado
River corridor of Grand Canyon: an integrative
approach: Report from Northern Arizona University
submitted to Grand Canyon Monitoring and Research
Center, U.S. Geological Survey, Flagstaff, Ariz., 218 p.
Lock, P.A., and Naiman, R.J., 1998, Effects of stream
size on bird community structure in coastal temper-
ate forests of the pacific northwest, U.S.A.: Journal of
Biogeography, v. 25, p. 773–782.
Martin, R., 1989, A story that stands like a dam—Glen
Canyon and the struggle for the soul of the west: New
York, Henry Holt and Company, 345 p.
Melis, T.S., Webb, R.H., Griffiths, P.G., and Wise, T.W.,
1995, Magnitude and frequency data of debris flows
in Grand Canyon National Park and vicinity, Arizona:
U.S. Geological Survey Water Resources Investigations
Report 94-4214.
Naiman, R.J., 1992, New perspectives for watershed
management: balancing long-term sustainability with
cumulative environmental change, in Naiman, R.J.,
ed., Watershed management: balancing sustainability
and environmental change: New York, Springer-Ver-
lag, p. 3–11.
Naiman, R.J., and Decamps, H., eds., 1990, The ecol-
ogy and management of aquatic-terrestrial ecotones:
Carnforth, England, Parthenon, and Paris, UNESCO.
National Drought Mitigation Center, 2005, Why the
drought monitor: Lincoln, Nebr., www.drought.unl.
edu/dm/about.html, accessed February 8, 2005.
National Research Council, 2002, Riparian areas: func-
tions and strategies for management: Washington,
D.C., National Academy Press, 436 p.
Nelson, C., 2001, Life history of the Kanab ambersnail
on native and nonnative host plants in Grand Canyon,
Arizona: Flagstaff, Northern Arizona University, M.S.
thesis, 98 p.
Nilsson, C., 1992, Conservation management of riparian
communities, in Hansson, L., ed., Ecological principles
of nature conservation: London, Elsevier Applied Sci-
ence, p. 352–372.
Nilsson, C., Grelsson, G., Johansson, M., and Sperens, U.,
1989, Patterns of plant species richness along river-
banks: Ecology, v. 70, p. 77–84.
Riparian Vegetation and Associated Wildlife 121
Pogue, M.G., 2004, A new species of Schinia HĦber from
riparian habitats in the Grand Canyon (Lepidoptera:
Noctuidae: Heliothinae): Zootaxa, v. 788, p. 1–4.
Saab, V., 1999, Importance of spatial scale to habitat
use by breeding birds in riparian forests: a hierarchical
analysis: Ecological Applications, v. 9, p. 135–151.
Schlosser, I.J., and Karr, J.R., 1981, Water quality in agri-
cultural watershed: impact of riparian vegetation during
base flow: Water Resources Bulletin, v. 17, p. 233–240.
Schmidt, J.D., 1990, Recirculating flow and sedimenta-
tion in the Colorado River in Grand Canyon, Arizona:
Journal of Geology, v. 98, p. 709–724.
Schmidt, J.D., and Rubin, D.M., 1995, Regulated stream-
flow, fine-grained deposits, and effective discharge in
canyons with abundant debris fans, in Costa, J.E., Miller,
A.J., Potter, K.W., and Wilcock, P.R., eds., Natural and
anthropogenic influences in fluvial geomorphology:
Washington, D.C., American Geophysical Union, Geo-
physical Monograph Series, v. 89, p. 177–194.
Schmidt, J.D., Topping, D.J., Grams, P.E., and Hazel, J.E.,
2004, System-wide changes in the distribution of fine
sediment in the Colorado River corridor between Glen
Canyon Dam and Bright Angel Creek, Arizona: Final
report from Utah State University to Grand Canyon
Monitoring and Research Center, U.S. Geological
Survey, Flagstaff, Ariz.
Skagen, S.K., Melcher, C.P., Howe, W.H., and Knopf,
F.L., 1998, Comparative use of riparian corridors and
oases by migrating birds in southeast Arizona: Conser-
vation Biology, v. 12, p. 896–909.
Stanton, R.B., 1965, Down the Colorado, edited by D.L.
Smith: Norman, Oklahoma University Press, 227 p.
Stevens, L.E., 1990, The Colorado River in Grand
Canyon: a guide (3d ed., 3d printing): Flagstaff, Ariz.,
Red Lake Books.
Stevens, L.E., and Ayers, T.J., 2002, Biogeographic pat-
terns among the nonnative flora and vertebrate fauna
of Grand Canyon, in Tillman, B., ed., Nonnative
species in the Sonoran Desert: Tucson, University of
Arizona Press.
Stevens, L.E., Meretsky, V.J., Nelson, C.B., and Sorensen,
J.A., 1998, The endangered Kanab ambersnail at
Vaseys Paradise, Grand Canyon, Arizona: 1997 final
report: U.S. Department of the Interior, Grand Canyon
Monitoring and Research Center report, Flagstaff, Ariz.
Stevens, L.E., Schmidt, J.C., Ayers, T.J., and Brown, B.T.,
1995, Flow regulation, geomorphology, and Colorado
River marsh development in the Grand Canyon, Ari-
zona: Ecological Applications, v. 5, p. 1025–1039.
Stromberg, J.C., 2001, Influence of stream flow regime
and temperature on growth rate of the riparian tree,
Platanus wrightii, in Arizona: Freshwater Biology, v. 46,
p. 227–239.
Stromberg, J.C., and Patten, D.T., 1991, Flood flows
and dynamics of Sonoran riparian forests: Rivers, v.
2, p. 221–235.
Stromberg, J.C., Tiller, R., and Richter, B., 1996, Effects
of groundwater decline on riparian vegetation of
semiarid regions: the San Pedro, Arizona: Ecological
Applications, v. 6, p. 113–131.
Toner, M., and Keddy, P., 1997, River hydrology and
riparian wetlands: a predictive model for ecological
assembly: Ecological Applications, v. 7, p. 236–246.
Topping, D.J., Rubin, D.M., and Vierra, L.E., Jr., 2000,
Colorado River sediment transport 1. Natural sediment
supply limitation and the influence of Glen Canyon
Dam: Water Resources Research, v. 36, p. 515–542.
122 The State of the Colorado River Ecosystem in Grand Canyon
Topping, D.J., Schmidt, J.C., and Vierra, L.E., Jr., 2003,
Computation and analysis of the instantaneous-dis-
charge record for the Colorado River at Lees Ferry,
Arizona—May 8, 1921, through September 30, 2000:
U.S. Geological Survey Professional Paper 1677, 118 p.
Turner, R.M., and Karpiscak, M.M., 1980, Recent
vegetation changes along the Colorado River between
Glen Canyon Dam and Lake Mead, Arizona: U.S.
Geological Survey Professional Paper 1132, 125 p.
U.S. Department of the Interior, 1995, Operation of Glen
Canyon Dam Final Environmental Impact Statement:
Salt Lake City, Utah, Bureau of Reclamation, 337 p.
U.S. Department of the Interior, 2004a, Draft supple-
mental environmental assessment proposed experi-
mental actions for water years 2005–2006 Colorado
River, Arizona, in Glen Canyon National Recreation
Area and Grand Canyon National Park: Salt Lake City,
Utah, Bureau of Reclamation, 28 p.
U.S. Department of the Interior, 2004b, Draft environ-
mental impact statement Colorado River manage-
ment plan, v. 1, Grand Canyon National Park, Grand
Canyon, Ariz., National Park Service, 244 p.
U.S. Geological Survey, 2005, Grand Canyon Monitoring
and Research Center, www.gcmrc.gov/products/flow_
data/flow_data.aspx, accessed February 8, 2005.
Waring, G.L., 1995, Current and historical riparian
vegetation trends in Grand Canyon, using multitempo-
ral remote sensing analyses of GIS sites: Final report,
National Park Service, 24 p.
Contact Information:
Barbara E. Ralston
General Biologist
U.S. Department of the Interior
U.S. Geological Survey
Southwest Biological Science Center
Flagstaff, AZ
bralston@usgs.gov
J
e
f
f
S
o
r
e
n
s
e
n
,
A
r
i
z
o
n
a
G
a
m
e
a
n
d
F
i
s
h
D
e
p
a
r
t
m
e
n
t
Warren, P.L., and Schwalbe, C.R., 1986, Lizards along
the Colorado River in Grand Canyon National Park:
possible effects of fluctuating river flows: National
Technical Information Service Report PB88-183538,
Glen Canyon Environmental Studies Report No.
26/87, 21 p.
Webb, R.H., Melis, T.S., and Valdez, R.A., 2002, Obser-
vations of environmental change in Grand Canyon,
Arizona: U.S. Geological Survey Water Resources
Investigations Report 02-4080.
Webb, R.H., Schmidt, J.C., Marzolf, G.R., and Valdez,
R.A., eds., 1999, The 1996 controlled flood in Grand
Canyon: Washington, D.C., American Geophysical
Union Geophysical Monograph, 367 p.
First page photograph credit: Jeff Sorensen, Arizona Game and Fish Department
Introduction
Riparian habitats, or vegetated areas along streams
and rivers, in the Western United States typically support
a disproportionately large number of birds compared
to adjacent nonriparian habitats, both in terms of bird
abundance and the number of species present (also
known as species richness). The Grand Canyon eco-
system is no exception and provides important habitat
to wintering, migrant, and breeding birds (Brown and
others, 1987; Carothers and Brown, 1991; Sogge and
others, 1998; Kearsley and others, 2004; Spence, 2004).
Importantly, the ecosystem also provides habitat for
several bird species of special concern, including the
southwestern willow flycatcher (Empidonax traillii extimus),
California condor (Gymnogyps californianus), bald eagle
(Haliaeetus leucocephalus), and American peregrine falcon
(Falco peregrinus anatum).
This chapter summarizes the considerable infor-
mation available from recent studies on the ecology of
Grand Canyon bird species and communities. Because
changes in riparian habitat undoubtedly influence the
abundance and distribution of Grand Canyon birds,
the chapter starts by briefly examining dam-induced
habitat alterations that may affect birds. The direct and
indirect effects of Glen Canyon Dam operations, includ-
ing the modified low fluctuating flow (MLFF) alternative
that was implemented starting in 1996, are considered
for how they influence specific bird species and com-
munities. Particular attention is given to the species of
special concern listed above. The chapter concludes
with a summary and a discussion of research priorities
within the context of the Glen Canyon Dam Adaptive
Management Program.
Background
The riparian vegetation of the Grand Canyon
ecoregion is complex and dynamic, changing in response
to flooding, the invasion of new nonnative species, long-
term successional patterns, and climate (Turner and
Karpiscak, 1980; Webb and others, 1999). The primary
driving variables in the terrestrial riparian ecosystem in
Grand Canyon are the flow characteristics and hydro-
graph of the Colorado River (Carothers and Aitchison,
1976; Stevens and others, 1995; Kearsley and Ayers,
2001). The distribution of the riparian vegetation in the
Chapter 7
Birds of the
Colorado River
in Grand Canyon:
a Synthesis of
Status, Trends,
and Dam
Operation
Effects
Jennifer A. Holmes
John R. Spence
Mark K. Sogge
124 The State of the Colorado River Ecosystem in Grand Canyon
Grand Canyon is also strongly influenced by the geogra-
phy and geology of the region. The river corridor follows
a 1,772-ft (540-m) elevation gradient through changing
vegetation communities ranging from Great Basin
desertscrub (classification per D.E. Brown, 1994) found
at Lees Ferry (RM 0) with an elevation of 3,117 ft (950
m) to Sonoran desertscrub at Diamond Creek (RM 226)
with an elevation of 1,345 ft (410 m). Type of bedrock
geology present and the presence of major side canyons
are perhaps the most important geological factors rel-
evant to the distribution of riparian vegetation along the
river (Turner and Karpiscak, 1980; Stevens and others,
1995; Spence, 2004). For example, where the bedrock
consists of Precambrian schist and granite, which are
hard and slow to erode, the river corridor is narrow and
tends not to support much riparian vegetation except at
the mouths of tributaries.
Before the construction of Glen Canyon Dam, the
hydrograph of the Colorado River in Grand Canyon
was driven by spring snowmelt floods and occasional
large tributary inflows produced by monsoonal late-
summer rains (Dolan and others, 1974; Carothers and
Aitchison, 1976; Topping and others, 2003). Spring
flooding controlled the abundance and distribution of
riparian vegetation, producing a distinct trim line at
about the 125,000 cfs level. Water-surface elevation, or
stage, is typically given in terms of rate of flow (cubic
feet per second (cfs)) because elevation varies over the
length of the river corridor depending on local channel
morphology. Above this line an extensive community
of old high-water zone (OHWZ) vegetation occurred,
including species such as Apache plume (Fallugia para-
doxa), net-leaf hackberry (Celtis laevigata var. reticulata),
mesquite (Prosopis glandulosa), and catclaw acacia (Acacia
greggii). Because the lower zone, below 50,000 cfs, was
flooded and scoured most years (Topping and oth-
ers, 2003), sparse vegetation was present below the
trim line and consisted of coyote willow (Salix exigua),
tamarisk (Tamarix ramosissima), rushes (Juncus sp.), and
grasses (Poaceace family) (Clover and Jotter, 1944). Most
of the plant species found today in lower vegetation
zones were present before the construction of the dam
(Clover and Jotter, 1944; Kearsley and Ayers, 2001).
Predam conditions resulted in varied riparian vegetation,
producing dense riparian stands more than 164 ft (50 m)
wide at some tributary mouths, but elsewhere riparian
stands were patchily distributed and generally between
10 and 66 ft (3–20 m) wide (Flowers, 1959; Kearsley and
Ayers, 2001).
The hydrograph of the Colorado River changed
dramatically with the completion of Glen Canyon Dam
(Webb and others, 1999; Topping and others, 2003). Of
the many changes wrought by the dam and discussed
elsewhere in this report, the most influential one in terms
of riparian vegetation is the reduction of peak annual
flows. In the absence of historical floods that removed
lower zone vegetation, perennial plant species were
able to move into and colonize these areas. These new
areas of riparian vegetation are referred to as the new
high-water zone (NHWZ) to distinguish them from the
higher predam riparian habitats. The amount of NHWZ
riparian vegetation greatly increased between 1963
and 1983 (Pucherelli, 1986), and much of the coloniza-
tion of the NHWZ was by nonnative species, especially
tamarisk (Turner and Karpiscak, 1980). Areas of marsh
also developed in return channel-eddy complexes (Cluer,
1997), covering approximately 1% of the NHWZ of the
river corridor by 1991 (Stevens and others, 1995).
The floods and subsequent high flows of 1983–85
produced considerable scour and an estimated 13%
(Waring, 1995) to 39% (Pucherelli, 1986) reduction in
area of the NHWZ. Following the floods, the NHWZ
gradually recovered. In 1991, interim flows were estab-
lished that caused further changes, primarily the sta-
bilization of marshes and riparian colonization of the
lower portion of the NHWZ between 25,000–33,000
cfs. The 1996 beach/habitat-building flow through the
river corridor was designed to scour tamarisk vegetation
in the lower portions of the NHWZ but had only short-
term burial impacts on the vegetation, which recovered
rapidly (Kearsley and Ayers, 1999; Stevens and others,
2001). Flows under the MLFF alternative have not
altered the areal extent of riparian vegetation patches
from that established during the period of interim flows
in the early 1990s (Kearsley and Ayers, 1996, 1999).
Dam-induced changes to riparian habitat undoubt-
edly affected the abundance and distribution of riparian
birds. These changes are not directly measurable because
Mark K. Sogge, U.S. Geological Survey
Birds of the Colorado River in Grand Canyon 125
the first surveys of breeding birds along the Colorado
River (Carothers and Sharber, 1976) were not conducted
until after the dam was completed. Since then, surveys
and studies have examined many aspects of the ecology
of Grand Canyon birds, including the direct and indi-
rect effects of the dam on specific bird species and bird
communities. Riparian breeding bird studies have been
conducted along the river corridor since the initiation of
the Glen Canyon Environmental Studies in 1982 (Brown
and others, 1987; Brown, 1987, 1989; Brown and
Trosset, 1989; Spence and Pinnock, 1993; Grahame and
Pinnock, 1995; Hualapai Tribe and SWCA, Inc., 1995;
Petterson and Spence, 1997; Spence, 1997, 2004; Sogge
and others, 1998; Kearsley and others, 2004; Yard and
others, 2004). These studies collected data on the com-
position of bird communities; patterns of species abun-
dance, richness, and diversity; and habitat distribution.
They also provided information on habitat associations
and identified riparian-dependent bird species.
Species-specific bird studies have also been con-
ducted along the river corridor. Focus was given to bird
species that were or are federally listed as endangered
or threatened, including bald eagle (Brown and others,
1989, 1998; Brown and Stevens, 1992, 1997; Brown,
1993; Leibfried and Montgomery, 1993; Spence and
others, 2002; van Riper and Sogge, 2004), peregrine
falcon (Ellis and Monson, 1989; Brown, 1991a; Ward,
2000), and southwestern willow flycatcher (Brown, 1988;
Sogge and others, 1997; Johnson, 2000; Yard, 2004a).
A number of riparian bird species were also the sub-
ject of research, including the Bell’s vireo (Vireo bellii)
(Brown and others, 1983), black-chinned hummingbird
(Archilochus alexandri) (Brown, 1992), and brown-headed
cowbird (Molothrus ater) (Brown, B.T., 1994). Studies of
the winter riparian bird community (Sogge and others,
1998; Spence, 2004; Yard, 2004b) and the aquatic bird
community (Stevens and others, 1997a; Spence, 2004;
Yard, 2004b) have also been conducted. These stud-
ies provide considerable information on the ecology of
Grand Canyon bird communities and the direct and
indirect effects of the dam on specific bird species and
bird communities.
Status and Trends
Breeding Riparian Birds
The breeding bird community associated with the
riparian habitat along the Colorado River is made up
of bird species generally restricted to riparian habitats
and species that can also be found in adjacent upland,
nonriparian habitats. More than 30 species have been
recorded breeding in the riparian patches along the
river within the study area. Most of these are songbirds
including warblers, wrens, finches, orioles, and sparrows
that nest and forage for insects within the NHWZ and
OHWZ vegetation. Of the 15 most common ripar-
ian breeding bird species (table 1), 10 are Neotropical
migrants that breed in the study area but winter primar-
ily south of the United States-Mexico border. The rest
of the breeding birds that use the canyon are year-round
residents or short-distance migrants that primarily winter
in the region or in nearby southern Arizona.
Repeated research since the mid-1970s has shown
that Glen Canyon Dam and its operation have few
direct flow-related effects upon the riparian breeding
bird community. The primary change influencing these
birds has been increased habitat availability caused by
the establishment of the NHWZ riparian vegetation in
areas that had relatively sparse vegetation before the
dam (Brown and others, 1987; Carothers and Brown,
1991; Sogge and others, 1998; Spence, 2004). Brown
and Johnson (1985, 1987) also found that flows directly
affected some birds that occupied this new habitat during
periods of high daily change in the river level or dur-
ing enormous seasonal fluctuations that occurred before
1991 and the establishment of interim flow operating
criteria. For example, they found that flows as high as
31,000 cfs, approximate powerplant capacity, flooded
only a few nests, including some common yellowthroat
(Geothlypis trichas) nests; however, flows of more than
40,000 cfs began flooding nests and nest plants of some
riparian breeding species, specifically the Bell’s vireo
U.S. Geological Survey
126 The State of the Colorado River Ecosystem in Grand Canyon
and yellow-breasted chat (Icteria virens) (Brown and
Johnson, 1987).
The interim operating criteria, which were in effect
from 1991–96, limited maximum releases to 20,000 cfs
and set minimum flows at 8,000 cfs during the day and
5,000 cfs at night. Daily fluctuations were also limited
to a maximum of 8,000 cfs. This change in river flows
promoted the establishment of a narrow band of veg-
etation near the edge of the river (Stevens and Ayers,
1994; Sogge and others, 1998). Sogge and others (1998)
examined the direct impact of interim operating crite-
ria on breeding birds in the hydrologically active zone
(HAZ), the area potentially inundated by flows between
5,000 and 20,000 cfs. They found that few species
nested either close to the ground or close to the river;
only one common yellowthroat nest was placed low
enough to be inundated at 20,000 cfs. The black phoebe
(Sayornis nigricans), however, places its nests just 3–6 ft (1–2
m) above the water, and some phoebe nests would be
inundated by any flows that raised the water level by as
little as 3 ft (1 m). For other breeding bird species, Sogge
and others (1998) found that nests were placed well away
from the HAZ and avoided inundation under the interim
operating criteria. Because the MLFF alternative is
similar to interim operating criteria (Kearsley and Ayers,
1999), most nests (other than some black phoebe nests)
are unlikely to be affected by changing water levels.
Sogge and others (1998) examined the potential long-
term effects of dam operations on the Grand Canyon
ecosystem bird community. They developed models and
identified habitat features that predicted bird abundance,
species richness, and diversity. They found that riparian
location along the river corridor, patch size, and volume
of larger woody species, especially of tamarisk, within a
riparian vegetation patch were positively correlated with
bird abundance, species richness, and diversity. Specifi-
cally, these models predicted the following:
Flow patterns that result in smaller, more isolated
habitat patches would decrease bird numbers, spe-
cies richness, and diversity.
Flow patterns that create larger and more contiguous
habitat patches would increase bird abundance and
richness within the constraints of local topography
and geomorphology.
Loss of mesquite vegetation would decrease bird
abundance.
Increases in the number of habitat patches would
increase overall number of birds and bird species.
Changes from tamarisk shrub/tree to willow shrub/
tree are not likely to greatly affect bird abundance
and species richness of the Grand Canyon ecosystem
bird community.
A subsequent study by Spence (2004) modeled
riparian bird habitat relationships but did not examine
the relationship between riparian patch size and charac-
teristics such as total bird abundance, species richness,
and species diversity. Despite the fact that data were
being derived from different riparian patches, this study
obtained results similar to Sogge and others (1998), con-
cluding that higher woody-species volume and river loca-
tion were the best predictors of breeding bird abundance
and richness.
These two studies (Sogge and others, 1998; Spence,
2004) demonstrate that riparian patch size, the volume
of woody species within a habitat patch, and the loca-
1.
2.
3.
4.
5.
Table 1. The 15 generally most common terrestrial breeding
bird species (in alphabetical order) found in riparian habitats
along the Colorado River in Grand Canyon.
Common Name Scientific Name
Ash-throated flycatcher Myiarchus cinerascens
Bell’s vireo Vireo bellii
Bewick’s wren Thryomanes bewickii
Black-chinned hummingbird Archilochus alexandri
Blue grosbeak Passerina caerulea
Blue-gray gnatcatcher Polioptila caerulea
Bullock’s oriole Icterus bullockii
Common yellowthroat Geothlypis trichas
House finch Carpodacus mexicanus
Lesser goldfinch Carduelis psaltria
Lucy’s warbler Vermivora luciae
Mourning dove Zenaida macroura
Song sparrow Melospiza melodia
Yellow warbler Dendroica petechia
Yellow-breasted chat Icteria virens
Birds of the Colorado River in Grand Canyon 127
tion of the patch along the river corridor are primary
factors that affect the abundance and species richness of
birds within a riparian patch. Also, the mix of NHWZ
and OHWZ vegetation within a patch probably affects
the distribution and abundance of specific bird species
because certain species have ecological preferences in
nesting and foraging in one or the other vegetation type
(Sogge and others, 1998; Spence, 2004). Collectively, the
body of research indicates that dam operations with the
greatest potential to impact breeding bird species within
the Grand Canyon ecoregion are those that would affect
the extent and amount of riparian vegetation along the
river, such as large-magnitude planned or unplanned
floods.
Overall, there has been relatively little change in
the distribution of riparian habitat since the initiation
of canyon bird studies in the mid-1970s. Likewise, the
riparian breeding bird community within the study area
appears not to have changed appreciably in species
composition during that 25-yr period (Spence, 2004).
For the most part, the bird species that were most com-
mon in the 1980s are the most common today (Kearsley
and others, 2004; Spence, 2004). Two exceptions are
the Bell’s vireo and the song sparrow (Melospiza melodia),
which have apparently expanded their breeding ranges
within the Grand Canyon ecoregion (Brown and others,
1983; Spence, 2004; Yard and Blake, 2004).
To track trends in riparian breeding bird popula-
tions, one long-term monitoring program was initiated
in 1996 and continued through 2000 (Spence, 2004) and
another from 2001 through 2004 (Kearsley and others,
2004). These studies included baseline monitoring of the
breeding riparian birds, southwestern willow flycatcher,
and riparian habitat in selected patches along the river
corridor. Several species, mostly Neotropical migrants,
showed consistent detection rates during the 1996–2000
time period (Spence, 2004). The blue-gray gnatcatcher
(Polioptila caerulea) showed a steady decline, with detection
rates dropping about 30%–50%. Two species, Bullock’s
oriole (Icterus bullockii) in Glen Canyon and yellow warbler
(Dendroica petechia) throughout the study area, showed sta-
tistically significant increases in detection rates (Spence,
2004).
Most birds found in the study area are not year-
round residents; therefore, other factors acting outside
the Grand Canyon ecosystem influence bird populations,
and this influence is especially true for migratory breed-
ing birds. Outside factors—changes in winter and migra-
tory habitat, winter weather events, and climate outside
the region—can affect bird survivorship and are inde-
pendent of the effects of adaptive management (Spence,
2004; Holmes and others, 2005).
Overwintering Aquatic Birds
Increases in abundance and species richness of the
aquatic bird community—loons, grebes, cormorants,
herons, ducks, rails, and sandpipers—in the Grand
Canyon ecoregion correspond with the increased river
clarity and productivity associated with the presence of
Glen Canyon Dam (Stevens and others, 1997a; Spence,
2004). These aquatic bird species use the Grand Canyon
ecoregion almost exclusively in the winter, nonbreeding
season. Two primary foraging guilds are represented: (1)
diving species that consume mostly fish and invertebrates
within the water column or on the river bed and (2)
dabbling species that forage in cobble bars and shal-
lower areas where they can reach aquatic vegetation and
associated invertebrates (table 2). These aquatic birds can
be directly affected by dam operations that change the
distribution of prey species in the water column of the
river or, in the case of dabbling species, cover or expose
foraging beds. Also, higher discharge rates increase river
velocity and potentially increase foraging costs for species
in both guilds (Spence, 2004).
Aquatic bird species are distributed fairly predict-
ably within the study area. The upper reaches of the
river, from Glen Canyon Dam to about 25 mi (40 km)
downstream, tend to be relatively clear and support
habitat and food, including an abundance of introduced
Table 2. The 10 generally most common overwintering aquatic
bird species (in alphabetical order) encountered during surveys
along the Colorado River below Glen Canyon Dam.
Common Name Scientific Name
American coot Fulica americana
American wigeon Anas americana
Bufflehead Bucephala albeola
Common goldeneye Bucephala clangula
Common merganser Mergus merganser
Gadwall Anas strepera
Green-winged teal Anas crecca
Lesser scaup Aythya affinis
Mallard Anas platyrhynchos
Ring-necked duck Aythya collaris
128 The State of the Colorado River Ecosystem in Grand Canyon
©
20
0
5
B
o
b
S
t
e
el
e
;
us
e
d
wi
th
p
er
m
i
ss
i
o
n
rainbow trout (Oncorhynchus mykiss) for diving species.
Dabblers can forage only in wider reaches with extensive
shallow, low-turbidity water. Accordingly, dabblers are
concentrated in wider reaches above the Little Colorado
River (Stevens and others, 1997a; Spence, 2004).
Spence (2004) found that species composition and
abundance of the aquatic bird communities within the
study area show considerable fluctuations among years.
Given similar flows, however, the resources available to
waterfowl in the Grand Canyon ecoregion are relatively
similar among years. Primary productivity is greatest in
the clear water below the dam to approximately 25 mi
(40 km) downstream and then drops rapidly as the river
becomes more turbid as sediment and organic matter
enter the river from tributaries. Hence, it is likely that the
large year-to-year fluctuations in aquatic birds noted by
Spence (2004) are due to factors outside the region. Out-
side factors may be numerous and potentially include
conditions on the breeding grounds, recreation activities,
changes in habitat availability, climate conditions, and
hunting. All these factors can interact in complex ways
in determining the composition and abundance of the
winter aquatic community in the Grand Canyon ecore-
gion (Spence, 2004).
Because of the high variability in abundance for
many species, the power to detect trends in overwinter-
ing aquatic birds is low. Comparing results of surveys
conducted between 1973 and 1994 by Stevens and oth-
ers (1997a) to data from surveys between 1998 and 2000,
Spence (2004) found strong similarities in the aquatic
bird communities, and the most common birds detected
during both periods were similar (Spence, 2004).
Species of Concern
Southwestern Willow
Flycatcher
The willow flycatcher
(Empidonax traillii) is a small Neo-
tropical migratory bird that breeds
across much of North America
and winters in portions of Cen-
tral America and northern South
America. The southwestern
subspecies (E. t. extimus) breeds only
in dense riparian habitats in the
Western United States, including por-
tions of Utah, Nevada, Colorado, New
Mexico, Arizona, and southern California.
Breeding generally occurs from late May through early
August (Sogge, 2000). When the southwestern willow
flycatcher was federally listed as an endangered species in
1995, fewer than 400 breeding territories for the subspe-
cies were known throughout the Southwest; however, by
2001, that number had increased to approximately 1,000
territories distributed among more than 200 breed-
ing sites (Sogge and others, 2003). By 2003, because
of increased survey effort and a population increase in
central Arizona, there were an estimated 410 territories
in Arizona alone (Smith and others, 2004).
Historically, southwestern willow flycatchers were
probably found within most major drainages in Arizona
(Paradzick and Woodward, 2003) but were uncommon
within the Grand Canyon ecoregion primarily because
periodic high flows limited dense riparian habitat. The
first record of a willow flycatcher in the ecoregion is from
Lees Ferry in 1909, but it is not known whether it was a
migrant or a breeding bird. The first nest was found in
1935. The next record was of a probable breeder col-
lected in 1953 (summarized in Sogge and others, 1997).
Flycatchers have consistently nested along the river cor-
ridor in recent years, as new riparian habitat, primarily
tamarisk, has developed in response to altered river flow
regimes. This expansion of riparian vegetation may have
provided additional habitat for the flycatcher. Migrant
willow flycatchers also occur along the river corridor,
typically in late May and early June, and most of these
migrants are probably of the nonendangered northern
subspecies (E. t. adastus).
There are no direct flow-related impacts to south-
western willow flycatchers because they nest high in tam-
arisk vegetation, which is well above the level of normal
fluctuating river flows. Indirect effects may occur as the
result of flow-related changes to riparian patch size, veg-
etation density, and invertebrate populations that form
the flycatcher prey base. The 1996 beach/hab-
itat-building flow did not adversely affect
southwestern willow flycatchers or their
breeding habitat structure (Stevens
and others, 2001). If future flood
flows enhance riparian habitat and
patch size, flycatchers may benefit.
Conversely, if they substantially
reduce riparian habitat at current
breeding sites, the flycatcher may
be impacted.
Wetland/marsh vegetation
has been proposed as important
flycatcher foraging habitat in the study
area (Stevens and others, 2001). The
Birds of the Colorado River in Grand Canyon 129
necessity of wetlands to flycatchers is difficult to evaluate
because the species often breeds at sites in the South-
west where extensive wetlands are absent (Sogge and
Marshall, 2000).
Other potential impacts to southwestern willow
flycatchers include human-related disturbance. South-
western willow flycatchers are not apparently sensitive to
disturbances such as rafts or boats floating past breed-
ing sites; however, people moving through occupied
flycatcher habitat can damage habitat, disturb the birds,
or impact a nest. During the mid-1990s, visitor closures
were instituted at known flycatcher breeding sites in
Grand Canyon. To date, there is no evidence of direct,
human-related impact to flycatchers along the river
corridor. Potential and indirect human-related impacts
include the eradication of tamarisk, which the flycatch-
ers use for nesting. Its removal, particularly from known
breeding sites, would adversely affect flycatchers.
Brown-headed cowbirds are nest parasites and lay
their eggs in the nests of other birds, which then incu-
bate the cowbird eggs and raise the young cowbirds as
if they were their own young. B.T. Brown (1994) and
Sogge and others (1997) reported that flycatchers in the
Grand Canyon ecoregion experienced high rates (>25%
of nests) of brown-headed cowbird parasitism, which
reduced flycatcher nest success and productivity. There is
no evidence, however, that dam operation or river flows
affect cowbird populations or nest parasitism rates.
Because southwestern willow flycatchers migrate
southward each winter, they are affected by many factors
during the migration and wintering periods. Further-
more, flycatchers will regularly disperse long distances
and move to different sites between years; however, the
reasons behind these movements are not well known
(U.S. Fish and Wildlife Service, 2002). These reali-
ties complicate interpretation of population trends in
the study area and the evaluation of poten-
tial impacts of the operations of Glen
Canyon Dam.
Over the last 30 yr, the popula-
tion of breeding southwestern
willow flycatchers in upper Grand
Canyon has been very small
and limited to riparian patches
between approximately RM 28
and 71. From 1982 to 1991, 2 to
11 male flycatchers were detected
annually, with a maximum of 4
nests in any 1 yr (Brown, 1988,
1991b). Between 1992 and 2003, only
1 to 5 territories were found in any year
(Sogge and others, 1997; Johnson, 2000;
Paradzick and Woodward, 2003; Yard, 2004a). Flycatch-
ers bred only in the relatively larger patches, and breed-
ing patch size ranged from 1.5 to 2.2 acres (0.6–0.9 ha).
Breeding patches were dominated by tamarisk, and all
nests had been placed in tamarisk (Sogge and others,
1997; Yard, 2004a). Overall, the southwestern willow fly-
catcher population in the upper river corridor continues
to persist at a very low level, at only one or two sites.
In 1995, breeding flycatchers using one territory
were first noted in newly developed native riparian
habitat in the Lake Mead delta area, immediately down-
stream of the Grand Canyon-Lake Mead boundary. The
following year this population reached 10 territories,
but the delta was flooded during the next 2 yr by rising
reservoir levels, and flycatchers were no longer present
by 1998. Beginning in 1998, breeding southwestern wil-
low flycatchers were discovered at a variety of upstream
sites within lower Grand Canyon between RM 246 and
273 (Paradzick and Woodward, 2003). It is possible that
birds found before 1998 breeding downstream, in what
is now inundated delta habitat, moved upstream to the
lower Grand Canyon reach. Between 1998 and 2001,
7–12 flycatcher territories were recorded in lower Grand
Canyon; however, recent surveys in 2002 and 2003 found
no breeding flycatchers in lower Grand Canyon (Smith
and others, 2003, 2004) and only 2 territories in 2004
(McLeod and others, 2005; Munzer and others, 2005).
Nesting success in the upper Grand Canyon fly-
catcher population is generally low, and the population
is probably not self-sustaining (Sogge and others, 1997).
Breeding success in lower Grand Canyon is not well
documented, but the lack of detections in 2002 and
2003 suggests that productivity from 1998 to 2001 was
probably too low to provide for continued population
persistence. Recent habitat changes along the lower
river corridor caused by the changing reservoir levels in
Lake Mead, however, have probably affected
flycatcher site occupancy.
California Condor
The California condor is one
of the rarest birds in the world
and was federally listed as endan-
gered in the United States in
1967. In Arizona, reintroduction
was conducted beginning in 1996
under a special provision of the
Endangered Species Act of 1973.
Condors are opportunistic scav-
engers that feed primarily on large, dead
mammals such as deer, elk, bighorn sheep
©
2
00
5
J
e
ff
C
o
l
e
m
a
n;
u
s
e
d
w
it
h
p
e
r
m
is
s
io
n
130 The State of the Colorado River Ecosystem in Grand Canyon
(Ovis canadensis), range cattle, sheep, and horses. Condors
can soar and glide up to 50 mi/h (80 km/h) and travel
100 mi (161 km) or more per day in search of food. They
are long lived, living up to 60 yr, with low reproductive
rates. Most nest sites have been found in caves, on rock
ledges, or in tree cavities (Snyder and Schmitt, 2002;
Arizona Game and Fish Department, 2004).
In prehistoric times, condors ranged from Canada to
Mexico, across the Southern United States to Florida,
and to the east coast in New York. Based on
evidence from bones, feathers, and egg-
shells found in caves, condors were a
resident of Grand Canyon. A dramatic
range reduction occurred about
10,000 yr ago, coinciding with the
late Pleistocene extinction of large
mammals that condors depended
on for food (Arizona Game and
Fish Department, 2004). Settlement
of the Western United States, shoot-
ing, poisoning from lead and DDT,
egg collecting, and general habitat
degradation resulted in further dramatic
population reductions (Snyder and Schmitt,
2002). Between the mid-1880s and 1920s,
there were scattered reports of condors in Arizona,
with the last sighting near Williams, Ariz., in 1924. By
the late 1930s, all remaining condors were found only in
California, and by 1982 the total population had dwin-
dled to just 22 birds (Snyder and Schmitt, 2002; Arizona
Game and Fish Department, 2004).
The U.S. Fish and Wildlife Service, The Peregrine
Fund, Arizona Game and Fish Department, National
Park Service, and other collaborators established
a condor captive-breeding and release program in
Arizona. Vermilion Cliffs National Monument is the
main reintroduction site, and birds released at this site
frequent Grand Canyon. Since December 1996, the
Arizona restoration project has released approximately
6–8 birds per year. There are now over 30 condors flying
free in Arizona, and natural reproduction is occurring in
the Grand Canyon region: in late 2004, 2 wild-hatched
chicks stretched their wings and successfully fledged. As
part of the continuing reintroduction project, individual
condors will continue to be monitored daily (Arizona
Game and Fish Department, 2004).
Many of the reintroduced condors have been
observed within the Grand Canyon ecoregion. Although
typically seen soaring overhead, condors regularly
bathe and sun themselves along the banks of the
Colorado River (Andi Rogers, Arizona Game and Fish
Department, oral commun., 2005). It is likely that the
population of condors in the region will continue to
increase because of continued reintroduction efforts and
natural increase. Since condors make little use of ripar-
ian habitat and are not typically found along the edge of
the river, there are no likely effects of MLFF operations.
The only dam management actions likely to affect this
species would be those that resulted in available carrion
such as dead fish along the river corridor, which could
attract concentrations of feeding condors.
Bald Eagle
Bald eagles are common breed-
ers in Alaska and parts of Canada
but are far less numerous in the
contiguous United States, where
they were once critically endan-
gered. Because of extensive
and successful recovery efforts
since the 1960s, many bald eagle
populations have increased, and
in 1995 the U.S. Fish and Wildlife
Service downlisted the species from
endangered to threatened in the lower
48 States.
Although still somewhat rare as a breeder
in Arizona, hundreds of bald eagles migrate into the
State each winter; eagle numbers in Arizona increased
from 225 in 1992 to 440 in 2001 (Beatty, 2001). Winter-
ing eagles typically concentrate along rivers, lakes, and
reservoirs where preferred prey, including fish, waterfowl,
and carrion, is readily available (Grubb and Kennedy,
1982; Brown, 1993). The Grand Canyon ecoregion is
one such concentration area, and eagles are generally
present from November through March, which coincides
with trout spawning and an abundance of waterfowl
within the corridor. Within the study area, bald eagles
are found primarily from Lees Ferry downstream to the
confluence of the Little Colorado River. From 1991 to
1995, the maximum daily number of eagles detected
during helicopter surveys of this reach ranged from
11 to 24 individuals (van Riper and Sogge, 2004). The
Colorado River corridor in Grand Canyon hosted only
5%–10% of the wintering eagles present in Arizona on
any given day during this time, but the total number of
eagles using the corridor over an entire season may be
substantially more (van Riper and Sogge, 2004). System-
atic corridor-wide surveys were not conducted before or
after the 1991–95 period.
Bald eagles often congregate at Nankoweap Creek
(RM 52), a small tributary to the Colorado River in
which rainbow trout sometimes spawn in large numbers.
G
e
or
g
e
A
n
d
re
j
ko
,
A
ri
z
on
a
G
a
m
e
a
n
d
Fi
s
h
D
e
p
a
rt
m
e
n
t
Birds of the Colorado River in Grand Canyon 131
Eagles have concentrated here since the early 1980s
(Brown and others, 1989; Leibfried and Montgomery,
1993). From 1986 to 1995, maximum daily eagle counts
ranged from 4 to 26, with the number of eagles vary-
ing directly with the abundance and availability of
trout in the creek (Brown and Stevens, 1992; Leibfried
and Montgomery, 1993; van Riper and Sogge, 2004).
Neither the size of trout spawn nor eagle abundance at
Nankoweap Creek was related to dam release levels
(van Riper and Sogge, 2004). There is also no
evidence that eagle abundance throughout
the river corridor is affected by river
flow, although it is likely that river tur-
bidity affects the ability of eagles to
forage for fish along the mainstem.
Human disturbance can affect
bald eagles. Brown and Stevens
(1997) and van Riper and Sogge
(2004) documented disturbance of
wintering bald eagles by humans,
including flushing of eagles by
hikers, rafters, anglers, and research
activity. Hikers in the Nankoweap
Creek delta area caused the greatest
disturbance to the eagles there, but such
disturbances were reduced in years when a visitor-use
closure was instituted.
Evaluating the effects of the operations of Glen
Canyon Dam and other management activities on bald
eagles is complicated by the fact that eagle abundance
in Grand Canyon is influenced by both local conditions
and regional factors. Furthermore, bald eagles will travel
long distances in search of abundant, easily available
prey (Stalmaster, 1987) and can move readily between
food concentrations at Grand Canyon, Lake Powell, and
other regional lakes and rivers. Nevertheless, habitat use
by foraging eagles is strongly influenced by fluctuating
river flows; high flows reduce eagle foraging habitat diver-
sity, lower foraging success in river habitat, and restrict
foraging opportunities (Brown and others, 1998).
Management changes that alter prey availability
could alter eagle abundance and distribution within the
Grand Canyon ecosystem. For example, if the selective
withdrawal of warmer water from Lake Powell increases
the numbers of carp, catfish, and suckers (Hunt and
others, 1992), more food resources may be available to
eagles. Eagles may have more difficulty foraging along the
river, however, if trout numbers decrease and/or spawn-
ing is reduced, either through water temperature/turbid-
ity changes or through nonnative fish removal efforts.
Peregrine Falcon
Dramatic declines in peregrine falcon populations
led to the addition of the peregrine to the Federal List of
Endangered and Threatened Wildlife in 1970, where it
was listed as endangered. Following successful recovery
efforts, the peregrine falcon was delisted in 1999. The
Endangered Species Act requires a minimum 5 yr of
post-delisting monitoring in cooperation with State
agencies to confirm recovery.
The peregrine commonly breeds
in cliffs and uses open landscapes for
foraging. Nest sites are usually associ-
ated with water (White and others,
2002). In winter, some breeders
stay in their nesting areas, and
others may migrate.
Diet of the peregrine consists
mostly of birds, from songbirds to
small geese. They also occasionally
eat mammals, especially bats (White
and others, 2002). During the breed-
ing season, peregrine falcons in the
Grand Canyon ecoregion feed on white-
throated swifts (Aeronautes saxatalis), swallows,
and bats (Brown, 1991a). In winter, they feed mainly on
waterfowl. Many of their prey items feed on invertebrate
species, especially flies (Diptera), that emerge out of the
Colorado River (Stevens and others, 1997b).
Given these life-history traits, any impacts to pere-
grine falcons from dam operations are likely to be indi-
rect, possibly through influences on the distribution and
abundance of aquatic macroinvertebrate populations,
which in turn would influence the availability of the
peregrine’s prey items such as swifts, bats, and ducks.
The Grand Canyon peregrine population was
thought to be low in the mid-1970s and apparently
increased dramatically in the 1980s (U.S. Fish and
Wildlife Service, 1984; Ellis and Monson, 1989; Ward,
2000). In 1981 and 1982, two nests or “eyries” were
found during surveys between the Tanner and Bright
Angel Trails. In 1998 and 1999, 12 eyries were found in
these same areas (R.V. Ward, Grand Canyon National
Park, oral commun., 2005). During the same period, the
National Park Service conducted surveys throughout
appropriate habitat within Grand Canyon National Park,
including along the river corridor, and concluded that
the peregrine population in Grand Canyon appeared
stable since 1988 (Ward, 2000).
G
eo
r
g
e
A
n
dr
e
jk
o
,
A
ri
zo
n
a
G
a
m
e
a
nd
Fi
s
h
D
e
p
ar
t
m
e
n
t
132 The State of the Colorado River Ecosystem in Grand Canyon
Recent Findings
Kearsley and others (2004) examined the interrela-
tionships between vegetation and animal life, including
birds, as part of a monitoring project for terrestrial ripar-
ian resources that took place from May 2001 to May
2003. Preliminary findings regarding terrestrial breed-
ing birds showed patterns similar to those of previous
studies. More breeding pairs and higher species diversity
were detected at larger sites (Yard and Blake, 2004).
Vegetation density was found to be an important com-
ponent of habitat quality for riparian breeding birds in
the Grand Canyon ecoregion, with the densities of most
bird species positively correlated with the abundance of
mesquite and acacia (Kearsley and Lightfoot, 2004). No
difference was found in the abundance of birds over the
3 yr of the study although sample sizes were too low to
analyze trends (Yard and Blake, 2004). The most com-
monly detected breeding species were the same as those
in previous studies (Brown and others, 1987; Sogge and
others, 1998; Spence, 2004).
Discussion and Future
Research Needs
The construction of Glen Canyon Dam and the
subsequent changes in the hydrograph of the Colorado
River resulted in dramatic changes in the amount of
available habitat for both the riparian breeding and
the overwintering aquatic bird communities within the
Colorado River ecosystem. Perennial plant species, espe-
cially tamarisk, colonized areas previously scoured by
floods, creating new riparian patches in the high-water
zone. These areas provide habitat for over 30 species of
breeding birds, including many Neotropical migrants
and the endangered southwestern willow flycatcher.
Increased river clarity and productivity below the dam
provide suitable habitat for many aquatic bird species
such as ducks, loons, grebes, and cormorants.
Patterns of abundance and distribution of ripar-
ian breeding birds and overwintering riparian and
aquatic birds within the study area are now well known.
Less well known are the long-term effects of adaptive
management and the management activities needed to
ensure the continued conservation of riparian resources,
their associated avian communities, and bird species of
conservation concern. Continued monitoring would be
required to address these information needs.
Monitoring riparian breeding birds to detect popula-
tion changes requires considerable commitments of both
time and effort to obtain sufficient data for biological
and statistical significance. Data from the 1996 to 2000
breeding bird monitoring program were used by Spence
(2004) to determine the adequacy of the monitoring
program to detect changes in bird populations. He found
that trends could not be detected for 24 of 32 (75%)
riparian breeding species and that 5 to 30 yr of sampling
were required to detect a 10% change in species abun-
dance. Half of the 16 most common species included in
the analysis would require over 10 yr of monitoring to
detect a 10% population change, while 5 rarely detected
species cannot be monitored by using the sampling pro-
tocols tested in the analyses (Spence, 2004).
An alternative approach to continued monitoring
of riparian birds would be to use aerial photography,
remote sensing, and geographic information systems
(GIS) in order to measure habitat variables within the
study areas that have been shown to predict bird num-
bers, richness, and diversity (Sogge and others, 1998).
Key variables for monitoring would include the size and
distribution of riparian patches, area of NHWZ and
OHWZ woody species, and measures of total vegeta-
Mark K. Sogge, U.S. Geological Survey
Birds of the Colorado River in Grand Canyon 133
tion volume; however, simply monitoring habitat quality
and extent may miss potential changes in selected bird
species caused by factors within the study area other than
riparian vegetation dynamics and may miss potentially
profound changes in some bird species (Spence, 2004).
Existing data, collected in previous studies, could be
used to better model and predict how future changes in
the riparian vegetation will affect changes in terrestrial
avifauna populations within the Grand Canyon ecore-
gion. In particular, it would be valuable to extend the
current models of bird community and patch-level habi-
tat variables (Sogge and others, 1998; Spence, 2004) to
the level of individual bird species. This extension would
allow the development of more useful conceptual models
and more detailed predictions regarding avian resources
in the Grand Canyon ecoregion. For example, models
developed by Sogge and others (1998) predicted that
changes from tamarisk shrub/tree to willow shrub/tree
are not likely to greatly affect overall bird abundance and
species richness within riparian patches in Grand Can-
yon, yet individual species have specific behavior, physiol-
ogy, and ecology, and some may decline in response to
such habitat changes. Extending models to individual
species would allow identification of species that may be
sensitive to future changes in the riparian vegetation.
Riparian woodlands, such as those within the Grand
Canyon ecoregion, provide vital habitat for bird species
of conservation concern and support the highest diver-
sity of landbird species of all habitats in the Southwest
(Rich and others, 2004). Dam operations affect birds
within the ecoregion primarily through effects on breed-
ing habitat. Under the MLFF alternative, these impacts
are likely to be fairly minor compared with climate and
habitat changes outside the Colorado River corridor.
Thus, the Grand Canyon ecoregion is likely to continue
to be an important resource for riparian birds. A well-
designed monitoring program that takes into account
sampling design and statistical power can be used to
establish baseline values regarding the distribution and
abundance of specific species from which future com-
parisons can be made over time. If monitoring data are
linked to information regarding ecological resources and
habitat requirements for specific species and the moni-
toring is conducted in conjunction with more regional,
large-scale monitoring, insight into the causes of popula-
tion changes and the effects of management actions may
result (Holmes and others, 2005).
Dam operations have been shown to be directly
linked to overwintering aquatic birds through effects on
primary and secondary productivity; thus, they may be a
useful resource to monitor. Overwintering aquatic birds
can be monitored relatively easily because more than
50% of the aquatic birds occur at or above Lees Ferry
in a typical winter. Further study is necessary, however,
to determine how to structure any future aquatic bird
monitoring program (Spence, 2004).
Continued monitoring of species of special concern
would require continuation or development of monitor-
ing protocols specific to each species. In particular, the
southwestern willow flycatcher population in Grand
Canyon is extremely small relative to the current range-
wide population, which encompasses approximately
1,400 territories (U.S. Geological Survey, unpub. data,
2004). Because flycatcher monitoring must follow a
standard, multivisit protocol, conducting such surveys
within the study area requires substantial resources.
Overlaying this protocol is the challenge of relating river
flows to any direct or indirect impacts to the flycatcher
and its habitats. The potential impacts of tamarisk
removal associated with riparian restoration projects
should also be considered. Therefore, the nature and
extent of future flycatcher monitoring, and the ability to
interpret its results, may be worthy of discussion within
the Glen Canyon Dam Adaptive Management Program.
Bald eagles have not been systematically or intensively
monitored along the Colorado River since 1995; never-
theless, eagles are still noted during some winter research
raft trips (Yard, 2004b). van Riper and Sogge (2004)
evaluated various monitoring techniques and noted that
helicopter-based surveys would be the most effective
method in terms of coverage and ability to detect the
eagles. On the other hand, aerial surveys would have to
be considered in light of potential recreation issues and
current and future Grand Canyon National Park policies.
Logistical difficulties associated with access make sur-
veying for peregrine falcons in Grand Canyon National
Park extremely difficult, and a thorough sample using
unbiased or random methodologies has been impossible
U.S. Geological Survey
134 The State of the Colorado River Ecosystem in Grand Canyon
(Ward, 2000). Despite these difficulties, the National
Park Service will monitor at least five territories within
the park (R.V. Ward, Grand Canyon National Park, oral
commun., 2005), and Glen Canyon National Recreation
Area will monitor one or more territories above Lees
Ferry starting in 2005.
Many factors and processes apart from dam opera-
tions affect the structure and functioning of the Grand
Canyon ecosystem, such as changes in regional climatic
and atmospheric conditions, natural disturbances, adja-
cent land uses, the spread of invasive species, and fire
suppression. These natural and human-caused events,
along with adaptive management actions, have affected
and will continue to affect the abundance, distribution,
and composition of the Grand Canyon bird communi-
ties and their habitats.
References
Arizona Game and Fish Department, 2004, Unpub-
lished abstract compiled and edited by the Heritage
Data Management System: Phoenix, Arizona Game
and Fish Department, 5 p.
Beatty, G.L., 2001, Arizona bald eagle winter counts:
Arizona Game and Fish Department special report, 43
p.
Brown, B.T., 1987, Monitoring bird population densi-
ties along the Colorado River in Grand Canyon: 1987
breeding season: U.S. Department of Commerce,
National Technical Information Service Report PB89-
103311, 26 p.
Brown, B.T., 1988, Breeding ecology of a willow fly-
catcher population in Grand Canyon, Arizona: West-
ern Birds, v. 19, p. 25–33.
Brown, B.T., 1989, Breeding ecology of riparian birds
along the Colorado River in Grand Canyon, Ari-
zona. Technical Report No. 25: Tucson, University of
Arizona, Cooperative National Park Resources Studies
Unit.
Brown, B.T., 1991a, Abundance, distribution, and ecol-
ogy of nesting peregrine falcons in Grand Canyon
National Park, Arizona: U.S. Department of the Inte-
rior, National Park Service, Grand Canyon National
Park, Ariz., 45 p.
Brown, B.T., 1991b, Status of nesting willow flycatcher
along the Colorado River from Glen Canyon Dam to
Cardenas Creek, Arizona. Endangered Species Report
20: Phoenix, Ariz., U.S. Fish and Wildlife Service.
Brown, B.T., 1992, Nesting chronology, density and
habitat use of black-chinned hummingbirds along the
Colorado River, Arizona: Journal of Field Ornithol-
ogy, v. 63, p. 393–400.
Brown, B.T., 1993, Winter foraging ecology of bald
eagles in Arizona: Condor, v. 95, p. 132–138.
Brown, B.T., 1994, Rates of brood parasitism by brown-
headed cowbirds on riparian passerines in Arizona:
Journal of Field Ornithology, v. 65, p. 160–168.
Brown, B.T., Carothers, S.W., and Johnson, R.R., 1983,
Breeding range expansion of Bell’s vireo in Grand
Canyon, Arizona: Condor, v. 85, p. 499–500.
Brown, B.T., Carothers, S.W., and Johnson, R.R., 1987,
Grand Canyon birds: Tucson, University of Arizona
Press, 302 p.
Brown, B.T., and Johnson, R.R., 1985, Glen Canyon
Dam, fluctuating water levels, and riparian breeding
birds: the need for a management compromise on the
Colorado River in Grand Canyon: USDA Forest Ser-
vice General Technical Report RM-120, p. 76–80.
Brown, B.T., and Johnson, R.R., 1987, Fluctuating flows
from Glen Canyon Dam and their effect on breed-
ing birds of the Colorado River: U.S. Department of
Commerce, National Technical Information Service,
NTIS PB88-183512/AS, 95 p.
Brown, B.T., Mesta, R., Stevens, L.E., and Weisheit, J.,
1989, Changes in winter distribution of bald eagles
along the Colorado River in Grand Canyon, Arizona:
Journal of Raptor Research, v. 23, p. 110–113.
Brown, B.T., and Stevens, L.E., 1992, Winter abun-
dance, age structure, and distribution of bald eagles
along the Colorado River, Arizona: Southwestern
Naturalist, v. 37, p. 404–408.
Brown, B.T., and Stevens, L.E., 1997, Winter bald eagle
distribution is inversely correlated with human activity
along the Colorado River, Arizona: Journal of Raptor
Research, v. 31, p. 7–10.
Brown, B.T., Stevens, L.E., and Yates, T.A., 1998, Influ-
ences of fluctuating river flows on bald eagle foraging
behavior: Condor, v. 100, p. 745–748.
Birds of the Colorado River in Grand Canyon 135
Brown, B.T., and Trosset, M.W., 1989, Nesting-habitat
relationships of riparian birds along the Colorado
River in Grand Canyon, Arizona: Southwestern
Naturalist, v. 34, p. 260–270.
Brown, D.E., 1994, Biotic communities: southwestern
United States and northwestern Mexico: Salt Lake
City, University of Utah Press.
Carothers, S.W., and Aitchison, S.W., eds., 1976, An eco-
logical survey of the riparian zones of the Colorado
River between Lees Ferry and the Grand Cliffs: final
report. Grand Canyon National Park Technical Report
Series No. 10: Washington, D.C., U.S. Department of
the Interior, National Park Service (Grand Canyon
National Park, Colorado River Research Series, contri-
bution no. 38, xiii + 251 p., Grand Canyon, Ariz.).
Carothers, S.W., and Brown, B.T., 1991, The Colorado
River through Grand Canyon: Tucson, University of
Arizona Press, 235 p.
Carothers, S.W., and Sharber, N.J., 1976, Birds of the
Colorado River, in Carothers, S.W., and Aitchison,
S.W., eds., An ecological survey of the riparian zone of
the Colorado River between Lees Ferry and the Grand
Wash Cliffs, Arizona. Colorado River Research Tech-
nical Report No. 10: Grand Canyon National Park,
Ariz., p. 109–122.
Clover, E.U., and Jotter, L., 1944, Floristic studies in the
canyon of the Colorado and tributaries: American
Midland Naturalist, v. 32, p. 591–642.
Cluer, B.L., 1997, Eddy bar responses to the sediment
dynamics of pool-riffle environments: Final report to
U.S. Department of the Interior, National Park Ser-
vice, Grand Canyon National Park, Ariz.
Dolan, R., Howard, A., and Gallenson, A., 1974, Man’s
impact on the Colorado River in the Grand Canyon:
American Scientist, v. 62, p. 392–401.
Ellis, D.H., and Monson, G., 1989, A historical review of
peregrine falcon breeding and summering records for
Arizona and the Navajo Indian Reservation: South-
western Naturalist, v. 34, p. 565–569.
Flowers, S., 1959, Vegetation of Glen Canyon, in
Woodbury, A.M., ed., Ecological studies of flora and
fauna in Glen Canyon. University of Utah Anthropo-
logical Papers No. 40 (Glen Canyon Series No. 7): Salt
Lake City, University of Utah Press, p. 21–61.
Grahame, J.D., and Pinnock, C.A., 1995, Breed-
ing birds along the Colorado River through
Glen Canyon—past and present: final report to
Glen Canyon Environmental Studies, Bureau of
Reclamation: Resource Management Division, Glen
Canyon National Recreation Area, 43 p.
Grubb, T.G., and Kennedy, C.E., 1982, Bald eagle
winter habitat on southwestern national forests. USDA
Forest Service Research Paper RM-237: Fort Collins,
Colo., Rocky Mountain Forest and Range Experimen-
tal Station, 13 p.
Holmes, J.A., Johnson, M.J., and van Riper, C., III,
2005, Evaluation of the breeding riparian birds moni-
toring program for the Colorado River ecosystem:
Final report to U.S. Department of the Interior, U.S.
Geological Survey, Southwest Biological Science Cen-
ter, Grand Canyon Monitoring and Research Center,
Flagstaff, Ariz.
Hualapai Tribe and SWCA, Inc., 1995, Breeding bird
status along the Colorado River in lower Grand
Canyon: Final report submitted to Glen Canyon Envi-
ronmental Studies.
Hunt, W.G., Driscoll, D.E., Bianchi, E.W., and Jackson,
R.E., 1992, Ecology of bald eagles in Arizona: Santa
Cruz, Calif., Biosystems Analysis, Inc.
Johnson, M.J., 2000, Final report: southwestern willow
flycatcher surveys conducted along the Colorado River
corridor in Grand Canyon National Park (1998–2000):
U.S. Geological Survey, Biological Resources Dis-
cipline, Forest and Rangeland Ecosystem Science
Center, Colorado Plateau Field Station/Northern
Arizona University, 18 p.
Kearsley, M.J.C., and Ayers, T.J., 1996, Second year
transition monitoring of riparian vegetation from Glen
Canyon to Pearce Ferry: Final report submitted to
the Grand Canyon Monitoring and Research Center,
Flagstaff, Ariz., 35 p.
Kearsley, M.J.C., and Ayers, T.J., 1999, Riparian veg-
etation responses: snatching defeat from the jaws of
victory and vice versa, in Webb, R.H., Schmidt, J.C.,
Marzolf, G.R., and Valdez, R.A., eds., The controlled
flood in Grand Canyon: Washington, D.C., American
Geophysical Union, Geophysical Monograph Series, v.
110, p. 309–327.
136 The State of the Colorado River Ecosystem in Grand Canyon
Kearsley, M.J.C., and Ayers, T.J., 2001, Review assess-
ment and recommendations regarding terrestrial
riparian vegetation monitoring in the Colorado River
corridor of Grand Canyon: Final report submitted to
the Grand Canyon Monitoring and Research Center,
Flagstaff, Ariz., 79 p.
Kearsley, M.J.C., Cobb, N., Yard, H., Lightfoot, D.,
Brantley, S., Carpenter, G., and Frey, J., 2004, Inven-
tory and monitoring of terrestrial riparian resources
in the Colorado River corridor of Grand Canyon: an
integrative approach: Draft final report to the U.S.
Geological Survey Grand Canyon Monitoring and
Research Center, 218 p.
Kearsley, M.J.C., and Lightfoot, D., 2004, Integration
and interpretation of vegetation and faunal abundance
patterns, in Kearsley, M.J.C., Cobb, N., Yard, H.,
Lightfoot, D., Brantley, S., Carpenter, G., and Frey, J.,
2004, Inventory and monitoring of terrestrial riparian
resources in the Colorado River corridor of Grand
Canyon: an integrative approach: Draft final report to
the U.S. Geological Survey Grand Canyon Monitoring
and Research Center, p. 127–156.
Leibfried, W.C., and Montgomery, W.L., 1993, Regu-
lated flows, trout spawning, and abundance of bald
eagles on the Colorado River, Grand Canyon National
Park, in Rowlands, P., van Riper, C., III, and Sogge,
M.K., eds., Proceedings of the First Biennial Confer-
ence on Research in Colorado Plateau National Parks:
National Park Service Transactions and Proceedings
Series NPS/NRNAU/NRTP-92/10, p. 37–48.
McLeod, M.A., T.J. Koronkiewicz, B.T. Brown, and S.W.
Carothers. 2005. Southwestern Willow Flycatcher
surveys, demography, and ecology along the lower
Colorado River and tributaries, 2004. Annual report
submitted to U.S. Bureau of Reclamation, Boulder
City, NV, by SWCA Environmental Consultants,
Flagstaff, AZ. 155 pp.
Munzer, O.M., English, H.C., Smith, A.B., Woodward,
A.A., and Tudor, A.T., 2005, Southwestern willow
flycatcher 2004 survey and nest monitoring report:
Arizona Game and Fish Department Technical Report
244, 65 p.
Paradzick, C.E., and Woodward, A.A., 2003, Distribu-
tion, abundance, and habitat characteristics of south-
western willow flycatchers (Empidonax traillii extimus) in
Arizona, 1993–2000: Studies in Avian Biology, v. 26,
p. 22–29.
Petterson, J., and Spence, J.R., 1997, 1996 avian com-
munity monitoring in the Grand Canyon: Final report
submitted to Grand Canyon Monitoring and Research
Center, Flagstaff, Ariz., National Park Service, 34 p.
Pucherelli, M.J., 1986, Evaluation of riparian vegeta-
tion trends in the Grand Canyon using multitem-
poral remote sensing techniques. Final report: U.S.
Department of the Interior, Bureau of Reclamation,
Glen Canyon Environmental Studies Report
GCES/18/87, NTIS PB88-183470.
Rich, T.D., Beardmore, C.J., Berlanga, H., Blancher, P.J.,
Bradstreet, M.S.W., Butcher, G.S., Demarest, D.W.,
Dunn, E.H., Hunter, W.C., Inigo-Elias, E.E., Ken-
nedy, J.A., Martell, A.M., Panjabi, A.O., Pashley, D.N.,
Rosenberg, K.V., Rustay, C.M., Wendt, J.S., and Will,
T.C., 2004, Partners In Flight North American Land-
bird Conservation Plan: Ithaca, N.Y., Cornell Lab of
Ornithology.
Smith, A.B., Dockens, P.E.T., Tudor, A.A., English,
H.C., and Allen, B.L., 2004, Southwestern willow
flycatcher 2003 survey and nest monitoring report:
Arizona Game and Fish Department Technical Report
233, 65 p.
Smith, A.B., Woodward, A.A., Dockens, P.E.T., Martin,
J.S., and McCarthey, T.D., 2003, Southwestern willow
flycatcher 2002 survey and nest monitoring report:
Arizona Game and Fish Department Technical Report
210, 49 p.
Snyder, N.F.R., and Schmitt, N.J., 2002, California
condor (Gymnogyps californianus), in Poole, A., and
Gill, F., eds., The Birds of North America, No. 610:
Philadelphia, Pa., The Birds of North America, Inc.
Sogge, M.K., 2000, Breeding season ecology, in Finch,
D.M., and Stoleson, S.H., eds., Status, ecology, and
conservation of the southwestern willow flycatcher:
USDA Forest Service Rocky Mountain Research Sta-
tion General Technical Report RMRS-GTR-60, p.
57–70.
Birds of the Colorado River in Grand Canyon 137
Sogge, M.K., Felley, D., and Wotawa, M., 1998, Ripar-
ian bird community ecology in the Grand Canyon,
final report: U.S. Geological Survey, Colorado Plateau
Field Station report, 57 p.
Sogge, M.K., and Marshall, R.M., 2000, A survey of
current breeding habitats, in Finch, D.M., and Stole-
son, S.H., eds., Status, ecology, and conservation of
the southwestern willow flycatcher: USDA Forest
Service Rocky Mountain Research Station General
Technical Report RMRS-GTR-60, p. 43–56.
Sogge, M.K., Sferra, S.J., McCarthey, T.D., Williams,
S.O., and Kus, B.E., 2003, Distribution and character-
istics of southwestern willow flycatcher breeding sites
and territories: Studies in Avian Biology, v. 26, p. 5–11.
Sogge, M.K., Tibbitts, T.J., and Petterson, J., 1997, Sta-
tus and breeding ecology of the southwestern willow
flycatcher in the Grand Canyon: Western Birds, v. 28,
p. 142–157.
Spence, J.R., 1997, Breeding bird surveys along the
Colorado River, Glen Canyon, Arizona. 1996 sum-
mary progress report and evaluation of the long-term
monitoring program. Report to Grand Canyon Moni-
toring and Research Center: National Park Service,
Resource Management Division, Glen Canyon
National Recreation Area, 37 p.
Spence, J.R., 2004, The riparian and aquatic bird
communities along the Colorado River from Glen
Canyon Dam to Lake Mead, 1996–2002: Draft 2.0,
final report submitted to Grand Canyon Monitoring
and Research Center, U.S. Geological Survey, Flag-
staff, Ariz.
Spence, J.R., and Pinnock, C.A., 1993, Breeding bird sur-
vey along the Colorado River from Glen Canyon Dam
to Lees Ferry. Final report to Bureau of Reclamation,
Glen Canyon Environmental Studies: National Park
Service, Resource Management Division, Glen Canyon
National Recreation Area, 26 p.
Spence, J.R., Russell, B.K., and Kangus, W.B., 2002,
Bald eagle surveys at Glen Canyon National Rec-
reation Area, Utah and Arizona, 1991–2002: Utah
Birds, v. 16, p. 12–22.
Stalmaster, M.V., 1987, The bald eagle: New York,
Universe Books, 227 p.
Stevens, L.E., and Ayers, T.J., 1994, The effects of
interim flows from Glen Canyon Dam on riparian
vegetation along the Colorado River in Grand Canyon
National Park, Arizona: draft 1994 annual technical
and administrative report: U.S. Department of the
Interior, National Biological Service, Colorado Plateau
Research Station, 45 p.
Stevens, L.E., Ayers, T.J., Bennett, J.P., Christensen, K.,
Kearsley, M.J.C., Meretsky, V.J., Phillips, A.M., III,
Parnell, R.A., Spence, J., Sogge, M.K., Springer, A.E.,
and Wegner, D.L., 2001, Planned flooding and Colo-
rado River riparian trade-offs downstream from Glen
Canyon Dam, Arizona: Ecological Applications, v. 11,
p. 701–710.
Stevens, L.E., Buck, K.A., Brown, B.T., and Kline, N.C.,
1997a, Dam and geomorphological influences on
Colorado River waterbird distribution, Grand Can-
yon, Arizona, USA: Regulated Rivers: Research &
Management, v. 13, p. 151–169.
Stevens, L.E., Schmidt, J.C., Ayers, T.J., and Brown, B.T.,
1995, Flow regulation, geomorphology and Colorado
River marsh development in the Grand Canyon,
Arizona: Ecological Applications, v. 5, p. 1025–1039.
Stevens, L.E., Shannon, J.P., and Blinn, D.W., 1997b,
Colorado River benthic ecology in Grand Canyon,
Arizona: dam, tributary and geomorphological influ-
ences: Regulated Rivers: Research & Management, v.
13, p. 129–149.
Topping, D.J., Schmidt, J.C., and Vierra, L.E., Jr., 2003,
Computation and analysis of the instantaneous-dis-
charge record for the Colorado River at Lees Ferry,
Arizona—May 8, 1921, through September 30, 2000:
U.S. Geological Survey Professional Paper 1677, 118 p.
Turner, R.M., and Karpiscak, M.M., 1980, Recent
vegetation changes along the Colorado River between
Glen Canyon Dam and Lake Mead, Arizona: U.S.
Department of the Interior, U.S. Geological Survey
Professional Paper 1132.
U.S. Fish and Wildlife Service, 1984, American peregrine
falcon recovery plan (Rocky Mountain/Southwest
population). Prepared in cooperation with the Ameri-
can Peregrine Falcon Recovery Team: Denver, Colo.,
U.S. Fish and Wildlife Service, 105 p.
138 The State of the Colorado River Ecosystem in Grand Canyon
U.S. Fish and Wildlife Service, 2002, Southwestern wil-
low flycatcher final recovery plan: Albuquerque, N.
Mex., U.S. Fish and Wildlife Service.
van Riper, C., III, and Sogge, M.K., 2004, Bald eagle
abundance and relationship to prey base and human
activity along the Colorado River in Grand Canyon
National Park, in van Riper, C., III, and Cole, K.,
eds., The Colorado Plateau: cultural, biological and
physical research: Tucson, University of Arizona
Press, p. 163–185.
Ward, R.V., 2000, Abundance and distribution of per-
egrine falcons in Grand Canyon National Park: Final
report, U.S. Department of the Interior, National Park
Service, Grand Canyon National Park, Ariz.
Waring, G.L., 1995, Current and historical riparian
vegetation trends in Grand Canyon, using multitempo-
ral remote sensing analyses of GIS sites: Final report,
National Park Service, 24 p.
Webb, R.H., Wegner, D.L., Andrews, E.D., Valdez, R.A.,
and Patten, D.T., 1999, Downstream effects of Glen
Canyon Dam on the Colorado River in Grand Can-
yon: a review, in Webb, R.H., Schmidt, J.C., Marzolf,
G.R., and Valdez, R.A., eds., The controlled flood in
Grand Canyon: Washington, D.C., American Geo-
physical Union, Geophysical Monograph Series, v.
110, p. 1–22.
White, C.M., Clum, N.J., Cade, T.J., and Hunt, W.G.,
2002, Peregrine falcon (Falco peregrinus), in Poole, A., and
Gill, F., eds., The Birds of North America, No. 660:
Philadelphia, Pa., The Birds of North America, Inc.
Yard, H., 2004a, Southwestern willow flycatcher surveys,
in Kearsley, M.J.C., Cobb, N., Yard, H., Lightfoot, D.,
Brantley, S., Carpenter, G., and Frey, J., eds., Inven-
tory and monitoring of terrestrial riparian resources
in the Colorado River corridor of Grand Canyon: an
integrative approach: Draft final report to the U.S.
Geological Survey Grand Canyon Monitoring and
Research Center, p. 185–189.
Yard, H., 2004b, Overwintering birds, waterfowl, and
raptors, in Kearsley, M.J.C., Cobb, N., Yard, H.,
Lightfoot, D., Brantley, S., Carpenter, G., and Frey, J.,
eds., Inventory and monitoring of terrestrial riparian
resources in the Colorado River corridor of Grand
Canyon: an integrative approach: Draft final report to
the U.S. Geological Survey Grand Canyon Monitoring
and Research Center, p. 173–184.
Yard, H., and Blake, J.G., 2004, Breeding bird assess-
ment and surveys and monitoring, in Kearsley, M.J.C.,
Cobb, N., Yard, H., Lightfoot, D., Brantley, S.,
Carpenter, G., and Frey, J., eds., Inventory and moni-
toring of terrestrial riparian resources in the Colo-
rado River corridor of Grand Canyon: an integrative
approach: Draft final report to the U.S. Geological
Survey Grand Canyon Monitoring and Research
Center, p. 97–122.
Yard, H., van Riper, C., III, Brown, B.T., and Kearsley,
M.J., 2004, Diets of insectivorous birds along the
Colorado River in Grand Canyon, Arizona: Condor, v.
106, p. 106–115.
Contact Information:
Jennifer A. Holmes
Wildlife Biologist, Senior
Northern Arizona University
Colorado Plateau Research Station
Flagstaff, AZ
Jennifer.Holmes@nau.edu
John R. Spence
Ecologist
U.S. Department of the Interior
National Park Service
Glen Canyon National Recreation Area
Page, AZ
John_Spence@nps.gov
Mark K. Sogge
Ecologist
U.S. Department of the Interior
U.S. Geological Survey
Southwest Biological Science Center
Flagstaff, AZ
mark_sogge@usgs.gov
First page photograph credit: George Andrejko,
Arizona Game and Fish Department
Chapter 8
Debris Flows
in Grand
Canyon and
the Rapids
of the
Colorado River
Robert H. Webb
Peter G. Griffiths
Christopher S. Magirl
Thomas C. Hanks
Introduction
Coarse sediment—gravel, cobbles, and boulders—is
transported to the Colorado River almost exclusively
by debris flows, which are irregularly occurring types of
flash flood events. By supplying boulders that exceed the
capacity of the river to move them at most discharges,
debris flows create and maintain the hundreds of debris
fans and associated rapids that control the lengthwise
or longitudinal profile of the Colorado River in Grand
Canyon. Debris flows occur in 740 tributaries of the
Colorado River in Grand Canyon between Lees Ferry (at
RM 0) and the Grand Wash Cliffs (at RM 277), the phys-
ical feature that marks the western boundary of Grand
Canyon National Park and the end of Grand Canyon.
Coarse sediment is of interest within the Glen
Canyon Dam Adaptive Management Program because
of its relation to key components of the Colorado River
ecosystem. The deposition of coarse-grained sediment at
tributary junctures builds large debris fans that constrict
the river and form rapids. Debris fans and debris bars,
which develop below rapids, create the fan-eddy complex
that is the cornerstone of the physical framework of the
river in Grand Canyon (fig. 1). In addition, the pools
upstream and downstream of debris fans slow sedi-
ment movement or trap it for temporary storage. The
pool-drop system created by debris fans is prime habitat
for the endangered humpback chub (Gila cypha), while
coarse sediment injected into the river during debris
flows is used by other aquatic organisms, notably the alga
Cladophora glomerata. The navigation of the river by white-
water boaters also can be affected by debris-flow events.
Monitoring the input of coarse sediment into the
Colorado River ecosystem and its long-term redistribu-
tion by the river is critical to understanding how dam
operations affect coarse sediment deposition and, indi-
rectly, other ecosystem components. Scientists are able to
model debris-flow magnitude and frequency from exten-
sive data sets developed through long-term monitoring.
Also, this chapter estimates the amount of sediment
contributed by debris flows and models its deposition
at tributary junctures to evaluate the effects of debris
flows over several temporal and spatial scales, includ-
ing the recent period of operations of Glen Canyon
Dam. Data are combined with modeling to evaluate
long-term changes in rapids and to explain large-scale
features. The chapter also summarizes data from debris-
fan monitoring activities by the U.S. Geological Survey’s
140 The State of the Colorado River Ecosystem in Grand Canyon
(USGS) Grand Canyon Monitoring and Research
Center (GCMRC) and research by Water Resources and
Geology Discipline scientists. Finally, the chapter consid-
ers the role of experimental high flows and the modi-
fied low fluctuating flow (MLFF) alternative on coarse-
sediment reworking.
Background
Distributed along 277 mi (446 km) of river between
the Paria River and Grand Wash Cliffs, the 740 tributar-
ies that produce debris flows drain 4,600 mi2 (12,000
km²) of steep terrain between the north and south rims
of Grand Canyon (Webb and others, 2000). Debris
flows, which are typically more than 80% sediment by
weight, are slurries of clay to boulder-sized sediment
mobilized during periods of intense or sustained pre-
cipitation. The exposed bedrock landscape of Grand
Canyon National Park provides an ideal setting for the
initiation of debris flows: high relief combines with
differential rock strength to create a high potential for
slope failure (Griffiths and others, 2004). Most slope
failures that become debris flows (75%) occur in the
Hermit Formation and Esplanade Sandstone of the
Supai Group and in the Muav Limestone and Bright
Angel Shale of the Tonto Group (fig. 2a). Other promi-
nent sources include the Dox Sandstone, Cardenas Lava,
Vishnu Schist, and Quaternary Basalts in western Grand
Canyon. Tributaries are documented to have produced
debris flows throughout the Holocene (Melis and others,
1994; Hereford and others, 1998).
In Grand Canyon, debris flows are initiated by a
combination of intense precipitation and subsequent
Figure 1. Diagram showing a fan-eddy complex in Grand Canyon. Debris flows from tributary canyons carry coarse sediment that
is deposited at the juncture with the Colorado River, forming deposits called debris fans. Debris fans constrict the Colorado River
and raise its bed elevation, creating rapids. Especially during floods, the river entrains the sediment on the debris fan and transports
it downstream through the pool, where the larger particles become lodged on debris bars that form secondary rapids. Between the
constrictions of the primary and secondary rapids, pools and eddies form, creating a depositional setting for sandbars.
Debris Flows in Grand Canyon and the Rapids of the Colorado River 141
slope failure (Cooley and others, 1977; Webb and
others, 1988, 1989, 1999b, 2000, 2004; Griffiths and
others, 2004). The most common type of slope failure
is termed the “firehose effect” (Melis and others, 1994)
(fig. 2b), where streamflow falling over cliffs, typically
in the Redwall Limestone, strikes bedrock and accu-
mulated colluvium and causes slope failure and mixes
these materials and water to form a slurry. Debris flows
that reach the Colorado River deposit their material
on debris fans. These enlarged or aggraded debris fans
constrict the river and raise the riverbed elevation until
mainstem flows rework coarse-grained deposits (Webb
and others, 1989). “Reworking” is a term describing river
entrainment and transport of particles from debris fans,
including the winnowing of fine-grained particles (clay
to cobble size) and the movement of boulders, either
on the fan surface or into the river. The large boulders
that remain after this reworking form the core of rap-
ids that modify the longitudinal profile of the river and
locally control the physical framework of the present-
day Colorado River in Grand Canyon (Webb, 1996).
Debris-fan reworking was extensive before construc-
tion of Glen Canyon Dam. Reduced peak flow on the
regulated river represents a fourfold decrease in its sedi-
ment-transport potential compared to predam conditions
(Howard and Dolan, 1981). As a result, ability of the
river to erode newly deposited sediment from debris fans
has been reduced. Reworking still occurs on a limited
basis, typically during maximum powerplant releases or
intentional flood releases from Glen Canyon Dam (Webb
and others, 1999a). Today, because the reworking by the
Colorado River is limited, debris flows from unregulated
tributaries are now an effective agent of change in the
river corridor (Howard and Dolan, 1981), affecting the
water-surface profile, hydraulics through rapids, and the
associated pools and eddies downstream.
Status and Trends
Debris-flow Frequency
Debris flows in Grand Canyon were relatively well
documented in the 20th century (Webb, 1996; Webb
and others, 2000; Griffiths and others, 2004). We use
the term Grand Canyon loosely to collectively refer to
the river corridor from Lees Ferry to the Grand Wash
Cliffs, merging Marble and Grand Canyons and their
respective subreaches. Direct observations provided a
complete record of debris flows from 1984 to 2004 (fig.
3), which was augmented with repeat photography, such
as that shown in figure 4, that provides a separate record
(1890 through 1983) of debris flows from 147 tributar-
ies. In this analysis, we only documented debris flows
that reached debris fans and/or the Colorado River; we
did not include debris flows that occurred upstream in
tributaries but did not reach the river corridor.
Direct Observations (1984–2004)
Debris flows, rockfalls, and significant streamflow
floods were directly observed or compiled from the
accounts of river runners along the river in Grand
Canyon from 1984 through 2004. These data provide
a complete record of debris flows that reached the
Colorado River from all Grand Canyon tributaries for
more than 21 yr (fig. 3). During this period, a total of
104 events occurred in 88 tributaries for an average of
4.95 debris flows per year. A total of 14 debris flows
occurred in 2001 and again in 2002, the most prolific
period in the record. Webb and others (2000) analyzed
Figure 2. A. Relative frequency of the location at which slope
failures in bedrock or colluvium resulting in debris flows that
reach the river (n = 101, 1939 through 2003) have occurred in
Grand Canyon. B. Relative frequency of initiation mechanisms for
selected debris flows from 1939 through 2003 in Grand Canyon (n =
68) (from Griffiths and others, 2004).
142 The State of the Colorado River Ecosystem in Grand Canyon
precipitation records around Grand Canyon and found
that the annual number of debris flows is not related to
total summer precipitation (as illustrated by the precipi-
tation record shown in fig. 3 in conjunction with annual
debris flows). This suggests that antecedent moisture
has little effect on debris-flow occurrence (Griffiths and
others, 2004).
Most debris flows occurred in Marble Canyon or
eastern Grand Canyon, with notable exceptions at Lava
Falls Rapid (RM 179) in 1995 (Webb and others, 1999b)
and between RM 189 and RM 209 from 1999 through
2001. Several tributaries delivered more than one debris
flow to the river between 1984 and 2004. For example,
Seventyfive Mile Creek had four debris flows, and
Monument Creek (RM 93.9) had three. Multiple debris
flows within a drainage basin suggest that slope and chan-
nel destabilization caused by the initial event may lead
to repeated events until either the loosened sediment is
removed or sufficient time elapses between severe storms
to allow healing of hillslopes and channel margins.
Repeat Photography and
Debris Flows (1890–1983)
Repeat photography (fig. 4) has been used in numer-
ous studies in Grand Canyon to document long-term
changes in both terrestrial ecology and geomorphology
(Turner and Karpiscak, 1980; Stephens and Shoemaker,
1987; Webb and others, 1989, 1999a; Melis and oth-
ers, 1994; Webb, 1996; Griffiths and others, 2004). This
type of scientific photography is particularly useful for
Figure 3. Observational record of debris flows in Grand Canyon,
from 1984 to 2004, compared with total summer rainfall (July
through September) at Grand Canyon National Park airport. No
data are available for summer precipitation in 1997.
A.
B.
Figure 4. Repeat photographs of Crystal Rapid. A. (February
9, 1890) This downstream view from the right scout point at
Crystal Rapid (RM 98) shows a wide, gentle rapid during the
second expedition through Grand Canyon. This expedition,
led by Robert Brewster Stanton, occurred in winter 1890. The
deepest water in the rapid is on river right, and emergent rocks
are on the left side (R.B. Stanton, courtesy of the National
Archives and Records Administration, College Park, Maryland).
B. (February 1, 1990) A debris flow in 1966 constricted the river
by more than 80%, creating what was considered the most
formidable rapid in Grand Canyon. Floods between 1966 and
1986 widened out the constriction, reducing the navigational
hazard this rapid posed. Although Crystal Rapid has lost some
of its ferocious reputation, it remains one of the largest in Grand
Canyon (T. Brownold, stake 1471).
Debris Flows in Grand Canyon and the Rapids of the Colorado River 143
evaluating the types of landscape changes associated
with debris flows. Most of the information for historical
debris flows was obtained through comparison of repeat
photography and historical photographs taken between
1871 and 1964. Between 1989 and 2002, 1,365 historical
photographs of the river corridor were matched to deter-
mine significant changes to tributary channels, debris
fans, and rapids throughout the canyon. The year with
the most abundant and widespread coverage is 1890,
when the well-documented Stanton expedition occurred
(Webb, 1996). Several sets of low-altitude aerial photo-
graphs taken between 1935 and 1984 were also analyzed
for evidence of debris flows at the river.
To determine the frequency of debris flows at the
river from 1890 through 1983, the 1890 photographs
and their matches were interpreted for evidence of
debris-flow occurrences at 147 debris fans. This process
revealed that debris flows occurred at 84 of 147 tributar-
ies (Griffiths and others, 2004), indicating that 57% of
the tributaries generated one or more events from 1890
through 1983. Because any of these 84 tributaries could
have delivered more than one debris flow, additional
data, such as written accounts, were used to identify a
total of 93 debris flows from the 84 tributaries over a
period of a century. From 1890 through 1983, 6% of
tributaries produced two or more debris flows, includ-
ing five at Lava Falls Rapid (RM 179) (Webb and oth-
ers, 1999b).
Analysis of aerial photograpy identified an addi-
tional 23 debris flows for a total of 107 debris flows that
occurred between 1890 and 1983 at the mouths of 167
tributaries from Glen Canyon Dam to Separation Rapid
(RM 240), the head of Lake Mead. Using this data set
as an unbiased sample of the entire population of 740
tributaries, the rate of debris-flow occurrence at the
river is estimated at 5.0/yr for all tributaries from 1890
through 1983. This rate is essentially identical to the
4.95/yr frequency observed between 1984 and 2004.
If the results for both records are combined, 211 debris
flows are known to have occurred along the Colorado
River in Grand Canyon between 1984 and 2003.
Net Observed Effects of Debris Flows
Of the documented 211 historical debris flows in
172 tributaries, 55 significantly affected the Colorado
River by creating rapids or increasing constrictions
during the past century (Webb, 1996; Webb and others,
2000; Griffiths and others, 2004). From 1984 through
2004, 8 rapids were created, and 15 were constricted
by debris flows. The observational evidence indicates
that the occurrence of debris flows is not spatially
random in Grand Canyon. Debris flow activity is par-
ticularly concentrated in Marble Canyon and other
reaches where the river trends towards the southwest or
south-southwest. The findings indicate that about 10%
of tributaries had two or more debris flows in the last
century, with a maximum of six debris flows at Lava Falls
Rapid (Webb and others, 1999b) and five debris flows at
Seventyfive Mile Creek during the 20th century.
Modeling Debris-flow Frequency
Griffiths and others (1996) developed a model of
debris-flow frequency (1890–1990) in Grand Canyon
between Lees Ferry and Diamond Creek (RM 0 to RM
226), and Webb and others (2000) extended that model
to the Grand Wash Cliffs (RM 277). The model identi-
fied several parameters that are significantly related
to the occurrence of debris flows that reach the river,
including the presence and location of shale in the basin,
drainage-basin area, mean drainage-basin gradient, and
the aspect of the river corridor. Drainage-basin variables
that are the most significant in influencing the occur-
rence of debris flows are suggested in a map showing the
distribution of debris-flow probabilities (fig. 5). One ten-
dency is for debris-flow frequency to decrease when the
river corridor trends away from a southwesterly course;
Griffiths and others (2004) attributed this to the regional
trajectory of summer storms, which tend to move from
the southwest. The effect of drainage-basin area is
evident in Marble Canyon, where the largest tributar-
ies have a higher probability of debris-flow occurrence.
The height of the Hermit Formation and the gradi-
ent from this unit to the Colorado River appear to be
especially important in Marble Canyon and reflect the
dominant contribution of shale units to debris flows in
Grand Canyon.
In eastern Grand Canyon, a greater variety of
source materials, combined with structural variability,
resulted in a mosaic of probabilities (fig. 5). The presence
of and gradient below shales strongly affect debris-flow
probability. River aspect and drainage-basin area are sig-
nificant but less influential in this reach. Certain sections
of the river corridor that trend northwesterly generally
have tributaries with low probabilities. In western Grand
Canyon, the presence of three source formations and
the overall gradient of each tributary from headwaters
to river strongly influenced debris-flow probability. The
height of the Hermit Formation is less influential in
western than eastern Grand Canyon or Marble Canyon
144 The State of the Colorado River Ecosystem in Grand Canyon
because that formation is farther from the river. Debris-
flow probability is lowest downstream from Diamond
Creek, where the river trends northwesterly, except in
the reach immediately upstream from the Grand Wash
Cliffs, where debris-flow probabilities are high because
of the proximity of shales to the river corridor.
Our observations and statistical analyses show that
(1) all 740 Grand Canyon tributaries produce debris
flows, albeit some at a low frequency; (2) about 60% of
tributaries produce one or more debris flows per century;
(3) about 10% of tributaries produce two or more debris
flows per century; and (4) no tributary has produced
more than six debris flows in the last century.
Debris-flow Sediment Yield
Data on debris-flow frequency, volume, and particle-
size distributions were combined to create a model of
debris-flow sediment yield in Grand Canyon (Webb and
others, 2000). Using this model, it is estimated that debris
flows contribute between 155,000 and 325,000 tons/yr
(141,000 and 295,000 Mg/yr) of sediment to debris
fans in Grand Canyon. Marble Canyon contributes the
greatest amount of debris-flow sediment, which is con-
sistent with both empirical observations and the modeled
distribution of debris-flow occurrence in Grand Canyon
(Griffiths and others, 2004).
Modeling debris-flow sediment yield requires a
number of important assumptions. In this case, it was
assumed that all debris flows from a given tributary were
the same size, which means the model does not realisti-
cally depict a magnitude-frequency relation. Further-
more, the sediment-yield model does not account for
extreme events not included in the historical record and
small events that are inadequately represented. Some of
these problems could be resolved by using a fully sto-
chastic model of debris-flow frequency, but objectively
determining model constraints based on the limited data
would be difficult.
Figure 5. Debris-flow probabilities of 740 tributaries of the Colorado River in Grand Canyon (Griffiths and others, 2004). This map
depicts group probabilities of the occurrence of one or more debris flows in a tributary during the century between 1890 and 1990.
Debris Flows in Grand Canyon and the Rapids of the Colorado River 145
Incorporating an average boulder content of 14%
of debris-flow volumes (Melis and others, 1994; Webb
and others, 1999b, 2000), the total boulder delivery from
all 740 tributaries is 1.1 billion ft3 (31 million m3) per
thousand years. Distributing these boulders evenly along
the river corridor without removal, dissolution, or ero-
sion raises the bed by 2.4 ft (0.7 m) per thousand years,
which we consider to be a reasonable order of magni-
tude. To distribute boulders more realistically, deposition
was limited to the areas of recent debris fans, calculating
the area of deposition at each tributary confluence as a
rectangle defined by the length and average width of the
rapid. For each confluence, local bed rise was calculated
by dividing the total volume of sediment delivered by
debris flow by the estimated area of deposition; these
results were reported in Webb and others (2000, 2004).
River Reworking of
Aggraded Debris Fans
In the years immediately following a debris flow,
Grand Canyon rapids are known to be unstable because
of reworking by the Colorado River (Howard and
Dolan, 1981; Kieffer, 1985; Webb and others, 1989,
1999b; Melis and others, 1994). Before closure of Glen
Canyon Dam in 1963, the Colorado River removed most
debris-flow deposits during the early summer floods,
which averaged 82,000 cubic feet per second (cfs) and
were as large as 220,000 cfs. Those flood events swept
all but the largest particles downstream and redepos-
ited cobbles and small boulders on debris bars that
constrained the extent of eddies and controlled second-
ary rapids. The interaction between the frequency and
magnitude of tributary debris flows and mainstem floods
resulted in debris fans and rapids that were relatively
stable in the intervening time periods between debris
flows.
Lava Falls Rapid offers one of the best documented
cases of debris-fan reworking, which occurred during
the 1996 beach/habitat-building flow (fig. 6) and dur-
ing other floods in the predam and postdam periods
(Webb and others, 1999b). Most of the reworking that
occurred during the 1996 event happened as the dis-
charge increased to its peak; reworking slowed markedly
during the first day of peak discharge. Nine radio-tagged
boulders traveled an average distance of 262 yards (240
m) from their initial positions on the Prospect Canyon
debris fan during the 1996 event (Pizzuto and others,
1999; Webb and others, 1999b). Debris-flow deposits in
1939, 1954, 1955, 1963, and 1966 were also reworked
by subsequent floods; some of the aggraded debris fans
(1954, 1963, 1966) were completely removed, while some
of the deposition (1939, 1955) remained to cause persis-
tent changes in Lava Falls Rapid.
While Kieffer (1985) stated that exceptionally
large floods (>400,000 cfs) are required to completely
rework some aggraded debris fans, Magirl and others
(2005) found several examples of debris flows that were
effectively removed by modest floods. For example, an
8 to 10 ft (2.4 to 3.0 m) drop at Doris Rapid was nearly
completely removed by a 220,000-cfs flood in 1921. Also,
a 3.0 ft (0.91 m) riffle at To Hajisho Wash (RM 28.5) in
1923 was completely removed by the 127,000-cfs flood
in 1927. While we do not dispute that the amount of
reworking increases with the magnitude of floods, effec-
tive reworking and redistribution of coarse sediment can
occur at a variety of flood discharges and are heavily
dependent on particle size of the aggraded debris fan,
the elapsed time between debris flow and reworking
flood, and the stream power available to transport sedi-
ment from a specific debris fan.
Webb and others (1999a) reported that reworking
decreases with the time elapsed between the debris flow
and the flood event because average releases may inter-
lock particles into an overlapping network, significantly
increasing the force necessary to dislodge and carry par-
ticles from debris fans. This process that leads to inter-
locked particles, which is the net result of physical rear-
rangement, abrasion of particle-particle contact points,
and differential dissolution at contact points, is termed
“suturing.” Suturing is common on debris fans that have
not had debris-flow aggradation historically, and we have
observed some suturing on recently aggraded debris
fans, such as 18 Mile Wash (1987 debris flow) and 127.6
Mile Wash (1989 debris flow), on the distal margin where
submergence occurs frequently. The documented occur-
rence of suturing provides compelling reason to decrease
the elapsed time between controlled flood releases, if
reworking of aggraded debris fans is a priority.
To understand how rapids have changed over time,
water-surface profiles from 1923 and 2000 were com-
pared to detect geomorphic change (Magirl and others,
2005). Magirl and others (2005) compared the longitu-
dinal profile surveyed in 1923 (U.S. Geological Survey,
1924) with a profile constructed from lidar data taken in
2000. Ninety-one tributary junctures along the Colo-
rado River in Grand Canyon between Lees Ferry and
Diamond Creek were evaluated for change. These sites
represent 39% of all rapids and 67% of named rapids.
At these 91 locations, 11 rapids were known to have not
changed between 1923 and 2000, 6 rapids exhibited a
rise in the elevation at the head of the rapid of 4.6 ft (1.4
m) or more, and the elevation at the head of 2 rapids
146 The State of the Colorado River Ecosystem in Grand Canyon
decreased more than 4.6 ft (1.4 m). The net change at 91
rapids is shown in figure 7. More rapids aggraded (18)
than degraded (7) for elevation changes greater than or
equal to 2.3 ft (0.7 m), which is the threshold of detect-
able elevation change. Moreover, elevation increases
were consistently larger than elevation decreases. Of
the 10 debris fans associated with the largest elevation
increases, 8 were aggraded by one or more known debris
flows since 1923 (table 2 in Magirl and others, 2005). Of
the five debris fans associated with the largest elevation
decreases, only two had debris flows since 1923 (Magirl
and others, 2005). Finally, the average elevation of pools
at the heads of rapids was 0.85 ft (0.26 m) higher in 2000
than in 1923.
Comparison of the 1923 and 2000 profiles also
reveals the interaction between rapids as a result of
debris-flow deposition (fig. 8). The 1966 debris flow
at Crystal Rapid (RM 98) caused a rise in river level
several miles upstream, which drowned out the tailwaves
of Boucher Rapid to create what river runners refer to
as “Lake Crystal.” The 1951 debris flow at Boucher
Rapid had the opposite effect on the tailwaves of
Hermit Rapid. The result was one of the larger hydrau-
lic features in Grand Canyon, the notorious fifth wave in
B.A.
Figure 6. Changes in Lava Falls Rapid (RM 179) during the 1996 beach/habitat-building flow. The river flows from bottom to top in
these views. A. (March 24, 1996) The 1995 debris flow from Prospect Canyon (left side of the views) constricted the river by about 60%.
This view shows the freshly deposited sediments with no vegetation on river left (left side of the view). B. (April 9, 1996) Reworking by
the 1996 beach/habitat-building flow, which had a peak discharge of about 47,500 cfs at Lava Falls Rapid, removed 208,000 ft³ (5,900
m3) of the aggraded debris fan, increasing the width of the rapid by an average of 16 ft (5 m) (photographs courtesy of Grand Canyon
Monitoring and Research Center, U.S. Geological Survey).
Debris Flows in Grand Canyon and the Rapids of the Colorado River 147
to transport particles more than 3 ft (0.91 m) in diameter,
except during maximum powerplant releases or inten-
tional flood releases and even then only at the largest
rapids. Cobbles and boulders carried from debris fans
by the regulated Colorado River appear to be redepos-
ited in the pool immediately downstream of the debris
fans instead of on the debris bar farther downstream
(Pizzuto and others, 1999; Webb and others, 1999b).
This altered pattern of redeposition reflects a change in
the geomorphic framework of the Colorado River; in the
case of Granite Rapid, repeated debris flows and mod-
est reworking from 1984 through 2003 have resulted in a
lengthening of the rapid tailwaves through the pool and
into the secondary riffle.
Many rapids in the Colorado River system have
become larger during the last 30 yr because debris-fan
constrictions and individual boulders cannot be totally
removed by typical dam releases (Graf, 1979; Howard
and Dolan, 1981; Melis and others, 1994; Webb, 1996).
Continued deposition of coarse sediment into the river
channel by debris flows will likely fill deeper pools above
and below rapids while also enhancing the size of eddies;
however, with the notable exception of the Crystal Rapid
debris flow of 1966, most rapids affected by recent
debris-flow aggradation are less hazardous to navigation.
Exposed or shallowly submerged rocks are significant
navigational obstacles, and debris-fan aggradation tends
to narrow the channel, increasing flow depths and either
removing or submerging existing rocks. At the same time,
Figure 7. Net elevation change in 91 rapids of the Colorado River
in Grand Canyon between 1923 and 2000. The locations of rapids
that did not change between 1923 and 2000 and were therefore
used to anchor the change in longitudinal profiles are shown as
inverted triangles.
Figure 8. Comparison of 1923 and 2000 water-surface elevation
profiles in upper Granite Gorge in Grand Canyon. Despite three
debris flows, the head at Granite Rapid has changed little in 77 yr,
but its drop has steepened. In contrast, aggradation from the 1966
debris flow at Crystal Rapid is clearly visible in the comparison, and
this aggradation affects Boucher Rapid upstream. Likewise, a 1951
debris flow at Boucher Creek has affected Hermit Rapid upstream.
Hermit Rapid, which is a compressional wave associated
with the jet of the rapid entering a backwater controlled
by an aggraded Boucher Creek debris fan.
Leopold’s (1969) analysis of the pool-and-rapid
morphology of the Colorado River, which is one of the
more widely referenced figures, presents the cumula-
tive vertical drop of the river as a function of distance
as measured in 1923 for the first 150 mi (241 km) below
Lees Ferry. Leopold concluded that 50% of the total
drop occurred in only 9% of the length of the river. In
2000, 66% of the total drop in river occurred in 9% of
the length over the 227 mi (365 km) below Lees Ferry.
When only the first 150 mi (241 km) of river is consid-
ered for direct comparison with Leopold (1969), 71%
of the total rapid occurs in 9% of the length (fig. 9),
reflecting the greater amount of aggradation in Marble
Canyon and eastern Grand Canyon compared to west-
ern Grand Canyon.
Impacts of Dam Operations
on Aggradation
Operations of Glen Canyon Dam have long been
hypothesized to increase aggradation of the riverbed
by limiting the reworking of debris fans (Howard and
Dolan, 1981; Kieffer, 1985; Webb and others, 1989;
Melis and others, 1994; Griffiths and others, 2004).
Glen Canyon Dam has reduced peak discharges on the
Colorado River, which now has insufficient stream power
148 The State of the Colorado River Ecosystem in Grand Canyon
the drops through these rapids are steeper, and sizes of
waves are typically larger; these changes could lead to
increased incidence of boat flips.
The Longitudinal Profile
of the River
Hanks and Webb (in press) interpreted the longi-
tudinal profile of the Colorado River through Grand
Canyon in relation to debris-flow sedimentation at
rapids. Rapids represent short-wavelength (about 0.6
mi (1 km)), small-amplitude (less than about 16 ft (5 m))
convexities (areas that round outward from the riverbed)
in the longitudinal profile of the river, arising from the
shallow gradient in the upstream pool and the steep
gradient through the rapid itself. The kinds of changes
detected in the comparison of the 1923 and 2000 pro-
files (Magirl and others, 2005) discussed previously can-
not be as easily detected when the longitudinal profile of
the Colorado River is displayed for the length of Grand
Canyon (fig. 10).
Analysis of the entire longitudinal profile through
Grand Canyon reveals two river-profile convexities
that are long-wavelength (about 62 mi (about 100 km)),
large-amplitude (49 to 98 ft (15 to 30 m)) river-profile
convexities (Hanks and Blair, 2003; Hanks and Webb, in
press): the eastern canyon convexity between RM 30 and
RM 80 and the western canyon convexity between RM
160 and RM 250. Both of these convexities are easily
discernable in figures 11 and 12. These large-amplitude
convexities have strong spatial correlations with high
probabilities of debris-flow occurrence, high densities of
debris fans, and the largest debris fans along the river.
Convexities of intermediate scale are also identified in
the longitudinal profile. River-profile convexities require
an active and powerful geologic process to maintain
them, in this case the abundant, frequent, and volumi-
nous debris-flow activity in Grand Canyon. Presumably
for all of the Holocene and at least some of the late
Pleistocene, the Colorado River has been expending its
energy transporting sediment within Grand Canyon,
integrating short-wavelength convexities into long-wave-
Figure 9. Comparison between the cumulative vertical drop
of the river in 1923 (data from Leopold, 1969) and the cumulative
vertical drop in 2000 (Magirl and others, 2005).
Figure 10. The longitudinal profile of the Colorado River through
Marble and Grand Canyons (Lees Ferry to Diamond Creek) as
surveyed in 1923 (U.S. Geological Survey, 1924) and as measured by
using lidar in 2000 (from Magirl and others, 2005; Hanks and Webb,
in press). At this scale, differences in the 1923 and 2000 profiles (see
fig. 8) are not apparent. The straight line represents the average
river gradient and illustrates profile convexities that appear to be
related to debris-flow deposition in the Colorado River.
Debris Flows in Grand Canyon and the Rapids of the Colorado River 149
Hanks and Webb (in press) compared debris-flow
probabilities (calculated from fig. 5) averaged over 12.4-
mi (20-km) intervals with this difference profile and
found that the highest debris-flow probabilities have a
strong association with intermediate and long-wave-
length convexities. The falling limbs of the Eastern and
Western Canyon convexities are both associated with
decreasing debris-flow probabilities as they enter the
upper and lower Granite Gorge, respectively, reflecting
the decrease in the rate of tributary sediment delivery
in these reaches. As shown in figure 12, the reach-aver-
aged sediment yield from the debris-flow sediment-yield
model is also associated with profile convexities, although
the association is less than for the probabilities alone.
Several overall characteristics of the river corri-
dor appear to be associated with the characteristics of
long-wavelength convexities. The presence of abundant
emergent islands in the river is associated with the tops
of the convexities where overall slopes are relatively low.
Cultural sites on fine-grained sediment deposits appear
to be most common on the tops of large-scale convexities
as well. The largest rapids on the river appear to be asso-
ciated with the falling limbs of convexities (Hanks and
Webb, in press). This latter characteristic may result from
both the increased reach-scale gradient on the falling
limbs of the convexities and a greater spacing between
rapids, which minimizes interaction between the drops.
Summary and Management
Implications
Debris flows transport poorly sorted sediment onto
debris fans in the Colorado River at a frequency that
varies through Grand Canyon. Historically, an aver-
age of 5.0 debris flows per year has occurred in Grand
Canyon. The occurrence of these debris flows does not
appear to be related to seasonal precipitation amounts.
Modeling debris-flow frequency in Grand Canyon based
on the interpretation of 1,365 photographs of the river
corridor yielded frequency information in 167 of 740
tributaries (23%). Of the 167 tributaries, 98 (59%) had
debris flows during the last 100 yr.
Frequency estimates indicate that 57% of the
tributaries had at least one debris flow per century, while
about 10% of the tributaries had a frequency of more
than two debris flows per century. Estimates of sediment
yield to the Colorado River in Grand Canyon by debris
flow are as high as 3.3 million tons (3.0 million Mg) of
sediment per decade, of which 452,000 tons (410,000
Mg) are boulders larger than 10 inches (>256 mm) in
Figure 11. A profile of the difference in elevation (ft) between the
longitudinal profile of the Colorado River between Lees Ferry and
the Grand Wash Cliffs and its average gradient (both are shown in
fig. 10). The Western Canyon offset shows that this convexity likely
has a larger magnitude if the true bedrock profile were known.
Convexities discussed in this chapter are shown on the profile.
Figure 12. A profile of the difference in elevation (ft) between the
Colorado River and an average gradient (fig. 11) compared with a
reach-averaged sediment yield for debris flows as calculated by
using a stochastic model (Webb and others, 2000).
length convexities. This suggests that little or no bedrock
incision has occurred during about the last 11,000 yr.
Detrending of the longitudinal profile (fig. 11)
reveals the magnitude of convexities and allows com-
parison with other features along the river corridor. For
example, the largest convexities coincide with reaches
that contain 13 of the 14 largest debris fans (Hanks and
Webb, in press). In addition, the locations and areas of
the 444 Holocene debris fans between Lees Ferry and
Diamond Creek (Melis, 1997) coincide with fluctuations
in the difference profile (Hanks and Webb, in press).
150 The State of the Colorado River Ecosystem in Grand Canyon
diameter. Distributed evenly throughout the river corri-
dor, these boulders would raise the bed in Grand Canyon
by 2.3 ft per thousand years (0.72 m per thousand years).
If deposition is limited to existing debris fans at tribu-
tary mouths, these boulders would raise the bed by an
average of 8.66 ft (2.64 m) per thousand years at each
confluence.
By combining the frequency model with relations
for debris-flow volume and particle-size distribution,
debris-flow sediment yields were calculated for several
time periods. On average, debris flows deliver between
0.15.106 and 0.33.106 tons/yr (0.14.106 and 0.30.106 Mg/
yr) of sediment to the main channel. Although debris
flows deliver only 23,142 to 48,488 tons/yr (21,000 to
44,000 Mg/yr) of boulders to the river, these boulders
control the longitudinal profile and geomorphic frame-
work of the river, defining debris fans, rapids, and related
sandbars, and are unlikely to be removed by regulated
flows. Moreover, the effects of debris flows are shown to
affect the river on several length and temporal scales.
Comparison of the two water-surface profiles (one
surveyed in 1923 and one in 2000) showed a change in
80 rapids. The average elevation of pools at the heads
of rapids was 0.85 ft (0.26 m) higher between 1923 and
2000, indicating net aggradation of the coarse-grained
sediment forming the rapids throughout Grand Canyon.
Furthermore, comparison of the two water-surface pro-
files showed enhanced pool-and-rapid structure; while
50% of the total drop of the river occurred in just 9%
of the river distance in 1923, that figure increased to
66% by 2000. Reconstruction of water-surface profiles
showed that debris-flow deposition can also have large
upstream effects, particularly in the cases of reducing
gradients between rapids and reducing the fall in the
upstream rapid.
Analysis of the entire longitudinal profile through
Grand Canyon reveals convexities that reflect sustained
debris-flow deposition. Specifically, there are two long-
wavelength (about 62 mi (about 100 km)), large-ampli-
tude (49 to 98 ft (15 to 30 m)) river-profile convexities:
the eastern canyon convexity between RM 30 and RM
80 and the western canyon convexity between RM 160
and RM 250. These large-amplitude convexities have
strong spatial correlations with high probabilities of
debris-flow occurrence, high densities of debris fans, and
the largest debris fans along the river. These convexi-
ties appear to be maintained by debris-flow activity in
Grand Canyon, presumably for all of the Holocene. In
this period, the Colorado River has been expending its
energy transporting sediment within Grand Canyon and
integrating short-wavelength convexities into long-wave-
length convexities, with little or no bedrock incision.
Coarse-sediment fill in the channel of the Colorado
River at any wavelength has its origins in the rapids,
which result from the tributary debris flows and fans that
feed them. In contrast, the amount of fill in the rap-
ids, either individually or collectively, is a small volume
compared to the fill that creates the longer wavelength
convexities. Periodic channel maintenance floods are
not likely to perform the transport necessary to create
the long-wavelength convexities, although reworking of
locally aggraded debris fans is clearly feasible. The river-
reworking processes by which point-source contribu-
tions of debris-flow sediment are aggregated into longer
wavelength convexities are as yet unknown and certainly
involve a more detailed understanding of the flood
dynamics of the predam river. Nevertheless, the close
spatial associations of the longer wavelength convexi-
ties with the locations and sizes of debris fans and with
the frequency of debris flows lead to the conclusion that
unusual accumulations of debris fill in the channel are
their principal cause, just as they are for the rapids.
In terms of adaptive management and operations of
Glen Canyon Dam, reworking of aggraded debris fans
has been shown to be feasible. Reworking has been docu-
mented during modified low fluctuating flow releases,
maximum powerplant releases, and flood releases up
to 47,500 cfs. Both reworking and transport capac-
ity increase with increasing discharge, which suggests
that flood releases larger than powerplant capacity of
about 33,500 cfs are more efficient than smaller events.
Because reworking mostly occurred in the rising limb
of flood hydrographs, large-magnitude floods designed
for debris-fan reworking do not have to have signifi-
cant duration. As shown in figure 13, a flood designed
Figure 13. Hypothetical Glen Canyon Dam release and flow
hydrographs designed to create a peak discharge of 100,000
cfs for 1 min at Diamond Creek, followed by a beach-building
discharge of 45,000 cfs for a duration of 2 d.
Debris Flows in Grand Canyon and the Rapids of the Colorado River 151
to rework debris fans could have a peak discharge of
100,000 cfs for only 1 min at Diamond Creek then drop
rapidly to a beach-building discharge for several days.
This type of management prescription may be only used
every 5 to 10 yr, with smaller intervening releases.
References Cited
Cooley, M.E., Aldridge, B.N., and Euler, R.C., 1977,
Effects of the catastrophic flood of December, 1966,
North Rim area, eastern Grand Canyon, Arizona:
U.S. Geological Survey Professional Paper 980, 43 p.
Graf, W.L., 1979, Rapids in canyon rivers: Journal of
Geology, v. 87, p. 533–551.
Griffiths, P.G., Webb, R.H., and Melis, T.S., 1996, Initia-
tion and frequency of debris flows in Grand Canyon,
Arizona: U.S. Geological Survey Open-File Report
96-491, 35 p.
Griffiths, P.G., Webb, R.H., and Melis, T.S., 2004, Initia-
tion and frequency of debris flows in Grand Canyon,
Arizona: Journal of Geophysical Research, Surface
Processes, v. 109, F04002, doi:10.1029/2003JF000077,
14 p.
Hanks, T.C., and Blair, J.L., 2003, Comment (on dif-
ferential incision of the Grand Canyon related
to Quaternary faulting–constraints from U-series
and Ar/Ar dating by J. Pederson, K. Karlstrom,
W. Sharp, and W. McIntosh): GEOLOGY Online
Forum, www.gsajournals.org, DOI 10.1130/0091-
7613(2003)31 <e16:DIOTGC> 2.0.CO; 2.
Hanks, T.C., and Webb, R.H., in press, Effects of tribu-
tary debris on the longitudinal profile and incision
of the Colorado River in Grand Canyon: Journal of
Geophysical Research, Surface Processes.
Hereford, R., Thompson, K.S., and Burke, K.J., 1998,
Numerical ages of Holocene tributary debris fans
inferred from dissolution pitting on carbonate boul-
ders in the Grand Canyon of Arizona: Quaternary
Research, v. 50, p. 139–147.
Howard, A., and Dolan, R., 1981, Geomorphology of
the Colorado River in Grand Canyon: Journal of
Geology, v. 89, p. 269–297.
Kieffer, S.W., 1985, The 1983 hydraulic jump in Crystal
Rapid: implications for river-running and geomorphic
evolution in the Grand Canyon: Journal of Geology, v.
93, p. 385–406.
Leopold, L.B., 1969, The rapids and the pools, in The
Colorado River region and John Wesley Powell: U.S.
Geological Survey Professional Paper 669, p. 131–145.
Magirl, C.S., Webb, R.H., and Griffiths, P.G., 2005,
Long-term change in the water-surface profile of the
Colorado River in Grand Canyon, Arizona: Water
Resources Research, v. 41, Wo5021, doi:10.1029/
2003WR002519, 10 p.
Melis, T.S., 1997, Geomorphology of debris flows and
alluvial fans in Grand Canyon National Park and their
influence on the Colorado River below Glen Canyon
Dam, Arizona: Tucson, University of Arizona, unpub-
lished Ph.D. dissertation, 495 p.
Melis, T.S., Webb, R.H., Griffiths, P.G., and Wise,
T.J., 1994, Magnitude and frequency data for his-
toric debris flows in Grand Canyon National Park
and vicinity, Arizona: U.S. Geological Survey Water
Resources Investigations Report 94-4214, 285 p.
Pizzuto, J.E., Webb, R.H., Griffiths, P.G., Elliott, J.G,
and Melis, T.S., 1999, Entrainment and transport of
cobbles and boulders from debris fans, in Webb, R.H.,
Schmidt, J.C., Marzolf, G.R., and Valdez, R.A., eds.,
The controlled flood in Grand Canyon: scientific
experiment and management demonstration: Wash-
ington, D.C., American Geophysical Union, Geophysi-
cal Monograph Series, v. 110, p. 53–70.
Stephens, H.G., and Shoemaker, E.M., 1987, In the foot-
steps of John Wesley Powell: Boulder, Colo., Johnson
Books, 286 p.
Turner, R.M., and Karpiscak, M.M., 1980, Recent
vegetation changes along the Colorado River between
Glen Canyon and Lake Mead, Arizona: U.S. Geologi-
cal Survey Professional Paper 1132, 125 p.
U.S. Geological Survey, 1924, Plan and profile of Colo-
rado River from Lees Ferry, Ariz., to Black Canyon,
Ariz.-Nev. and Virgin River, Nev.: U.S. Geological
Survey, 1924, 21 p.
Webb, R.H., 1996, Grand Canyon, a century of change:
Tucson, University of Arizona Press, 290 p.
152 The State of the Colorado River Ecosystem in Grand Canyon
Webb, R.H., Griffiths, P.G., and Melis, T.S., 2004,
Debris flows and rapids in Grand Canyon: implica-
tions for erosion rates and evacuation of sediment
from tributary canyons, in Young, R.A., and Spamer,
E.E., eds., The Colorado River: origin and evolution:
Grand Canyon, Ariz., Grand Canyon Association
Monograph 12, p. 183–188.
Webb, R.H., Griffiths, P.G., Melis, T.S., Elliott, J.G.,
and Pizzuto, J.E., 1999a, Reworking of debris fans
by the 1996 controlled flood in Grand Canyon, in
Webb, R.H., Schmidt, J.C., Valdez, R.A., and Marzolf,
G.R., eds., The 1996 flood in Grand Canyon: sci-
entific experiment and management demonstration:
Washington, D.C., American Geophysical Union,
Geophysical Monograph, p. 37–51.
Webb, R.H., Griffiths, P.G., Melis, T.S., and Hartley,
D.H., 2000, Sediment delivery by ungaged tributar-
ies of the Colorado River in Grand Canyon, Arizona:
U.S. Geological Survey Water Resources Investigations
Report 00-4055, 67 p.
Webb, R.H., Melis, T.S., Griffiths, P.G., Elliott, J.G.,
Cerling, T.E., Poreda, R.J., Wise, T.W., and Pizzuto,
J.E., 1999b, Lava Falls Rapid in Grand Canyon, effects
of late Holocene debris flows on the Colorado River:
U.S. Geological Survey Professional Paper 1591, 90 p.
Webb, R.H., Pringle, P.T., Reneau, S.L., and Rink, G.R.,
1988, Monument Creek debris flow, 1984: implica-
tions for formation of rapids on the Colorado River
in Grand Canyon National Park: Geology, v. 16, p.
50–54.
Webb, R.H., Pringle, P.T., and Rink, G.R., 1989, Debris
flows in tributaries of the Colorado River in Grand
Canyon National Park, Arizona: U.S. Geological
Survey Professional Paper 1492, 39 p.
Contact Information:
Robert H. Webb
Research Hydrologist
U.S. Department of the Interior
U.S. Geological Survey
Water Resources Discipline
Tucson, AZ
rhwebb@usgs.gov
Peter G. Griffiths
Research Hydrologist
U.S. Department of the Interior
U.S. Geological Survey
Water Resources Discipline
Tucson, AZ
pggriffi@usgs.gov
Christopher S. Magirl
Hydrologist
U.S. Department of the Interior
U.S. Geological Survey
Water Resources Discipline
Tucson, AZ
magirl@usgs.gov
Thomas C. Hanks
Geophysicist
U.S. Department of the Interior
U.S. Geological Survey
Geology
Menlo Park, CA
thanks@usgs.gov
First page photograph credit: R.B. Stanton, courtesy of the National Archives
and Records Administration, College Park, Maryland
Chapter 9
Recreation
Use Values and
Nonuse Values
of Glen and
Grand Canyons
John Loomis
Aaron J. Douglas
David A. Harpman
Introduction
Grand Canyon and Glen Canyon are unique natu-
ral environments valued worldwide, providing world-
class recreational activities and habitat for wildlife and
endangered fish. This chapter discusses the various eco-
nomic values of the Grand Canyon and Glen Canyon
ecosystems, how these values are measured, and what we
know about them. The economic value of the services
provided by these ecosystems is also discussed. Other
chapters in this report address related topics, including
the potential effects of Grand Canyon campsite availabil-
ity on recreation (chapter 12) and the hydropower values
of Glen Canyon Dam (chapter 10).
This chapter focuses on how recreation use and
economic values are influenced by alternative river flow
regimes and Glen Canyon Dam operations. Recreation
is of interest within the Glen Canyon Dam Adaptive
Management Program for several reasons. First, the
Grand Canyon Protection Act (GCPA) of 1992 expressly
mentioned recreation as one of the three elements to be
monitored. Specifically, the act stated that “long-term
monitoring . . . shall include any necessary research and
studies to determine the effect of the Secretary’s actions
. . . on the natural, recreational and cultural resources
of Grand Canyon National Park and Glen Canyon
National Recreation Area” (GCPA, sec. 1805(b)). Sec-
ond, because Glen Canyon is a national recreation area
and recreation is one of the dual mandates of national
parks (Loomis, 2002), the effects of alternative dam oper-
ations on recreation should be expected to receive sig-
nificant attention. Third, recreation use and value are of
interest in an integrated scientific research effort because
of potentially important linkages between various eco-
system components. For example, it becomes important
to understand how changes in river flows influence trout
size and abundance and how changes in fish population
in turn influence recreation use and value.
As noted previously by the National Research
Council (1999, p. 13), however, there has not been
regular, systematic, or comprehensive monitoring of the
effects of alternative flow regimes and other operational
changes on recreation use and public values. Therefore,
this chapter draws from research conducted over the last
two decades to summarize the available information on
recreation use, benefits, and public values of Grand and
Glen Canyons. This partial information is the best avail-
able at the present time to inform adaptive management
154 The State of the Colorado River Ecosystem in Grand Canyon
(see Overview, this report) of Grand and Glen Canyons
about the consequences of operational changes on recre-
ation use and public values.
Background
Types of Economic Values
Provided by Glen and
Grand Canyons
Economic values can be divided into three main cat-
egories: (1) visitor use values such as recreation, (2) local
economic effects associated with visitor spending that sup-
ports commercial outfitters, hotels, restaurants, and towns
such as Page, Arizona, and (3) public nonuse values.
Use Values to Visitors
Most people are familiar with the local economic
effects of visitor spending and the positive economic
impacts of recreation and tourism in the form of jobs
and local income. In addition, the visitors themselves
also receive economic benefits from the actual recre-
ation experience in the form of what is termed “con-
sumer surplus.” This consumer surplus represents the
monetary measure of the increase in economic well
being that a visitor receives, and would be willing to
pay, over and above the existing cost of the recreational
trip. Consumer surplus, or net willingness to pay, is the
federally recommended measure of benefits for use
by agencies such as the Bureau of Reclamation when
conducting benefit-cost analyses (U.S. Water Resources
Council, 1983).
As an example of consumer surplus, consider the
case of private whitewater boating on the Colorado
River through Grand Canyon. A private boater expe-
riencing a long awaited and possibly once-in-a-lifetime
trip would likely pay more for the trip than just the
travel costs to Grand Canyon in order to enjoy a 2-week
float trip. This extra value or consumer surplus realized
on these trips may be influenced by the river flows, as
minimal flows will reduce the size of some rapids, as
well as the time available for stopping at side canyons
and visiting cultural sites.
Local Economic Effects
of Visitor Spending
While visitor expenditures are a cost to the visitor,
they create positive, direct economic effects in the local
area in the form of additional income and employment
in the retail sector. The ripple, or multiplier effects, of
such spending reverberates to the wholesale and distri-
bution sectors that supply the retail sector. This second
round of economic stimulation is known as indirect
effects. Further, a portion of the money received by
employees as wages and profits to business owners gets
respent in the local economy on other consumer goods
and services. These respent portions are often referred to
as induced effects. The regional multiplier is the sum of
these direct, indirect, and induced effects divided by the
direct effect (Douglas and Harpman, 1995; Loomis and
Walsh, 1997; Minnesota Implan Group, 1997). Thus, the
local economic effects of recreation include employment
and income that originate from the economic sectors
directly catering to tourists but also include many other
sectors indirectly affected by recreation spending.
Nonuse Values
Nonuse values include the benefits that people
derive from simply knowing that a unique and irreplace-
able natural environment or species exists even if the
individual does not visit it or see it. This component of
nonuse value is often called “existence value.” Nonuse
values also include a bequest value from knowing that
protection today provides the unique natural resources
to future generations. The unique natural environment
of Grand Canyon was used as an example in the semi-
nal article on existence value by Krutilla (1967, p. 778).
Krutilla also noted that continued existence of a threat-
ened and endangered species (e.g., humpback chub (Gila
cypha)) would also generate existence values.
The U.S. Department of the Interior officially
recognized existence values in 1986 when it included
these values in procedures to calculate natural resource
damages from hazardous substances (U.S. Department
of the Interior, 1986). The inclusion of nonuse values
was broadened by the U.S. District Court of Appeals
decision (1989), which referred to nonuse values as pas-
sive-use values, and indicated that both use and pas-
sive-use values must be considered when performing a
natural resource damage assessment. The concept of
Recreation Use Values and Nonuse Values of Glen and Grand Canyons 155
nonuse values and its empirical measurement by using
surveys were given a qualified endorsement by a blue
ribbon panel commissioned by the National Oceanic
and Atmospheric Administration and chaired by two
Nobel laureates (Arrow and others, 1993). The poten-
tial role of nonuse values in policy analysis for the Glen
Canyon Environmental Studies program is discussed in
Harpman and others (1993). The growing acceptance
of including nonuse values in valuation of ecosystems is
evident in the latest National Research Council (2005, p.
6) report, which states, “economic valuation of changes
in ecosystem services should be based on . . . both use
and nonuse values.”
Of course, what is relevant for this chapter is how
the nonuse values of Grand Canyon change with alter-
native dam operations that affect the native species in
Grand Canyon for current and future generations. We
will discuss the one study, performed in 1994, that mea-
sured nonuse values for changes in natural resources in
Grand Canyon (Welsh and others, 1995). Unfortunately,
there have been no follow-up studies to measure how
nonuse values have changed as a result of stabilization of
river flows, beach building, and recovery efforts for the
humpback chub.
Empirical Measurement of Use
and Nonuse Values
Recreation Use Values
To estimate visitor benefits or consumer surplus,
there are two broad categories of methods: (1) those
that rely upon actual behavior or what are called
“revealed preference methods” and (2) those that rely
upon intended behavior or what are called “stated
preference methods.” The travel cost method (TCM)
for estimating recreation demand is an example of a
commonly used revealed preference method. The basic
TCM uses variations in visitors’ travel costs as a proxy
for the price of a trip and the number of trips taken
as a measure of quantity to trace out a demand curve
for recreation at the particular site. From the demand
curve, the consumer surplus or net willingness to pay
(WTP) can be measured (Loomis and Walsh, 1997).
More recent TCM models involve a random utility
model of site choice to reveal the net WTP of visitors
for access to a recreation site.
Within stated preference methods is the frequently
used contingent valuation method (CVM), as well as
newer conjoint and choice experiment techniques. These
stated preference methods use a survey to construct a
simulated or hypothetical market in which an individual’s
consumer surplus is elicited. For example, the analyst
often asks a visitor what is the maximum increase they
would pay to visit a particular site. In most of the CVM
studies reviewed below, a popular willingness-to-pay
question format, called “dichotomous choice,” is used.
In this question format, individuals are asked if they
would pay a higher trip cost of $X. The amount of $X
varies across the sample. At higher dollar amounts, a
lower percentage of visitors would pay that amount, and
conversely at lower dollar amounts, a higher percentage
would pay that amount. By plotting the percentage of
people who would pay each dollar amount, a demand-
like relationship is traced out from which consumer
surplus can be calculated. Conjoint and choice experi-
ments also use a survey to present alternative trips with
different levels of trip characteristics (e.g., flow levels, size
of beaches) to directly estimate how consumer surplus
changes with changes in trip characteristics.
Both methods, the TCM and CVM, are recom-
mended for use by Federal agencies such as the Bureau
of Reclamation for valuing recreation (U.S. Water
Resources Council, 1983). There have been hundreds
of applications of both methods worldwide. Extensive
comparison of the consumer surplus estimated by using
the TCM and CVM suggests that the two have conver-
gent validity, as estimates of consumer surplus by the two
approaches are nearly identical and are highly correlated
with each other (Carson and others, 1996).
Measurement of Regional Economic
Effects of Recreation
To calculate the local economic impacts associ-
ated with visitor expenditures, an input-output model
is used to calculate the multiplier effects (Douglas and
Harpman, 1995; Loomis, 2002). The general eco-
nomic approach of combining expenditure data with
an input-output model is called “regional economic
analysis” because it usually measures the effect on a
local economy (county) or a region (group of counties).
The input-output model translates the visitor expen-
ditures into total local income and employment. The
regional economic analyses reported in this chapter use
IMPLAN® software (Minnesota Implan Group, 1997)
156 The State of the Colorado River Ecosystem in Grand Canyon
to construct an input-output model of the local region.
The input-output model captures the ripple, or multi-
plier effects, of the direct spending on indirect spending
of related industries and induced spending of workers
receiving additional wages.
Measurement of Nonuse Values
Stated preference methods such as the CVM or con-
joint/choice experiments are the only methods currently
available for measuring nonuse values. Much like the
recreation application of the CVM, the application of
these methods to measure the willingness of a consumer
to pay for nonuse values involves using a survey to con-
struct a simulated or hypothetical market or referendum.
In this case, however, a random sample of households is
interviewed or mailed a survey. They are asked whether
they would vote in favor of or against a particular man-
agement action or program involving protection of the
resource at a specific cost to their household in the form
of higher taxes or prices.
Status, Trends,
and Recent Findings
Recreation Use Values
of Glen Canyon
Fishing Use
Because of the clear, cold water released from Glen
Canyon Dam, the 15-mi (24-km) stretch of the Colorado
River below the dam provides a significant trout fishery
in Arizona. As shown in figure 1, use has fluctuated over
the past 10 yr. In part, fluctuating use may be due to
increasing populations of trout reaching carrying capac-
ity of the stream, which results in a reduction of the size
of fish from their earlier trophy size. Changes in fishing
regulations (including greater restrictions in permissible
fishing gear and reductions in catch limits) may also be
partially responsible for fluctuating use patterns. Shore
anglers are less affected than boat anglers because there
are threshold minimum flows for floating the boats over
the rocks in critical reaches. The downturn in Lees
Ferry angler use may be due to a variety of factors and,
therefore, suggests the need for an integrated multivari-
ate statistical analysis of the relationship between angler
use, catch-per-unit effort, and fish size. Chapter 2 of
this report presents a catch-per-unit effort that suggests
that the decline in angler days may be due to decreasing
angler catch rate, which fell from a peak of nearly 1.5
fish per hour in the 1998–99 time period to 0.5 fish per
hour in 2002. At the same time, the size of fish contin-
ued to decrease through 2001.
Fishing Values
There are only two economic valuation studies of
recreational fishing in the Lees Ferry reach, both of
which are fairly old. The first study was by Richards
and others (1985) and used the TCM to estimate the
consumer surplus of trophy trout anglers and nontrophy
trout anglers. The authors calculated net willingness to
pay over and above their trip cost of $381 for trophy
trout anglers and $272 for nontrophy trout anglers.
Using the percentages of trophy and nontrophy anglers
that existed at the time of the study (now more than 20
yr ago, and before the change in fishing regulations),
the weighted average value was $304 in 1983 dollars. In
2004 dollars, the consumer surplus is $580 per trip.
The second study, and the only one we are aware
of that related angler value to river flow levels, was
conducted by Bishop and others (1987). During 1985,
anglers were sampled at Lees Ferry. The investigators
used a dichotomous choice, willingness-to-pay format
in their CVM survey of anglers. The net willingness to
pay for the anglers’ actual trips ranged between $130
for those experiencing constant flows and $104 for those
experiencing fluctuating flows. In 2004 dollars this is
Figure 1. Angler use at Lees Ferry, 1995–2004 (developed from
data found at http://www2.nature.nps.gov/mpur/ for Glen Canyon
National Recreation Area).
Recreation Use Values and Nonuse Values of Glen and Grand Canyons 157
about $250 per trip for those experiencing constant flows
and nearly $200 for those experiencing fluctuating flows.
While the 20%–25% difference in value per trip
with constant versus fluctuating flows gives some insight
into the effect of Glen Canyon Dam operations on the
value of fishing, the study by Bishop and others (1987)
also asked anglers their net willingness to pay for a wide
range of flow scenarios. The results indicated that angler
benefits (WTP) peaked at a flow of 10,000 cubic feet per
second (cfs) with constant flows at $126 per trip ($240 in
2004 dollars). This finding is in the same range of ben-
efits as the actual trip experience, where 75% of anglers
experienced a flow between 5,000 and 15,000 cfs on
their most recent trip (Bishop and others, 1987, p. 121).
At low flows, such as 3,000 cfs, the values per trip fell to
half at $60 per trip ($114 in 2004 dollars). At high flows
of 25,000 cfs, the value per trip was $94 ($178 in 2004
dollars). Thus, there is a nonlinear relationship between
angler benefits and flows.
Combining the values per angler trip and the num-
ber of anglers yields estimates of annual recreational
fishing benefits that range from $2.4 to $4.8 million
(2004 dollars) at optimum flow levels, corresponding to
the range of past use of 10,000 to 20,000 angler days.
At low flows these values drop by more than half to $1.1
million to $2.3 million annually.
While these values suggest that the recent flow
experiments such as the low summer steady flows (LSSF)
in 2000, which held flows steady at 8,000 cfs, probably
did not result in a large reduction in angler benefits from
optimum flows, it is not possible to know for sure without
conducting surveys of anglers during this period. Judging
solely by angler use levels, the LSSF was associated with
angler use equivalent to the previous year (1999). This is
corroborated by the findings of Hjerpe and Kim (2003),
who interviewed fishing guides and found that the 2000
season had slightly higher angler use but in line with the
growth in angler use in the past 5 yr.
Day-use Rafting
The 15 RM of Glen Canyon below the dam pro-
vide an opportunity for scenic day-use rafting. At low to
moderate river flows, the half-day raft trips launch below
Glen Canyon Dam and float downstream to Lees Ferry
(Douglas and Harpman, 1995). At very high flows or at
times when the national security level reaches red and
visitors are not allowed to launch from the dam, they
must launch from Lees Ferry and motor most of the way
upstream and then float down. As is shown in figure 2,
this is a popular float trip, averaging around 40,000 visi-
tors each year.
Economic Values of Day-use Rafting
The only study of the economic value of day-use
rafting was performed nearly 20 yr ago in 1986 by
Bishop and others (1987). About half of their sample
were rafters starting their trips below Glen Canyon
Dam and then floating down, and the other half were
motoring up and then floating down. Dichotomous
choice CVM was used to estimate the consumer surplus
that day-use rafters received. Statistical tests indicated
that the WTP functions were not statistically different
between the rafters starting their float trip below the
dam and those who motored up and then floated down
(Bishop and others, 1987, p. 145). Overall, the net WTP
or consumer surplus was $26 per day, or nearly $50
in 2004 dollars. These figures translate into an annual
value of $2 million.
Based on the attribute survey and the lack of dif-
ference in WTP values for the two departure locations,
Bishop and others (1987, p. 145–146) concluded that
day-use rafting trip values were not sensitive to river
flow. This insensitivity to flows may be because visi-
tors lacked prior knowledge regarding river flows and
because of the lack of rapids in this section of the river.
Obviously, it would be desirable to update these values,
especially with the new flow regimes since 1986, and
revisit the Bishop and others (1987) finding of no appar-
ent relationship between flow and recreation benefits for
the day-use rafting.
Figure 2. User days per year at Glen Canyon National Recreation
Area for day-use rafting, 1995–2004 (source: fax from Jacki Blais,
Glen Canyon National Recreation Area, to Lara Schmit, U.S.
Geological Survey, January 12, 2005).
158 The State of the Colorado River Ecosystem in Grand Canyon
Recreational Rafting Use
of Grand Canyon
Grand Canyon National Park is world renowned
for its whitewater rafting opportunities (Behan, 2000).
Typically, commercial motorized rafting trips navigate
the entire 226 RM from Lees Ferry to Diamond Creek
(one of the takeout points) in 1 week, although there are
also 2-week commercial oar trips. Private boaters typi-
cally take about 18 d for this same trip (Hjerpe and Kim,
2003). Passengers of commercial trips and some private
oar trips navigate the river either from Lees Ferry to
Phantom Ranch (88 RM) or meet their trip at Phantom
Ranch and end at Diamond Creek.
The popularity of rafting in Grand Canyon sky-
rocketed during the late 1960s and 1970s, prompting
the National Park Service to place limits on the number
of commercial and private boaters. During the 1970s
to 1990s, 21 commercial rafting companies took visitors
down Grand Canyon. Today, the limits are 115,000 user
days for commercial rafting companies and about 55,000
for noncommercial, or private, users. The binding nature
of these limits on use can be seen in figure 3.
The number of user days, however, masks the large
difference in the number of passengers permitted for the
two groups. Since the private trips are often much longer
than commercial trips, there are nearly six times the
number of commercial passengers (18,500–19,600) as
there are noncommercial passengers (3,400–3,600).
Over time the demand for private trips has increased
substantially relative to commercial trips. This increase
has resulted in a substantial and growing wait to obtain
a private permit. The wait has grown from about 5 yr
in 1991 to 12 yr in 2003, resulting in the National Park
Service revisiting its allocation by issuing in October 2004
the Draft Environmental Impact Statement Colorado
River Management Plan. The preferred alternative in
the management plan nearly doubles the noncommercial
user days and the number of private passengers permit-
ted (Grand Canyon National Park, 2004).
Because of excess demand and user limits, one
cannot cite visitor use as an indicator of visitor response
to alternative flow regimes (Hjerpe and Kim, 2003).
Rather, one must conduct surveys to ascertain how visi-
tor experience and economic benefits (i.e., consumer
surplus) change with alternative flow regimes and natural
resource conditions.
There are two surveys that provide some insights
into how visitor satisfaction in nonmonetary terms varies
with flow levels. The first survey was performed as part
of the original Glen Canyon Environmental Studies
program during 1984 and 1985 by Bishop and oth-
ers (1987). Surveys of river guides, private trip leaders,
commercial passengers, and private rafters found that
constant flows in the range of 20,000–25,000 cfs yielded
the highest satisfaction ratings. In essence, flows in this
range allowed for larger rapids, more time for stopping at
side canyons and attraction sites, and less time motoring
or rowing. Low flow levels frequently resulted in visitors
having to walk around certain rapids, which decreased
trip satisfaction.
The second survey was by Stewart and others
(2000), who during 1998 and 1999 replicated the por-
tion of the Bishop and others (1987) study that dealt
with whitewater rafting trips. This more recent study
found the same pattern of visitor satisfaction ratings with
regard to flow. In particular, the study found that white-
water rafting satisfaction was highest at constant flows
of 20,000–25,000 cfs and that several flow-related trip
characteristics such as large rapids and time to stop and
hike the side canyons were highly important (Stewart and
others, 2000). This study also found that large beaches
with shade from trees for stopping and camping was
rated as moderately important.
Related to these satisfaction surveys are trip diary
data collected by Roberts and others (2002) and Roberts
and Bieri (2001) to develop a Grand Canyon river trip
simulator. In their simulation model, river flows are a key
determinant of boat speed and, hence, the amount of
time for trip-related activities such as swimming, visiting
cultural sites, hiking, campsite selection, and decisions to
Figure 3. Number of private and commercial boater user days,
Grand Canyon National Park, 1998–2003 (source: www.nps.gov/
grca/crmp/documents/stats/1998-2003ByMonth.pdf).
Recreation Use Values and Nonuse Values of Glen and Grand Canyons 159
layover part of or an entire day. Roberts and Bieri (2001)
also used trip diaries to study the LSSF of 2000. They
found that the low flows reduced the amount of time
visitors had for swimming, hiking, and visiting cultural
sites from 7 h/d on a typical trip (flow averaging 19,000
cfs) to 3.5 h with the 8,000 cfs of the LSSF. Thus, with
the 8,000-cfs flows, an additional 3.5 h/d were spent
motoring or rowing to make up for lost time because of
the slower flows (Roberts and Bieri, 2001, p. 13).
Economic Values of Private and
Commercial Rafting in Grand Canyon
The price of a commercial Grand Canyon rafting
trip is substantial, averaging around $215 per person
per day (Hjerpe and Kim, 2003). For private trips, the
National Park Service requires a $100 payment to get on
the waiting list, coupled with a payment of $100 for every
person taking the trip (Hjerpe and Kim, 2003). One way
to think of the $100 payment to get on the waiting list is
as an “option value,” a willingness to pay to maintain the
opportunity to raft Grand Canyon in the future.
The payment of commercial fees and related visitor
expenditures of commercial and private rafters creates
significant positive economic effects in the region in
the form of income and employment. These effects are
detailed in the regional economic effects section of this
chapter.
Our interest here is the amount of benefits received
by the whitewater rafters and kayakers themselves. This
is their net economic value, or consumer surplus. The
first and only primary data study on whitewater rafting
economic values and how they change with flow was
conducted by Bishop and others in 1987 (see also Boyle
and others, 1993). These authors used a dichotomous
choice CVM survey of commercial and private boat-
ers. They found that commercial passengers’ willingness
to pay for their actual trip experience rose from $127
($228 in 2004 dollars) at 5,000 cfs to a maximum value
of $888 ($1,598 in 2004 dollars) at higher flows and then
declined only slightly at 40,000 cfs (Boyle and others,
1993). For private boaters the value of their actual trips
rose from $111 ($200 in 2004 dollars) at 5,000 cfs to a
maximum of $637 ($1,147 in 2004 dollars) at 28,000 cfs.
This value falls to $455 ($819 in 2004 dollars) at 40,000
cfs (Boyle and others, 1993). The larger decline in value
at the highest flows is likely because of less experienced
private trip leaders compared with professional guides
on commercial trips. Nonetheless, these are very high
recreation trip values, although less dramatic on a per-
day basis (about 7 d for a commercial trip). The relatively
high value is commensurate with the high-quality experi-
ence and high satisfaction received by the majority of
visitors to Grand Canyon.
Summing these updated 2004 dollar values over the
nearly 19,000 commercial passengers and 3,500 private
boaters, these values at optimum flows are $30 million
and $4 million annually in economic value to commer-
cial passengers and private boaters, respectively. Based
on Bishop and others (1987), there would be a drop in
total whitewater rafting benefits from $34 million to $5
million, or a loss of $29 million, at low flows such as
5,000 cfs (a flow level close to the LSSF of 2000). Unfor-
tunately, the opportunity to evaluate the actual loss in
recreation benefits with the LSSF was missed.
There is only one more recent estimate of white-
water river recreation in Grand Canyon that has been
made. Hammer (2001) used the TCM with travel costs
calculated by using the U.S. Postal Service ZIP Code™
data from the 1998 recreation preference surveys of Hall
and Shelby (2000) and Stewart and others (2000). Since
these surveys were not originally designed for economic
analysis, however, they do not allow for analysis of how
the economic value of the rafting trips changes with flow
levels. Nonetheless, Hammer (2001) did a very careful
job of using ZIP Codes to calculate travel costs of visitors
and using this information to estimate a TCM demand
curve. This calculation is somewhat challenging because
each visitor usually takes just one trip. But by using a
zonal or visits-per-capita type demand model, Hammer
(2001) was able to calculate an estimate of the value in
1998–99. The value per trip for private boaters was $134
or ($148 in 2004 dollars) and $314 ($351 in 2004 dollars)
for commercial trip passengers (Hammer, 2001).
The inability to link recreation benefits from
Hammer’s (2001) study to flows is unfortunate. If there is
an insufficient budget to fund separate recreation eco-
© 2005 Geoff Gourley; used with permission
160 The State of the Colorado River Ecosystem in Grand Canyon
nomic surveys, then an integrated social science survey
that would allow for data to provide both preference and
valuation in relation to flows would be desirable. Such
recreation valuation monitoring should be planned into
future flow experiments.
Regional Economic Effects of
Water-based Recreation in Glen
and Grand Canyons
As discussed previously, recreational fishing and
rafting involve tens of thousands of visitors each year to
the Glen Canyon and Grand Canyon areas. Commercial
rafting on Grand Canyon currently involves 16 differ-
ent companies. One of these companies also provides
day-use rafting in Glen Canyon, and there are support
companies providing visitor and shuttle transportation
to day-use rafters. Many anglers in Glen Canyon also
employ professional fishing guides, who often use boats
for guiding. Even the anglers and private rafters who do
not use commercial outfitters spend a significant amount
of money in the local areas on hotels, restaurants, gro-
ceries, ice, gasoline, and supplies.
The local economic effects of recreation on the
retail, supporting wholesale, and distribution sectors and
on induced spending elsewhere in the economy from
river-based recreation expenditures have been calcu-
lated in two studies. The most comprehensive regional
economics study was by Douglas and Harpman (1995),
who used data collected in 1985. Table 1 presents the
average trip expenditures, total trip expenditures, the
amount spent in the local region (defined by Douglas and
Harpman as Coconino County, where Page, Arizona,
is located, and Mohave County), and the resulting total
employment. The employment effects reflect only visitor
spending in the region, and the multiplier effects are cal-
culated by using IMPLAN® (see Douglas and Harpman,
1995). The monetary amounts used in Douglas and
Harpman (1995) have been updated to 2004 dollars, but
the employment estimates are calculated directly from
their data (see Douglas and Harpman, 1995, tables 3 and
6 using 1990 relationships) and are consistent with calcu-
lations developed by Douglas (2005) of 438 total jobs in
whitewater boating in Grand Canyon National Park.
As indicated in table 1, the total of nearly 600 jobs
provided by river-based recreation is a substantial num-
ber of jobs. Commercial rafting, particularly the Grand
Canyon National Park segment, contributes the majority
of the jobs, 438.
Unfortunately, there has been little systematic
comprehensive research that links changes in flows to
angler use or other visitor use and, hence, to associ-
ated changes in economic impacts. There has been one
study on the effect of the lower summer steady flows on
economic impacts to the rafting and angling outfitters
by Hjerpe and Kim (2003). They found that these low
flows had minor effects, mostly related to damage to
equipment in Grand Canyon and angler boats in Glen
Canyon. Day-use rafters in the Glen Canyon reach were
apparently not affected (Hjerpe and Kim, 2003). The
direct economic consequences of higher outfitter costs
because of commercial boat damage were in the range
of $25,000, and the loss in fishing guide services during
the spike flows that were part of LSSF was in the range
of $33,000 (Hjerpe and Kim, 2003). As the result of
low flows stranding commercial whitewater trips, there
were also three rescue operations, which cost $30,000.
Hjerpe and Kim (2003) aggregated these as losses to
Coconino County for the regional economic analysis.
The IMPLAN® estimated total effects on Coconino
County were $117,705 in losses. This estimated total is
relatively minor to the local economy; however, Hjerpe
and Kim (2003) acknowledged that they may have mixed
local economic losses from reduced fishing guide income
with increases in economic activity from boat repair
and helicopter rescue. There is a problem with using
regional economic impact analysis as an indicator of the
economic effects of changes in flows: higher outfitter
expenses, which are costs, have the potential to actually
increase economic activity. Thus, a benefit-cost analysis
that would treat higher costs to outfitters as the actual
losses that they are to society would probably be more
appropriate for monitoring the economic consequences
of changes in the flow regime.
© 2005 Geoff Gourley; used with permission
Recreation Use Values and Nonuse Values of Glen and Grand Canyons 161
The comparison of these relatively minor local eco-
nomic effects from the LSSF illustrates the limitation of
relying on regional economic analyses: if visitor use does
not change much because of the National Park Service
limits on passenger days, then the effects of river flow
will not be manifested in changes in regional economic
effects (Hjerpe and Kim, 2003). Rather, any effects
of river flow changes are more likely to be manifested
as changes in user satisfaction and use value. Without
conducting a valuation survey, the most likely economic
effects of changes in river flow regimes in Grand Canyon
may be missed.
Nonuse Values
As noted above, Grand Canyon is also a source
of nonuse or existence/bequest values to people who
may never or no longer visit Grand Canyon (Harpman
and others, 1993). The same is true with the continued
existence of endangered species, such as the humpback
chub, in their natural habitat. A given household’s non-
use values for improving river flow management in Glen
Canyon and Grand Canyon can be obtained by using the
CVM. To date, there has been only one nonuse CVM
survey that asked willingness to pay to improve native
vegetation (and associated birds and other wildlife),
native fishes, game fish (such as trout), river recreation,
and cultural sites in Glen Canyon National Recreation
Area downstream of Glen Canyon Dam and in Grand
Canyon National Park. This study was conducted in late
1994 and early 1995 by Welsh and others (1995).
There were two main sampling strata in the Welsh
and others (1995) study: a sample called market area,
reflecting Western U.S. households receiving hydropower
from Glen Canyon Dam, and then a national sample,
reflecting the rest of the Nation. These sampling strata
received different payment vehicles: the market area
would pay for protection via higher utility bills, while
the national sample would pay via higher taxes. There
were three main river-flow scenarios and accompanying
changes in five main environmental attributes (beaches,
risk of erosion to cultural sites, vegetation/bird habi-
tat, native fish, and trout). At the time the survey was
conducted, the main flow regimes under consideration
involved degrees by which the fluctuating flows previ-
ously associated with peaking power production would
be moderated. The three main flow scenarios were (1)
moderate fluctuating flows, (2) low fluctuating flows, and
(3) seasonally adjusted steady flows. While all scenarios
would maintain beaches at current conditions, decrease
the erosion risk to cultural sites substantially, and increase
streamside vegetation by 10%, the steady flow scenario
was described as providing improvements for native
fish and trout.
The response rates in the two sample strata were
quite high by comparison to most CVM studies in the
literature. The national sample had an overall response
rate of 74%, and the marketing area sample had a
response rate of 83% (Welsh and others, 1995).
To provide a conservative estimate of nonuse
values, only those responses that indicated “definitely
yes” as yes responses to the dichotomous choice CVM
responses were counted and yielded the values per
household shown in table 2. Counting only “definitely
yes” responses has been shown in actual cash validity
test comparisons to yield a valid measure of actual cash
willingness to pay (Champ and others, 1997; Ethier
and others, 2000).
Table 1. Local economic effects of river recreation in northern Arizona (2004 dollars).
[Source: Douglas and Harpman (1995). The monetary amounts used in Douglas and Harpman (1995) have been
updated to 2004 dollars, but the employment estimates are calculated directly from their data]
Glen Canyon National Grand Canyon Total
Recreation Area National Park
Day -use Anglers Commercial Private
rafting rafting rafting
Average trip expenditures $115.90 $355.91 $2,711.30 $981.84
Nonresident total expenditures $3,803,374 $3,655,174 $36,542,901 $2,872,876 $46,874,326
Nonresident total expenditures in region $3,803,374 $1,810,320 $15,351,554 $1,195,997 $22,161,245
Total jobs 100 48 406 32 586
162 The State of the Colorado River Ecosystem in Grand Canyon
The steady flow scenario, which was supposed to
be the most beneficial for fish, had the highest willing-
ness to pay, especially by households living closest to
Grand Canyon (i.e., the market area sample strata). The
annual per-household willingness-to-pay values are quite
reasonable and, yet, when aggregated up to the number
of households in the population, produce estimates in
the $3 billion to $4 billion range. Thus, the nonexclud-
able and nonrival nature of protection of the Grand
Canyon environment is evident in the fact that modest
willingness to pay per household across the country adds
up to a substantial total. In some sense this response is
not unusual; after all, Grand Canyon is a national park
and Glen Canyon a national recreation area. In fact,
Grand Canyon was designated by the United Nations
Educational, Scientific and Cultural Organization
(UNESCO) as a World Heritage Site in 1979, suggest-
ing that limiting the aggregation of benefits to just U.S.
households probably significantly understates the total
economic value received from Grand Canyon by people
around the world.
This survey is now nearly 10 yr old, and much
has been learned in that decade about the response
of the Grand Canyon environment to changes in flow
regime, including the decreasing trends in humpback
chub populations. Several of the ecological responses to
moderate or steady flows described in the survey, while
based on the best available science at the time, have
not been fully realized. As the name “contingent valua-
tion method survey” suggests, the values obtained from
such a survey are contingent upon the description of the
environmental effects. Since the actual and anticipated
environmental effects associated with different flow
regimes are somewhat different from those envisioned at
the time of survey, it may be appropriate to update the
nonuse value survey to more current ecological estimates
that are associated with different flow regimes. The facts
that the alternative flow regimes currently being imple-
mented are somewhat different from those described in
the survey and that quite different management actions,
such as temperature control devices, are being consid-
ered suggest that to keep nonuse valuations current, it
would be worthwhile to update this survey. The National
Park Service Organic Act of 1916 has been interpreted
as a dual mandate of recreation and preservation of
the national parks (Loomis, 2002). This dual mandate
suggests that it is insufficient to only measure economic
values of a national park by recreation use value.
More recent legislation such as the Redwoods Act—as
amended in 1978 to the General Authorities Act of
1970 (16.U.S.C. 1a-1), which governs the National Park
Service—provides that, when there is a conflict between
recreation use and preserving the parks unimpaired,
preservation is the primary objective. The preservation
value of a national park, therefore, is also reflected in its
nonuse values. As suggested previously by the National
Research Council (1999), these nonuse values reflect
values to nonrecreation stakeholders and should be used
by managers in decisionmaking.
Discussion and Future
Research Needs
Based on past studies, it appears that economic val-
ues for natural and cultural resources in Grand Canyon
and Glen Canyon amount to nearly $40 million of use
values and several billion dollars worth of nonuse values
each year. With 12-yr waiting lists to raft Grand Canyon
National Park and tens of thousands of anglers and
boaters using Glen Canyon National Recreation Area
each year, recreation is obviously an important public
use of these two areas. The now nearly 10-yr-old study
Table 2. Estimates of nonuse value for three flow scenarios (2004 dollars) (updated from Welsh and others, 1995).
National sample Market area sample
Flow scenario Per Annual value Per Annual value
household (millions) household (millions)
Moderate fluctuations $17.06 $2,858 $27.94 $79
Low fluctuations $25.19 $4,219 $27.17 $77
Steady flow $25.69 $4,303 $36.57 $103
Recreation Use Values and Nonuse Values of Glen and Grand Canyons 163
suggested that preservation of natural environments in
Grand Canyon provides nonuse values (e.g., existence
and bequest values) to citizens nationwide. Given that
nonuse values are nonrival goods available to all, the
rapid population growth in the Southwestern United
States suggests that nonuse values have probably risen
substantially as well.
Unfortunately, studies of the recreation and nonuse
values are between 10 and 20 yr old and reflect the flow
regimes in place before the Grand Canyon Protection
Act of 1992. Recreation use values and nonuse values
in the current flow regime and how they change with
experimental flow regimes have not been studied. With
proposed increases in whitewater rafting use levels in
Grand Canyon National Park, it is time to begin regu-
larly monitoring recreation satisfaction and use value.
The Grand Canyon Protection Act specifically lists recre-
ation monitoring, and the omission of systematic recre-
ation use value monitoring has been previously pointed
out by the National Research Council (1999). Behan
(2000) provided detailed suggestions on the type of recre-
ation monitoring that is needed. If the spirit of adaptive
management and the function of the Grand Canyon
Monitoring and Research Center are to be more fully
realized, the call in the Grand Canyon Protection Act of
1992 to monitor recreation needs to be given the same
consideration that has been given to other resources dis-
cussed in this report. If integrated adaptive management
of all resources is to be performed, then it is important
to monitor recreation satisfaction and value since recre-
ation is directly affected by flow regimes, changes in sand
deposits for beaches, and trout size and abundance. For
example, a better understanding of changes in angler use
in Lees Ferry could result from an integrated analysis of
use in relation to catch rate and fish condition class.
The importance of recreation use and nonuse values
is becoming more apparent as trade-offs are explored
between desired flows to recover endangered humpback
chub and those desired for recreation and hydropower.
Although the Endangered Species Act of 1973 requires
that all reasonable and prudent recovery alternatives be
considered, having economic valuation information on
recreation, endangered fish, and hydropower may aid in
finding a suitable balance. Federal oversight agencies such
as the Office of Management and Budget may require
benefit-cost information on temperature control devices
being considered at Glen Canyon Dam to aid recovery of
humpback chub. Adaptive management requires putting
monitoring in place prior to new management actions.
Now is the time to do the same for recreation use eco-
nomic values so that the intended and unintended effects
of future management actions can be assessed.
References
Arrow, K., Solow, R., Portney, P., Leamer, E., Radner, R.,
and Schuman, H., 1993, Report of the NOAA Panel
on Contingent Valuation: Federal Register, v. 58, no.
10, p. 4602–4614.
Behan, J., 2000, Recreation in the Colorado River
ecosystem and the Adaptive Management Program:
Flagstaff, Ariz., Grand Canyon Monitoring and
Research Center.
Bishop, R.C., Boyle, K.J., Welsh, M.P., Baumgartner,
R.M., and Rathbun, P.R., 1987, Glen Canyon Dam
releases and downstream recreation: an analysis of
user preferences and economic values. Glen Canyon
Environmental Studies Report No. 27/87: Washington,
D.C., Bureau of Reclamation (available from NTIS,
5285 Port Royal Road, Springfield, VA 22161, as NTIS
report no. PB88-182546/AS).
Boyle, K., Welsh, M., and Bishop, R., 1993, The role of
question order and respondent experience in contin-
gent-value studies: Journal of Environmental Econom-
ics and Management, v. 25, p. S80–S99.
Carson, R., Flores, N., Martin, K., and Wright, J., 1996,
Contingent valuation and revealed preference meth-
odologies: comparing estimates for quasi-public goods:
Land Economics, v. 72, p. 151–163.
Champ, P., Bishop, R., Brown, T., and McCollum, D.,
1997, Using donation mechanisms to value non-use
benefits from public goods: Journal of Environmental
Economics and Management, v. 33, no. 2, p. 151–162.
Douglas, A., 2005, Colorado River environmental and
recreational values below Glen Canyon Dam: Fort
Collins, Colo., U.S. Geological Survey.
Douglas, A.J., and Harpman, D.A., 1995, Estimating
recreation employment effects with IMPLAN for the
Glen Canyon Dam region: Journal of Environmental
Management, v. 44, p. 233–247.
Ethier, R., Poe, G., Schulze, W., and Clark, J., 2000, A
comparison of hypothetical phone and mail contingent
valuation responses for a green-pricing electricity pro-
gram: Land Economics, v. 76, no. 1, p. 54–67.
Grand Canyon National Park, 2004, Draft Environmen-
tal Impact Statement Colorado River Management
Plan, v. 1.
164 The State of the Colorado River Ecosystem in Grand Canyon
Hall, T., and Shelby, B., 2000, 1998 Colorado River
boater study: Grand Canyon, Ariz., Grand Canyon
National Park.
Hammer, M., 2001, Applying the TCM with second-
ary data to white water boating in Grand Canyon
National Park: Fort Collins, Colorado State University,
Department of Agricultural and Resource Economics,
M.S. thesis.
Harpman, D., Welsh, M., and Bishop, R., 1993, Nonuse
economic value: emerging policy analysis tool: Rivers,
v. 4, no. 4, p. 280–291.
Hjerpe, E., and Yeon-Su Kim, 2003, Economic impacts
of summer steady flows on recreation: Flagstaff,
Northern Arizona University, School of Forestry.
Krutilla, J., 1967, Conservation reconsidered: American
Economic Review.
Loomis, J., 2002, Integrated public lands management:
principles and applications to national forests, parks,
wildlife refuges and BLM lands (2d ed.): New York,
Columbia University Press.
Loomis, J., and Walsh, R., 1997, Recreation economic
decisions: comparing benefits and costs (2d ed.): State
College, Pa., Venture Publishing.
Minnesota Implan Group, 1997, IMPLAN Profes-
sional: social accounting and impact analysis software:
Stillwater, Minn.
National Research Council, 1999, Downstream: adaptive
management of Glen Canyon Dam and the Colorado
River ecosystem: Washington, D.C., National Academy
Press.
National Research Council, 2005, Valuing ecosystem
services: toward better environmental decision-making:
Washington, D.C., National Academies Press.
Richards, M., Wood, B., and Caylor, D., 1985, Sport-
fishing at Lees Ferry, Arizona: user differences and
economic values: Flagstaff, Ariz., report to Northern
Arizona University Organized Research Committee.
Roberts, C., and Bieri, J., 2001, Impacts of low flow
rates on recreation rafting traffic on the Colorado
River in Grand Canyon National Park: Study per-
formed for Bureau of Reclamation and Grand Canyon
Monitoring and Research Center.
Roberts, C., Stallman, D., and Bieri, J., 2002, Modeling
complex human-environment interactions: the Grand
Canyon River Trip Simulator: Ecological Modelling, v.
153, p. 181–196.
Stewart, W., Larkin, K., Orland, B., Anderson, D.,
Manning, R., Cole, D., Taylor, J., and Tomar, N., 2000,
Preferences of recreation user groups of the Colorado
River in Grand Canyon: Flagstaff, Ariz., Grand
Canyon Monitoring and Research Center.
U.S. Department of the Interior, 1986, Natural resource
damage assessment regulations. 51. Federal Register
27674. Codified at 43 C.F.R. 11.10-11.93 (1987).
U.S. District Court of Appeals, 1989, State of Ohio vs.
U.S. Department of the Interior. Case no. 86-1575.
District of Columbia, July 14.
U.S. Water Resources Council, 1983, Economic and
environmental principles for water and related land
resource implementation studies: Washington, D.C.,
U.S. Government Printing Office.
Welsh, M.P., Bishop, R.C., Phillips, M.L., and
Baumgartner, R.M., 1995, Glen Canyon Dam,
Colorado River Storage Project, Arizona—non-
use value study final report: Madison, Wis., Hagler
Bailly Consulting, Inc., September 8, 1995, 400 p.:
Springfield, Va.: National Technical Information Ser-
vice, NTIS no. PB98-105406.
Contact Information:
John Loomis
Professor
Colorado State University
Department of Agriculture and Resource Economics
Fort Collins, CO
John.Loomis@colostate.edu
Aaron J. Douglas
Natural Resource Economist
U.S. Department of the Interior
U.S. Geological Survey
Fort Collins, CO
aaron_douglas@usgs.gov
David A. Harpman
Natural Resource Economist
U.S. Department of the Interior
Bureau of Reclamation
Denver, CO
dharpman@do.usbr.gov
First page photograph credit: © 2005 Geoff Gourley; used with permission
Chapter 10
Status and
Trends of
Hydropower
Production at
Glen Canyon
Dam
David A. Harpman
Aaron J. Douglas
Introduction
Approximately 7% of the electricity in the United
States is generated by hydroelectric powerplants (Energy
Information Administration, 2004, p. 2). These plants
are an invaluable component of the Nation’s inter-
connected electric power system in which generation
resources consist of thermal, nuclear, solar, wind, and
other sources. Hydroelectric powerplants are character-
ized by exceptionally low costs of operation, are highly
reliable, and produce electricity without burning fos-
sil fuels and producing air pollution. In addition, they
provide voltage control, system regulation, and other
ancillary services which help ensure the reliability and
electrical integrity of the system.
Although they play an important role in the electric
power system, hydroelectric powerplants, such as the
one at Glen Canyon Dam, have some widely recognized
environmental effects. Large hydro facilities have blocked
the spawning of anadromous and migratory aquatic spe-
cies, eliminated the downstream transport of sediment,
fundamentally altered the seasonal hydrograph, affected
water chemistry, and changed the downstream tempera-
ture regime (Collier and others, 1996; Poff and others,
1997; Van Steeter and Pitlick, 1998 a, b). Furthermore,
the operation of these plants, particularly those used to
produce peaking, or variable, power, causes hourly vari-
ations in stream flow and elevation, thereby adversely
affecting downstream aquatic and riparian communi-
ties (Nilsson and others, 1997; Parasiewicz and others,
1998) and recreation (Bishop and others, 1987; Kearsley
and others, 1994; Welsh and Poe, 1998). The unveil-
ing of plans to construct Glen Canyon Dam spurred
a nationwide protest (Bradley, 1964; Martin, 1989).
Construction of the dam started a public environmental
discourse which continues to this day (McPhee, 1971;
Elfring, 1990; Brower, 1997; Long and Essick, 1997;
Jacobs and Wescoat, 2002).
Glen Canyon Dam
Glen Canyon Dam, shown in figure 1, is a 710-ft-
high (216 m) concrete thick arch dam. It is the second
highest dam in the United States (Hoover Dam is 16 ft
(5 m) higher). Construction of Glen Canyon Dam began
on October 1, 1956; Lake Powell started filling on March
166 The State of the Colorado River Ecosystem in Grand Canyon
13, 1963; and the first electric power was generated on
September 4, 1964.
The outlet works at Glen Canyon Dam are com-
posed of four hollow “jet tubes” and two spillways.
These outlet works are used only under special condi-
tions, primarily to accommodate releases from the dam
that exceed the amount of water which can be released
through the powerplant. Such releases may occur when
the reservoir is full and tributary inflows exceed the
capacity of the powerplant or they may be ordered for
environmental purposes such as the 1996 beach/habitat-
building flow.
The hollow jet tubes consist of four 96-inch-diame-
ter (244 cm) pipes. The combined release capacity of the
four hollow jet tubes is 15,000 cubic feet per second (cfs).
The intake elevation of the jet tubes is approximately
3,374 ft (1,028 m), or 326 ft (99 m) below the surface
when the reservoir is full. These elevations are illustrated
in figure 2.
Each of the two spillways consists of an intake struc-
ture with two 40- by 52.5-ft (12- by 16-m) radial gates
and a concrete-lined spillway tunnel. These spillways
are located on both sides of the dam, each of which is
capable of releasing 104,000 cfs when the reservoir is full
(3,700 ft (1,128 m)). The elevation of the spillway crest
is 3,648 ft (1,112 m). The spillways cannot be used to
release water from the reservoir when the lake elevation
falls below 3,648 ft (1,112 m).
Powerplant
The powerplant at Glen Canyon Dam is made up
of eight hydroelectric generation units. Since 1964, these
units have been uprated and rewound several times. As
of August 2003, the combined generation capability of
the powerplant (at unity power factor) is 1,320.0 mega-
watts (MW) (Seitz, 2004). A separate penstock feeds each
of the eight Francis type turbines, which each produce
approximately 155,000 horsepower. Current operat-
ing rules require at least 40 ft (12 m) of submergence to
prevent the entrainment of air into the penstocks, which
would cause damage to the turbines. As a result, the
powerplant cannot be operated at lake elevations below
3,490 ft (1,064 m). Each turbine has a release capacity of
approximately 4,150 cfs when the reservoir is full. The
nominal powerplant release capacity is approximately
33,200 cfs.
Electricity Background
Electricity cannot be efficiently stored on a large
scale by using currently available technology. It must be
produced as needed. Consequently, when a change in
demand occurs—such as when an irrigation pump or a
central air conditioner is turned on—somewhere in the
interconnected power system, the production of elec-
tricity must be increased to satisfy this demand. In the
language of the utility industry, the demand for electric-
ity is known as “load.” Load varies on a monthly, weekly,
daily, hourly, and even second-by-second basis. During
the year, the aggregate demand for electricity is highest
when heating and cooling needs are greatest. During a
given week, the demand for electricity is typically higher
on weekdays and lower on weekends, particularly holiday
weekends.
The maximum amount of electricity which can be
produced by a powerplant is called its capacity. Capac-
ity is typically measured in megawatts. The capacity
of thermal powerplants is determined by their design,
their location, and the ambient temperature. In the case
of hydroelectric powerplants, capacity varies over time
because it is a function of reservoir elevation, the amount
of water available for release, and the design of the facil-
ity. Because the capacity at hydroelectric powerplants is
highly variable, the amount of dependable or market-
able capacity is of particular significance. The amount
of dependable or marketable capacity is determined by
using various probabilistic methods (e.g., Ouarda and
others, 1997).
The large variation in loads has important implica-
tions for the electrical generation system. In particular,
it greatly influences the amount of generation capacity
required and, therefore, the capital cost of the system.
The implications of large variation in loads can be
Figure 1. Aerial view of Glen Canyon Dam (photograph courtesy
of the Bureau of Reclamation).
Status and Trends of Hydropower Production at Glen Canyon Dam 167
readily illustrated by two extreme cases. First, assume
the demand for electricity is constant and is 1.0 MW at
all times. This assumed demand would imply (ignoring
security and reliability concerns) that a utility supplier
could supply this demand by building a 1.0-MW pow-
erplant and operating it continuously. For a month (30
d), this situation would imply generation of 1.0 MW for
720 h, which would generate 720 megawatthours (MWh)
of electricity. Now assume that the demand for electric-
ity is more variable: assume that it is 1 MW for 1 h of
the month and 0.5 MW for the rest of the hours in the
month. In this case, the costs of constructing a 1-MW
powerplant must also be incurred, but the plant gener-
ates only 360.5 MWh of energy (1 MW * 1 h + 0.5 MW
* 719 h), or approximately 50% of its potential output.
The highly variable nature of the demand for electricity
results in the following observable characteristics of the
electrical power system: (1) some powerplants are idle
for part, or all, of the day or season, and (2) the capital
costs of electricity production are quite high relative to
operational costs.
Electric energy is most valuable when it is most
in demand—during the day when people are awake
and when industry and businesses are operating. This
period, when the demand is highest, is called the
“onpeak period.” In the West, the onpeak period is
typically defined as the hours from 7 a.m. to 11 p.m.,
Monday through Saturday. All other hours are consid-
ered to be offpeak.
Hydroelectric Power and the
Interconnected Power System
The two most commonly encountered types of
hydroelectric powerplants are run-of-river plants and
peaking plants. Run-of-river plants typically have little
water storage capability. Consequently, generation at
run-of-river plants is proportional to water inflow, result-
ing in little variation in electrical output during the day.
Peaking hydroelectric powerplants, such as Glen Canyon
Dam, often have significant water storage capability and
are designed to rapidly change output levels to satisfy
changes in the demand for electricity. Peaking hydroelec-
tric powerplants are particularly valuable because they
can be used to generate power during onpeak periods,
thereby avoiding the cost of operating more expensive
thermal plants such as gas turbine units.
In addition to furnishing capacity and energy,
hydroelectric powerplants play an important role in
the interconnected electric power system by supplying
ancillary services. They contribute to system reliability
by furnishing reactive power, voltage support, and system
regulation services. These facilities also fulfill part of the
regional reserve requirements and provide backup gen-
eration in the event of unexpected outages. In addition,
they provide extra energy during extremely hot or cold
weather periods and help maintain transmission stability
during system disturbances.
Figure 2. Important operating elevations of Glen Canyon Dam and Lake Powell in feet above mean sea level (msl).
168 The State of the Colorado River Ecosystem in Grand Canyon
The Economic Value
of Hydropower
The economic value of operating an existing hydro-
electric powerplant is measured by the avoided cost of
doing so. In this context, avoided cost is the difference
between the total power system cost of satisfying the
demand for electricity “with” and “without” operating
the hydroelectric powerplant. Conceptually, avoided
cost is the savings realized by supplying electricity from
a low-cost hydroelectric power source rather than from a
higher cost thermal source. These savings arise, in part,
because the cost of operating a hydroelectric power-
plant is relatively low in comparison to thermal units.
For example, the average operating expense for a typical
hydroelectric powerplant in 2003 was $7.51/MWh. In
contrast, the average cost of operating a typical fossil-fuel
steam plant was $22.59/MWh, and the average cost of
operating a typical gas turbine unit was approximately
$48.93/MWh (Energy Information Administration,
2004, p. 49, table 8.2).
The economic value of operating an existing
hydroelectric powerplant varies considerably with time
of day. The variable cost of meeting demand varies on a
second-by-second basis depending on the load, the mix
of plants being operated to meet load, and their output
levels. During offpeak periods, demand is typically satis-
fied with lower cost coal, run-of-river hydropower, and
nuclear units. During onpeak periods, the additional load
is met with more expensive sources such as gas turbine
units. Consequently, the economic value of hydropower
is greatest during hours when the demand for electricity,
and the variable cost of meeting demand, is the highest.
If the cost of purchasing an additional
megawatthour of electricity from a least cost source were
observable in the market, then the economic value of
producing hydroelectricity could be readily determined.
For example, assume that in a particular hour the cost
of purchasing a megawatthour of electricity from the
least cost source was $30 and that the cost of produc-
ing a megawatt of hydroelectricity was $6. The avoided
cost, or economic value, of producing an additional
megawatthour of hydroelectric power at that time would
be $24 ($30 - $6).
In addition to operating costs, the fixed investment
(or capital) costs of alternative sources of electricity sup-
ply may contribute to the economic value or avoided cost
of an existing hydroelectric powerplant. If a hydroelectric
powerplant were decommissioned or its operations were
restricted, the generation capacity in the system would be
reduced. A new powerplant, probably a thermal plant,
would need to be constructed to replace this lost capacity.
If there were initially excess capacity in the system, the
construction of a new powerplant could be deferred until
a future date but would be constructed sooner than would
otherwise be the case. If there were little or no excess
capacity in the system, the need would be more imme-
diate. All other factors being the same, the time of the
required replacement capacity has a significant effect on
the present worth of these additional costs. The sooner
the replacement capacity is required, the higher the pres-
ent worth of the costs incurred. The further out in the
future these construction expenditures occur, the smaller
the present worth of the costs incurred.
Federal Power
There is a long history of Federal involvement in
the provision, operation, and regulation of the electric
power system. The foundations for this involvement are
based on three factors: first, the electric power industry is
a natural monopoly (produces a product most efficiently
supplied by one supplier in a given area) and hence is
subject to regulation under the Sherman Antitrust Act
of 1890, the Interstate Commerce Act of 1887, and
other applicable statutes; second, the Federal Govern-
ment owns most of the Nation’s large-scale hydroelectric
resources; and third, Federal economic development
programs facilitated the provision of electricity to large
areas of the rural United States (Energy Information
Administration, 1996).
During the Great Depression and the Dust Bowl
years (ca. 1930s), the Federal Government became much
more active in the provision and regulation of electric-
ity. This involvement was consistent with the widespread
belief that electricity should be inexpensive and readily
available to the public. Federal power- and water-devel-
opment projects were also promulgated for purposes of
creating employment, resettling the landless, improv-
ing agricultural production, and fostering business and
industry. A number of large public works projects were
begun during this period; among these were the con-
struction of Hoover Dam and the construction of Grand
Coulee Dam, which remains the Nation’s largest hydro-
electric facility. A hallmark of the era was creation of the
Rural Electrification Administration (REA). The Rural
Electrification Act of 1936 (Public Law 74-605) estab-
lished the REA to provide loans and assistance to organi-
zations providing electricity to rural areas and towns with
populations under 2,500. REA-backed cooperatives were
instrumental in tripling the proportion of rural homes
Status and Trends of Hydropower Production at Glen Canyon Dam 169
and farms served by electricity between 1932 and the
entry of the United States into the Second World War in
1941 (Energy Information Administration, 1996).
The Bureau of Reclamation constructed and
operates Glen Canyon Dam. The Western Area Power
Administration (hereafter Western), an entity established
by the Department of Energy Organization Act of 1977,
now markets and transmits the electricity produced by
the dam.
In compliance with the Colorado River Storage
Project (CRSP) Act of 1956, power generated at Glen
Canyon Dam and other CRSP facilities is first provided
to CRSP-participating projects. These are typically
Bureau of Reclamation irrigation projects, and genera-
tion is used to meet their pumping needs. Generation
that is surplus to these “project uses” is then marketed
by Western to about 200 wholesale power customers
entitled to preference allocations. These preference
customers are generally municipal and county utilities,
rural electric cooperatives, Federal reservations, Indian
Tribes, and certain other authorized entities (see General
Accounting Office, 2001, for further details on prefer-
ence). A list of current preference customers and their
allocations can be found in Western Area Power Admin-
istration (2004a). These preference customers, in turn,
sell electricity to approximately 1.7 million residential,
commercial, industrial, and agricultural users located
predominantly in a six-State region comprising Nevada,
Utah, Arizona, Wyoming, Colorado, and New Mexico.
Each of the preference customers is allocated an
amount of energy and capacity based on Western’s
determination of the marketable power resource. The
most recent such determination, the “Post-2004 Deter-
mination of Marketable Resource,” is described in
Western Area Power Administration (2004b). If these
preference customers require additional energy and there
is additional energy available, Western may sell addi-
tional power to them on a short-term basis. If energy is
available beyond the needs of the preference customers,
Western may exchange energy with other suppliers or
make sales on the spot market. If generation is insuf-
ficient to meet the allocations of preference customers,
Western must exchange energy with another supplier or
purchase additional energy on the market.
Western’s rate-setting procedure for power differs
from that of a non-Federal utility (General Accounting
Office, 2000). By statute, power must be marketed at
the lowest possible rates that are consistent with sound
business practice. Administratively, Glen Canyon Dam
is located in Western’s Salt Lake City Area/Integrated
Projects (SLCA/IP) region. The preferred customer rate
is known as the SLCA/IP rate. The SLCA/IP rate is set
to ensure that revenues are sufficient to repay all assigned
costs within a prescribed period. These costs include
annual power operation and maintenance costs, power
facility construction costs including interest, certain envi-
ronmental costs, and other nonpower-related costs that
power users were assigned by Congress to repay (includ-
ing irrigation costs that water users are unable to repay).
The current SLCA/IP (F7) rate is $20.72/MWh.
The SLCA/IP rates charged by Western are
designed to recover the taxpayer investment. They are
lower than current wholesale market rates for electric
power. Comparisons of the electric power rates are
relatively difficult because of differences in contractual
terms, commitment lengths, products provided, and fluc-
tuations in electricity market conditions. Because of these
complexities, the average revenue per megawatthour of
wholesale electricity sold is widely used in the industry as
a metric for comparison purposes. Two studies compared
Western’s average revenue received per megawatthour
of wholesale electricity sold against those of other utili-
ties. In a 1994 study, the General Accounting Office
(1996b, p. 105, table V.1.) estimated Western’s average
revenue per megawatthour of wholesale power sold to be
47% of the revenue received by publicly owned utili-
ties and 52% of the revenue received by investor-owned
utilities. A 1999 assessment by the Energy Information
Administration estimated Western’s average revenue per
megawatthour of wholesale power sold to be 42% of the
average revenue received by utilities in the West (Energy
Information Administration, 2000, p. 44, table 13).
Monitoring of the
Hydropower Resource
The hydropower resource at Glen Canyon Dam
is monitored rather intensively. Since the plant’s super-
visory control and data acquisition (SCADA) system
was installed in the early 1980s, it has monitored and
recorded a voluminous amount of information about
power production and operations. At each of the eight
units, these data include generation, release, reservoir
afterbay water-surface elevations, voltage, frequency, cir-
cuit breaker status, gate positions, bearing temperatures,
transformer settings, and the temperatures and opera-
tional status of a variety of appurtenant equipment.
170 The State of the Colorado River Ecosystem in Grand Canyon
These data are sampled at time increments ranging
from 1 to 4 s. Selected data are recorded and archived at
larger time intervals. This level of monitoring is standard
business practice in the electric power industry.
The Effects of
Environmental Constraints
The single most important determinant of hydro-
power production and economic value at Glen Can-
yon Dam is the amount of water released during the
month. Based on projected hydrologic conditions,
monthly and annual release volumes for Glen Canyon
Dam and all major CRSP facilities are established by
the annual operating plan (AOP) at the beginning of
the water year (see www.usbr.gov/uc/water/rsvrs/ops/
aop/aop05draft.pdf for an example). Monthly release
volumes under the modified low fluctuating flow (MLFF)
operating regime are identical to those under historical
operations (1963–91) except for water years in which
an experimental flow occurs. In years when experimen-
tal releases occur, monthly release volumes may vary
substantially from the historical pattern, and there may
be differences in monthly releases across the entire water
year.
The environmental constraints that are part of the
MLFF operating regime affect hydropower production
at Glen Canyon Dam to varying degrees under differ-
ent circumstances. Typically, these constraints are most
limiting for moderate monthly release volumes and
less constraining at either very high or very low release
volumes. The maximum release constraint of 25,000 cfs
is binding or constraining only when the reservoir eleva-
tions and monthly release volumes are sufficiently high
to permit releases of 25,000 cfs or greater. The upramp
rate of 4,000 cfs/h and the downramp rate of 1,500
cfs/h limit the hour-to-hour ability of the powerplant to
respond to changes in load the majority of the time. At
extreme high and low monthly release volumes, these
ramp-rate restrictions have less effect. The maximum
daily change in flow constraint limits the 24-h change
in flow to 5,000; 6,000; or 8,000 cfs depending on the
monthly release volume. This constraint greatly reduces
the ability of the powerplant to respond to load changes
within any given 24-h period. At very high and very low
release volumes, the maximum daily change constraint
has less of an effect on hydropower operations. Readers
wishing to explore the effects of the MLFF constraints
at Glen Canyon Dam in conjunction with different
monthly release volumes and reservoir elevations can do
so by using an interactive computer model developed by
Harpman (2002). A rigorous mathematical exposition of
hourly hydropower models is presented in Edwards and
others (1999). A Microsoft Excel spreadsheet example of
such a model is described in Edwards (2003).
The relative effects of the hourly constraints at
Glen Canyon Dam and changes in monthly release
volumes are illustrated in figure 3, which is constructed
from Harpman (1997, 1999b) and from unpublished
results. The interim low fluctuating flow (iLFF) opera-
tion regime (interim operating criteria elsewhere in
this report) was a precursor to the MLFF. The iLFF
maximum release constraint was 20,000 cfs, and the
iLFF upramp rate was 2,500 cfs. Otherwise, these two
operational regimes are identical. Figure 3 compares the
monthly economic value of hydroelectricity produced
at Glen Canyon Dam in water year 1996 under three
conditions: historical operations of Glen Canyon Dam,
operation under the iLFF, and operation under the iLFF
with the beach/habitat-building flow (BHBF) of 1996.
Although the 1996 BHBF experiment started on March
22 and ended on April 8, to facilitate this 7-d high-
release experiment, changes in water-release volumes
were required over much of the water year. As shown in
figure 3, the changes in monthly release volumes neces-
sitated by the 1996 BHBF created a significant effect.
Relative to historical operations, the economic effects of
the iLFF hourly constraints are less pronounced.
The Costs of
Environmental Constraints
Relative to historical operations, the MLFF hourly
environmental constraints on hydropower operations (see
Overview, this report) have both short-run and long-run
effects. In the short run, the MLFF reduces the maxi-
mum generation ability of the powerplant to respond to
changes in load. As a result, more of the load must be
met by other generators in the system, typically thermal
generators. Since operation of thermal powerplants is
more expensive than hydropower, additional costs are
incurred. In the long run, new or replacement power-
plants are needed earlier than would otherwise be the
case. Increased capital costs associated with the con-
struction of new and replacement powerplants may be
substantial.
A number of economic analyses of changes in the
operation of Glen Canyon Dam have been undertaken.
Only three of these contain analyses of the MLFF alter-
native. Of these, only one study contains an estimate of
Status and Trends of Hydropower Production at Glen Canyon Dam 171
both the short-run and long-run power system impacts
of the MLFF. The Bureau of Reclamation Power
Resources Committee (PRC) estimated the long-run cost
of several alternative operating regimes at Glen Canyon
Dam by using a production expansion model. Using
regression analysis to interpolate between modeled alter-
natives, the PRC estimated that the annualized economic
cost of changing from historical operations to the MLFF
was $36.1 million (annualized value, 1996 dollars) per
year. Because of excess capacity in the system, most of
these costs were projected to be incurred late in the 50-yr
analysis period (Power Resources Committee, 1995).
Two very detailed short-run studies of MLFF effects
at Glen Canyon have been undertaken to date. These
studies differ in conceptual approach and intended appli-
cation. To correctly interpret their results, it is important
to distinguish between the approaches they employ.
A study by Harpman (1999a) estimated the short-
run economic effect of the MLFF relative to historical
operations. This effect is conceptually equivalent to the
difference between the historical line and the iLFF line
in figure 3. The purpose of this study was to estimate the
economic cost to all power users in the interconnected
system. Using an hourly constrained nonlinear optimiza-
tion model and spot market prices, Harpman estimated
that the economic cost of the MLFF was $6.173 million
(1996 dollars) for a representative 11.3 million acre-feet
(maf) (13,933 million m3) water year.
A short-run, ex post study of the 2000 low steady
summer flow (LSSF) experiment was undertaken by
Palmer and others (2004). Their approach was to esti-
mate the difference in hydropower value between the
MLFF and the 2000 LSSF relative to MLFF operations
in the absence of the LSSF. The approach employed
is conceptually equivalent to estimating the difference
between the BHBF and the iLFF in figure 3. The goal of
the Palmer and others (2004) study was to estimate the
ex post financial cost of the LSSF to Federal power users,
a subset of the power users in the interconnected system.
Early in water year 2000, additional water was released
from Lake Powell to create storage space in advance of
expected spring inflows and to facilitate low steady flows
during the summer experiment period. Because of the
onset of the drought, approximately 605,000 acre-feet
(af) (745,965,000 m3) of this water was not replaced by
inflows during water year 2000. Although some of the
inflow deficit was subsequently recovered as of Decem-
ber 2004, a 228,000-af (281,124,000 m3) storage deficit
remains, and the reservoir elevation is 2.9 ft (0.9 m) lower
than it would have been without the LSSF (Thomas
Ryan, Bureau of Reclamation, written commun., 2005).
Technically, the effects of the LSSF experiment are
still ongoing. In order to complete their study, Palmer
and others (2004) were forced to make several assump-
tions about when this inflow deficit would be recovered
and the LSSF experiment would be concluded. Using
these assumptions, observed prices, and an hourly linear
programming model, Palmer and others (2004) estimated
that the ex post cost of the LSSF experiment to Federal
power users was approximately $32 million (2000 dollars).
Status and Trends in
Hydropower Production
The average release at Glen Canyon Dam from
water year 1978 through 1999 was approximately 10.93
maf (13,477 million m3). Since the onset of the drought
in 2000, releases have been much lower than average,
and hydropower production has fallen annually. In water
year 2000, the annual release was 9.38 maf (11,566
million m3). The annual release from 2001 to 2004 has
reflected the minimum objective release of approxi-
mately 8.23 maf (10,148 million m3). As illustrated in
figure 4, diminished inflows to Lake Powell combined
with this nearly constant annual release have resulted in
markedly lower reservoir elevation levels.
The average annual hydropower production from
1978 to 1999 was approximately 5,196,113 MWh. As
shown in table 1, in recent years the production of
hydropower at Glen Canyon Dam has been consider-
ably below average. Generation has diminished since
the onset of the drought in 2000. Although the annual
Figure 3. Comparison of the monthly economic value in water
year 1996 of hydroelectricity produced at Glen Canyon Dam
under different operating regimes (interim low fluctuating
flows, iLFF; beach/habitat-building flow, BHBF; and historical
operations, 1963–91).
172 The State of the Colorado River Ecosystem in Grand Canyon
release over the last 4 yr has been approximately the
same (8.23 maf (10,148 million m3)), decreasing head,
or the distance water falls, has adversely affected annual
generation. In 2001, approximately 3,940,247 MWh
were generated. By water year 2004, only 3,320,196
MWh were generated.
Water year 2004 unregulated inflows to Lake Powell
were approximately 51% of the 30-yr average (1971–
2000). By the end of December 2004, the elevation of
Lake Powell had fallen to 3,564.6 ft (1,086.5 m), and
reservoir contents were approximately 8,678.0 thou-
sand acre-feet (kaf) (10,699,974 thousand m3), or 36%
of capacity. This decrease in storage capacity is 135.4
ft (41.3 m) below full pool and 74.6 ft (22.7 m) above
the minimum elevation necessary for power generation.
These conditions last occurred in 1969 when the reser-
voir was being filled. Hydropower generation at Glen
Canyon Dam is a function of both the amount of water
released through the turbines and the head (for example,
see Harpman, 2002, appendix 5).
As the reservoir elevation has fallen, the head avail-
able for hydropower production has declined, and this
decrease has adversely affected generation capacity.
When the reservoir is full (elevation 3,700 ft (1,128 m)),
at a release level of 25,000 cfs the generation capability
at Glen Canyon Dam is approximately 1,017 MW. At
an elevation of 3,564.6 ft (1,086.5 m) and a release level
of 25,000 cfs, the generation capacity at Glen Canyon
Dam is approximately 749 MW, a reduction of 268
MW or approximately 26.4% (John Brooks, Bureau of
Reclamation, written commun., 2005).
The Basin Fund
The Colorado River Storage Project, authorized by
the CRSP Act of 1956 (Public Law 84-485), is a pro-
gram to develop, and make available for use, the water
resources of the upper Colorado River Basin. The CRSP
is composed of 4 multipurpose storage units—Flaming
Gorge Dam, Wayne N. Aspinall Unit, Navajo Unit, and
Glen Canyon Dam, often referred to as the “mainstem
units”—and 21 authorized participating projects in the
States of Colorado, New Mexico, Utah, and Wyoming.
Financial exchanges necessary to the operation
of the CRSP are facilitated by the Upper Colorado
River Basin Fund (Basin Fund). The Basin Fund was
established by section 5 of the CRSP Act. The Basin
Fund is a financial instrument that obviates the need for
congressional actions to pay for project operation and
maintenance. It is a dedicated Treasury account for the
deposit of project revenues, which would otherwise be
deposited in the general fund, and a source of funds for
the payment of project expenses, which would otherwise
need to be appropriated. Existence of the Basin Fund
greatly streamlines the financial operations of CRSP and
participating projects.
As described in the CRSP Act, all revenues col-
lected in connection with the operation of the CRSP
and participating projects are credited to the Basin Fund.
Revenues are available, without further appropriation
for defraying the cost of operation, maintenance, and
replacements of and emergency expenditures for all
CRSP facilities, with the exception that for participating
projects, such costs are paid for with revenues gener-
ated from that project. Funds in excess of those needed
for project operation and maintenance needs, certain
environmental programs, and assigned costs of the
salinity control program are paid annually to the general
Figure 4. Lake Powell end of month (EOM) elevation in feet
above mean sea level (msl).
Table 1. Average and recent net hydropower
production at Glen Canyon Dam.
Water year Net generation (MWh)
1978–99 average 5,196,113
2000 4,600,453
2001 3,940,247
2002 3,772,544
2003 3,518,297
2004 3,320,196
Status and Trends of Hydropower Production at Glen Canyon Dam 173
fund of the Treasury to repay CRSP investment costs
with interest. Funds in excess of those requirements are
then apportioned to each of the upper Colorado River
Basin States to help defray investment costs associated
with developing the irrigation components of the 21
authorized participating projects. There are a number
of complex provisions, clauses, and details associated
with the CRSP Act, the Basin Fund, and project repay-
ment, but they are all outside the scope of this chapter. A
description of these aspects of the CRSP Act and their
implications for the recovery of the taxpayer investment
can be found in General Accounting Office (1996a).
As dictated by prudent business practice, the
SLCA/IP power rate is set at a level sufficient to meet
operational and repayment needs while accommodating
foreseeable variations in generation and resultant rev-
enue. Sometimes unforeseen, adverse hydrologic condi-
tions occur, necessitating greater than expected replace-
ment power purchases. Purchases of replacement power,
in excess of revenues, are made with funds from the
Basin Fund. If adverse hydrologic conditions continue,
the Basin Fund may be depleted and an increase in the
SLCA/IP rate would be required. Conversely, favorable
hydrologic conditions can result in greater than expected
revenues from hydropower sales; these additional rev-
enues are deposited into the Basin Fund. Funds in excess
of those needed for project operations are treated as
described in the CRSP Act. If favorable hydrologic
conditions persist for a number of years, repayment on
investment is accelerated, and a downward adjustment to
the SLCA/IP rate may be initiated.
The status of the Basin Fund fluctuates monthly,
thereby reflecting the timing of project revenues and
expenditures. Exclusive of replacement power purchases,
about $95 million is needed to fund CRSP project
operational needs on an average annual basis. The vast
majority of these revenues are derived from hydropower
sales in the CRSP system. Since 2000, extensive and
unplanned purchases of replacement power have been
required because of the drought. These purchases have
drawn down the cash reserve balance in the Basin Fund.
In mid-January 2005, the Basin Fund held approxi-
mately $50 million (David Taylor, Western Area Power
Administration, oral commun., 2005).
In addition to replacement power purchases, the
Basin Fund is used to meet CRSP payroll and other
operational and maintenance expenses. Certain envi-
ronmentally related expenditures, including the costs of
the Glen Canyon Dam Adaptive Management Program,
are also made from the Basin Fund. To the extent that
the Basin Fund is depleted during a period of adverse
hydrologic conditions, funding for all of these purposes
could be jeopardized. Western has taken two steps to
bolster CRSP revenues and improve the status of the
Basin Fund. First, Western has lowered their determi-
nation of marketable power resource (Western Area
Power Administration, 2004b). The lower determination
reduces the amount of replacement energy that they are
required to purchase and shifts more of the risk of being
energy-short in times of adverse hydrologic conditions to
power users. Second, Western has initiated a substantial
rate-increase action (Western Area Power Administra-
tion, 2005). As proposed, the new rate is $25.77/MWh,
a 24.4% increase over the current rate. Western’s rate
proposal also includes a provision for cost recovery
charge, which can be implemented when revenue short-
falls are projected. These steps will increase the amount
of revenue derived from hydropower sales and bolster
the position of the Basin Fund.
The Grand Canyon Protection Act of 1992 con-
tains special provisions establishing and funding the
Glen Canyon Dam Adaptive Management Program.
Section 1805 of the act establishes the long-term moni-
toring program now carried out by the Glen Canyon
Dam Adaptive Management Program, and section 1807
of the act declares expenditures for this program to be
nonreimbursable. These provisions shift the burden of
paying these costs from project beneficiaries, such as
water and power users, to taxpayers in general. Expendi-
tures by the Glen Canyon Dam Adaptive Management
Program are drawn from the Basin Fund. These pay-
ments are then credited towards project repayment,
thereby reducing the repayment obligation of all project
beneficiaries. As a result, these environmentally related
expenditures are indirectly paid for by all taxpayers in
the United States.
Outlook for the Future
In the years to come, the amount of hydropower
generated at Glen Canyon Dam is largely dependent on
future hydrologic conditions. While future hydrologic
conditions can never be known with certainty, probabi-
listic forecasts can provide some important insights. The
Bureau of Reclamation uses the RiverWare™ modeling
system (Zagona and others, 2001) for CRSP multiyear
planning studies. These multiyear planning studies
employ the indexed sequential method (described in
Ouarda and others, 1997) for simulating future hydro-
logic conditions. RiverWare™ modeling runs completed
in January 2005 use December 2004 CRSP end-of-
month reservoir elevations as starting conditions and
174 The State of the Colorado River Ecosystem in Grand Canyon
simulate monthly operations for 90 different monthly
inflow sequences, each of which is 90 yr long. Statistical
analysis of these simulations suggests that the median
time required to refill Lake Powell (achieve an eleva-
tion of 3,695 ft (1,126 m) in July) is approximately 19 yr
(Thomas Ryan, Bureau of Reclamation, written com-
mun., 2005). This evidence suggests that several decades
may pass before hydropower production at Glen Canyon
returns to the level of the late 1990s.
The scheduled replacement of the turbine run-
ners at Glen Canyon Dam will increase hydropower
generation from 1% to 7% (John Brooks, Bureau of
Reclamation, written commun., 2005). Turbine run-
ners are the large fan-like blades turned by the force of
water falling through the penstocks. The existing turbine
runners at Glen Canyon Dam have reached the end of
their design life and are now scheduled for replacement.
Improvements in runner technology, design, and fabrica-
tion methods now allow for improvements in efficiency
(more electricity produced for a given amount of water
released), greater generation capacity (higher genera-
tion level for a given head and water release level), or
both. Analysis of alternate turbine runner designs, their
costs, and benefits is now underway. Installation of new
turbine runners on two of the generation units at Glen
Canyon Dam is expected to begin within 2 yr, and all of
the turbine runners are expected to be replaced within
the next 10 yr.
The potential installation of temperature control
devices (TCDs) at Glen Canyon Dam is expected to
reduce generation by less than 1% (Bureau of Reclama-
tion, 1999). Thermal and chemical stratification develops
in Lake Powell during the summer months. The penstock
intakes are located at an elevation of 3,476 ft (1,059 m)
and are typically within the cold, hypolimnetic strata (see
chapter 4, this report). As a result, releases from Glen
Canyon Dam remain at about 50ºF (10ºC) all year long.
Native fish populations persist at these relatively low
temperatures, but it is believed that their spawning and
rearing success is drastically reduced (see chapter 2, this
report). The Bureau of Reclamation is studying the fea-
sibility of installing TCDs at Glen Canyon Dam (Bureau
of Reclamation, 1999). The purpose of these TCDs is to
allow for the management of downstream temperatures
to benefit existing and remnant populations of native
fish. Current plans call for installation of TCDs on two
of the eight generation units at Glen Canyon Dam.
Given current drought conditions, the outlook for
hydropower production at Glen Canyon Dam is some-
what bleak in the near term; however, conditions are
expected to improve in the future. Easing of the cur-
rent drought and subsequent gradual improvements in
hydrologic conditions in the Colorado River Basin are
expected to refill reservoirs and increase the amount of
hydropower that can be generated at Glen Canyon Dam
and other CRSP units. At any given reservoir elevation
and release, the planned installation of new turbine
runners will result in an increase in hydropower genera-
tion. Although the amount of CRSP generation has
been reduced by the drought in recent years and CRSP
rates are slated to increase, the hydroelectric energy
produced at Glen Canyon Dam has been, and continues
to be, one of the lowest cost sources of electric energy
available in the West.
Acknowledgments
This effort was supported by the U.S.
Geological Survey, Grand Canyon Monitoring and
Research Center, interagency agreement DCN-
5IA93953031. The assistance of Michael W. Loring,
regional economist, Bureau of Reclamation, Salt Lake
City, Utah, is gratefully acknowledged.
References
Bishop, R.C., Boyle, K.J., Welsh, M.P., Baumgartner,
R.M., and Rathbun, P.C., 1987, Glen Canyon Dam
releases and downstream recreation: an analysis of
user preferences and economic values. Glen Canyon
Environmental Studies Report No. 27/87: Madison,
Wis., Hagler Bailly Consulting, Inc. Springfield, Va.:
National Technical Information Service, NTIS no.
PB88-183546/AS, 188 p.
Bradley, R.C., 1964, Requiem for a canyon: Sacramento,
California Academy of Sciences, Pacific Discovery, v.
17, no. 3, p. 1–9.
Brower, D.R., 1997, Let the river run through it: Sierra,
v. 82, no. 2, p. 42–43.
Bureau of Reclamation, 1999, Glen Canyon Dam modi-
fications to control downstream temperatures—plan
and draft environmental assessment. U.S. Department
of the Interior, Bureau of Reclamation: Salt Lake City,
Utah, Bureau of Reclamation.
Status and Trends of Hydropower Production at Glen Canyon Dam 175
Collier, M.P., Webb, R.H., and Schmidt, J.C., 1996,
Dams and rivers: a primer on the downstream effects
of dams: Tucson, Ariz., U.S. Geological Survey
Circular 1126.
Edwards, B.K., 2003, The economics of hydroelectric
power. New horizons in environmental economics:
Northampton, Mass., Edward Elgar Publishing, Inc.
Edwards, B.K., Flaim, S.J., and Howitt, R.E., 1999,
Optimal provision of hydroelectric power under envi-
ronmental and regulatory constraints: Land Econom-
ics, v. 75, no. 2, p. 267–283.
Elfring, C., 1990, Conflict in the Grand Canyon—how
should Glen Canyon be managed?: BioScience, v. 40,
no. 10, p. 709–711.
Energy Information Administration, 1996, The chang-
ing structure of the electric power industry: an
update. DOE/EIA-D562 (96): Washington, D.C.,
U.S. Department of Energy, Energy Information
Administration.
Energy Information Administration, 2000, Federal
financial interventions and subsidies in energy mar-
kets 1999: energy transformation and end use. SR/
OIAF/2000-02: Washington, D.C., U.S. Department
of Energy, Energy Information Administration.
Energy Information Administration, 2004, Elec-
tric power annual 2003. DOE/EIA-0348(2003):
Washington, D.C., U.S. Department of Energy,
Energy Information Administration.
General Accounting Office, 1996a, Information on
the allocation and repayment of costs of construct-
ing water projects. Report no. GAO/RCED-96-109:
Washington, D.C., U.S. General Accounting Office.
General Accounting Office, 1996b, Power marketing
administrations: cost recovery, financing, and com-
parison to non-Federal utilities. Report no. GAO/
AIMD-96-145: Washington, D.C., U.S. General
Accounting Office.
General Accounting Office, 2000, Power marketing
administrations—their ratesetting practices compared
with those of non-Federal utilities. Report no. GAO-
00-114: Washington, D.C., U.S. General Accounting
Office.
General Accounting Office, 2001, Federal power—the
evolution of preference in marketing Federal power.
Report no. GAO-01-373: Washington, D.C., U.S.
General Accounting Office.
Harpman, D.A., 1997, Glen Canyon Dam beach/habi-
tat-building test flow—an ex post analysis of hydro-
power cost. U.S. Department of the Interior, Bureau
of Reclamation report no. EC-97-01: Springfield, Va.,
National Technical Information Service, NTIS no.
PB97-159321, 53 p.
Harpman, D.A., 1999a, Assessing the short-run eco-
nomic cost of environmental constraints on hydro-
power operations at Glen Canyon Dam: Land Eco-
nomics, v. 75, no. 3, p. 390–401.
Harpman, D.A., 1999b, The economic cost of the
1996 controlled flood, in Webb, R.H., Schmidt, J.C.,
Marzolf, G.R., and Valdez, R.A., eds., The controlled
flood in Grand Canyon: Washington, D.C., American
Geophysical Union, Geophysical Monograph Series, v.
110, p. 351–357.
Harpman, D.A., 2002, Understanding hourly opera-
tions at Glen Canyon Dam, Colorado River Storage
Project, AZ-GCPSE95 version 2.0. U.S. Department
of the Interior, Bureau of Reclamation report no.
EC-2002-01: Springfield, Va., National Technical
Information Service, NTIS no. PB2003-101395, www.
du.edu/~dharpman/glenpage.htm, 50 p.
Jacobs, J.W., and Wescoat, J.L., Jr., 2002, Managing river
resources: lessons from Glen Canyon Dam: Environ-
ment, v. 44, no. 2, p. 8–19.
Kearsley, L.H., Schmidt, J.C., and Warren, K.D., 1994,
Effects of Glen Canyon Dam on sand deposits used as
river campsites in the Grand Canyon of the Colorado
River, U.S.A.: Regulated Rivers: Research & Manage-
ment, v. 9, p. 137–149.
Long, M.E., and Essick, P., 1997, The Grand Man-
aged Canyon: National Geographic, v. 192, no. 1, p.
117–135.
Martin, R., 1989, A story that stands like a dam—Glen
Canyon and the struggle for the soul of the West (1st
ed.): New York, Henry Holt & Company.
McPhee, J., 1971, Encounters with the archdruid: New
York, Farrar, Straus & Giroux.
176 The State of the Colorado River Ecosystem in Grand Canyon
Nilsson, C., Jansson, R., and Zinko, U., 1997, Long-term
responses of river-margin vegetation to water-level
regulation: Science, v. 276, no. 5313, p. 798–800.
Ouarda, T.B.M.J., Labadie, J.W., and Fontane, D.G.,
1997, Indexed sequential hydrologic modeling for
hydropower capacity estimation: Journal of the
American Water Resources Association, v. 33, no. 6, p.
1337–1349.
Palmer, S.C., Burbidge, C., and Patno, H., 2004, The
financial impacts of the low summer steady flow
experiment at Glen Canyon Dam: Salt Lake City,
Utah, U.S. Department of Energy, Western Area
Power Administration, 32 p.
Parasiewicz, P., Schmutz, S., and Moog, O., 1998, The
effect of managed hydropower peaking on the physical
habitat, benthos, and fish fauna in the River Bregen-
zerach in Austria: Fisheries Management and Ecology,
v. 5, no. 5, p. 403–418.
Poff, N.L., Allan, J.D., Bain, M.B., Karr, J.R.,
Prestegaard, K.L., Richter, B.D., Sparks, R.E., and
Stromberg, J.C., 1997, The natural flow regime—a
paradigm for river conservation and restoration: Bio-
Science, v. 47, no. 11, p. 769–784.
Power Resources Committee, 1995, Power system
impacts of potential changes in Glen Canyon power
plant operations. Phase III final report: Denver, Colo.,
Stone and Webster Management Consultants, Inc.
Springfield, Va., National Technical Information
Service, NTIS no. PB96-114004, 305 p.
Seitz, J., 2004, Output capacity of Glen Canyon pow-
erplant 1964–2004. Technical investigation (D-8440):
Denver, Colo., Bureau of Reclamation, Technical
Service Center.
Van Steeter, M.M., and Pitlick, J., 1998a, Geomor-
phology and endangered fish habitats of the upper
Colorado River 1. Historic changes in streamflow,
sediment load, and channel morphology: Water
Resources Research, v. 34, no. 2, p. 287–302.
Van Steeter, M.M., and Pitlick, J., 1998b, Geomor-
phology and endangered fish habitats of the upper
Colorado River 2. Linking sediment transport to
habitat maintenance: Water Resources Research, v. 34,
no. 2, p. 303–316.
Welsh, M.P., and Poe, G.L., 1998, Elicitation effects
in contingent valuation: comparisons to a multiple
bounded discrete choice approach: Journal of Envi-
ronmental Economics and Management, v. 36, no. 1,
p. 170–185.
Western Area Power Administration, 2004a, Annual
report 2003: U.S. Department of Energy, Western
Area Power Administration.
Western Area Power Administration, 2004b, Salt Lake
City integrated projects—notice of determination of
the post-2004 marketable resource: Federal Register, v.
69, no. 98, Thursday, May 20, 2004, p. 29135–29136.
Western Area Power Administration, 2005, Salt Lake
City integrated projects—rate order WAPA-117: Fed-
eral Register, v. 70, no. 11, Tuesday, January 18, 2005,
p. 2858–2859.
Zagona, E.A., Fulp, T.J., Shane, R., Magee, T., and
Goranflo, H.M., 2001, RiverWare: a generalized tool
for complex reservoir system modeling: Journal of the
American Water Resources Association, v. 37, no. 4, p.
913–929.
Contact Information:
David A. Harpman
Natural Resource Economist
U.S. Department of the Interior
Bureau of Reclamation
Denver, CO
dharpman@do.usbr.gov
Aaron J. Douglas
Natural Resource Economist
U.S. Department of the Interior
U.S. Geological Survey
Fort Collins, CO
aaron_douglas@usgs.gov
First page photograph credit: Bureau of Reclamation
Chapter 11
Cultural
Resources in
the Colorado
River Corridor
Helen C. Fairley
Introduction
Cultural resources along the corridor of the
Colorado River include archaeological sites and other
types of historic properties, as well as resources that are
of traditional concern to Native American peoples such
as springs, landforms, sediment and mineral deposits,
native plant concentrations, and various animal species.
All of these resources have the potential to be affected
indirectly, and in some cases directly, by the operations
of Glen Canyon Dam.
The principal cultural resource goal of the Glen
Canyon Dam Adaptive Management Program is to
“preserve, protect, manage and treat cultural resources
for the inspiration and benefit of past, present and
future generations” (Glen Canyon Dam Adaptive
Management Program, 2001). The National Park
Service (NPS) goal for managing archaeological and
historic resources in the Colorado River corridor is in-
place preservation with minimal impact to the integrity
of the resources. When in-place preservation is not pos-
sible, the NPS and other Federal agencies consider data
recovery through excavation of archaeological remains
to be an appropriate alternative in certain cases. The six
Native American tribes who actively participate in the
Glen Canyon Dam Adaptive Management Program and
have long-standing traditional ties to the Grand Canyon
region—Hopi Tribe, Hualapai Tribe, Kaibab Band of
Paiute Indians, Paiute Indian Tribe of Utah, Navajo
Nation, and Pueblo of Zuni—are generally supportive
of in-place preservation goals for cultural resources,
but they have widely varying opinions regarding the
appropriateness of undertaking intervention measures to
mitigate dam and visitor impacts, such as installing check
dams to control erosion or conducting excavations to
recover information from archaeological sites.
This chapter describes research, monitoring, and
mitigation activities during the past 15 yr that have
evaluated and addressed ongoing impacts to cultural
resources in the Colorado River corridor because of
dam operations and other agents of deterioration, such
as visitation and rainfall-induced erosion. The chap-
ter begins with a summary of research and inventory
activities prior to the early 1990s, which is followed by
a summary of the monitoring and research activities
initiated in response to the Operation of Glen Canyon
Dam Final Environmental Impact Statement (EIS) and
the Secretary of the Interior’s Record of Decision (ROD)
178 The State of the Colorado River Ecosystem in Grand Canyon
(U.S. Department of the Interior, 1995, 1996). The chap-
ter ends with some recommendations for the future.
Background
Current information concerning cultural resources is
based on a number of previous investigations within the
Colorado River corridor in Glen and Grand Canyons.
Comprehensive overviews of previous investigations are
included in Ahlstrom and others (1993), Fairley and oth-
ers (1994), and Fairley (2003). These studies documented
evidence of human occupation in the Grand Canyon
region extending back as far as the Paleo-Indian period,
at least 11,000 yr before present. Starting around 4,500
yr ago, petroglyphs, spear points, and uniquely crafted
artifacts known as “split twig figurines” were placed at
various locations along and near the Colorado River
in Glen and Grand Canyons, indicating intermittent
use of the area by Late Archaic hunters and gatherers.
Sparse and somewhat controversial evidence for use of
the canyon by early farming cultures before 1000 B.C.
is present in eastern Grand Canyon (Davis and others,
2000); however, the best documented and most inten-
sive use of Grand Canyon by farmers occurred during
the 11th and early 12th centuries A.D., during a time
period known as Pueblo II. During this time, people with
ancestral ties to modern Puebloan cultures built numer-
ous small masonry dwellings, irrigation ditches, erosion
control features, and granaries for storing corn, squash,
and cotton throughout Grand Canyon (fig. 1). For rea-
sons not yet fully understood (but that are likely due at
least in part to climate-induced stressors), these ancestral
Puebloan farmers moved away from Grand Canyon
around the end of the 12th century, although small
groups continued to visit the canyon for seasonal hunt-
ing, plant gathering, trading, and ceremonial pilgrimages
into the first decades of the 20th century. Following the
departure of the ancestral Puebloan occupants, ancestors
of the Hualapai, Havasupai, and Southern Paiute moved
into the region. Numerous campsites and food process-
ing areas dating to the late prehistoric (A.D. 1300–1540),
proto-historic (A.D. 1540–1776), and early historic (A.D.
1776–1850) periods testify to their extensive, and at
times intensive, seasonal use of inner Grand Canyon.
Descendants of these Yuman and Numic speakers were
still residing in and around Grand Canyon when the
first European settlers arrived in the region in 1848,
and Havasupai, Hualapai, and Southern Paiute people
continue to live in and near Grand Canyon to this day.
After 1880, however, as Euro-Americans became increas-
ingly enamored with the spectacular scenery and eco-
nomic potential of the region, the archaeological record
becomes increasingly dominated by the material remains
of Euro-American miners, trappers, homesteaders, gov-
ernment explorers, and tourists (fig. 2).
Previous Cultural Resource
Research and Inventories
Euro-Americans first noted archaeological remains
in the river corridor during the Powell expeditions
of 1869 and 1871–72 (Powell, 1875). Powell and his
crew found traces of previous human occupation in
the canyon in the vicinity of the Little Colorado River,
Unkar Creek, Bright Angel Creek, and Shinumo Creek.
In the 1950s and 1960s, investigations of archaeologi-
cal remains in the river corridor became more focused
under the direction of the NPS, in part because of
anticipated dam developments in Grand Canyon (Taylor,
1958; Euler, 1967a). In the late 1960s and early 1970s,
researchers affiliated with the School of American
Research, New Mexico, and Prescott College, Arizona,
conducted surveys and excavations in the river corri-
dor and adjacent areas to investigate prehistoric settle-
ment patterns (Schwartz, 1965; Euler and Taylor, 1966;
Euler, 1967b; Schwartz and others, 1979, 1980, 1981).
Figure 1. Pottery dating to the Pueblo II period is commonly
found at archaeological sites along the Colorado River. These
formerly buried pot sherds have become exposed by wind
(photograph by Amy Draut, U.S. Geological Survey).
Cultural Resources in the Colorado River Corridor 179
Together, these studies provided the foundational infor-
mation concerning the numerous and diverse cultural
resources existing within the river corridor.
In 1984, NPS archaeologists conducted test exca-
vations at five archaeological sites along the Colorado
River in advance of their stabilization (Jones, 1986).
Three of these sites were deteriorating primarily because
of visitor use, one was deteriorating because of both
human and natural impacts, and one was being dam-
aged by erosion from a side channel, all resulting in the
need for stabilization measures. At all of these sites, exca-
vations revealed the presence of subsurface stratigraphic
cultural deposits extending back many centuries earlier
than surface evidence indicated. For example, at one site
(AZ B:10:4) near Deer Creek, a buried roasting feature
produced calibrated radiocarbon dates ranging between
A.D. 610 and 380 B.C., whereas the overlying surface
structure contained ceramics dating to the Pueblo I-II
period (about A.D. 800–1150) (Jones, 1986, p. 105). At
another site (AZ B:16:1) near Whitmore Wash, surface
materials of Southern Paiute affiliation were found in
association with a fire pit that was radiocarbon dated to
A.D. 1230–1340, overlying a roasting feature with a cali-
brated radiocarbon age of 1365–905 B.C. (Jones, 1986,
p. 51). This project was important for highlighting the
presence of deeply buried cultural deposits at numerous
archaeological sites in the river corridor (fig. 3). These
older, underlying deposits are often invisible on the sur-
face because floods, slope wash, and aeolian (wind) pro-
cesses have deposited sediment on top of earlier cultural
remains, obscuring them from view.
In 1990–91, an intensive archaeological inven-
tory was conducted by NPS archaeologists (Fairley
and others, 1994) in preparation for the writing of
the EIS (U.S. Department of the Interior, 1995).
This inventory located 475 sites within the assessed
area, which extended from Glen Canyon Dam to
Separation Canyon, about 255 RM, and up to the
estimated 300,000 cubic feet per second (cfs) flood
level. The sites ranged in type from isolated hearths and
dispersed lithic scatters to complex multiple-component
habitations, many with associated roasting features or
masonry structures. The sites ranged in age from Late
Archaic, about 2500–1500 B.C., to the mid-20th century.
Many sites date to the ancestral Puebloan occupation
between A.D. 950 and 1200, while many other sites are
affiliated with the ancestral Pai and Paiute use of Grand
Canyon from about A.D. 1250 to 1870.
Of the sites within the surveyed area, approximately
336 were considered to be situated within the area of
potential effect from dam operations, and many of these
sites had identifiable impacts that were believed to be
related to dam operations (Fairley and others, 1994, p.
148). Dam-related impacts were categorized as direct,
indirect, or potential. Direct impacts included sites
where inundation or bank cutting from dam-controlled
Figure 2. The remains of Bert Loper’s boat came to rest on
the banks of the Colorado River in 1948, shortly after Loper
drowned in 24 and 1/2-mile Rapid. This boat is an example of the
varied historic artifacts found along the Colorado River in Grand
Canyon (photograph by Jeff Sorensen, Arizona Game and Fish
Department).
Figure 3. The curved masonry wall of a deeply buried prehistoric
structure was uncovered during excavations at site AZ C:13:10 in
April 1984. No evidence of this structure was visible on the site
surface prior to excavation (photograph by Helen Fairley, U.S.
Geological Survey).
180 The State of the Colorado River Ecosystem in Grand Canyon
river flows had occurred within the site in recent years.
Indirect impacts included (1) bank slumpage or slope
steepening from river flows immediately adjacent to the
site, (2) arroyo cutting or other erosion phenomena tied
to the effects of dam-controlled flows, and (3) effects of
visitor impacts at sites because of changes in recreational
use patterns related to recent dam operations. Poten-
tially impacted sites included all those located within the
estimated area of inundation from a 300,000-cfs flood.
This flood level reflected the former estimated volume of
the highest historical flood on record (in 1884) (Hereford
and others, 1993; cf. Topping and others, 2003, p. 31)
and also the maximum release level possible from Glen
Canyon Dam, estimated at 256,000 cfs, combined with
a hypothetical 40,000-cfs flood event from the Little
Colorado River and other tributary streams.
Of the 336 sites considered to be within the area of
potential effect, 33 showed evidence of direct impacts,
138 revealed evidence of indirect impacts, and 238 were
categorized as potentially impacted based on their loca-
tion below the estimated 300,000-cfs level or based on
their location in or on unconsolidated Holocene sedi-
mentary deposits (Fairley and others, 1994, p. 148). (The
impact numbers exceed total number of sites because
many sites exhibited more than one category of impact.)
Participating Native American tribes also conducted
cultural resource inventories in the early to mid-1990s
to identify resources with important traditional cul-
tural values, including but not limited to prehistoric
Native American archaeological sites. These studies
were conducted by the Hopi Tribe, the Hualapai Tribe,
the Navajo Nation, the Southern Paiute Consortium
(comprising the Kaibab Band of Paiute Indians and the
Paiute Indian Tribe of Utah), and the Pueblo of Zuni
(Stoffle and others, 1994, 1997; Hart, 1995; Roberts
and others, 1995; Ferguson, 1998; Stevens and Mercer,
1998). Numerous locations of cultural importance were
identified and evaluated by the individual tribes, includ-
ing areas with culturally important biological resources,
significant landscape features, mineral locations, and
specific archaeological resources. Assessments were
conducted by the tribes to identify potential impacts
resulting from dam operations and to formulate possible
treatment options. These studies have subsequently been
used by the Bureau of Reclamation for the identification
and preliminary evaluation of traditional cultural proper-
ties within the area of potential effect as defined by the
Programmatic Agreement for Cultural Resources (see dis-
cussion below) (U.S. Department of the Interior, 1995, p.
Att-23); however, studies to formally define and evaluate
traditional cultural properties have yet to be completed.
Monitoring and Research
Monitoring of Cultural Resources
Before the 1990s, the NPS annually monitored a
sample of archaeological sites in the Colorado River
corridor. These initial monitoring efforts focused primar-
ily on sites prone to impacts from visitors. After the com-
prehensive inventory of the river corridor was completed
in 1991, the NPS expanded its monitoring program to
encompass the full suite of resources located within the
hypothetical area of potential effect from dam opera-
tions. In the mid-1990s, the Southern Paiute Consortium
and the Hualapai and Hopi Tribes also initiated their
own monitoring programs to track changes in resource
conditions at culturally important locations in the river
corridor (Stoffle and others, 1995; Ferguson and others,
1997).
Since 1994, monitoring of historic properties that
are eligible to be listed in the National Register of
Historic Places—nationally, regionally, and locally sig-
nificant prehistoric and historic sites, structures, objects,
and places of traditional cultural importance—has
been conducted under the auspices of a Programmatic
Agreement for Cultural Resources. The agreement exists
between the Bureau of Reclamation, NPS, Advisory
Council on Historic Preservation, Arizona State Historic
Preservation Office, and six affiliated Native American
tribes. Current monitoring protocols, established under
interim guidelines of the Programmatic Agreement for
Cultural Resources, document the presence of all types
of impacts occurring to archaeological resources in the
river corridor, regardless of ultimate cause.
In addition to the NPS monitoring efforts, the Hopi
Tribe, Hualapai Tribe, and Southern Paiute Consortium
conduct annual monitoring trips to assess changes to
their traditional cultural resources and to assess the
general health of the ecosystem through their own tradi-
tional value system. Tribal monitoring has been con-
ducted both through and outside of the programmatic
agreement, as not all resources of tribal concern meet
the established definitions of National Register-eligible
historic properties.
The Programmatic Agreement for Cultural
Resources is concerned with tracking and mitigating
dam effects at approximately 318 National Register-
eligible archaeological sites in the river corridor: 54 in
Glen Canyon National Recreation Area and 264 in
Grand Canyon National Park (Leap and others, 2000,
p. I-8). Approximately 160 of these sites are actively
Cultural Resources in the Colorado River Corridor 181
monitored at the present time. All sites currently moni-
tored fall within the affected environment as defined by
the EIS (U.S. Department of the Interior, 1995). The
monitoring is carried out by staff from the NPS, work-
ing with cooperators from Northern Arizona University.
National Park Service archaeologists conduct monitor-
ing trips several times each year and produce annual
monitoring reports, which are submitted to the Bureau
of Reclamation in partial fulfillment of an ongoing coop-
erative agreement (Leap and others, 2000). Currently,
archaeological sites above Lees Ferry are not being moni-
tored (Chris Kincaid, Glen Canyon National Recreation
Area, oral commun., 2004); only sites downstream of
Lees Ferry are routinely monitored.
Archaeological sites are currently selected for
monitoring and remedial treatments based on interim
protocols established under the programmatic agree-
ment. These protocols include judgmentally selecting
sites for monitoring based on perceived susceptibility or
likely vulnerability to erosion or visitor impacts. Sites are
monitored on a cycle that varies from semiannually to
annually, biennially, or once every 3, 4, or 5 yr (Leap and
others, 2000). Monitoring cycles are assigned on the basis
of perceived levels of stability or visitor use, but the mon-
itoring cycles are not rigidly adhered to and frequently
change. For example, of the 91 sites monitored in fiscal
year 2001 (FY01), 55% (n = 50) were monitored more or
less frequently than their assigned monitoring cycle, and
18% were reassigned to a new monitoring cycle based on
perceived changes in their stability. Sites that are stable
or show no signs of visitor use are not included in the
current monitoring program. This bias in site selection
was intentionally designed to focus attention on those
sites that were theoretically at greatest risk from dam-
age from visitor use and erosion and were most likely to
require preservation treatments in the foreseeable future;
however, the deliberate emphasis placed on monitoring
sites that are assumed to be most threatened limits the
usefulness of the resulting data for drawing systemwide
conclusions about status and trends of site condition,
rates of impacts, or overall effects of dam operations on
historic properties.
The main goals of the current monitoring program
are to document site impacts and evaluate the need for
site protection measures such as erosional control check
dams. Changes in the numbers, types, and locations of
site impacts are documented in yearly reports prepared
by the NPS and Northern Arizona University coopera-
tors (see Leap and others, 2000, for a listing of annual
reports through 1999; see also Leap and Kunde, 2000;
Dierker and others, 2001, 2002; Leap and others, 2003).
These reports discuss the results of site-specific evalua-
tions, identify specific changes occurring at individual
sites, and make recommendations about future protec-
tion measures, including data recovery. The reports do
not track systemwide trends in site condition or evaluate
changes in site condition relative to dam-controlled flows.
Currently, archaeological site-monitoring activities
conducted under the programmatic agreement involve
repeat site visits, visual assessments of site impacts, and
qualitative assessments of overall condition, which are
documented through the use of repeat photography
and completion of a two-page checklist of impacts. As
described in the FY02 annual monitoring report (Dierker
and others, 2002, p. 2),
Archaeologists qualitatively assess impacts to sites
via repeat observations. The degree of impact is
categorized as “present” or “absent,” with physi-
cal erosion further categorized as “active” or
“inactive.” Active erosion is defined as obvious
recent movement, disturbance, or rearrange-
ment of sediment or artifacts onsite. Inactive
erosion is defined as a (less obvious) perception
that past geophysical processes are discernable
at the site, but are not presently at work.
Visitor impacts are recorded as present or absent
in five categories: social trails, artifact collection piles,
evidence of onsite camping, criminal vandalism, and
other impacts (fig. 4). Physical impacts are recorded as
present or absent and either active or inactive within
the following eight categories: surface erosion, gullying,
Figure 4. Visitors frequently remove artifacts from their original
locations and concentrate them in “collection piles,” resulting
in loss of information about the original context of the artifacts
(photograph by Helen Fairley, U.S. Geological Survey).
182 The State of the Colorado River Ecosystem in Grand Canyon
arroyo cutting, bank slumpage, aeolian/alluvial ero-
sion or deposition, side canyon erosion, animal-caused
erosion, and other erosion (fig. 5). Impacts that the NPS
views as being directly related to dam operations include
bank slumpage and gullying/arroyo cutting in locations
where drainage systems are actively entrenching to
achieve grade with the present-day “highest discharge”
terrace levels formed under dam-controlled flows. The
precise role of dam operations relative to other erosional
forces—precipitation events, human trampling, wind,
and other “natural” and “cultural” agents of erosion—in
causing or exacerbating erosion of archaeological sites in
the river corridor remains a topic of continuing contro-
versy in the scientific community.
Because the current archaeological site monitor-
ing program does not measure or otherwise attempt
to quantify impacts or rates of change in either a rela-
tive or absolute sense, it is difficult to draw any specific
conclusions about overall trends in resource condition in
relation to either the interim operating flows of 1991–95
or the modified low fluctuating flow (MLFF) alternative
implemented in 1996 (U.S. Department of the Interior,
1996). Two conclusions can be drawn, however, by
using the currently available monitoring information: (1)
archaeological sites continue to receive impacts from visi-
tor use and erosion, and (2) archaeological site conditions
are likely to continue to deteriorate (at an unknown rate)
because impacts from visitor use and erosion are ongoing
and not likely to diminish in the foreseeable future.
Erosion Control with Check Dams
Beginning in 1995, the NPS began installing rock
and brush check dams at selected archaeological sites in
the river corridor in an attempt to control erosion. Check
dams were first piloted as an erosion control measure
in an area below the Little Colorado River known as
“Palisades,” where gullies bisect two archaeological sites
(AZ C:13:99 and AZ C:13:100). National Park Service
archaeologists had monitored and documented a pro-
gressive deepening and widening of the drainages in the
Palisades area since 1978. Continuing channel erosion
caused the collapse and disappearance of numerous
slab-lined cists and portions of masonry structures in the
late 1980s and early 1990s, hence the decision to initiate
erosion control measures. With assistance and supervi-
sion from the Zuni Conservation Project (a team of soil
conservation experts from the Pueblo of Zuni), NPS
archaeologists installed 70 check dams at the two sites by
using a variety of local materials and construction styles
(Leap and Coder, 1995; Leap and others, 2000).
Since the initial pilot project in 1995, NPS archae-
ologists have installed approximately 280 check dams at
29 different archaeological sites (Leap and others, 2000).
Currently, 260 erosion-control features are actively moni-
tored and maintained at 27 sites (Leap and others, 2003,
p. 58). An evaluation of check dam effectiveness con-
ducted in 2002 (Pederson and others, 2003) found that
the brush checks built with a “basket weave” technique
seemed to work best and caused less damage to sur-
rounding terrain upon failure than did check dams built
Figure 5. An erosional gully cutting headward into a
prehistoric roasting feature (photograph by Amy Draut, U.S.
Geological Survey).
Figure 6. Members of the Zuni Soil Conservation Project
constructing a check dam by using the basket-weave technique
(photograph courtesy of Grand Canyon National Park).
Cultural Resources in the Colorado River Corridor 183
with rocks or logs (fig. 6). The researchers observed that
brush checks tended to fail in their central sections or get
ripped out as a woody mass, whereas the more rigid rock
and log checks were often flanked by the gullies via lat-
eral slope erosion, further exacerbating erosional impacts
to the sites. Pederson and others (2003) concluded that
brush checks were less damaging than stone checks and
that check dams, without routine maintenance, could
cause more harm than good; however, check dams could
temporarily slow rates of erosion provided they were
routinely maintained. These findings are considered
somewhat tentative and in need of further verification
because they were based on observations conducted over
the course of a single monsoon season in 2002, which
was one of the driest monsoon seasons on record.
Test Flow Impacts on
Cultural Resources
Many of the archaeological resources along the cor-
ridor of the Colorado River are situated on or contained
within the Holocene sedimentary deposits, which form
dunes and terraces (fig. 7). The sediment resource has
declined, and the alluvial terraces have eroded since the
completion of Glen Canyon Dam. A systemwide method
for regenerating the river terraces and redistributing sedi-
ment is considered an essential component to maintain-
ing integrity of cultural resources in place.
The 1996 beach/habitat-building flow, or controlled
flood, presented an opportunity to study the effects of
high-flow discharge from Glen Canyon Dam on alluvial
terraces and margin deposits along the river corridor.
Although the effects of the 1996 beach/habitat-building
flow of 45,000 cfs on archaeological sites could not be
predicted, the hope was that it could provide systemwide
mitigation to most cultural sites in the Colorado River
corridor through the accumulation of additional sedi-
ment at higher elevations than normally would occur
under the MLFF alternative (Balsom and Larralde, 1996,
p. 3). Mitigation and monitoring of archaeological sites,
ethnobotanical resources, and sediment accumulation at
the mouths of arroyos were undertaken to evaluate the
effects of this experimental high flow. In addition, rates
of terrace retreat were studied in the Glen Canyon reach
to determine whether terraces containing archaeological
sites were negatively affected by the high flows (Balsom
and Larralde, 1996).
The overall findings of the cultural resource studies
done in conjunction with the 1996 beach/habitat-build-
ing flow were that the 45,000-cfs flow had either no
effect, no adverse effect, or in some instances a beneficial
effect on cultural resources (Balsom and Larralde, 1996,
p. 25). In a few locations, however, especially in the Glen
Canyon reach, loss of sediments occurred in a manner
that, in the long run, could be detrimental to cultural
resources. Follow-up studies conducted by the Hopi
Tribe (Yeatts, 1998) and by Northern Arizona University
researchers (Hazel and others, 2000) found that sediment
deposited in arroyo mouths by the 1996 beach/habi-
tat-building flow persisted in some locations for several
years, especially where brush check dams had been
installed in the lower reaches of the drainages. These
studies, however, did not specifically evaluate whether
the sediment plugs diminished the rate of down-cutting
in upper reaches of the affected gullies. Nevertheless, the
studies demonstrated that backfilling of some erosional
channels could be accomplished by periodic high-flow
events and that high, sediment-enriched flows offer one
potential means of conducting systemwide mitigation for
effects of dam operations on cultural resources.
Since 1996, several additional test flows have taken
place, including the 2000 low summer steady flows
experiment, the 2003–05 fluctuating nonnative fish
suppression flows, and the November 2004 experimen-
tal high flow. No specific cultural resource monitoring
programs were conducted in conjunction with the 2000
low summer steady flows or the 2003–05 fluctuating
nonnative fish suppression flows; however, analysis of
the sediment mass balance under the 2003 and 2004
winter nonnative fish suppression flows showed that the
Figure 7. A buried Pueblo II structure is becoming exposed
by erosion near Unkar Delta (photograph by Amy Draut, U.S.
Geological Survey).
184 The State of the Colorado River Ecosystem in Grand Canyon
fluctuating flows eroded material above and beyond
the amount supplied by tributaries within the previ-
ous year. Also, analysis showed that higher levels of
very fine grained material were entrained during the
initial few days of the 2003 and 2004 high fluctuating
flows, suggesting the possibility that fine sediments that
were derived from predam deposits contributed to the
sediment being transported out of the system (David
Topping, U.S. Geological Survey, oral commun., 2005).
Although not conclusive, this information suggests that
there may be some additional loss of predam terrace
deposits, where most archaeological sites are situated,
occurring under the experimental fluctuating nonnative
fish suppression flows.
During the 2004 experimental high flow, one of the
areas monitored during the 1996 beach/habitat-build-
ing flow and two additional sites upstream in Marble
Canyon were monitored to determine whether sandbars
would be created in the vicinity of archaeological sites, in
locations where the new sandbars could serve as sources
of sediment for windborne redeposition on downwind
archaeological sites. An additional aim of this research
was to document whether sediment would be depos-
ited in the mouths of arroyos that currently bisect some
archaeological sites and whether these sediment “plugs”
would be retained long enough to help reduce the rate of
downcutting in affected arroyos because of the tempo-
rary elevation increase at the arroyo mouths. Prelimi-
nary observations indicate that large sandbars did form
upwind of the site areas and that sediment did backfill
arroyos; however, at the time this report was being writ-
ten, it was still too early to determine whether the 2004
experimental high flow would benefit the archaeological
sites over the long term. U.S. Geological Survey scien-
tists will be monitoring the fate of the new sandbars
and associated archaeological sites over the next year to
determine whether and to what degree the newly formed
sandbars contribute sediment to the windborne deposits
that blanket sites located at higher elevations.
The Role of Windborne Sediment
in Preserving Archaeological Sites
In addition to the monitoring activities previously
described, several research projects have been initiated
and supported through the Glen Canyon Dam Adaptive
Management Program over the past decade to improve
overall understanding of how Glen Canyon Dam
operations may be affecting archaeological sites in the
Colorado River corridor below the dam (Thompson and
others, 2000; Draut, 2003; Pederson and others, 2003;
Draut and others, 2005; Wiele and Torizzo, 2005).
Of particular interest is whether and how postdam
changes in overall sediment supply and flow regimes
downstream of the dam may be contributing to the ero-
sion of high-elevation dunes and terraces bordering the
Colorado River, where many archeological sites occur.
These terraces are bisected by numerous arroyos and
gullies draining to the river, and many of the erosion
channels pass through or by archaeological features.
Pieces of anecdotal and empirical (aerial photography)
evidence indicate that these gullies have been increasing
in size, depth, and abundance over the past four decades
(Hereford and others, 1993). The question that both dam
managers and NPS managers are seeking to answer is,
“To what extent is the ongoing erosion of these higher
elevation terraces related to dam operations?”
Scientists generally agree that gullies are formed in
response to specific, and often unusually intense, pre-
cipitation events (e.g., Webb, 1985). If precipitation is
responsible for the establishment of gullies and arroyos,
what might dam operations have to do with the ongo-
ing erosion of the predam terraces? The answer lies in
understanding the dynamic nature of the predam fluvial
system in contrast with the present-day, dam-controlled
hydrologic system, which is also very dynamic but in
fundamentally different ways (Topping and others,
2003). The predam system was generally characterized
by high seasonal variability and low daily variability.
Flows during the winter months were typically quite low,
often running at less than 3,000 cfs. The flows usually
stayed low until late April or early May, when runoff
from the Rocky Mountains started making its way to the
Gulf of California. Spring flows typically peaked in June
or early July, with additional spikes in late summer in
response to localized monsoon storm events. The annual
spring snowmelt floods ranged between about 35,000 and
120,000 cfs and averaged around 55,000 cfs, with peak
flows of 120,000 cfs reoccurring about once every 6 yr
(Topping and others, 2003). The highest known flood in
historic times occurred in 1884 with an estimated flow of
210,000 ± 30,000 cfs. In 1921, a flood of 170,000 cfs was
measured at the Grand Canyon gage (Topping and others,
2003, p. 31). Floods even larger than this are known from
the geomorphic record (O’Connor and others, 1994).
The high spring flows typically carried huge sedi-
ment loads. An analysis of the historical predam sedi-
ment transport records from the Lees Ferry gage and
Grand Canyon gage shows that the monthly sediment
loads during May averaged around 13.9 to 17.6 million
tons (12.6 and 16 million Mg) per month, respectively,
Cultural Resources in the Colorado River Corridor 185
which is close to 20% of the annual amount transported
each year (Topping and others, 2000b). As flood waters
receded, sand was deposited at and below the flood stage
along the river banks. After these flood deposits dried
out, wind transported the fluvial sediment farther inland,
where some of it covered archaeological sites and formed
coppice dune fields around mesquite thickets.
Today, virtually 100% of the sediment load that
used to be transported by the river through Grand
Canyon is trapped upstream in Lake Powell. Two major
tributaries below Glen Canyon Dam, the Paria River
and the Little Colorado River, contribute the bulk of the
current sediment coming into the river system. Taken
together, the contributions of sand from various sources
provide Grand Canyon with approximately 16% of its
predam levels (see chapter 1, this report). Compound-
ing the effects of this drastic reduction in sediment, the
dam is operated to meet peak power demands, so in the
postdam era prior to the Record of Decision (1963–96)
the Colorado River fluctuated by as much as 25,000 cfs
on a daily basis, and the daily discharge range exceeded
10,000 cfs on 43% of all days (before the dam, daily
ranges in excess of 10,000 cfs occurred on only about
1% of all days) (Topping and others, 2003). Further-
more, flows higher than 9,000 cfs essentially guarantee
that any fine-grained sediment coming into the system
will be transported downstream to Lake Mead in a
period of a few weeks to a few months (Topping and
others, 2000 a, b; Rubin and others, 2002), and analy-
sis of the continuous discharge record from Lees Ferry
gage and the Grand Canyon gage demonstrates that
flows greater than this level have dominated the postdam
record (Topping and others, 2003, p. 48).
The reduction in sand supply translates into a
reduction in the size, height, and volume of sandbars
throughout the river corridor (Hazel and others, 1999;
Schmidt and others, 2004). Furthermore, because dam-
controlled flows are generally constrained below 25,000
cfs, sand and silt are no longer being deposited at higher
elevations where fine sediment would be less susceptible
to riverine transport, more readily available for inland
transport by wind, and able to backfill the lower reaches
of arroyos and gullies that dissect the terraces.
Aeolian (windborne) sediment has previously been
shown to play an important role in the formation and
subsequent reworking of terraces where many archaeo-
logical sites are located (Hereford and others, 1993,
1996), and aeolian deposition has been hypothesized
to play an important role in mitigating the effects of
runoff erosion (Lucchitta, 1991). In order to improve
our understanding of how changes in sediment supply
and river flow dynamics are affecting the archaeological
sites, a study was initiated in 2003 to examine the role of
aeolian sediment in preserving archaeological sites from
several different perspectives: (1) the relative importance
of aeolian sedimentation in the past compared to today,
both as a terrace-forming process and also in backfilling
incipient rills and gullies; (2) the extent to which aeolian
sand cover may be diminishing throughout the ecosystem
under current sediment-limited conditions; (3) the extent
to which current rates of aeolian transport vary at differ-
ent locations under varying ecological and geomorphic
parameters throughout the river corridor; and (4) the
extent to which aeolian transport rates and downwind
aeolian sand cover could potentially increase when new
bars are formed in optimal locations relative to the areas
where archaeological sites occur.
To study these issues, Draut (2003; Draut and
Rubin, in press) established wind-transport monitoring
instruments at six locations throughout the corridor of
the Colorado River (fig. 8). These monitoring stations
measure the amount of sand being transported by vary-
ing wind speeds at different times of the year and under
different sediment supply conditions. Preliminary results
from the first year of data (Draut and Rubin, in press)
indicated that wind speeds and predominant directions
vary widely throughout the river corridor and that trans-
Figure 8. U.S. Geological Survey scientists installing a weather
station near RM 65. These devices measure wind velocity and
direction at six locations along the Colorado River (photograph by
Amy Draut, U.S. Geological Survey).
186 The State of the Colorado River Ecosystem in Grand Canyon
port conditions are also highly variable, limiting efforts to
model sediment-transport rates for the system as a whole.
To determine the extent to which aeolian sedi-
ment formed the material in which archaeological sites
are embedded, Draut worked collaboratively with NPS
archaeologists, U.S. Geological Survey scientists, and
other researchers to evaluate subsurface deposits at a
nonrandom sample of archaeologically rich locations
throughout the river corridor (Draut and others, 2005).
These investigations focused on describing the vari-
ous geomorphic processes that have contributed to the
formation and preservation of archaeological sites by
closely examining the sedimentary structures preserved
in subsurface contexts (fig. 9). Preliminary findings indi-
cated that aeolian deposits were common throughout
the prehistoric landscape of the river corridor and were
important factors in the formation of many sites but that
wind deposition was clearly not the only matrix-form-
ing process at work in the past. Aeolian sediment blan-
kets the surface of many archaeological sites, but often
these windborne deposits cover substrates that are both
fluvially (from the river) and colluvially (from the slope)
derived (Draut and others, 2005).
In terms of understanding how dam operations
could be altered to enhance the sediment supply available
for redeposition by the wind, Draut is tracking the fate of
several sandbars that formed during the November 2004
experimental high flow near previously established sedi-
ment-transport monitoring stations (Draut and Rubin, in
press). During 2005–06, Draut will evaluate the extent
to which aeolian processes may contribute to the erosion
of the newly formed sandbars and track the amount of
sediment transported by the wind from the sandbars to
nearby archaeological features. This information will
allow researchers to assess the relative importance of
fluvial and aeolian processes in maintaining the sedimen-
tary matrices of archaeological sites located above the
level of dam-controlled flows in the river corridor.
Ethnobotanical Resources and
Other Tribally Valued Resources
The Hopi and Hualapai Tribes and Southern
Paiute Consortium have monitored a variety of culturally
important resources in the Colorado River corridor since
the mid-1990s. The monitored resources of concern
include culturally valued plants and plant gathering loca-
tions, traditionally valued mineral resources, landscape
features, traditional use areas, and archaeological sites.
Since 1995, the Southern Paiute Consortium has
monitored culturally important resources in the corridor
of the Colorado River to assess their condition relative to
Glen Canyon Dam operations and visitor use, to edu-
cate and train tribal members in resource monitoring,
and to educate tribal members and the general public
about the traditional importance of Grand Canyon to
the Southern Paiute people (Stoffle and others, 1995;
Drye and others, 2001; Bullets and others, 2003, 2004).
Approximately 20 individual locations are monitored by
the Southern Paiute Consortium on a 6-yr cycle. Some
locations are visited every year, while others are visited
only once or twice over the 6-yr period. Locations of
importance include traditional plant areas, rock art,
specific perennial tributaries, and archaeological sites
with evidence of use by Southern Paiute ancestors. The
general assessment from Southern Paiute Consortium
monitoring during the past 5 yr is that most of the moni-
tored resources appear to be in relatively good condition,
although concerns about visitor trails at archaeological
sites, visitor behavior around certain traditionally signifi-
cant locations in Grand Canyon, and drought stress on
plants have been noted. To date, no specific recommen-
dations regarding Glen Canyon Dam operations have
been forthcoming from these monitoring efforts.
The Hualapai Tribe monitored traditional cultural
resources in conjunction with the 1996 beach/habitat-
building flow (Jackson and others, 1997) and again in
2001, 2002, 2003, and 2004 (Jackson and others, 2004
a, b). Beginning in 2001, the baseline conditions of 15
previously documented traditional cultural locations
in Grand Canyon and of 5 previously undocumented
Figure 9. A U.S. Geological Survey scientist examines
stratigraphy exposed in the wall of an arroyo near RM 209
(photograph by Amy Draut, U.S. Geological Survey).
Cultural Resources in the Colorado River Corridor 187
locations were evaluated by Hualapai tribal members
and consultants using a numerical condition index rat-
ing system. Several additional sites were added to the
monitoring program in 2002–04, resulting in a total
of 28 traditional cultural locations receiving 1 or more
years of monitoring. Natural and cultural impacts were
ranked on a 5-point scale, from 0 (absent) to 4 (severe).
Impact ratings averaged over all 20 sites indicated that
human impacts, both visitor- and dam-related, were
more problematic than were natural impacts. Negative
dam-related impacts included water stress on vegetation
because of the lack of periodic inundation of higher
elevation plant communities, continuing nonnative plant
encroachment, and the loss of beach area from dam-
controlled flows. Effects from the diminishing surface
elevation of Lake Mead, accompanied by vegetation
encroachment of nonnative plants, primarily tamarisk
(Tamarix ramosissima), were also noted. Throughout the
corridor of the river, human impacts from trailing,
artifact movement, and onsite camping were observed
to be a problem, with the latter impacts rated as heavy
to severe in several cases. On Lake Mead, the wakes of
motor boats are also thought to contribute significantly
to beach erosion ( Jackson and others, 2004b). Based only
on the 2001 monitoring results, human impacts were the
most significant impacts observed at most sites, with 13
of the 20 locations rated as having heavy (3) to severe (4)
human impacts. In contrast, only 5 of these 20 loca-
tions monitored in 2001 were assigned similar natural
impact ratings, and most of the impacts were from
rodent burrowing and side canyon flash floods. Overall
in 2001, the average rating for all natural impacts was
1.8, whereas the average human impact rating was 2.6.
In future years, repeated analyses of the same sites will
allow Hualapai tribal members to determine whether or
not conditions are improving or deteriorating relative to
2001 baseline conditions, but at this time (2005), com-
parative data are insufficient for conducting this analysis.
The Hopi Tribe initiated an ethnobotanical proj-
ect in 1998 to evaluate traditional plant resources in
the corridor of the Colorado River (fig. 10). This study,
completed in 2001, identified over 128 plant species in
Grand Canyon that were traditionally used for ceremo-
nies, medicines, food, and other necessities of daily life
(Lomaomvaya and others, 2001). Beginning in 2002,
the Hopi Tribe initiated a multiyear project to evalu-
ate whether terrestrial ecosystem data currently being
collected by cooperating university scientists (Kearsley
and others, 2002) could be useful to the Hopi Tribe for
assessing resource conditions from a Hopi perspective.
The results of this initial study indicated that the terres-
trial ecosystem data could be usefully interpreted from a
Hopi perspective, provided that the data were translated
from scientific categories and terminology into Native
American categories and terminology (Huisinga and
Yeatts, 2003). In 2003, the Hopi Tribe initiated a pilot
study to begin assessing the terrestrial ecosystem data
from a native Hopi perspective. Results of this pilot
study are anticipated to be available in spring 2005.
Summary
Resource monitoring of archaeological and tra-
ditional cultural resources suggests that archaeological
resources continue to be impacted both by physical
processes such as surface erosion and gullying and by
recreational visitors. Although surface erosion and visitor
impacts would undoubtedly be occurring without the
presence or continuing operation of Glen Canyon Dam,
the manner in which the dam is currently operated
prohibits the retention of sediment within the river cor-
ridor. The diminishing supply of sediment appears to be
contributing to and exacerbating the rate and amount
of erosion occurring at all levels within the ecosystem.
In addition to impacts from Glen Canyon Dam opera-
tions, visitor impacts such as trailing, trampling, and col-
lection of artifacts are contributing to the degradation
of many archaeological sites in the river corridor and
of several locations of traditional importance to Native
American people.
Figure 10. Hopi elders discuss and document the uses of
culturally valued plants in the Colorado River corridor. The
photograph was taken at RM 43 below the 100,000-cfs water-
surface elevation (photograph courtesy of Michael Yeatts and the
Hopi Tribe).
188 The State of the Colorado River Ecosystem in Grand Canyon
Tribal assessments of dam-related impacts to ethno-
botanical resources in the river corridor offer somewhat
mixed results. For the most part, the Southern Paiute
Consortium has identified satisfactory conditions for tra-
ditional plant resources, although there is some evidence
of plant deterioration, probably from ongoing drought
conditions. Meanwhile, the Hualapai Tribe has expressed
concern over the condition of certain key botanical
resources (e.g., the willow tree (Salix sp.) at Granite Park),
and they have noted an apparent increase in nonna-
tive plants such as Bermudagrass (Cynodon dactylon) and
camel thorn (Alhagi maurorum) at specific locations in the
river corridor.
Under the current NPS monitoring program,
frequency of monitoring is tied to perceived levels of
erosion or visitor use, with those sites showing more
evidence of active erosion or more frequent visitor use
being monitored more frequently than those showing less
impact. As one would expect from a monitoring program
that is weighted towards tracking impacts at the more
threatened and heavily visited sites, annual monitoring
results show relatively high levels of physical and visitor-
related impacts.
Because the current data set is lacking measure-
ments related to rates or degrees of erosion, it is not
possible to determine whether or to what degree rates
of erosion at archaeological sites may have changed in
recent years under the MLFF regime. The NPS monitor-
ing program is undergoing a reevaluation and redesign in
2005 to better meet the needs of the Glen Canyon Dam
Adaptive Management Program for information related
specifically to effects of Glen Canyon Dam operations
on National Register-eligible historic properties. Once
revised monitoring protocols are implemented in 2006, it
will be possible to track rates and trends in gully forma-
tion and downcutting relative to different flow regimes
and to make systemwide assessments of resource condi-
tion over time.
The limited monitoring, in conjunction with the
1996 beach/habitat-building flow and 2004 experimen-
tal high flow, indicates that the creation of sandbars
above the level of normal dam operations may have
beneficial effects on archaeological sites in two respects:
(1) by creating sources of sediment for subsequent wind-
borne redeposition at archaeological sites located upwind
of the newly formed sandbars and (2) by temporarily
raising the effective base level to which terrace channels
are downcutting in the short term, thereby temporarily
slowing the rates of downcutting and headward migra-
tion of erosional gullies. For these measures to be effec-
tive over the long term, however, periodic high flows
under sediment-enriched conditions would need to be
repeated at relatively frequent intervals.
References
Ahlstrom, R.V.N., Purcell, D.E., Zyniecki, M., Gilpin,
D.A., and Newton, V.L., 1993, An archaeological over-
view of Grand Canyon National Park. Archaeological
Report No. 93-92: Flagstaff, Ariz., SWCA.
Balsom, J.R., and Larralde, S., eds., 1996, Mitigation
and monitoring of cultural resources in response to the
experimental habitat building flow in Glen and Grand
Canyons, spring 1996: Salt Lake City, Utah, Bureau of
Reclamation, Upper Colorado Region.
Bullets, I., Snow, T., Posvar, E., Rogers, K., Piekielek, J.,
Rogers, M., Snow, M., and Phillips, A., III, 2004, 2004
Southern Paiute Consortium Colorado River Corridor
Resource Evaluation Program annual report of activi-
ties: Prepared for the Bureau of Reclamation, Upper
Colorado Region, Salt Lake City, Utah.
Bullets, I., Snow, T., Posvar, E., Snow, R., Bow, J., Dean,
E., and Phillips, A., III, 2003, 2003 Southern Paiute
Consortium Colorado River Corridor Resource
Evaluation Program annual report of activities:
Prepared for the Bureau of Reclamation, Upper
Colorado Region, Salt Lake City, Utah.
Davis, S.W., Davis, M.E., Lucchitta, I., Finkel, R., and
Caffee, M., 2000, Early agriculture in the eastern
Grand Canyon of Arizona, USA: Geoarchaeology, v.
15, no. 8, p. 783–798.
Dierker, J.L., Leap, L.M., and Andrews, N., 2001, Fiscal
year 2001 archaeological site monitoring and manage-
ment activities along the Colorado River in Grand
Canyon National Park. RCMP Report No. 70: Submit-
ted to Bureau of Reclamation, Salt Lake City, Utah.
Dierker, J.L., Leap, L.M., and Andrews, N., 2002, Fiscal
year 2002 archaeological site monitoring and manage-
ment activities along the Colorado River in Grand
Canyon National Park. RCMP Report No. 87: Submit-
ted to Bureau of Reclamation, Salt Lake City, Utah.
Draut, A., 2003, The role of eolian sediment transport
in the preservation of archeological features: a new
research initiative in Grand Canyon National Park:
Grand Canyon Monitoring and Research Center
October 2003 Symposium, Tucson, Ariz.
Cultural Resources in the Colorado River Corridor 189
Draut, A.E., and Rubin, D.M., in press, Measurements
of wind, aeolian sand transport, and precipitation
in the Colorado River corridor, Grand Canyon,
Arizona—November 2003 to December 2004: U.S.
Geological Survey Open-File Report.
Draut, A.E., Rubin, D.M., Dierker, J.L., Fairley, H.C.,
Griffiths, R.E., Hazel, J.E., Jr., Hunter, R.E., Kohl, K.,
Leap, L.M., Nials, F.L., Topping, D.J., and Yeatts, M.,
2005, Sedimentology and stratigraphy of the Palisades,
Lower Comanche, and Arroyo Grande areas of the
Colorado River corridor, Grand Canyon, Arizona:
U.S. Geological Survey Scientific Investigations Report
2005-5072, http://pubs.usgs.gov/sir/2005/5072, 68 p.
Drye, B., Bulletts, I., Phillips, A., III, Snow, T.,
Stanfield, G., 2001, 2000 Southern Paiute
Consortium Colorado River Corridor Resource Evalu-
ation Program: annual report of activities: Submitted
to Bureau of Reclamation.
Euler, R.C., 1967a, The canyon dwellers: American
West, v. 4, no. 2, p. 22–27, 67–71.
Euler, R.C., 1967b, Helicopter archaeology: American
West Review, v. 1, p. 24.
Euler, R.C., and Taylor, W.W., 1966, Additional archaeo-
logical data from upper Grand Canyon: Nankoweap to
Unkar revisited: Plateau, v. 39, p. 26–45.
Fairley, H.C., 2003, Changing river: time, culture, and
the transformation of landscape in Grand Canyon:
a regional research design for the study of cultural
resources along the Colorado River in lower Glen
Canyon and Grand Canyon, National Park, Arizona:
Tucson, Ariz., Statistical Research, Inc., SRI Press,
Technical Series 79.
Fairley, H.C., Bungart, P.W., Coder, C.M., Huffman, J.,
Samples, T.L., and Balsom, J.R., 1994, The Grand
Canyon river corridor survey project: archaeological
survey along the Colorado River between Glen Canyon
Dam and Separation Rapid: Flagstaff, Ariz., Bureau
of Reclamation Glen Canyon Environmental Studies
Program, cooperative agreement no. 9AA-40-07920.
Ferguson, T.J., 1998, Ongtupka niqw Pisisvayu (Salt
Canyon and the Colorado River): the Hopi People
and the Grand Canyon. Final ethnohistoric report
for the Hopi Glen Canyon Environmental Studies
Project: Submitted by Hopi Cultural Preservation
Office, Kykotsmovi, Ariz., to Bureau of Reclama-
tion Glen Canyon Environmental Studies Program,
Flagstaff, Ariz.
Ferguson, T.J., Yeatts, M., Kuwanwisiwma, L.J., and
Dongoske, K., 1997, The Hopi People, the opera-
tion of Glen Canyon Dam, and management of
cultural resources in the Grand Canyon: Paper pre-
sented at the Ninth Conference on Research and
Resource Management in Parks and on Public Lands,
The George Wright Society Biennial Conference,
Albuquerque, N. Mex.
Glen Canyon Dam Adaptive Management Program,
2001, AMP Strategic Plan.
Hart, R.E., 1995, Zuni Glen Canyon Environmen-
tal Studies ethnohistorical report. Institute of the
North American West, Seattle & Washington:
Unpublished report submitted by Zuni Heritage and
Historic Preservation Office to the Glen Canyon
Environmental Studies Office, Bureau of Reclamation,
Flagstaff, Ariz.
Hazel, J., Kaplinski, M., Parnell, R., Manone, M., Dale,
A., 1999, Topographic and bathymetric changes at
thirty-three long-term study sites, in Webb, R.H.,
Schmidt, J.C., Marzolf, G.R., and Valdez, R.A., eds.,
The Controlled Flood in Grand Canyon: Washington,
D.C., American Geophysical Union, Geophysical
Monograph Series, v. 110, p. 161–184.
Hazel, J.E., Jr., Kaplinski, M., Manone, M., and Parnell,
R., 2000, Monitoring arroyo erosion of pre-dam river
terraces in the Colorado River ecosystem, 1996–1999,
Grand Canyon National Park, AZ: Report on file at
the Grand Canyon Monitoring and Research Center,
Flagstaff, Ariz., 29 p.
Hereford, R., Fairley, H.C., Thompson, K.S., and
Balsom, J.R., 1993, Surficial geology, geomorphol-
ogy and erosion of archeologic sites along the
Colorado River, eastern Grand Canyon, Grand
Canyon National Park, Arizona. Grand Canyon
National Park in cooperation with the Bureau of
Reclamation, Glen Canyon Environmental Studies,
Flagstaff, Ariz.: U.S. Geological Survey Open-File
Report 93-517.
Hereford, R., Thompson, K.S., Burke, K.J., and Fairley,
H.C., 1996, Tributary debris fans and Late Holocene
alluvial chronology of the Colorado River, eastern
Grand Canyon, AZ: GSA Bulletin, v. 108, no. 1, p.
3–19.
190 The State of the Colorado River Ecosystem in Grand Canyon
Huisinga, K., and Yeatts, M., 2003, Soosoy Himu
Naanamiwiwyungwa: an analysis of the Grand
Canyon Monitoring and Research Center’s Terrestrial
Monitoring Program and the development of a Hopi
long-term plan. Final report for cooperative agreement
no. 01WRSA0358: Submitted by the Hopi Cultural
Preservation Office to U.S. Geological Survey, Grand
Canyon Monitoring and Research Center, Flagstaff,
Ariz.
Jackson, L., Kennedy, D.J., and Phillips, A.M., III,
2004a, Evaluating Hualapai cultural resources along
the Colorado River, 2002: Final report submitted to
Bureau of Reclamation, Upper Colorado Region, Salt
Lake City, Utah, and Grand Canyon Monitoring and
Research Center, Flagstaff, Ariz.
Jackson, L., Kennedy, D.J., and Phillips, A.M., III,
2004b, Evaluating Hualapai cultural resources along
the Colorado River, FY 2003 and FY 2004: Final
report submitted to Bureau of Reclamation, Upper
Colorado Region, Salt Lake City, Utah, and Grand
Canyon Monitoring and Research Center, Flagstaff,
Ariz.
Jackson, L., Osife, C., and Phillips, A.M., III, 1997,
Effects of Colorado River test flow experiment on
Hualapai and Southern Paiute traditional ethnobo-
tanical resources: Report submitted to Grand Canyon
Monitoring and Research Center.
Jones, A.T., 1986, A cross-section of Grand Canyon
archeology: excavations at five sites along the
Colorado River. Publications in Anthropology No.
28. Western Archeological and Conservation Center:
Tucson, Ariz., National Park Service.
Kearsley, M.J.C., Cobb, N., Yard, H., Lightfoot, D.,
Brantley, S., Carpenter, G., and Frey, J., 2002, Inven-
tory and monitoring of terrestrial riparian resources
in the Colorado River corridor of Grand Canyon: an
integrative approach. 2002 annual report. Cooperative
agreement no. 01WRAG0044 and 01WRAG0034:
Submitted to U.S. Geological Survey Grand Canyon
Monitoring and Research Center.
Leap, L.M., and Coder, C., 1995, Erosion control project
at Palisades Delta along the Colorado River corri-
dor, Grand Canyon National Park. RCMP Report
No. 29: Submitted to Bureau of Reclamation, Upper
Colorado Region, Salt Lake City, Utah.
Leap, L.M., Dierker, J.L., and Andrews, N., 2003, Fiscal
year 2003 archaeological site monitoring and manage-
ment activities along the Colorado River in Grand
Canyon National Park. RCMP Report No. 89: Sub-
mitted to Bureau of Reclamation, Upper Colorado
Region, Salt Lake City, Utah.
Leap, L.M., and Kunde, J.L., 2000, Fiscal year 2000
summary report: archaeological site monitoring and
management activities along the Colorado River in
Grand Canyon National Park. RCMP Report No.
68: Submitted to Bureau of Reclamation, Upper
Colorado Region, Salt Lake City, Utah.
Leap, L.M., Kunde, J.L., Hubbard, D.C., Andrews,
N., Downum, C.E., Miller, A., and Balsom, J.R.,
2000, Grand Canyon monitoring project 1992–1999:
synthesis and annual monitoring report FY99. Grand
Canyon National Park River Corridor Monitoring
Report No. 66: Submitted to Bureau of Reclamation,
Upper Colorado Region, Salt Lake City, Utah.
Lomaomvaya, M., Ferguson, T.J., Yeatts, M., 2001,
Ongtuvqava Sakwtala: Hopi ethnobotany in the
Grand Canyon: Submitted to the Grand Canyon
Monitoring and Research Center.
Lucchitta, I., 1991, Quaternary geology, geomorphol-
ogy, and erosional processes, eastern Grand Canyon,
Arizona. U.S. Geological Survey administrative
report: Submitted to Bureau of Reclamation, Upper
Colorado Region, Salt Lake City, Utah.
O’Connor, J.E., Ely, L.L., Wohl, E.E., Stevens, L.E.,
Melis, T.S., Kale, V.S., and Baker, V.R., 1994, A 4500-
year long record of large floods on the Colorado River
in the Grand Canyon, Arizona: Journal of Geology, v.
102, p. 1–9.
Pederson, J.L., Petersen, P.A., MacFarlane, W.W.,
Gonzales, M.F., and Kohl, K., 2003, Mitigation, moni-
toring, and geomorphology related to gully erosion of
cultural sites in Grand Canyon. Final report in fulfill-
ment of CA-01-WRAG-0074: On file, Grand Canyon
Monitoring and Research Center, Flagstaff, Ariz.
Powell, J.W., 1875, Explorations of the Colorado River
of the West and its tributaries, explored in 1869, 1870,
1871 and 1872 under the direction of the Secretary of
the Smithsonian Institution: Washington, D.C., U.S.
Government Printing Office.
Cultural Resources in the Colorado River Corridor 191
Roberts, A., Begay, R.M., and Kelley, K.B., 1995, Bits’iis
Nineeze (The River of Neverending Life), Navajo his-
tory and cultural resources of the Grand Canyon and
the Colorado River: Submitted by the Navajo Nation
Historic Preservation Department to the Glen Canyon
Environmental Studies Office, Bureau of Reclamation,
Flagstaff, Ariz.
Rubin, D.M., Topping, D.J., Schmidt, J.C., Hazel, J.,
Kaplinski, M., and Melis, T.S., 2002, Recent sedi-
ment studies refute Glen Canyon Dam EIS hypothesis:
Washington, D.C., American Geophysical Union,
EOS, v. 83, no. 25, p. 273, 277–278.
Schmidt, J.C., Topping, D.J., Grams, P.E., and Hazel,
J.E., 2004, System-wide changes in the distribution of
fine sediment in the Colorado River corridor between
Glen Canyon Dam and Bright Angel Creek, Arizona:
Final report submitted to the Grand Canyon Moni-
toring and Research Center, U.S. Geological Survey,
Flagstaff, Ariz.
Schwartz, D.W., 1965, Nankoweap to Unkar: an archae-
ological survey of the upper Grand Canyon: American
Antiquity, v. 30, p. 278–296.
Schwartz, D.W., Chapman, R.C., and Kepp, J., 1980,
Archaeology of Grand Canyon: Unkar Delta: Santa
Fe, N. Mex., School of American Research.
Schwartz, D.W., Kepp, J., and Chapman, R.C., 1981,
Archaeology of Grand Canyon: Walhalla Plateau:
Santa Fe, N. Mex., School of American Research.
Schwartz, D.W., Marshall, M.P., and Kepp, J., 1979,
Archaeology of Grand Canyon: the Bright Angel site:
Santa Fe, N. Mex., School of American Research.
Stevens, R.H., and Mercer, J.A., 1998, Hualapai Tribe’s
traditional cultural properties in relation to the
Colorado River, Grand Canyon, Arizona: final report:
Prepared for Bureau of Reclamation, Upper Colorado
Regional Office, Salt Lake City, Utah. Copy on file,
Hualapai Tribe’s Office of Cultural Resources, Peach
Springs, Ariz.
Stoffle, R.W., Austin, D.E., Fulfrost, B.K., Phillips,
A.M., III, and Drye, T.F., 1995, Itus, Auv, Te’ek, (Past,
Present, Future): managing Southern Paiute resources
in the Colorado River corridor: Report prepared for
Glen Canyon Environmental Studies Office, Bureau
of Reclamation, Flagstaff, Ariz.
Stoffle, R.W., Halmo, D.B., and Austin, D.E., 1997,
Cultural landscapes and traditional cultural properties:
a Southern Paiute view of the Grand Canyon and the
Colorado River: American Indian Quarterly, v. 21, p.
219–249.
Stoffle, R.W., Halmo, D.B., Evans, M.J., and
Austin, D.E., 1994, Piapaxa ‘Uipi (Big River Canyon):
Southern Paiute ethnographic resource inventory and
assessment for Colorado River corridor, Glen Canyon
National Recreation Area, Utah and Arizona, and
Grand Canyon National Park, Arizona: Submitted
to the Glen Canyon Environmental Studies Office,
Bureau of Reclamation, Flagstaff, Ariz.
Taylor, W.W., 1958, Two archaeological studies in
northern Arizona: the Pueblo ecology study: hail and
farewell and a brief survey through the Grand Canyon
of the Colorado River. Bulletin 30: Flagstaff, Ariz.,
Museum of Northern Arizona.
Thompson, K.S., Potochnik, A.R., Ryel, R., O’Brien,
G., Neal, L.A., 2000, Development of a geomorphic
model to predict erosion of pre-dam Colorado River
terraces containing archaeological resources: SWCA,
Environmental Consultants, Inc., submitted to Grand
Canyon Monitoring and Research Center, Flagstaff,
Ariz.
Topping, D.J., Rubin, D.M., Nelson, J.M., Kinzel, P.J.,
III, and Corson, I.C., 2000a, Colorado River sedi-
ment transport, pt. 2: systematic bed-elevation and
grain-size effects of supply limitation: Water Resources
Research, v. 36, p. 543–570.
Topping, D.J., Rubin, D.M., and Vierra, L.E., Jr., 2000b,
Colorado River sediment transport, pt. 1: natural
sediment supply limitations and the influence of Glen
Canyon Dam: Water Resources Research, v. 36, p.
515–542.
Topping, D.J., Schmidt, J.C., and Vierra, L.E., Jr., 2003,
Computation and analysis of the instantaneous-dis-
charge record for the Colorado River at Lees Ferry,
Arizona—May 8, 1921, through September 30, 2000:
U.S. Geological Survey Professional Paper 1677, 118 p.
U.S. Department of the Interior, 1995, Opera-
tion of Glen Canyon Dam Final Environmental
Impact Statement: Salt Lake City, Utah, Bureau of
Reclamation, Upper Colorado Region.
192 The State of the Colorado River Ecosystem in Grand Canyon
U.S. Department of the Interior, 1996, Record of
Decision, Operation of Glen Canyon Dam Final Envi-
ronmental Impact Statement.
Webb, R.H., 1985, Late Holocene flooding on the
Escalante River, south-central Utah: University of
Arizona, Department of Geosciences, unpublished
Ph.D. dissertation.
Wiele, S.M., and Torizzo, M., 2005, Modeling of sand
deposition in archeologically significant reaches of the
Colorado River in Grand Canyon, USA, in Bates, P.D.,
Lane, S.N., and Ferguson, R.I., eds., Computational
fluid dynamics: applications in environmental hydrau-
lics: Hoboken, N.J., Wiley, p. 357–394.
Yeatts, M., 1998, High elevation sand retention following
the 1996 spike: Report on file at the Grand Canyon
Monitoring and Research Center, Flagstaff, Ariz., 15 p.
Contact Information:
Helen C. Fairley
Social Scientist
U.S. Department of the Interior
U.S. Geological Survey
Southwest Biological Science Center
Flagstaff, AZ
hfairley@usgs.gov
First page photograph credit: Amy Draut, U.S. Geological Survey
C
o
u
r
t
e
s
y
o
f
M
i
c
h
a
e
l
Y
e
a
t
t
s
a
n
d
t
h
e
H
o
p
i
T
r
i
b
e
Chapter 12
Recreational
Values and
Campsites in
the Colorado
River
Ecosystem
Matt Kaplinski
Jeff Behan
Joseph E. Hazel, Jr.
Roderic A. Parnell
Helen C. Fairley
Introduction
Grand Canyon National Park is one of the best-
known wildland preserves in the world. Its designation
as a national park in 1919 sought to protect it for the
benefit of human visitors as well as to safeguard the
physical, biological, and cultural resources contained
within its borders. Interest in recreation on the Colorado
River has risen dramatically since the mid-1960s, and a
226-mi (364 km) journey through Grand Canyon by boat
is now regarded as one of the world’s premier wild-river
experiences. Recreational use of the Colorado River
corridor through Grand Canyon is closely regulated by
the National Park Service (NPS), and demand for the
corridor, particularly for river trips, greatly exceeds avail-
ability.
Beginning with the initial explorations of John
Wesley Powell in 1869, river runners and hikers have
used sandbars along the Colorado River below present-
day Glen Canyon Dam as campsites. These camps, and
their associated activities, make up an important element
of the modern-day recreational experience within Glen
and Grand Canyons. Because of their crucial role, the
relative size, distribution, and quality of campsites along
the Colorado River are of particular concern to river
managers (Bureau of Reclamation, 1995; Stewart and
others, 2000; Glen Canyon Dam Adaptive Management
Program, 2001; National Park Service, 2004).
This chapter presents an assessment of the cur-
rent state of knowledge concerning the impacts of Glen
Canyon Dam operations on the changing condition of
campsite areas and sandbars and the implications of
physical changes of the Grand Canyon ecosystem for
visitor capacity and quality of experience. After defin-
ing the study area and some key concepts, the chapter
briefly reviews the relationships between the condition
and extent of Colorado River sandbars and the qual-
ity of the visitor recreation experience. An overview of
historical status and trends of the number and size of
campsites along the Colorado River is followed by a
summary of recent findings. Discussion focuses on the
effects of the modified low fluctuating flow (MLFF) alter-
native and high-volume experimental flows on campsite
area. The chapter concludes with an evaluation of these
results relative to the stated recreation goals and man-
agement objectives of the Glen Canyon Dam Adaptive
Management Program (GCDAMP).
194 The State of the Colorado River Ecosystem in Grand Canyon
Background
The Colorado River flows approximately 293 RM
from Glen Canyon Dam to the Grand Wash Cliffs, the
physical feature that marks the western boundary of
Grand Canyon National Park. The focus in this chap-
ter is on the portion of the river from Lees Ferry to
Diamond Creek (RM 0–226), even though recreational
use of the river corridor extends another 50 mi (80 km)
downstream to Lake Mead. Lees Ferry is the launch-
ing point for river trips through Grand Canyon, and
Diamond Creek, on the Hualapai Indian Reservation, is
the typical takeout point.
Geomorphic Characteristics
of Campsites
Debris fans are sloping deposits of boulders, gravel,
and sand that form at the mouth of a tributary as the
result of flash flood events that constrict the main chan-
nel and increase the local bed elevation of the river
(Schmidt and Graf, 1990). At most constrictions, recircu-
lation zones or eddies (currents of water moving against
the main current in a circular pattern) are formed in the
river, and because of lower flow velocities, sand is depos-
ited within eddies. Flow patterns within an eddy define
the configuration of sand that is deposited (Schmidt and
Graf, 1990). Typically, eddies contain a primary recircu-
lating zone and often have secondary zones of separated
flow where the current rotates in the opposite direction
of the primary zone or is virtually stagnant. Sand depos-
its are classified based upon where they are deposited
in relation to the primary and secondary recirculating
patterns. Two types of deposits are the highest in eleva-
tion and are most typically associated with campsites:
separation deposits and reattachment deposits. Separa-
tion deposits mantle the downstream part of the debris
fan near the point where the main current separates to
form the eddy. Reattachment deposits are located at the
downstream end of the primary recirculating zone where
the main-channel current reattaches to the bank.
Channel-margin deposits are not associated with
tributary debris fans and occur along the channel banks.
These deposits form within small eddies associated with
bank irregularities caused by talus and rock outcropping.
A small number of these channel margin deposits are
used as campsites. Within some reaches of the corridor,
flat-lying units of rock crop out along the river, and the
ledges are also used as campsites.
Visitor Capacity and
Wilderness Experience
In the context of recreation, carrying capacity is
now referred to as “visitor capacity,” defined recently
as “. . . a prescribed number and type of people that
an area will accommodate, given the desired natural/
cultural resource conditions, visitor experiences, and
management program” (Haas, 2001). Ecological aspects
of visitor capacity usually relate to cumulative ecological
impacts, but in places such as Colorado River campsites,
actual physical space available, impacted or not, is also
a critical factor. These categories have been referred
to as “ecological” capacity (e.g., plant, animal, and soil
impacts) and “physical” capacity (e.g., people per unit
area of flat sleeping area; camping parties per beach)
(Shelby and Heberlein, 1986). This distinction is use-
ful for understanding visitor capacity for the Colorado
River, where both physical space available and resource
impacts are important management considerations.
In addition to resource protection, primary objec-
tives for recreation management include minimizing
impacts on, enhancing, and preserving the quality of rec-
reation experiences. Experience quality is complex and
affected by an array of factors, some of which are social
rather than ecological or physical, so visitor capacity also
has a “social” component. Social variables that affect
experience quality include the number of people visible
at one time in a given area and the number of encoun-
ters of one group with other parties or with groups of
a particular type or size (Shelby and Heberlein, 1986).
Resource protection will always be a primary concern
in highly sensitive areas, such as desert riparian zones,
where even moderate human activity may cause sig-
nificant ecological impacts, and in settings where heavy
use produces damage. Thus, in ecologically sensitive
areas, visitor capacity is limited by the need to protect
resources. For a broad range of less sensitive areas,
however, social variables that affect recreation experience
quality may limit visitor capacity at levels below those at
which unacceptable resource impacts occur.
On the Colorado River in Grand Canyon, legal
guidelines call for not only resource protection and a
quality recreational experience but also an undeveloped,
uncrowded, wilderness-type experience. It is clear that
on the Colorado River, outstanding opportunities for
wilderness experiences are indeed a key factor in the
river’s popularity and something that the general public
has come to expect when visiting Grand Canyon. This
popularity was confirmed during the initial phase of
river recreation research in the 1970s (Shelby, 1976) and
Recreational Values and Campsites in the Colorado River Ecosystem 195
in subsequent studies (Bishop and others, 1987; Hall and
Shelby, 2000).
The National Park Service has explicitly expressed
an intention to manage for wilderness-type experi-
ences for Colorado River visitors within Grand Canyon
National Park (National Park Service, 1995, p. 11). In
addition to the opportunity to experience natural eco-
logical conditions, one of the most important attributes
of a wilderness experience is solitude (Hendee and oth-
ers, 1990). Because the operation of Glen Canyon Dam
has eroded sandbars used for camping and has reduced
the sand available for maintaining them (see chapter 1,
this report), the area available for camping in the river
corridor has declined significantly since construction of
Glen Canyon Dam in 1963 (Kearsley and others, 1994;
Kaplinski and others, 2005). The decrease in campsite
area can affect solitude by increasing the level of crowd-
ing along the corridor. Crowding reduces the ability of
separate river trips to camp out of sight and hearing of
one another and also reduces the ability of individuals or
small groups within a particular trip to camp out of sight
and hearing of one another.
Because tracking changes in recreation resources
and experience quality means identifying quantifiable
parameters, campsite area has emerged as the preferred
parameter for measuring these attributes for recreational
monitoring programs in Grand Canyon. Campsite area
is not the only factor in Colorado River recreation that
affects experience quality, but it is a readily measurable
factor that has arguably changed more than any other
facet of the river experience in the past 40 yr.
Status and Trends
Studies of campsite area have been conducted by
Weeden and others (1975), Brian and Thomas (1984),
Kearsley and Warren (1993), Kearsley and others (1994),
Kearsley (1995), and Kearsley and Quartaroli (1997).
These studies evolved from qualitative estimates of
campsite carrying capacity to quantitative aerial photo-
graphic measurements. Weeden and others (1975) and
Brian and Thomas (1984) focused on developing an
inventory of the size and number of campsites through-
out the river corridor. Both of these studies estimated
the capacity of each site with dam releases above the
24,000–28,000 cubic feet per second (cfs) stage elevation,
with capacity defined as the number of campers that
could occupy a campsite for an overnight stay. Research-
ers have focused on high-elevation campsites because
summer demand for energy produces medium to high
releases. As a result, lower sandbar elevations are inun-
dated during the height of the commercial rafting season
(mid-May through mid-September) and therefore are
not available for camping during the time of year when
campsites are in highest demand. Kearsley and Warren
(1993) repeated the inventory and improved the campsite
area measurements by developing techniques to quanti-
tatively measure camp area from aerial photography and
videography. Subsequent studies by Kearsley and others
(1994), Kearsley (1995), and Kearsley and Quartaroli
(1997) improved upon the aerial photographic mapping
by using geographic information system software.
Kearsley and Warren (1993) studied camps between
Lees Ferry and Diamond Creek and divided them into
critical and noncritical reaches. A critical reach was
defined as any contiguous stretch of the river in which
the number of available campsites is limited because of
geologic characteristics, high demand, or other logistical
factors. Noncritical reaches were defined as any stretch
of the river in which campsites are plentiful and little
competition for the majority of sites occurs. These reach
definitions closely parallel the geomorphic reach defini-
tions of Schmidt and Graf (1990).
Kearsley and Warren (1993) found that campsites
had decreased dramatically in both number and size
since Weeden’s team completed its initial survey in 1973.
Reaches designated as critical because of limited avail-
ability of suitable campsites by Kearsley and Warren
(1993) (Marble Canyon, RM 11–40.8; upper Granite
Gorge, RM 76.5–116; and Muav Gorge, RM 139–164)
are nearly the same as the critical sections identified by
Weeden and others (1975). Campable area decreased
primarily because of erosion in critical reaches; in non-
critical reaches, decrease in campsite area was primarily
the result of vegetation encroachment (Kearsley and
Warren, 1993). An overall trend of increased campsite
size and number between 1973 and 1983 was attributed
to the high releases in 1983 needed to keep Lake Powell
from spilling over Glen Canyon Dam. The 1983 high
releases forced sand from the river channel onto sand-
bars, but the change was temporary; sandbars signifi-
cantly decreased in size and number less than 1 yr later.
Moreover, campsites in the upper Marble Gorge and
upper Granite Gorge decreased between 1973 and 1983
and between 1983 and 1991. The inventory documented
226 campsites above 25,000 cfs, which represented a
32% decrease in the number of campsites between 1973
and 1991. The inventory also found a 51% decrease in
large camps, resulting in a 44% decrease in campsite
area between 1973 and 1991. Campsite area decreased
an average of 9% between 1991 and 1994, with dispro-
portionately larger decreases at camps in critical reaches
196 The State of the Colorado River Ecosystem in Grand Canyon
(Kearsley, 1995). River-induced changes accounted for
80% of lost campsite area above the Little Colorado
River and 32% of loss below the Little Colorado River
confluence.
Kearsley and others (1994) concluded that loss of
Colorado River campsites was an ongoing process that
was initiated with the installation of Glen Canyon Dam
more than 30 yr ago and that the rate of decline had
slowed over time. The overall pattern of change was one
of initial systemwide decrease in sites (1965–73), variable
change during years of regulated high flows because of
high levels of precipitation (1983–86), and a systemwide
decrease in campsites between 1984 and the mid-1990s.
They noted that not all sandbars in Grand Canyon
respond in the same manner to high flows, fluctuating
flows, or vegetation encroachment and that campsite
availability in critical reaches had decreased the most.
To monitor changes in campsite availability resulting
from the 1996 beach/habitat-building flow, 53 camping
sandbars were randomly selected from the 218 remain-
ing from the 1991 inventory by Kearsley and Warren
(1993). The sites were physically measured 2 weeks
before, 2 weeks after, and 6 mo after the 7-d, 45,000-
cfs experimental flow (Kearsley and Quartaroli, 1997).
Float-by assessments were made of 200 sandbars, includ-
ing the 53 that were also measured. Results showed a
systemwide increase in campsite area. Half (100/200) of
the sites assessed were at least 10% larger, 39% (77/200)
were the same, and 12% (23/200) were smaller than
before the experimental flow. For 53 sites directly mea-
sured, 62% (33/53) increased in size, 17% (9/53) were
the same, and 21% (11/53) decreased in area. Float-by
assessments were less sensitive to measuring change but
not biased toward increase or decrease. At many sites,
sand was deposited directly on top of existing campable
areas and did not increase campsite area. At some sites,
new sand was deposited as a mound over previously
usable space, and the increase in slope angle resulted in
decreased camping area.
Eighty-two new sites were created, in the sense that
these sites were not usable just before the 1996 beach/
habitat-building flow. Although 33 of these “new” sites
were included in previous campsite inventories, all had
degraded to being unusable by the time of the experi-
ment. Many new sites consisted of deposition on low-ele-
vation sandbars with little sun or wind protection. These
sandbars were theoretically usable but not highly valued
as camps and were subject to rapid erosion. Forty of the
new sites were between RM 40 and RM 65. More than
twice as many sites were created in noncritical reaches
than in critical reaches. Six months after the beach/
habitat-building flow, only 55% (45/82) of the new sites
were still considered usable. The high flow obliterated
three previously inventoried campsites. Overall, the 1996
beach/habitat-building flow increased the number, size,
capacity, and aesthetic qualities of campsites. These ben-
efits were substantial, but degradation occurred quickly:
within 6 mo, nearly half of the new campsites were
unusable, remaining new sites were half their initial size,
and most of the increased area on measured established
sites had eroded. Relatively high-flow releases immedi-
ately following the 1996 beach/habitat-building flow in
late 1996 and 1997 probably exacerbated erosion.
Recent Monitoring
Following the Record of Decision in 1996 (U.S.
Department of the Interior, 1996) and the establish-
ment of the Glen Canyon Dam Adaptive Management
Program, a new campsite monitoring program was
initiated by the U.S. Geological Survey’s Grand Canyon
Monitoring and Research Center. The following discus-
sion focuses on the results of 6 yr of campsite area moni-
toring beginning in 1998 and ending in 2003 (Kaplinski
and others, 2005).
Monitoring Objectives
and Methods
The 1998–2003 campsite area monitoring program
focused on describing changes in the size of camp-
ing areas in the Colorado River corridor. Monitoring
included annually measuring campsite area at a series of
long-term monitoring sites and evaluating the changes in
campsite area among years and as the result of different
dam releases.
Annual surveys were conducted every October from
1998 through 2003 by crews from Northern Arizona
University’s Department of Geology to quantify camp-
site area change. Surveys at the selected study sites
were conducted by using standard total station survey
techniques (U.S. Army Corps of Engineers, 1994). The
surveyors adopted the criteria of Kearsley (1995) and
Kearsley and Quartaroli (1997) to identify campable
area. Campable area was defined as a smooth substrate,
preferably sand, with no more than 8º of slope and
with little or no vegetation. Not all campable areas were
mapped at every site. Instead, representative camp spots
were selected across a range of stage elevations. Camping
areas not represented in the mapping were typically far
(>328 ft (>100 m)) from the main mooring/cooking areas.
Recreational Values and Campsites in the Colorado River Ecosystem 197
Study Sites
The study sites are located throughout the Colorado
River corridor between Lees Ferry and Diamond Creek
(fig. 1) and were selected to coincide with a subset of
the long-term study sites used by the Northern Arizona
University sandbar monitoring project, which monitors
change in sandbar area and volume (Beus and others,
1992; Kaplinski and others, 1995, 1998; Hazel and
others, 1999, 2001, 2002). These sites were originally
selected on the basis of (1) distribution throughout the
geomorphic reaches, (2) size sufficient to guarantee
persistence through the period of study, (3) geomor-
phic diversity within and between sites, (4) availability
of historical data, and (5) variation in recreational use
intensity and vegetation cover (Beus and others, 1992).
Given these criteria, only a subset of the sandbar moni-
toring sites could be used to monitor campsite areas.
These sites, although not chosen randomly, have proven
to be representative of systemwide changes in terms of
changes in sand volume and area at campsites located
above high normal flows (above 20,000 cfs) (Schmidt
and others, 2004). Therefore, it is reasonable to assume
that changes to campsite areas at these sites are also rep-
resentative of changes to campsite area systemwide.
The study began with 31 study sites. In 2002, 6 sites
were added, for a total of 37 sites. Only the original 31
sites, which have been measured since 1998, were used
to summarize the campsite areas, while all sites were
used to calculate average percent change between years.
Sixteen of these sites are located in Marble Canyon
between the Paria River and the Little Colorado River
confluence, and 21 are located in Grand Canyon below
the Little Colorado confluence. There are 18 sites within
critical reaches as defined by Kearsley and Warren
(1993), and 19 are in noncritical reaches.
Figure 1. The Colorado River corridor below Glen Canyon Dam and locations of 31 study sites from the 1998–2003 campsite area
monitoring program. The shaded area represents Grand Canyon National Park. Study site locations are noted by river mileage.
198 The State of the Colorado River Ecosystem in Grand Canyon
Dam Releases Before and
During the Study Period
Dam releases during the 1998–2003 study period
included normal modified low fluctuating flow (MLFF)
operations, plus a low summer steady flow (LSSF) experi-
ment during 2000, the habitat maintenance flows (HMF)
in May 2000 and September 2000, and fluctuating non-
native fish suppression flows from January to March in
2002 and 2003 (fig. 2). Normal MLFF dam releases fluc-
tuate diurnally and seasonally, based on power demand
and water-delivery schedules. Typically, flow releases are
higher in winter and summer months and lower during
spring and fall months. In 1998 and 1999, daily mean
flow releases ranged from an average of approximately
19,400 cfs in high-volume months to approximately
12,400 cfs in low-volume months. The LSSF experiment
in 2000 consisted of two high-flow releases in spring
and fall and a period of low steady (no diurnal fluctua-
tion) flow during summer. The low steady flow during
summer was lowered to a constant 8,000 cfs. The high
flows were short-duration (4 d) dam releases of 31,000
cfs. These were the only two flows large enough to reach
above the 25,000-cfs stage elevation, or the upper limit
for nonexperimental MLFF operation releases, during
the study period.
River flow levels during the 1998 and 1999 survey
trips fluctuated from 10,000 to 18,000 cfs. Therefore,
surveyors were only able to measure camp areas consis-
tently at every site above the 15,000-cfs stage elevation.
Subsequent analysis of campsite area below 25,000-cfs
stage elevation excluded the measurements made dur-
ing 1998 and 1999. During the 2000 to 2003 surveys,
low-volume releases allowed measurement of camp area
above the 10,000-cfs stage elevations at some sites and
above 15,000-cfs stage elevation at all sites. Fluctuating
nonnative fish suppression flows were conducted from
January through March in 2002 and 2003. During these
experiments the flows fluctuated from 5,000 to 20,000
cfs. Comparison of camp area change between surveys
was conducted by using area measured only above the
25,000-cfs stage elevation.
Findings
Recent analysis of the 1998–2003 monitoring results
by Kaplinski and others (2005) demonstrated that the
total camp area above the 25,000-cfs stage elevation
significantly decreased during the study period (fig. 3).
Total campsite area changes were derived by summing
all of the campsite area measurements in a particular
reach. Between 1998 and 2003, the total campsite area
decreased by 55%. The average decrease was 15%
between each survey (fig. 3).
Longitudinal changes were examined by compar-
ing the total campsite area above and below the Little
Colorado River (LCR) confluence (fig. 4). In the follow-
ing discussion, the term Marble Canyon refers to sites
above the LCR, while the term Grand Canyon refers to
sites below the LCR. Campsite areas in Marble Canyon
and Grand Canyon decreased at a similar rate and
showed an overall loss of 57% and 53%, respectively.
There was a longitudinal difference in the response to
the powerplant capacity flows conducted as part of the
2000 LSSF experiment. Camp area in Grand Canyon
increased slightly (4%) following the high flows of the
LSSF experiment, while campsites in Marble Canyon
decreased by 24%. Area increases in Grand Canyon
camps are possibly related to greater deposition down-
stream of the LCR where the sediment supply is presum-
ably greater.
The pattern of campsite area change was different
in critical and noncritical reaches (fig. 5). Total campsite
area within critical reaches decreased by 37% during
the study period for an average decrease of 8% per year.
In noncritical reaches the change was greater, with a
total decrease of 63% and an average decrease of 18%
per year. Campsite area increased slightly in the critical
reaches (7%) following the LSSF experiment, whereas
sites in noncritical reaches decreased by 18%. Critical
reaches are generally narrower than noncritical reaches,
and the campsites tend to be smaller and less vegetated
Figure 2. Daily mean discharge hydrograph from the USGS
gaging station on the Colorado River near Lees Ferry during the
period of study. Note the daily and seasonal fluctuations in flow
volume during 1998 and 1999 and during the low summer steady
flow experiment in 2000 that included two high-flow events.
Recreational Values and Campsites in the Colorado River Ecosystem 199
25,000-cfs stage elevation, as well as above the 25,000-cfs
stage elevation. In the most recent monitoring study, sur-
veyors measured all campsite areas exposed at the time
of the visit, allowing campsite area changes to be divided
between discrete ranges of stage elevation (fig. 6).
High-elevation campsite area (above 25,000 cfs) has
progressively decreased during the study period, with the
exception of a short-lived increase within the 25,000-
cfs to 30,000-cfs range following the LSSF experiment.
Repeat surveys after 2000 showed that this slight increase
in campsite area decreased to levels equivalent to those
measured in 1998.
Camp area at lower elevations has increased because
of the deposition from high-flow events associated with
the LSSF experiment in 2000, the fluctuating nonnative
fish suppression flows from January to March 2003, and
medium- to high-volume (10,000 to 25,000 cfs) summer
dam operations. In fact, the amount of campsite area
available at lower elevations is now greater than that
available at higher elevations (fig. 7). Since the lower
elevation areas are within the zone of flow fluctuation,
these increases may not persist because lower elevation
sandbars are more susceptible to bank erosion than sand
at higher elevations (Hazel and others, 1999).
Campsite area and sandbar volume both decreased
during the study period; however, campsite area
decreased at a greater rate than did sandbar volume (fig.
8), which indicates that other factors contributed to the
Figure 4. Total camp area above the 25,000-cfs stage elevation in
Marble and Grand Canyons.
Figure 3. Total camp area above the 25,000-cfs stage elevation.
Figure 5. Total camp area above the 25,000-cfs stage elevation in
critical and noncritical reaches.
because the steep bedrock channels provide little space
for sediment deposition.
Campsite area exists across the entire range of
normal Glen Canyon Dam releases (5,000 to 25,000
cfs), and the amount of camp area available is greatly
dependent on flow levels. Some GCDAMP manage-
ment objectives are specifically concerned with measur-
ing sandbar area and volume between the 5,000-cfs and
Figure 6. Distribution of total campsite area above the 25,000-cfs
stage elevation in three different stage ranges.
200 The State of the Colorado River Ecosystem in Grand Canyon
loss of high-elevation campsite area. These factors pre-
sumably include vegetation growth, surface water runoff,
aeolian processes (wind-caused sediment movement), and
human impact.
Although not quantitatively addressed in this study,
visual observations and photographic documentation
compiled by the Grand Canyon River Guides, Inc.,
Adopt-a-Beach Program during the same period (1998–
2003) indicate that, excluding sandbar erosion, vegeta-
tion growth contributes most significantly to the loss of
high-elevation campsite area (Thompson and others,
1997; O’Brien and others, 1999, 2000; Thompson, 2001,
2002). Unfortunately, a direct comparison of campsite
area change and vegetation colonization during the
1998–2003 study period was not possible because of the
incompatibility of vegetation monitoring protocols (M.
Kearsley, Northern Arizona University, oral commun.,
2004). Kaplinski and others (2005) recommended that,
in the future, vegetation coverage changes at camp-
sites should be tracked by using remotely sensed aerial
imagery to provide a quantified assessment of the role of
vegetation in reducing total campsite area.
Surface runoff events that significantly decreased
campsite area were observed at only three sites during
the 6 yr of monitoring. Human impacts were generally
minimal, except at locations where vegetation prun-
ing and removal had increased or maintained campsite
areas. Aeolian reworking of sandbars did not appear to
be a significant factor in affecting campsite area because
of the amount of vegetation established along higher
elevation sandbar areas.
Discussion and Conclusions
Campsites within the Colorado River ecosystem
exist primarily on sandbars. The size and capacity of
camping area are directly related to the areal extent of
sandbars and the amount of vegetation colonizing the
sandbars (Kearsley and others, 1994). Previous studies
by Kearsley and Warren (1993) and Kearsley and others
(1994) established that substantial losses in open areas
used for camping had occurred because of sandbar ero-
sion and colonization by vegetation. Although both ero-
sion and vegetation reduce campsite area, the processes
and their effects are not identical.
Erosion of sandbars is caused primarily by opera-
tions of Glen Canyon Dam. The magnitude of daily
fluctuations, the ramping rates, and the increased ability
of clearwater releases to transport sediment have all been
identified as contributing factors (Beus and others, 1992;
Rubin and others, 2002).
Vegetation encroachment is leading to higher rates
of campsite area decrease than can be attributed to
erosion alone. Encroachment by nonnative species such
as tamarisk (Tamarix ramosissima) and camel thorn (Alhagi
maurorum), as well as by native species such as arrowweed
(Pluchea sericea) and coyote willow (Salix exigua), has led to
colonization on previously open sections of sandbars,
thus further decreasing campsite area. In some larger
and less frequently visited sites, dense patches of vegeta-
tion now make the sites essentially unusable for camping
activities. Although this process has substantially reduced
available space at many campsites, the effects of these
changes on visitor capacity are somewhat less clear than
when area is lost to erosion. For example, in some fre-
quently used camps, individual sleeping sites are cleared
Figure 7. Total campsite area above and below the 25,000-cfs
stage elevation.
Figure 8. Percent change between surveys of sandbar volume
and campsite area in Marble and Grand Canyons.
Recreational Values and Campsites in the Colorado River Ecosystem 201
of and kept free of vegetation by constant use. These
individual sites are also often separated by vegetation
“screens” that may actually serve to reduce the distance
that recreationists feel they need to be separated from
one another in order to achieve privacy. Additionally,
clumps of mature tamarisk trees along steep riverbanks
may serve to reduce sandbar erosion by anchoring sedi-
ment in place.
Rainfall-induced flash flooding also reduces camp-
site area. Flash flood impacts were transient before
completion of the dam because the effects of tributary
erosion were erased every year or two as flood flows from
high spring runoff deposited sediment and reworked
sandbars. Today, the effects of tributary flash floods are
cumulative and tend to be long lasting because periodic
high-flow events that are capable of transferring tribu-
tary-derived sediment to higher elevations are infrequent.
The monitoring results of Kaplinski and others
(2005) showed that between 1998 and 2003 more than
half of the available campsite area at the study sites was
lost. Camping area above the 25,000-cfs stage eleva-
tion decreased by 55% during this 6-yr period, and the
average rate of change was 15% per year. The decrease
in high elevation campsite area occurred both in Marble
Canyon and in Grand Canyon (above and below the
LCR) as well as within critical and noncritical reaches.
Notably, lower elevation campsite areas increased after
2000, and the total campsite area below the 25,000-cfs
stage elevation now exceeds the area available at higher
elevations. The rate of decrease in high-elevation camp-
site area greatly exceeds the decrease in sandbar volume.
This difference indicates that other factors—probably
vegetation encroachment—have contributed to the
recent loss of high-elevation campsite area. Unfortu-
nately, Kaplinski and others (2005) could not undertake
a quantitative comparison of campsite area change and
vegetation colonization from 1998 to 2003 because the
vegetation study is designed to detect systemwide, rather
than site-specific, changes in vegetation cover (Mike
Kearsley, Northern Arizona University, oral commun.,
2004).
In order to construct a longer term view of changes
to campsites in Grand Canyon, Kaplinski and others
(2005) compiled the percent change between surveys
from the campsite inventories conducted by Brian
and Thomas (1984), Kearsley and others (1994), and
Kearsley and Quartaroli (1997) and combined them
with the results from the 1998–2003 campsite monitor-
ing program. Between 1973 and 2003, the only observ-
able periods of increases in either the number of camps
or the size of camps occurred after the high flows of
1983–84, which were needed to keep Lake Powell from
overtopping Glen Canyon Dam, and the 1996 beach/
habitat-building flow, when flows were greater than pow-
erplant capacity. During years between flood events, both
before and after the implementation of the 1996 Record
of Decision (U.S. Department of the Interior, 1996),
sandbars declined in area, volume, and total number.
The campsite monitoring results showed that cur-
rent operations of Glen Canyon Dam are not meeting
the goals of the GCDAMP with respect to the recre-
ational resources of the Colorado River corridor. Spe-
cifically, the GCDAMP seeks to “maintain or improve
the quality of recreational experiences for users of the
Colorado River Ecosystem, within the framework of
the GCDAMP ecosystem goals” (Glen Canyon Dam
Adaptive Management Program, 2001). With this goal in
mind, the Adaptive Management Work Group devel-
oped the following management objectives to maintain
or improve recreational resources:
Maintain or improve the quality and range of rec-
reational opportunities in Glen and Grand Canyons
within the capacity of the Colorado River ecosystem
to absorb visitor impacts in ways consistent with NPS
and tribal river corridor management plans (objec-
tive 9.1) (Glen Canyon Dam Adaptive Management
Program, 2001).
Increase the size, quality, and distribution of camp-
ing beaches in critical and noncritical reaches in the
mainstem within the capacity of the Colorado River
ecosystem to absorb visitor impacts in ways consis-
tent with NPS and tribal river corridor management
plans (objective 9.3) (Glen Canyon Dam Adaptive
Management Program, 2001).
Maintain or enhance the wilderness experience in
the Colorado River ecosystem in consideration of
existing management plans (objective 9.4) (Glen
Canyon Dam Adaptive Management Program,
2001).
The significant decrease in campsite area during
the study period indicates that the second management
objective above (objective 9.3) is not being met. While
this study does not explicitly link changes in camp-
site area to the recreational/wilderness experience in
Grand Canyon, the significant decrease in campsite area
indicates that other management objectives are possibly
not being met. For example, a significant decrease in
campsite area may indicate a decrease in the range and
quantity of recreational opportunities. Also, because
existing campsites are smaller and thus more crowded,
the quality of campsites is not being maintained or
improved. The decrease in campsite area leads to more
1.
2.
3.
202 The State of the Colorado River Ecosystem in Grand Canyon
crowding and less choice for camps, which can negatively
affect the wilderness experience (Hendee and others,
1990). On the other hand, conclusions about the inher-
ently sociological aspects of the river experience based
solely on changes to campsite area at a limited number
of sites are tenuous. Future research and monitoring
should be expanded to include sociological aspects of
the recreational experience. In addition, future research
should investigate the linkages between the sociologi-
cal aspects of the recreational experience and physical
parameters, such as campsite area, in order to evaluate
whether GCDAMP goals are being achieved.
The continued existence of sandbars suitable for
camping in the Colorado River ecosystem will depend
on periodic high flows to redeposit sediment lost through
incremental erosion, scour, and vegetation encroach-
ment; therefore, the continuing availability of campsite
area is necessarily linked with the frequency and mag-
nitude of flood events from Glen Canyon Dam. Unless
vegetation is physically removed, and provided that
enough sediment is available for deposition, high-flow
events are the only mechanism by which sandbars used
as campsites above the 25,000-cfs stage elevation can be
built and maintained.
In order to properly address the management
objectives of the GCDAMP, the recreational monitoring
program should be expanded to include monitoring and
research of both physical and psychological parameters
of the river experience. Currently, the program consists
only of campsite area measurements at a limited number
of sites. This limited amount of information makes a
complete assessment of the stated goals and objectives
of the program currently impossible. One of the largest
gaps in the current knowledge base is a complete, sys-
temwide inventory of campsites in the Colorado River
ecosystem in Grand Canyon. Since the last inventory
in 1991, significant changes have occurred, and a new
measurement of the number and size of camps is essen-
tial to evaluate the current state of the resource. O’Brien
and Roberts (1999) and Roberts and Bieri (2001) used
a modified version of the 1991 campsite inventory to
develop a numerical river trip simulator model to predict
visitor-use dynamics on the Colorado River. This model
has been subsequently used by the NPS to assist in devel-
oping a management plan of the river corridor (National
Park Service, 2004). Unfortunately, an up-to-date mea-
surement of campsite carrying capacity was not available
for the planning efforts. O’Brien and Roberts (1999) and
Roberts and Bieri (2001) recommended that a method
be developed to convert campsite area measurements
to carrying capacity of a site; they suggested addressing
this key information need by developing a standardized
method of estimating carrying capacity for a beach so
that onsite estimates are performed consistently. They
also recommended that an interdisciplinary team com-
prising a statistician, a sociologist, a geologist, and sur-
veyors should be included in order to develop an empiri-
cally verifiable and repeatable method of measuring and
interpreting the campsite area, location, and abundance
in relation to other variables such as trip length, attrac-
tion sites, number of people, and social aspects of visitor
use. Campsite area measurements clearly indicate that
campsite area has declined. How does this measured
decline relate to carrying capacity of the river corridor?
Does the decline in campsite area relate directly to a
decline in the quality of the recreational and wilderness
experience? Are these parameters linked and, if so, how?
Future research aimed at addressing both the physical
parameters and the psychological aspects of recreation
along the Colorado River would help answer these types
of questions and more clearly assess whether or not the
goals and management objectives of the GCDAMP are
being achieved.
© 2005 Geoff Gourley; used with permission
Recreational Values and Campsites in the Colorado River Ecosystem 203
References
Beus, S.B., Avery, C.C., Stevens, L., Cluer, B.,
Kaplinski, M., Anderson, P., Bennett, J., Brod, C.,
Hazel, J., Gonzales, M., Mayes, H., Protiva, F., and
Courson, J., 1992, The influence of variable discharge
regimes on Colorado River sand bars below Glen
Canyon Dam, chap. 6, in The influence of vari-
able discharge regimes on Colorado River sand bars
below Glen Canyon Dam: Final report, submitted
to Glen Canyon Environmental Studies, Bureau of
Reclamation, 61 p.
Bishop, R.C., Boyle, K.J., Welsh, M.P., Baumgartner,
R.M., and Rathbun, P.R., 1987, Glen Canyon Dam
releases and downstream recreation: an analysis of user
preferences and economic values, executive summary:
Report to Glen Canyon Environmental Studies, Bureau
of Reclamation.
Brian, N.J., and Thomas, J.R., 1984, 1983 Colorado
River sand bar campsite inventory: National Park
Service, Division of Resources Management, Grand
Canyon National Park, Ariz.
Bureau of Reclamation, 1995, Operation of Glen
Canyon Dam Final Environmental Impact Statement.
Glen Canyon Dam Adaptive Management Program,
2001, Draft final strategic plan, August 17, 2001.
Haas, G.E., 2001, Visitor capacity in the national park
system: Social Science Research Review, Winter 2001,
U.S. Department of the Interior, National Park Service.
Hall, T.E., and Shelby, B., 2000, 1998 Colorado River
boater study: Grand Canyon National Park, report for
the National Park Service.
Hazel, J.E., Kaplinski, M., Parnell, R., and Manone,
M., 2002, Colorado River sand bar conditions in
2001: results from 12 years of monitoring, April 2002:
Fact sheet, Grand Canyon Monitoring and Research
Center, http://www.gcmrc.gov/library/reports/
physical/Fine_Sed/02_naufactsheet.pdf, 4 p.
Hazel, J.E., Jr., Kaplinski, M., Parnell, R., and Manone,
M., 2001, Monitoring the effects of the spring 2000
habitat maintenance flow on Colorado River ecosys-
tem sand bars: Fact sheet, Grand Canyon Monitoring
and Research Center, http://www.gcmrc.gov/library/
reports/physical/Fine_Sed/parnell_2001_fact3.pdf, 2 p.
Hazel, J.E., Jr., Kaplinski, M., Parnell, R., Manone, M.,
and Dale, A., 1999, Topographic and bathymetric
changes at thirty-three long-term study sites, in Webb,
R.H., Schmidt, J.C., Marzolf, G.R., and Valdez,
R.A., eds., The controlled flood in Grand Canyon:
Washington, D.C., American Geophysical Union, Geo-
physical Monograph Series, v. 110, p. 161–183.
Hendee, J.C., Stankey, G.H., and Lucas, R.C., 1990,
Wilderness management: Golden, Colo., Fulcrum
Publishing.
Kaplinski, M., Hazel, J.E., and Parnell, R., 2005, Camp-
site area monitoring in the Colorado River ecosystem:
1998 to 2003: Final report submitted to Grand Canyon
Monitoring and Research Center, Flagstaff, Ariz., 39 p.
Kaplinski, M., Hazel, J.E., Jr., and Beus, S.S., 1995,
Monitoring the effects of interim flows from Glen
Canyon Dam on sand bars in the Colorado River
corridor, Grand Canyon National Park, Arizona:
Final report to Glen Canyon Environmental Studies,
Northern Arizona University, Flagstaff, 62 p.
Kaplinski, M., Hazel, J.E., Jr., and Beus, S.S., 1998,
Monitoring sand bar campsites along the Colorado
River in Grand Canyon: Report to the Grand Canyon
Monitoring and Research Center, August 1998.
Kearsley, L.H., 1995, Monitoring the effects of Glen
Canyon Dam interim flows on campsite size along the
Colorado River in Grand Canyon National Park (final
report): National Park Service, Division of Resources
Management, Grand Canyon National Park.
Kearsley, L.H., and Quartaroli, R., 1997, Effects of a
sand bar/habitat building flow on campsites in Grand
Canyon: Final report of Applied Technology Associates
for the Glen Canyon Environmental Studies, 18 p.
Kearsley, L.H., Schmidt, J.C., and Warren, K.D., 1994,
Effects of Glen Canyon Dam on Colorado River sand
deposits used as campsites in Grand Canyon National
Park, USA: Regulated Rivers, v. 9, p. 137–149.
Kearsley, L.H., and Warren, K., 1993, River campsites in
Grand Canyon National Park: inventories and effects
of discharge on campsite size and availability (final
report): National Park Service, Division of Resources
Management, Grand Canyon National Park.
204 The State of the Colorado River Ecosystem in Grand Canyon
National Park Service, 1995, Grand Canyon National
Park general management plan: Denver, Colo.,
National Park Service, Denver Service Center (sum-
mary document), 67 p.
National Park Service, 2004, Draft Environmental
Impact Statement, Colorado River Management Plan,
draft dated October 2004.
O’Brien, G., Burke, K., and Hamilton, L., 2000, Effects
of natural flow and controlled- flow releases from
Glen Canyon Dam on Grand Canyon beaches, 1999:
a continuation of a repeat photography study by the
Grand Canyon River Guides, Inc. (Adopt-a-Beach
Program): Administrative report submitted to the
Grand Canyon Monitoring and Research Center
by the Grand Canyon River Guides Adopt-a-Beach
Program.
O’Brien, G., Janssen, J., Thompson, K., Burke, K., and
Potochnik, A., 1999, Effects of continuous high flows
and daily fluctuating flows from Glen Canyon Dam on
Grand Canyon beaches, 1997 and 1998: a continua-
tion of repeat photography study by Grand Canyon
River Guides, Inc. (Adopt-a-Beach Program): Admin-
istrative report submitted by the Grand Canyon River
Guides Adopt-a-Beach Program to the Grand Canyon
Monitoring and Research Center, Flagstaff, Ariz.
O’Brien, G., and Roberts, C., 1999, Evaluation of river
beach carrying capacity information utilized by the
Grand Canyon River Trip Simulator: analysis and rec-
ommendations for future study: Final report submitted
to the National Park Service, Grand Canyon Science
Center (CA8210-99-002), May 1, 1999, 17 p.
Roberts, C.A., and Bieri, J.A., 2001, Impacts of low
flows on recreational rafting traffic on the Colorado
River in Grand Canyon National Park: Final report
submitted to the Grand Canyon Monitoring and
Research Center, Flagstaff, Ariz., 18 p.
Rubin, D.M., Topping, D., Schmidt, J.C., Hazel, J.E.,
Kaplinski, M., and Melis, T., 2002, Recent sediment
studies refute Glen Canyon Dam hypothesis: EOS,
Transactions of the American Geophysical Union, v.
83, no. 25, p. 273, 277–278.
Schmidt, J.C., and Graf, J.B., 1990, Aggradation and
degradation of alluvial sand deposits, 1965–1986,
Colorado River, Grand Canyon National Park,
Arizona: U.S. Geological Survey Professional Paper
1493.
Schmidt, J.C., Topping, D.J., Grams, P.E., and Hazel,
J.E., 2004, System-wide changes in the distribution
of fine sediment in the Colorado River between Glen
Canyon Dam and Bright Angel Creek, Arizona: Final
report submitted to Grand Canyon Monitoring and
Research Center, Flagstaff, Ariz., 107 p.
Shelby, B., 1976, Use levels and crowding in the Grand
Canyon: National Park Service, Colorado River
Research Program, River Contact Study, final report,
pt. 3.
Shelby, B., and Heberlein, T.A., 1986, Capacity in
recreation settings: Corvallis, Oregon State University
Press.
Stewart, W., Larkin, K., Orland, B., Anderson, D.,
Manning, R., Cole, D., Taylor, J., and Tomar, N.,
2000, Preferences of recreation user groups of the
Colorado River in Grand Canyon: Final report to
the Grand Canyon Monitoring and Research Center,
cooperative agreement no. 98-FG-40-0190.
Thompson, K., 2001, Effects of low steady sum-
mer flows on Grand Canyon camping sand bars,
2000: Report to the Grand Canyon Monitoring and
Research Center, Flagstaff, Ariz.
Thompson, K., 2002, Long term monitoring of camping
beaches in Grand Canyon: a summary of results from
1996–2001: Report to Grand Canyon Monitoring and
Research Center, Flagstaff, Ariz., 15 p.
Thompson, K., Burke, K., and Potochnik, A., 1997,
Effects of the sand bar-habitat building flow and
subsequent interim flows from Glen Canyon Dam on
Grand Canyon camping sand bars, 1996: a repeat
photography study by Grand Canyon River Guides
adopt-a-sand bar program: Report to the Grand
Canyon Monitoring and Research Center.
U.S. Army Corps of Engineers, 1994, Engineering and
design-topographic accuracy standards: EM1110-1-
1005, p. 2-1 to 2-12.
U.S. Department of the Interior, 1996, Record of
Decision, Operation of Glen Canyon Dam Final
Environmental Impact Statement.
Weeden, H., Borden, F., Turner, B., Thompson, D.,
Strauss, C., and Johnson, R., 1975, Grand Canyon
National Park campsite inventory, contract no. CX
001-3-0061 with the National Park Service: University
Park, Pennsylvania State University.
Recreational Values and Campsites in the Colorado River Ecosystem 205
Contact Information:
Matt Kaplinski
Research Associate
Northern Arizona University
Department of Geology
Flagstaff, AZ
Matt.Kaplinski@nau.edu
Jeff Behan
Faculty Research Assistant
Oregon State University
Institute for Natural Resources
Corvallis, OR
jeff.behan@oregonstate.edu
Joseph E. Hazel, Jr.
Research Associate
Northern Arizona University
Department of Geology
Flagstaff, AZ
Joseph.Hazel@nau.edu
Roderic A. Parnell
Professor, Director of the Center for Environmental
Sciences and Education
Northern Arizona University
Department of Geology
Flagstaff, AZ
Roderic.Parnell@nau.edu
Helen C. Fairley
Social Scientist
U.S. Department of the Interior
U.S. Geological Survey
Southwest Biological Science Center
Flagstaff, AZ
hfairley@usgs.gov
J
e
f
f
S
o
r
e
n
s
e
n
,
A
r
i
z
o
n
a
G
a
m
e
a
n
d
F
i
s
h
D
e
p
a
r
t
m
e
n
t
First page photograph credit: © 2005 Geoff Gourley; used with permission
Chapter 13
Lessons from
10 Years of
Adaptive
Management
in Grand
Canyon
Jeffrey E. Lovich
Theodore S. Melis
Introduction
Almost a decade ago, the Secretary of the Interior
issued a Record of Decision (ROD) (U.S. Department
of the Interior, 1996) regarding the operation of Glen
Canyon Dam and its impacts on natural, cultural, and
recreational resources of the Colorado River below Glen
Canyon Dam. This decision was largely in response to
mandates of the Grand Canyon Protection Act of 1992.
Adaptive management (see Overview, this report, for
discussion of adaptive management), sometimes known
as “learning by doing,” was deemed to be the method
of choice after a multiyear environmental impact state-
ment process that included extensive public involvement.
Practitioners of adaptive management intentionally see
management policies as experimental because of the
scientific uncertainties inherent in such large, complex
ecosystems (Walters and Holling, 1990; Clark, 2002).
The intent in selecting this style of management is
reflected in the Operation of Glen Canyon Dam Final
Environmental Impact Statement (hereafter EIS) with
the following language:
It is intended that the ROD will initiate a process
of adaptive management, whereby the effects of
dam operations on downstream resources would
be assessed and the results of those resource
assessments would form the basis for future mod-
ifications of dam operations. Many uncertainties
still exist regarding the downstream impacts of
water releases from Glen Canyon Dam. The
concept of adaptive management is based on
the need for operational flexibility to respond
to future monitoring and research findings and
varying resource conditions (U.S. Department of
the Interior, 1995, p. 34).
The very language in the EIS acknowledges that
adaptive management is a process of experimentation,
monitoring, and evaluation.
The end of a decade of research and monitoring
provides an important opportunity to evaluate the effects
of Glen Canyon Dam operations on resources of con-
cern and to determine if the desired outcomes are being
achieved and whether they are compatible with one
another or not. In this concluding chapter we present
a summary of adaptive management of the Colorado
River ecosystem below Glen Canyon Dam by review-
ing predictions contained in the EIS. In table II-7 of the
EIS (summarized here in table 1), resource categories
208 The State of the Colorado River Ecosystem in Grand Canyon
are listed and associated with predictions for how those
resources would respond under the preferred alterna-
tive of modified low fluctuating flows (MLFF) (see table
2 for generalized operating rules). During preparation
of the EIS, the best scientific data available were used
to generate those prognostications; however, a decade
later we have significant new information for evaluating
the operation of Glen Canyon Dam in relation to the
objectives of the 1995 EIS and the 1992 Grand Canyon
Protection Act.
This report is the first systematic attempt to con-
duct an assessment of the changing state of resources in
the Colorado River ecosystem in Grand Canyon over a
decadal timeframe. Our objectives are to (1) encapsulate
what we have learned about the Colorado River ecosys-
tem over a decade of scientific inquiry, summarizing the
most salient conclusions of this report, and (2) discuss, in
very general terms, research and monitoring challenges
and questions facing the program.
While this report is not meant to be prescriptive
with respect to future action for the Glen Canyon Dam
Adaptive Management Program, the authors of the
preceding chapters presented evidence that dam opera-
tions during the last 10 yr under the preferred alternative
of the MLFF have not restored fine-sediment resources
or native fish populations in Grand Canyon, both of
which are resources of significant importance to the
program. Some resources of concern, however, have
improved under the MLFF, as shown below and also in
the preceding chapters.
What Have We Learned
from 10 Years of Adaptive
Management?
Adaptive management is an integrated, multidisci-
plinary approach for confronting uncertainty in natural
resources issues. It is adaptive because it acknowledges
that managed resources will always change as a result of
human intervention, that surprises are inevitable, and
that new uncertainties will emerge. Active, experimental
learning is the way in which the uncertainty is minimized
(Walters and Holling, 1990). Adaptive management
acknowledges that policies must satisfy social objectives
but also must be continually modified and flexible for
adaptation to surprises. Adaptive management, there-
fore, views policy as hypothesis. That is, most policies are
really questions, and management actions become treat-
ments in an experimental sense. Our summary of what is
known about the influence of the MLFF on downstream
resources examines many facets of the Colorado River
ecosystem, especially those resources deemed as most
important during the EIS process. Substantial impor-
tance is also ascribed to those resources affected by the
Endangered Species Act of 1973.
Fish Response
The Grand Canyon population of the federally
endangered humpback chub (Gila cypha) has declined
during the past decade under MLFF operations. Only
eight native fish species were historically found in Grand
Canyon. Six of these were desert species endemic (not
found elsewhere) to the Colorado River ecosystem, mak-
ing this one of the most unusual fish communities in the
world (Mueller and Marsh, 2002). Of the original eight,
only four remain in Grand Canyon, namely the hump-
back chub, the bluehead sucker (Catostomus discobolus),
the flannelmouth sucker (Catostomus latipinnis), and the
speckled dace (Rhinichthys osculus). Of these four, only the
humpback chub is endangered, and its numbers have
dropped dramatically in the last decade. At the same
time, nonnative fish have increased in both diversity and
abundance. The reasons for the decline of native fish
are commonly cited to include dramatic changes in the
thermal, sediment, and hydrologic regimes of the river
because of the construction and operation of numerous
dams in the basin, the introduction of nonnative preda-
tory and competitive fishes, and the introduction of
diseases and parasites (Mueller and Marsh, 2002). The
actual mechanisms of decline and extirpation are poorly
known, in part because of a lack of early data on popu-
lation numbers.
Our knowledge about the cause and effect between
dam operations and chub decline is incomplete; we
do know, however, that the current MLFF operation
has not resulted in increased survival and recruitment
of humpback chub, despite the prediction of the EIS
(table 1). Although there is no basis for claiming that
the current operation at Glen Canyon Dam resulted in
recent and repeated low recruitment and the continued
decline of the humpback chub, it is clear that the restric-
tions on dam operations since 1991 have not produced
the hoped-for restoration and maintenance of this
endangered species (see chapter 2, this report). During
the MLFF, basin hydrology has varied from drought to
wet conditions and then back to drought conditions.
Through these conditions, the decline of the humpback
chub has continued. This trend leads to questions about
whether daily, monthly, or even annual patterns of dam
Lessons from 10 Years of Adaptive Management in Grand Canyon 209
Table 1. Natural and cultural resources of the Grand Canyon ecosystem and predictions from the Operation of Glen
Canyon Dam Final Environmental Impact Statement (EIS) table II-7 (U.S. Department of the Interior, 1995) on how
these resources would respond under the modified low fluctuating flow (MLFF) alternative, which is the preferred
alternative in the EIS.
[Glen Canyon Dam Adaptive Management Program (GCDAMP) is a federally authorized initiative to ensure that the
primary mandate of the Grand Canyon Protection Act of 1992 is met through advances in information and resource
management. Resources are ordered in this table as they appear in the EIS. A plus sign (+) indicates that the
prediction was correct or exceeded expectations, a minus sign (-) indicates that the prediction was not entirely
correct or did not achieve the desired outcome, and plus and minus signs together (+/-) indicate a mixed outcome.
Data unavailable are indicated by a question mark (?) and may imply a total absence of data or that the data are not
available to the U.S. Geological Survey’s Grand Canyon Monitoring and Research Center through the research and
monitoring program under the GCDAMP]
Resource Prediction Outcome Comments
Sediment and aquatics
Fine sediment (sandbars and
related physical habitats linked
to native fishes (backwaters),
terrestrial vegetation, marshes,
campsites for recreation, and in
situ preservation of archeologi-
cal resources)
Modest improvement through
implementation of constrained
daily powerplant operations
and periodic implementation
of experimental high flows
following accumulation of new
tributary sand supplies in the
main channel of the ecosystem.
Sand accumulation was pre-
dicted to occur under average-
to-below-average hydrology and
associated hydroelectric power
operations.
-
Sandbars continued to erode,
and new sand inputs were not
accumulated within the main
channel. Experimental high
flows were conducted, but the
lack of flexibility in the tim-
ing and frequency of these
controlled floods limited their
effectiveness.
Coarse sediment (debris flow
impacts from tributaries and
their influence on the naviga-
bility of rapids and terrestrial
sandbars)
This resource was not included
in table II-7 of the EIS (U.S.
Department of the Interior,
1995), but predictions regard-
ing the fate of this material
were given on p. 104–105 of
the document. It is included in
this tabulation for the sake of
completeness.
Inputs of coarse-grained sediment
from tributary debris flows will
continue to accumulate in the
main channel under con-
strained hydropower operations,
causing rapids to worsen and
burying sandbars under coarse
deposits. High-flow releases
may partially rework the new
deposits and improve navigation
within rapids.
+
The influence of ongoing, natu-
rally occurring debris flows, in
terms of aggradation of rapids
and burial of sandbars, has
been partially mitigated by
occasional experimental high
flows. The ability of high dam
releases to rework new debris
flow deposits is related more
to peak discharge and timing
after debris-flow events than it
is to the duration of the high
releases.
210 The State of the Colorado River Ecosystem in Grand Canyon
Aquatic food web “Potential major increase”
+/-
Increases were apparent in Lees
Ferry reach but not canyon-
wide. Fine-sediment inputs
from tributaries below the Lees
Ferry reach are most likely
the limiting factor in primary
productivity.
Native fish “Potential minor increase”
+/-
Recruitment and population of
adult humpback chub de-
creased; native suckers may be
stable or slightly increasing.
Nonnative fish “Potential minor increase”
+
Rainbow trout population in-
creased substantially following
the operational change in the
Lees Ferry reach and within
Marble Canyon.
Interactions between native
and nonnative fish
“Potential minor increase in
warm, stable microhabitats”
An increase in warm, stable mi-
crohabitats would favor native
fish and nonnative warmwater
fish.
-
Warmer dam releases because of
drought-lowered Lake Powell
levels may have increased warm
microhabitats, but this situation
is not directly related to dam
operations.
Trout “Increased growth potential,
stocking-dependent”
-
Rainbow trout numbers have
increased in the Lees Ferry
reach, but condition factor
has declined. Stocking is not
required.
Vegetation
Woody plants Modest increase
Exotic species included (tamarisk,
camel thorn (Alhagi maurorum)).
+
Woody vegetation has increased,
especially arrowweed (Pluchea
sericea) and nonnative tamarisk,
in the riparian zone that was
formerly inundated frequently
under the no action period
(1963–91) of hydropower
operations.
Emergent marsh plants “Same as or less than no action”
+/-
Wet marsh species decreased, and
dry marsh species increased,
likely because of the reductions
of daily inundation and without
periodic rejuvenation through
floods.
Table 1. Natural and cultural resources of the Grand Canyon ecosystem and predictions from the Operation of Glen
Canyon Dam Final Environmental Impact Statement (EIS) table II-7 (U.S. Department of the Interior, 1995) on how
these resources would respond under the modified low fluctuating flow (MLFF) alternative, which is the preferred
alternative in the EIS. —Continued
Lessons from 10 Years of Adaptive Management in Grand Canyon 211
Wildlife
Wintering waterfowl Potential increase
+/-
Trends vary by species and are
difficult to distinguish from
background variation.
Endangered and other
special status species
Native fish (humpback chub,
razorback sucker, flannelmouth
sucker)
“Potential minor increase”
+/-
Recruitment and population of
adult humpback chub de-
creased; native suckers may be
stable or slightly increasing.
Bald eagle “Potential increase” ?Numbers in Arizona have in-
creased overall.
Peregrine falcon No effect
+
Numbers have been stable in
Grand Canyon since 1988.
Kanab ambersnail “Some incidental take”
+/-
Snail habitat increased since 1998,
but not snail numbers, which
are relatively stable.
Southwestern willow
flycatcher
“Undetermined increase” -No increase, but the flycatcher is
uncommon in Grand Canyon.
Cultural resources
Archaeological sites affected “Moderate (less than 157)”
?
Subsequent analyses have not
been conducted to fully assess.
Traditional cultural
properties affected
“Moderate” ?Subsequent analyses have not
been conducted to fully assess.
Traditional cultural
resources affected
“Increased protection”
?
Subsequent analyses have not
been conducted to fully assess.
Air quality
Effect of emissions
on regional air quality
“Slight reduction”
?
Not Addressed by Glen Canyon
Dam Adaptive Management
Program (GCDAMP).
Recreation
Angler safety “Moderate improvement” ?No long-term monitoring data.
Day rafting “Major improvement”
?
Pre-EIS study suggests that net
willingness-to-pay values were
insensitive to flows. More stud-
ies are needed.
Table 1. Natural and cultural resources of the Grand Canyon ecosystem and predictions from the Operation of Glen
Canyon Dam Final Environmental Impact Statement (EIS) table II-7 (U.S. Department of the Interior, 1995) on how
these resources would respond under the modified low fluctuating flow (MLFF) alternative, which is the preferred
alternative in the EIS. —Continued
212 The State of the Colorado River Ecosystem in Grand Canyon
Whitewater boating safety “Minor improvement”
?
NPS responsibility—not moni-
tored as part of GCDAMP.
Whitewater boating camp-
ing beaches (average area at
normal peak stage)
“Minor increase”
-
Camping areas have been dimin-
ished because of vegetation
expansion and sandbar erosion,
despite the fact that the new op-
erating policy has limited daily
peaking release to 25,000 cfs.
Whitewater boating
wilderness values
“Moderate to potential to become
major increase” ?
Potential decrease and decline in
campable areas (see chapter 12).
Economic benefits (not related
to hydropower revenue)
Positive
+
Increase to both locally and
regionally.
Power
Annual economic cost
(foregone hydroelectric
power revenue)
Acceptable costs relative to other
alternatives ?
Subsequent studies are not avail-
able to fully assess.
Wholesale rate of power Acceptable costs relative to other
alternatives
?
Not monitored as part of
GCDAMP. See Western Area
Power Administration (WAPA)
for data.
Retail rate of power (70% of
end users)
“No change to slight decrease”
?
Not monitored as part of
GCDAMP. See WAPA for data.
Retail rate of power (23% of
end users)
“Slight decrease to moderate
increase” ?
Not monitored as part of
GCDAMP. See WAPA for data.
Retail rate of power (7% of
end users)
Acceptable costs relative to other
alternatives ?
Not monitored as part of
GCDAMP. See WAPA for data.
Nonuse value “No data”
+
Substantial nonuse value, $3–$4
billion, has been demonstrated
as willingness to pay for flows to
protect fish.
Table 1. Natural and cultural resources of the Grand Canyon ecosystem and predictions from the Operation of Glen
Canyon Dam Final Environmental Impact Statement (EIS) table II-7 (U.S. Department of the Interior, 1995) on how
these resources would respond under the modified low fluctuating flow (MLFF) alternative, which is the preferred
alternative in the EIS. —Continued
Lessons from 10 Years of Adaptive Management in Grand Canyon 213
Table 2. Operating limits and general likelihood of occurrence under the preferred alternative (U.S. Department of
the Interior, 1995) of modified low fluctuating flows (maf = million acre-feet; cfs = cubic feet per second).
General range of hydrologic conditions for Glen Canyon Dam
Dam releases under
operating rules, as well
as constrained by annual
hydrology
Dry
(minimum of 8.23–10 maf of
annual release)
Normal
(10–15 maf of
annual release)
Wet
(15–20 maf of annual release)
Minimum releases
7 a.m.–7 p.m. (cfs)
8,000
(likely only during weekends)
8,000
(unlikely to occur)
8,000
(very unlikely to occur)
Minimum releases
7 p.m.–7 a.m. (cfs)
5,000
(very likely to occur
on weekends)
5,000
(unlikely to occur)
5,000
(very unlikely to occur)
Maximum peak under
diurnal releases (cfs)
25,000
(daily peaks reach about
18,000–19,000 cfs, mainly
during summer)
25,000 1
(daily peaks reach about
20,000–24,000 cfs, mainly
during summer)
25,000
(steady flows at this level
occur for 1.5 maf
monthly releases)
Daily fluctuations (cfs/24 h) 5,000 (possible)
6,000 (possible)
8,000 2 (possible)
5,000 (unlikely)
6,000 (possible)
8,000 2 (most likely)
5,000 (unlikely)
6,000 (unlikely)
8,000 2 (most likely)
Ramp rate (cfs/h) 4,000 up (always)
1,500 down (always)
4,000 up (always)
1,500 down (always)
4,000 up (always)
1,500 down (always)
Monthly volume (maf) 480,000–900,000 700,000–1,200,000 800,000–2,000,000
1 Exceeded during habitat maintenance flows.
2 Maximums represent normal or routine limits and may necessarily be exceeded during high water years.
214 The State of the Colorado River Ecosystem in Grand Canyon
operation alone are relevant to native fish recruitment or
whether changes in the sediment and thermal regimes
of the river imposed by regulation have had the greatest
influence on native fishes. Further, the issue of nonnative
fishes and their potential to limit recruitment of native
fish through predation and competition (although highly
suspected by scientists as a significant factor) remains
unresolved in Grand Canyon.
Populations of both bluehead and flannelmouth
suckers appear to have remained relatively stable under
the MLFF operating policy. The reasons for this persis-
tence are also unknown. Conversely, the relatively stable
habitat conditions created under the MLFF during
protracted drought conditions, coupled with a coarsen-
ing of substrate in the river channel (see section below on
fine sediment), appear to have greatly favored rainbow
trout (Oncorhynchus mykiss), particularly in the Lees Ferry
reach, as reflected in their increasing numbers dur-
ing the last decade.
Sediment Response
Research and monitoring have conclusively dem-
onstrated a net loss of fine sediment from the Colorado
River ecosystem under the MLFF. Closure of Glen
Canyon Dam eliminated about 84% of the sand that
historically entered Grand Canyon (see chapter 1, p. 18,
this report). Managing the remaining supply below the
dam will apparently require carefully managed experi-
mental high flows that are strategically released imme-
diately following tributary sand inputs. It is not yet clear
whether even this strategy will succeed in sustainable
restoration of sand resources throughout Grand Canyon
(Rubin and others, 2002).
Because physical processes related to hydrology and
sediment transport were relatively well studied, a logi-
cal question is, “Why did the writers of the EIS predict
this outcome incorrectly?” Again, the EIS writers did
a commendable job of using the best science available,
but three critical monitoring programs for measuring
suspended-sediment flux throughout Grand Canyon
were discontinued in the early 1970s. This situation, and
a lack of analytical or conceptual models (Marzolf and
others, 1999), forced the EIS team to evaluate only a
limited set of sand-transport data after the dam was built
and then work with flawed assumptions rather than with
continuous data records of flow and sediment concen-
tration. Clearly, long-term monitoring efforts and good
models (see Conceptual Modeling text box, Overview,
this report) are essential to the success of adaptive man-
agement and accurate predictions.
The EIS assumption that sand would accumulate
on the bed of the river over multiple years has been
transformed through learning and adaptive manage-
ment experimentation. Recent research suggests that
future management of sediment should involve high-flow
releases immediately following inputs of sand and finer
sediment from tributaries below the dam. While such
releases may be controversial because they bypass the
hydroelectric powerplant, recent studies also suggest that
the duration of such flows may need to be only a small
fraction of what was originally suggested. Such fine tun-
ing in the prescription of experimental high flows that
are used for achieving habitat restoration could reduce
the financial impacts and controversy associated with
such management actions.
A physical habitat component of ecosystem resto-
ration tied to the EIS strategy for restoration of native
fish depended on the outcome of modest improvement
in fine-sediment resources. As originally proposed in
the EIS, restoration of sand-based, nearshore habitats,
termed “backwaters,” has also not been realized (see
chapter 1, this report) under the strategy of MLFF
and hydrologically triggered experimental high flows.
Detailed synthesis studies of sediment inputs and outputs
to the system (fine-sediment mass balance), intensive
field monitoring, and change detection analyses from
remote-sensing data all point to a decrease in fine-sedi-
ment resources in Glen, Marble, and Grand Canyons in
the time since the EIS was implemented. These changes
have resulted in smaller and coarser grained sediment
deposits that are associated with a net loss of systemwide
sand supply and no evidence of accumulating sand from
tributary inputs, even under protracted drought hydrol-
ogy and constrained hydropower operations.
We also know from research on coarse sediment
dynamics that there has been an overall trend for the
Grand Canyon reach to experience coarsening of the
substrate in the river channel since completion of Glen
Canyon Dam. As fine sediment is eroded because of
dam operations, gravel and larger material remain. The
impact of this “coarsening” of the river substrate has
two potential biological implications: first is the creation
of preferred habitat for benthic invertebrates, which are
an important component of the aquatic ecology of the
system, and second is the creation of spawning substrate
for the nonnative rainbow trout. Both of these changes
move the system farther from predam conditions and
potentially benefit nonnative species like trout at the
expense of natives.
Loss of sand habitats was documented under the
no action period (1963–91), and loss has continued since
dam operations were altered to reduce sandbar erosion.
Lessons from 10 Years of Adaptive Management in Grand Canyon 215
Sand-transport data collected from 1999 through 2004
indicate that whenever the monthly flow regime from
the dam forces daily peak discharges significantly above
10,000 cfs for extended periods, new and existing sand
and finer sediments are being exported relatively quickly
(weeks to months), rather than accumulating in the main
channel over multiple years.
One alternative test of the MLFF concept for mul-
tiyear accumulation of sand supply might be to equalize
monthly volumes during droughts in order to further
limit daily peaks over such periods. If such a test failed to
increase sand supply through accumulation of tributary
inputs, then objectives for sand-habitat restoration might
have to be reconsidered, or more proactive strategies,
such as sediment augmentation, might need to be imple-
mented. Meanwhile, release of short-duration, habitat-
building flows following significant tributary sand inputs
appears to be the most certain option for restoration of
sand habitats below the dam.
Water Quality and Climate
The presence and design of Glen Canyon Dam
caused major environmental changes to the Colorado
River ecosystem, including (1) alterations in the timing
and variability of the annual, seasonal, and daily flow
patterns of the river; (2) drastic reduction of fine-sedi-
ment supply to the reaches of Glen, Marble, and Grand
Canyons; and (3) reduced variability in water tempera-
ture. On the basis of current science information, the
MLFF operating alternative has not effectively mitigated
the influence of regulation with respect to either the
thermal and hydrologic changes or the fine-sediment
supply limitation of the downstream ecosystem.
Given the importance of Lake Powell as the major
source of water for the Grand Canyon ecosystem below,
the lake monitoring program serves as an early warning
system for changes in water quality. Although data from
Lake Powell indicate that dam operations affect some
resources downstream from Glen Canyon Dam, water
quality in this large reservoir appears to be largely unaf-
fected by the new dam operations since 1991. Overall,
the water quality of the reservoir appears to be strongly
linked to climatic annual to decadal variability govern-
ing spring inflow events and to the aging of the reservoir.
The reservoir does have the potential to exert substantial
impacts on downstream resources, however. For example,
the current drought that started in the late 1990s reduced
the level and volume of Lake Powell to elevations not
seen since the reservoir began filling in the 1960s. The
lower storage level of the reservoir has brought warmer
surface (epilimnetic) waters to the penstocks, causing
higher temperature water to be discharged downstream.
The effects of warmer water on downstream bio-
logical resources are currently difficult to predict with
certainty and potentially include both positive and nega-
tive ecological consequences. Potential consequences
include the creation of conditions that support the main-
stem spawning of native fish, the invasion and domi-
nance of warmwater fishes from Lake Mead, undesirable
alteration of the food base, and unknown effects on the
coldwater fishery in the Lees Ferry reach. This “natural”
warming of Glen Canyon Dam releases that result from
falling reservoir levels provides an important opportunity
to test system responses to the possible installation of a
temperature control device on the dam.
Current understanding of global climate drivers
provides little ability to predict the timing or extent of
droughts over much of the Colorado River Basin. The
current drought may or may not be a so-called “mega
drought” because of the limited duration to date. At the
time of this writing we are cautiously optimistic about
precipitation and runoff predictions for the basin, but the
final outcome will have little to do with dam operations.
Human Use of the River
During the latter part of the 20th century, societal
values associated with river regulation began shifting
away from a policy focused solely on water supply and
energy development to one in which preservation of
natural resources was also valued. This shift in values
occurred only after river regulation by mainstem dams
was well underway on the Colorado River.
River regulation in itself has facilitated the devel-
opment of an economically significant business associ-
ated with whitewater rafting (see chapter 9, this report)
by reducing the predam variability in flow extremes of
the river and thereby allowing such activities to con-
tinue with relative economic safety throughout the year.
Recreational use of the river is one of the resources of
concern that appears to have benefited most from the
stabilizing influence of the MLFF, relative to more vari-
able dam operations. By eliminating very high and very
low discharges, the MLFF favors year-round recreational
boating and fishing. Although the most comprehen-
sive regional economic study of river-based recreation
is now 10 yr old, the figures are impressive: over $46
million (2004 dollars) in nonresident total expenditures
and maintenance of 586 jobs, with 438 jobs in com-
mercial rafting alone. Presumably those figures are
even higher today.
216 The State of the Colorado River Ecosystem in Grand Canyon
Camping
Between 1998 and 2003, camping area above the
25,000-cfs stage elevation decreased by 55%, and the
average rate of change was 15% per year. The decrease
in high-elevation campsite area occurred in Marble
Canyon and in Grand Canyon as well as within criti-
cal (campsite-limited) and noncritical reaches. Losses
are thought to be attributable both to net sediment
exports under current dam operations (see chapters
1 and 12, this report) and to encroachment of woody
vegetation (see chapter 6, this report). Notably, lower
elevation campsite areas increased after 2000, and the
total campsite area below the 25,000-cfs stage elevation
now exceeds the area available at higher elevations. The
rate of decrease in high-elevation campsite area greatly
exceeds the decrease in sandbar volume. Vegetation
encroachment most likely contributed to the recent loss
of high-elevation campsite area.
The exact relationship and interaction among camp-
ing areas, vegetation expansion, and dam operations are
unknown and provide a challenge for future researchers.
For example, increased vegetation in sandbar areas may
also provide greater substrate stability and shade, both
limiting campsite erosion rates and enhancing camping
areas from an aesthetic perspective. Tradeoffs between
vegetation expansion and sandbar stability must also be
considered from the perspective that increased vegeta-
tion might also limit the potential for wind processes to
beneficially blow sand deposits upslope onto cultural sites
that are subject to rainfall and runoff erosion.
Overall, in terms of recreation, future research
should focus on detailed analyses of how fishing use,
catch rates, and fish condition in Glen Canyon National
Recreation Area are related to flows. Economic data on
fishing and recreational rafting need to be updated to
establish current baseline data before new flow regimes
are initiated. Finally, studies to quantify the wilderness
experiences of recreational users must be initiated so that
the benefit of eventually achieving ecosystem restoration
can be fully evaluated.
Nonuse Values
Survey efforts tied to the EIS process found that
households across the Nation, including those that might
never visit Grand Canyon, were willing to pay addi-
tional taxes for flows that benefited native fish and trout.
Although the amount that people were willing to pay for
these benefits was quite reasonable, when aggregated up
to the number of households in the population, it pro-
duced estimates in the $3 billion to $4 billion (2004 dol-
lars) range (see chapter 9, this report). The public at large
is willing to pay to have flows and other management
actions that benefit Glen Canyon National Recreation
Area, Grand Canyon National Park, and the resources
found in both. While there are those who question the util-
ity of nonuse valuation of Grand Canyon resources, these
dollar amounts reflect the iconic values that make Grand
Canyon National Park famous throughout the world.
Hydroelectric Power Generation
Hydroelectric power and revenue associated with
its production and marketing are also highly valued
resources of concern to society, and environmental
constraints on Glen Canyon Dam operations under the
MLFF policy have large, annual economic costs (see
chapter 10, this report). The acceptability of those costs
has to be interpreted within the context of societal values
associated with both electrical energy and environmen-
tal conservation objectives. Environmental constraints
on dam operations have regional economic impacts on
power revenue that is generated to pay back the cost of
Glen Canyon Dam and to fund related water-resource
and energy development. EIS studies on recreational use
and nonuse values, however, suggest that dam opera-
tions under the MLFF benefit both local and regional
economies through stabilization of flows, despite infor-
mation that suggests that the ecological objectives of the
program (e.g., retention of fine sediment, recovery of the
humpback chub) remain unfulfilled.
Because of the constraints imposed on dam opera-
tions under the MLFF, the economic value of hydro-
power that was foregone is unknown. Existing scientific
data suggest, however, that the policy, no matter how
costly, has not resulted in the level of environmental ben-
efits predicted or desired in the EIS for natural resources
below the dam.
An ex post facto cost-benefit analysis of Glen
Canyon Dam operations is needed to fully assess the eco-
nomic value of the MLFF operation versus documented
environmental benefits below the dam. Additional
experimental designs and the eventual implementation
of alternative, longer-term changes in the current oper-
ating strategy would benefit from such an assessment.
Water Resources
Water allocation in the Colorado River Basin is
governed by the Colorado River Compact of 1922 and
subsequent laws and treaties. None of the laws pertain-
Lessons from 10 Years of Adaptive Management in Grand Canyon 217
ing to water management in the basin were superseded
by the Grand Canyon Protection Act of 1992 or the
1996 Record of Decision; therefore, delivery of the vol-
ume of water required under these laws, including under
the MLFF, has remained unaffected by the Glen Canyon
Dam Adaptive Management Program.
Cultural Resources
Under the MLFF operation, sand that is being
exported from Grand Canyon is coming not only from
new tributary inputs but also from existing beaches and
river terraces that contain archaeological sites. Many
archaeological sites in Grand Canyon have been covered
with windborne (aeolian) sand for centuries. This sand
was transported from lower elevation beaches that were
frequently resupplied with new sand sources derived
from annual floods during the predam era. Before dam
operations were constrained, operations in the no action
period (1963–91) were optimized for maximum water
storage and power revenue (within the constraints of
existing law and policies) rather than for strategically
conserving limited sand supplies remaining downstream
for restoration of sandbars (and, presumably, long-term
preservation of cultural sites). The MLFF operation has
not mitigated sand export, and therefore the sand supply
remains critically limited. With more sand leaving the
ecosystem than being supplied, more and more of these
archaeological sites are being exposed to the ravages
of erosion. As sites are eroded, artifacts and structures
are exposed, making them more susceptible to visitor
impacts and destabilization because of the loss of the
surrounding sedimentary matrix in which they are bur-
ied. Such changes make it difficult, if not impossible, for
archaeologists to reconstruct and interpret the historical
and cultural information contained within these impor-
tant settings within Grand Canyon.
In addition to being valued by scientists for the
information that the archeological sites provide, the
sites are also valued by many Native American people
who have traditional affiliations with these sites and the
Grand Canyon area in general. Other resources along
the Colorado River that are also valued by the tribes of
the region include traditionally used plants, minerals,
water sources, and significant landscape features. The
effects of the MLFF on these tribally valued resources
remain uncertain; the Native Americans engaged in the
Glen Canyon Dam Adaptive Management Program
would like to see more emphasis placed on monitoring
the effects on these resources in the future.
Vegetation in the River Corridor
Glen Canyon Dam operations under the MLFF
have stabilized flow conditions that were in effect before
the EIS in Grand Canyon and that had significant
impacts on riparian vegetation. The EIS predicted a
modest increase in woody vegetation (table 1), and that
prediction has proven largely correct, if not understated.
The EIS also predicted that marsh communities would
be the same as or less than expected under the no action
alternative, which is also largely correct. Since imple-
mentation of the MLFF, there has been a decrease in wet
marsh vegetation and an increase in dry marsh vegeta-
tion.
The stability of flows has encouraged an increase in
vegetation density in and near the wetted zone. While an
increase in vegetation may appear to be desirable, one of
the impacts is a decrease in available camping area for
recreational users (see chapter 12, this report). Further-
more, the increase is partially attributable to expansion
of nonnative tamarisk (Tamarix ramosissima) and arrow-
weed into the riparian zone. In the terrestrial realm,
future research should focus on identifying the responses
of wildlife to this fundamental change in habitat struc-
ture, striving to understand the relationship between
riparian vegetation and insects as related to the food web
of the river, and examining the effects of human-medi-
ated removal of nonnative vegetation versus natural dis-
turbance. Understanding the complex interface between
dam operations and overlapping elements of both the
terrestrial and aquatic parts of the river ecosystem pro-
vides even greater challenges.
Integrated Ecological Factors
Although linkages between native fish recruitment
and backwaters are not well documented and strategies
for achieving sandbar habitat restoration are still being
investigated, it is clear that physical habitat availability
does not reduce or diminish the need by native fish for a
sustainable food supply. Dam operations under the MLFF
have resulted in steadier flows and greater minimum dis-
charges of clearer water than operations in the no action
period, and this situation has probably led to increases in
the standing mass of algae and invertebrates (table 1).
A critical future research need is to develop a better
understanding of the linkages between the organic mat-
ter and invertebrates and the actual prey base of fish,
both native and nonnative. A large amount of data has
been collected on the food items consumed by nonnative
218 The State of the Colorado River Ecosystem in Grand Canyon
rainbow trout, but only limited data are available for the
humpback chub because of the endangered status of
this fish. Use of stable-isotope analysis will be critical to
assessing the energy sources and trophic pathways that
are important to fish. Also, critical tests of the hypothesis
that competition between nonnative trout and humpback
chub is negatively affecting humpback chub populations
are not possible because of inadequate data. Eating the
same food items is only the first criterion to establish that
competition is negatively impacting a particular species.
Further research will be required to determine if this
hypothesis is supported by data.
Recent Management Experiments
With respect to native fishes, we have learned that,
under the MLFF, focused efforts are still required to
understand the importance of the sediment and ther-
mal aspects of physical habitat in the early life history
of humpback chub and other species. Focused efforts
are also needed to understand the influence of intro-
duced nonnative species on the successful recruitment
of humpback chub to the adult life stage. Additional
experiments in these areas will require even more com-
mitment to the adaptive management approach by using
repeated implementation of both flow and nonflow
treatments over an extended period of perhaps 10 or
more years coupled with long-term monitoring.
Mechanical removal of nonnative fish, especially
rainbow trout, is currently in the third year of a 4-yr
implementation strategy to test the hypothesis that reduc-
tion of predatory and competitive fish species will result
in an increase in survival and recruitment of humpback
chub (see U.S. Department of the Interior, 2002, for
details of the current experimental design). While the
adaptive management program has demonstrated that
mechanical removal is an effective way to significantly
reduce the number of nonnative trout in the removal
reach, it has yet to detect the desired increase in the
number of spawning-age humpback chub.
In addition to understanding how aquatic ecology,
dam operations, and fish populations of the Colorado
River ecosystem are interrelated, scientists need to focus
future research on the effects of warming discharges of
water from Glen Canyon Dam. An experimental tem-
perature control device has been proposed for the dam
later in this decade, and substantial questions remain
unanswered as to the efficacy of this experimental treat-
ment with respect to both its risk and its cost benefit.
As linkages between the aquatic ecology of the river
and its native and nonnative fishes are defined, new
efforts for tracking critical elements of water quality will
need to evolve to track bioenergetic pathways and fish
responses to flow treatments such as temperature control
device operations. Also, as more information becomes
available about the role of fine sediment in the preserva-
tion of cultural sites, aquatic ecology, and fish recruit-
ment, there will need to be a commitment to model-
ing and monitoring of long-term fluxes of suspended
organic and inorganic materials through the river system.
Conclusions and
Future Challenges
Research and monitoring conducted by U.S.
Geological Survey scientists and their cooperators have
conclusively demonstrated a net loss of sediment from
the system and have documented the decline of the
federally endangered humpback chub during the last
decade. At this first milestone, both findings are critical
pieces of information to assess conditions and adjust
management actions in the spirit of adaptive manage-
ment. It is important to note that water-delivery require-
ments continued to be met throughout the decade after
the EIS, despite increased costs associated with environ-
mental and experimental regulation of flows.
Although incomplete, a substantial body of knowl-
edge now exists for the Colorado River ecosystem in
Grand Canyon. The overarching question is, “What will
society do with the knowledge now available to move
into the next active phase of the Glen Canyon Dam
Adaptive Management Program?” The complexity of
the natural system presents enormous challenges for
determining how resources and population numbers
vary in time and space and underscores the importance
of long-term studies to describe patterns and pro-
cesses. The next critical phase of adaptive management
requires strategic action on the part of both managers
and scientists.
Along with future action come the continuing chal-
lenge and need for greater integration of monitoring and
research studies. As the complexity of issues in the Glen
Canyon Dam Adaptive Management Program becomes
more obvious, so does the need for interdisciplinary, not
just multidisciplinary, science. Good examples of recent
interdisciplinary science include (1) the interface between
fine-sediment studies and cultural resources as the result
of research efforts to understand the deposition of wind-
carried sediment, (2) water-quality studies related to
temperature and fish biology, and (3) the developing link
between aquatic ecology studies and fish diets. Contin-
Lessons from 10 Years of Adaptive Management in Grand Canyon 219
ued efforts will be required to integrate knowledge across
disciplines and scales and to develop a more robust con-
ceptual model for the Grand Canyon ecosystem.
A continued adaptive management approach below
Glen Canyon Dam, one focused on systematic experi-
mentation, is recommended as a more efficient strategy
for learning than the approach initially undertaken of
monitoring the MLFF operation without comparing it
to other flow and nonflow (e.g., mechanical removal of
nonnative fish) alternatives. Future experimental treat-
ments (flow or otherwise) must be evaluated within a
strategic framework of periodic milestones and with
rigorous scientific review so as to effectively identify
viable management options for achieving the desired mix
of resource responses. Success in this approach relies first
and foremost on managers and stakeholders identifying
what is desired, as well as determining whether identi-
fied objectives are measurable by science and attainable
through dam operations.
In the context of an adaptive ecosystem manage-
ment process, the information identified at this mile-
stone should lead to a dialog between managers and
scientists about what other flow or nonflow alternatives
might be considered to achieve the desired environmen-
tal outcomes.
Because of the uncertainty about the cause and
effect of MLFF operations on Grand Canyon resources
and the even greater uncertainty about other conserva-
tion options, such as mechanical removal or thermal
modification, the next steps in the process seem best
approached as ongoing management policy experiments
in the spirit of adaptive management, punctuated with
frequent milestones at which the state of knowledge
gained is assessed by all interested parties.
Critical Issues for
Further Research
The synthesis of knowledge from over a decade of
research and monitoring in Grand Canyon provides an
opportunity to identify critical research needs in the Glen
Canyon Dam Adaptive Management Program. The
purpose of this report is to establish another milestone
in our understanding of the state of resources in Grand
Canyon; however, we do not attempt herein to provide
an exhaustive or prescriptive list of management options.
That effort will require another set of processes and
products. Instead, we focus on the drivers, or major sci-
ence questions, that will need to be addressed in the next
phase of adaptive management:
Why is the humpback chub population in Grand
Canyon declining? Specifically, what factors or
combination of factors are most influential in this
downward trend?
What is the linkage between native and nonna-
tive fish population dynamics and the aquatic and
terrestrial food base, and how are these factors
related to dam operations?
Under a potentially continuing drought scenario
for the upper basin of the Colorado River, what
are the impacts of warmwater discharges on the
ecosystem?
If additional research demonstrates that sediment
inputs from tributaries below the dam cannot be
manipulated to achieve the desired conservation
of sandbars and backwaters with dam operations,
is sediment augmentation a viable option?
What are the specific linkages between dam
operations and archaeological site erosion, and
what are the options for preserving the significant
variety of culturally important resources and the
information values associated with nonrenewable
heritage resources?
Experimentation and research are needed to deter-
mine what role Glen Canyon Dam operations have had
in these issues and whether further changes in those
operations can benefit key resources.
References
Clark, M.J., 2002, Dealing with uncertainty: adaptive
approaches to sustainable river management: Aquatic
Conservation: Marine and Freshwater Ecosystems, v.
12, p. 347–363.
Marzolf, G.R., Jackson, W.L., and Randle, T.J., 1999,
Flood releases from dams as management tools: inter-
actions between science and management, in Webb,
R.H., Schmidt, J.C., Marzolf, G.R, and Valdez, R.A.,
eds., The controlled flood in Grand Canyon: Washing-
ton, D.C., American Geophysical Union, Geophysical
Monograph Series, v. 110, p. 359–367.
Mueller, G.A., and Marsh, P.C., 2002, Lost, a desert
river and its native fishes: a historical perspective of
the lower Colorado River: Denver, Colo., U.S. Gov-
ernment Printing Office, Information and Technology
Report USGS/BRD/ITR-2002-0010, 69 p.
•
•
•
•
•
220 The State of the Colorado River Ecosystem in Grand Canyon
Contact Information:
Jeffrey E. Lovich
Deputy Center Director
U.S. Department of the Interior
U.S. Geological Survey
Southwest Biological Science Center
Flagstaff, AZ
jeffrey_lovich@usgs.gov
Theodore S. Melis
Physical Scientist
U.S. Department of the Interior
U.S. Geological Survey
Southwest Biological Science Center
Flagstaff, AZ
tmelis@usgs.gov
J
e
f
f
S
o
r
e
n
s
e
n
,
A
r
i
z
o
n
a
G
a
m
e
a
n
d
F
i
s
h
D
e
p
a
r
t
m
e
n
t
Rubin, D.M., Topping, D.J., Schmidt, J.C., Hazel, J.,
Kaplinski, M., and Melis, T.S., 2002, Recent sediment
studies refute Glen Canyon Dam hypothesis: Eos,
Transactions of the American Geophysical Union, v.
83, p. 273, 277–278.
U.S. Department of the Interior, 1995, Operation of
Glen Canyon Dam Final Environmental Impact State-
ment: Salt Lake City, Utah, Bureau of Reclamation,
Upper Colorado Region, 337 p., appendices.
U.S. Department of the Interior, 1996, Record of Deci-
sion, Operation of Glen Canyon Dam: Washington,
D.C., Office of the Secretary of Interior, 13 p.
U.S. Department of the Interior, 2002, Proposed experi-
mental releases from Glen Canyon Dam and removal
of nonnative fish; environmental assessment: Salt Lake
City, Utah, Bureau of Reclamation, Upper Colorado
Region, 112 p., appendices.
Walters, C.J., and Holling, C.S., 1990, Large-scale man-
agement experiments and learning by doing: Ecology,
v. 7, p. 2060–2068.
First page photograph credit: © 2005 Christopher Taesali; used with permission

Navigation menu