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, Darwins 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 regions 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