Central Arizona Water Conservation District
CAWCD Report to the Secretary, United States Department of the Interior
U.S. Fish and Wildlife Service Final Biological Opinion on the Transportation and Delivery of Central Arizona Project Water to the Gila River Basin
May 1995 LG
JUN 1 9 1995
U.S. FISH & ES WILDLIFE SERVICE FIELD4FT10E - OENIX EXECUTIVE SUMMARY
The Central Arizona Water Conservation District ("CAWCD") has prepared this report in response to the April 1994 Biological Opinion issued by the U.S. Fish and Wildlife Service ("Service") concluding that continued Central Arizona Project ("CAP") water deliveries to central Arizona jeopardize threatened or endangered fish species in the upper Gila River basin and adversely modify critical habitat.
The CAP is a multi-billion dollar project authorized by Congress in the Colorado River Basin Project Act of 1968 and subsequently constructed by the United States Bureau of Reclamation ("Reclamation"). Reclamation started building the aqueduct in 1973 and declared the CAP water supply system substantially complete in 1993.
CAWCD is a political subdivision of the State of Arizona organized for the express purpose of contracting with the United States for repayment of CAP costs and subcontracting with non-Indian water users for delivery of CAP water. CAWCD's repayment obligation will approach $2 billion. CAWCD also operates and maintains the CAP, which provides water for agricultural, municipal and industrial, and Native American use to approximately eighty percent of Arizona's population.
The Biological Opinion. The biological opinion asserts that non-native fish in CAP water delivered to agricultural users in central Arizona could escape into mostly-dry natural watercourses and swim up to two hundred miles upstream--bypassing dams, fish barriers and dry stream segments along the way--where they would jeopardize four threatened or endangered fish species and adversely modify critical habitat for three of those species. The reasonable and prudent alternatives ("RPAs") specified in the Biological Opinion require Reclamation to: (1) transfer $500,000 a year for 25 years to the Service to fund research and conservation activities in Arizona and New Mexico, far beyond the possible area of CAP influence; (2) construct and maintain a series of fish barriers, even though the biological opinion stated that large mainstream dams were not effective barriers to fish movement; (3) monitor the presence and distribution of non- native fish in Arizona streams and waters for 100 years; and (4) develop and implement an information and education program on the conservation of native fish and their habitats. CAWCD estimates the cost of complying with the RPAs to be more than $150 million over the life of the CAP. Although the bulk of this financial burden will fall on CAWCD taxpayers and water users, CAWCD was not allowed to participate in the consultation process that led to the Biological Opinion.
The Biological Opinion contradicts previous consultations and biological opinions regarding the CAP. Millions of dollars were spent during construction of the CAP to minimize and mitigate any impact the project might have on the environment. To ensure that no endangered species would be harmed, Reclamation completed more than forty consultations with the Service on the CAP before the Biological Opinion was issued. Nevertheless, the conclusions of the Biological Opinion directly contradict
1 representations made by the Service in previous consultations and in a 1983 biological opinion on the CAP. The Biological Opinion did not acknowledge those previous consultations and opinions and offered no explanation for the contradictions.
The Biological Opinion ignored readily available data. The Biological Opinion chose to rely on unfounded assumptions and hypotheses instead of examining available scientific and commercial data, as required by law. This report shows that, contrary to the findings of the Biological Opinion, the data strongly supports the conclusion that CAP water deliveries will have no demonstrable impact on endangered fish. Numerous biological and hydrologic factors make it unreasonable to conclude that non-native fish from the CAP could reach areas occupied by the endangered fish. Even if an occasional non-native fish from the CAP did reach those areas, it would have no demonstrable impact in light of the millions of non-native fish that have been intentionally stocked in Arizona waters (often in or adjacent to endangered fish habitat) by the Service itself and state game and fish agencies operating with Service funds.
The Biological Opinion violated Service regulations. The Biological Opinion exceeded the scope of the consultation requested by Reclamation and improperly included the effects of actions that have nothing to do with the CAP in finding that the CAP causes jeopardy. The Biological Opinion also reflects a double standard: Although the Service has stated that its own stocking of thousands of rainbow trout--a known predator--directly into critical habitat for the razorback sucker and bonytail chub would not jeopardize those species, the Biological Opinion concludes that the accidental escape of any non-native fish from the CAP would jeopardize razorback sucker located nearly one hundred miles upstream, notwithstanding the intervening presence of large dams that block fish movement.
In short, under the auspices of the Endangered Species Act, the Service has declared that the CAP--a federal project that has been substantially completed at a cost of several billion dollars--cannot operate as intended unless certain conditions are met, chiefly the transfer of money to fund the Service's own research and conservation activities. As this report shows, the conclusions of the Biological Opinion are unfounded and contrary to available scientific and commercial data. Therefore, the Biological Opinion should be reconsidered and the jeopardy finding withdrawn.
2 TABLE OF CONTENTS
Index of Figures, Photos and Maps ...... iv
Introduction 1
I. Background ...... 2
A. The Central Arizona Project ...... 2
1. History ...... 2
2. Physical Features ...... 3
B. The Biological Opinion Contradicts Previous Section 7 Consultations on the CAP ...... 6
C. The Endangered Fish ...... 9
1. Spikedace ( Meda fulgida) ...... 9
2. Loach Minnow ( Tiaroga cobitis) ...... 10
3. Gila topminnow ( Poeciliopsis occidentalis occidentalis) ...... 11
4. Razorback sucker ( Xyrauchen texanus) ...... 11
D. Non-Native Fish in the CAP and Gila River Basin ...... 13
II. The Hassayampa River Sub-basin ...... 17
A. CAP Water Deliveries Cannot Jeopardize Endangered Fish in the Hassayampa River Sub-basin ...... 18
B. No Critical Habitat Has Been Designated in the Hassayampa River Sub-basin ...... 23
III. The Agua Fria River Sub-basin ...... 24
A The Service Previously Determined that the CAP Does Not Jeopardize Endangered Fish in the Agua Fria River Sub-basin ...... 25 B. No Critical Habitat Has Been Designated in the Agua Fria River Sub-basin ...... 29
IV. The Salt and Verde River Sub-basin ...... 29
A. CAP Water Deliveries Cannot Jeopardize Endangered Fish in the Salt and Verde River Sub-basin ...... 31
B. The CAP Will Not Adversely Modify Critical Habitat in the Salt and Verde River Sub-basin ...... 37
V. The Gila River Sub-basin ...... 37
A. Non-native Fish from the CAP are Not Likely to Reach the Gila River ...... 39
B. The CAP Cannot Affect Endangered Fish in the Gila River Above Coolidge Dam ...... 41
1. Non-native Fish from the CAP Cannot Reach the Gila River Above Coolidge Dam ...... 42
2. The CAP Will Not Adversely Modify Critical Habitat in the Upper Gila River ...... 43
C. The CAP Will Not Affect Endangered Fish in the San Pedro River ...... 44
1 The Endangered Fish were Extirpated from the San Pedro River Before CAP Water Deliveries Began ...... 44
2. No Critical Habitat Has Been Designated in the San Pedro River ...... 47
D. The CAP Cannot Reasonably Be Expected to Affect Native Fish or Critical Habitat in Aravaipa Creek ...... 47
VI. The Biological Opinion Violated Service Regulations ...... 49
A. The Biological Opinion Exceeded the Scope of the Consultation ...... 49
B. The Biological Opinion Improperly Determined the Effects of the CAP ...... 50
ii C. The Biological Opinion Did Not Consider the Best Available Scientific and Commercial Data ...... 53
D. The Biological Opinion Used a Double Standard in Evaluating the CAP ...... 53
Conclusion 55
Appendix 1 (Biological Opinion)
Appendix 2 (Sources not considered by Biological Opinion)
iii INDEX OF FIGURES, PHOTOS AND MAPS
Figures Follows page:
Figure 1: CAP Aqueduct System and Potential Locations of Endangered Fish ...... 1
Figure 2: Upper and Lower Colorado River Basin ...... 3
Figure 3: CAP Pumping Plants ...... 4
Figure 4: CAP/SRP Interconnection ...... 7
Figure 5: CAP Aqueduct System and Potential Locations of Spikedace ...... 9
Figure 6: CAP Aqueduct System and Potential Locations of Loach Minnow ...... 10
Figure 7: CAP Aqueduct System and Potential Locations of Gila Topminnow ...... 11
Figure 8: CAP Aqueduct System and Potential Locations of Razorback Sucker ...... 12
Figure 9: Electrofishing Data for CAP, 1987-1992 ...... 14
Figure 10: Flow Summary for Hassayampa River Near Morristown ...... 21
Figure 11: Temperature of Hassayampa River and Requirements of Striped Bass ...... 22
Figure 12: Profile of Salt and Verde Reservoir Systems ...... 30
Figure 13: Electric Fish Barrier--Plan View ...... 33
Figure 14: Electric Fish Barrier--Profile View ...... 33
iv Photos Follows page:
Photo 1: Box overchute to channel wash over CAP aqueduct ...... 4
Photo 2: Construction of CAP siphon under Salt River ...... 4
Photo 3: Flood protection dike adjacent to CAP aqueduct ...... 4
Photo 4: Wildlife crossing over CAP aqueduct ...... 4
Photo 5: Spikedace (Meda fulgida) ...... 9
Photo 6: Loach minnow (Tiaroga cobitis) ...... 9
Photo 7: Gila topminnow (Poeciliopsis occidentalis occidentalis) ...... 11
Photo 8: Razorback sucker (Xyrauchen texanus) ...... 11
Photo 9: CAP turnout for delivery of water to SRP canals ...... 13
Photo 10: Hassayampa River near CAP aqueduct, north of Gila River ...... 17
Photo 11: Buckeye Irrigation Company Canal dike in Gila River channel ...... 17
Photo 12: Tailwater discharge from BIC Canal into Hassayampa River ...... 17
Photo 13: Hassayampa River at USGS gauge near Arlington ...... 17
Photo 14: Arlington Canal dike in Gila River channel ...... 18
Photo 15: Tailwater from Arlington Canal at Gillespie Dam ...... 18
Photo 16: Gillespie Dam ...... 18
Photo 17: Painted Rock Dam ...... 18
Photo 18: Hassayampa River channel near USGS gauge at Morristown ...... 21
Photo 19: Hassayampa River channel about 2 miles south of Morristown ...... 21
Photo 20: Hassayampa River near CAP siphon crossing ...... 21
Photo 21: Enterprise Canal diversion at Gillespie Dam ...... 22
v Photo 22: Gila Bend Canal diversion at Gillespie Dam ...... 22
Photo 23: Palm Lake at The Hassayampa River Preserve ...... 22
Photo 24: Concrete dam on Cow Creek above Lake Pleasant ...... 25
Photo 25: Fish barrier on Tule Creek above Lake Pleasant ...... 25
Photo 26: Theodore Roosevelt Dam on the Salt River ...... 30
Photo 27: Modified Roosevelt Dam (under construction) ...... 30
Photo 28: Horse Mesa Dam on the Salt River ...... 30
Photo 29: Mormon Flat Dam on the Salt River ...... 30
Photo 30: Stewart Mountain Dam on the Salt River ...... 30
Photo 31: Horseshoe Dam on the Verde River ...... 30
Photo 32: Bartlett Dam on the Verde River ...... 30
Photo 33: Granite Reef Diversion Dam on the Salt River ...... 30
Photo 34: Electrical fish barrier on SRP's South Canal ...... 30
Photo 35: Electrical fish barrier on SRP's Arizona Canal ...... 30
Photo 36: Roosevelt Diversion Dam on the Salt River ...... 36
Photo 37: Plaque at Roosevelt Diversion Dam ...... 36
Photo 38: Coolidge Dam on the Gila River ...... 38
Photo 39: Ashurst-Hayden Dam on the Gila River ...... 38
Photo 40: Electrical fish barrier on Florence-Casa Grande Canal ...... 39
Photo 41: Headgate of Florence-Casa Grande Canal ...... 39
vi Maps Follows page:
Map 1: Hassayampa River Sub-basin ...... 18
Map 2: Agua Fria River Sub-basin ...... 25
Map 3: Salt and Verde River Sub-basin ...... 31
Map 4: Gila River ...... 39
Map 5: Gila River Above Coolidge Dam ...... 41
Map 6: San Pedro River and Aravaipa Creek ...... 44
vii Introduction
On April 20, 1994, the United States Fish and Wildlife Service (the "Service" or "USFWS") issued its Final Biological Opinion on the Transportation and Delivery of Central Arizona Project Water to the Gila River Basin (Hassayampa, Agua Fria, Salt, Verde, San Pedro, middle and upper Gila Rivers, and associated tributaries) in Arizona and New Mexico (the "Biological Opinion"). [See Appendix 1] The Biological Opinion concludes that continued delivery of Central Arizona Project ("CAP") water to CAP subcontractors in central Arizona will jeopardize four threatened or endangered fish species found in the upper reaches of the Gila River basin--the spikedace, loach minnow, Gila topminnow, and razorback sucker. [See Figure 1] The Biological Opinion also states that CAP water deliveries will adversely modify designated critical habitat for three of those species. As reasonable and prudent alternatives ("RPAs"), the Biological Opinion directs the United States Bureau of Reclamation ("Reclamation") to:
• Transfer $500,000 a year for 25 years to the Service to fund research and conservation activities in Arizona and New Mexico, far beyond the possible area of CAP influence;
• Construct and maintain a series of physical drop structures to serve as fish barriers, though the Biological Opinion states that even large mainstream dams are not effective barriers to fish movement;
• Monitor the presence and distribution of non-native fish in Arizona streams and waters for 100 years; and
• Develop and implement an information and education program on the conservation of native fish and their habitats.
The Central Arizona Water Conservation District ("CAWCD") estimates that the cost of complying with these RPAs will exceed $150 million over the life of the CAP.1 Although the bulk of this financial burden falls directly on CAWCD's taxpayers and water users, CAWCD was not allowed to participate in the Section 7 consultation process that led to the Biological Opinion.
After an extensive investigation and analysis of available data, most of which was not considered in the Biological Opinion, CAWCD has prepared this report addressing the effects of the CAP on endangered fish in the Gila River basin. The inescapable conclusion is that the findings of the Biological Opinion are contrary to the best available scientific and commercial data and represent arbitrary and capricious agency action.
Estimate of $152.056 million is based on Reclamation data and includes $6.264 million for construction of new fish barriers on Aravaipa Creek and the San Pedro River, $132.141 million to monitor non-native fish and operate existing electrical fish barriers for 100 years, $13.47 million for fund transfers to the Service and associated Reclamation overhead for 25 years, and $181,000 for public information and education. CENTRAL ARIZONA PROJECT AQUEDUCT SYSTEM AND POTENTIAL LOCATIONS OF ENDANGERED FISH
Legend of Endangered Fish
4 4■ Gila Topminnow Spikedace Razorback Sucker Loachminnow
Perennial/Regulated Streams Intermittent/Ephemeral Streams
Central Arizona Project Aqueduct
Figure 1 Millions of dollars were spent during construction of the CAP to minimize and mitigate any impact the project might have on the environment. Reclamation consulted with the Service on every aspect of the CAP to ensure that no endangered species would be harmed. Indeed, more than forty interagency consultations were successfully completed on the CAP before the Biological Opinion was issued. Nevertheless, the Service waited until the eleventh hour, after billions of dollars had been invested and the project was substantially complete, to raise the concerns expressed in the Biological Opinion. The Biological Opinion is especially troubling because it contradicts everything the Service had said in consultations and biological opinions regarding the CAP over the previous decade, yet without offering any explanation for the change in view.
As shown in this report, it is unreasonable to conclude that non-native fish in the CAP could reach those portions of the Gila River basin where endangered fish may be located. Contrary to the Biological Opinion, fish from the CAP would not be able to traverse the many dams, fish barriers and stretches of dry streambed that currently block upstream movement into native fish habitat. Even if a few non-native fish from the CAP did somehow reach native fish habitat, they could not reasonably be expected to have any demonstrable impact on endangered fish, particularly in light of the millions of non- native fish previously stocked in those waters by the Service and state game and fish agencies.
The Biological Opinion is also procedurally flawed. It failed to follow Service regulations in evaluating the effects of the CAP and exceeded the scope of the consultation requested by Reclamation. The Biological Opinion also ignored key scientific and commercial data relating to the biology of the fish in question and the hydrology of the Gila River basin. Worse, the Biological Opinion applied a different, tougher standard to the CAP than that applied by the Service in evaluating its own projects such as stocking of non-native sport fish.
The information presented in this report demands reconsideration of the Biological Opinion and compels a finding of no jeopardy.
I. Background
A. The Central Arizona Project.
1. History.
The Colorado River flows some 1400 miles from the mountains of Colorado to the Gulf of California in Mexico, draining an area comprising almost one-twelfth of the continental United States. Much of that drainage is so arid that its productivity and habitability are dependent on the controlled and managed use of the Colorado River system. That is especially true for the State of Arizona.
-2- Concern over the expanding development and use of the river led the seven Colorado River Basin states (Arizona, Nevada, California, New Mexico, Utah, Colorado and Wyoming) to sign the Colorado River Compact in 1922. The Compact divided the river system into an Upper Basin and a Lower Basin, with Lee's Ferry in northern Arizona as the dividing point, and apportioned 7.5 million acre-feet of Colorado River water annually to the Lower Basin states. [See Figure 2]
A long running dispute among the Lower Basin states over how to divide the 7.5 million acre-feet was ultimately resolved by a United States Supreme Court decree in 1964 that confirmed Arizona's right to 2.8 million acre-feet of Colorado River water annually and allocated 4.4 million acre-feet to California and 300,000 acre-feet to Nevada.
In 1968, Congress enacted the Colorado River Basin Project Act which authorized the Secretary of the Interior to construct the Central Arizona Project to enable Arizona to make use of its Colorado River allocation. The CAP serves many purposes. First and foremost, CAP water deliveries are necessary to reduce the dangerous overdraft of groundwater resources in central Arizona, while maintaining as much as possible of the area's irrigated farmland and providing a source of additional water to support anticipated municipal and industrial growth.2 The CAP also provides flood control, power management, recreation, and fish and wildlife benefits.3
Before the CAP, Arizona's groundwater resources were being depleted at the rate of two million acre-feet per year.4 In 1980, at the request of the Secretary of the Interior, Arizona enacted a Groundwater Management Act designed to achieve the conservation goals of the Colorado River Basin Project Act.5 The Groundwater Management Act requires that the withdrawal of groundwater and the natural recharge of Arizona's groundwater aquifers be brought into balance by 2025. Achievement of this goal is not possible without CAP water. Absent the CAP, Arizona's groundwater and surface water resources would be further depleted until they could no longer sustain the State's economy.
2. Physical Features.
The CAP consists of a 336-mile long water conveyance system of concrete-lined canals, inverted siphons, tunnels and pipelines that is designed to transport up to 2.2 million acre-feet of water per year from Lake Havasu on the Colorado River to the
2 S. Rep. No. 408, 90th Cong. 1st Sess. 27 (1967); H.R. Rep. No. 1312, 90th Cong., 2d Sess. 55 (1968).
3 43 U.S.C. § 1501(a).
4 S. Rep. No 408, 90th Cong. 1st Sess. 27 (1967); H.R. Rep. No. 1312, 90th Cong., 2d Sess. 55 (1968). One acre-foot is 325,851 gallons of water, approximately what a family of four or five consumes in a year.
5 A.R.S. § 45-401 et seq.
-3- COLORADO RIVER BASIN \ WYOMING
Flaming Gorge
Salt Lake City
env4r
UTAH Aspinall Unit
NEVADtk
r■ Clen r I `. ee Ferry" • AlrNavajo r Lake Reservoir C Las Vegas(, Da ■ Davis Dam and Santa Fe Lake Mohave . gep 4 ° Albuquerque ,4 Parker Dam 4 and Lake MatJasstt ARIZ NEW dimi ) *elK u 46 / MEXICO To Southern it *...... / • t wop. California user:s;:r ... S %Phoe 4e.■ V036 NIB. TOT Dam ° Mexicali •••" • Yuma #A. w • Tucson LI .00• WIPER COLORADO RIVER BA= A / 1 \ / LINER COLORADO RP/I2 BASIN No ales Culf of MEXICO California • DAM LOCATIONS Figure 2 southern boundary of the San Xavier Indian Reservation southwest of Tucson, Arizona. [Figure 3] Along the way, fourteen pumping plants lift the Colorado River water 2,900 feet in elevation. Construction of the CAP began in 1973, and water was first delivered to central Arizona in 1985 and the Tucson area in 1991. Reclamation declared the water supply system stage of construction substantially complete in 1993.
In addition to the water supply system, the CAP includes a regulatory storage system which allows Colorado River water transported by the CAP in winter months to be stored in central Arizona to satisfy higher demands by CAP customers during summer months. As originally authorized, the regulatory storage system called for the construction of Orme Dam near the confluence of the Verde and Salt rivers east of Phoenix or a "suitable alternative." Opposition to Orme Dam, which would have inundated much of the Fort McDowell Indian Community and may have affected bald eagle habitat, led Reclamation to consider alternative methods of providing regulatory storage and flood control for the Phoenix metropolitan area. The selected alternative, known as "Plan 6," called for the construction of New Waddell Dam on the Agua Fria River and Cliff Dam on the Verde River, and enlargement of Theodore Roosevelt Dam and modification of Stewart Mountain Dam on the Salt River. Cliff Dam was eliminated in 1987 in response to environmental concerns, and Plan 6 was amended to provide for modification of the existing Bartlett and Horseshoe dams on the Verde River.
The Colorado River Basin Project Act also authorized the construction of three additional dams in the Gila River basin: Hooker Dam on the Gila River in western New Mexico, Buttes Dam on the Gila River about four miles upstream from the existing Ashurst-Hayden Dam, and Charleston Dam on the San Pedro River. These three dams were intended to provide water conservation, sediment detention, flood control, recreation, and fish and wildlife benefits. In addition, an aqueduct was planned to carry San Pedro River water impounded by Charleston Dam to Tucson. Plans for Buttes and Hooker dams have been postponed indefinitely, and Charleston Dam is not expected to be built. As a substitute for the San Pedro aqueduct, the main canal of the CAP was extended about 36 miles to deliver Colorado River water to the San Xavier Indian Reservation.
The CAP was designed and constructed to minimize its effect on the environment. The CAP does not connect to any existing, natural watercourse other than the Colorado River and the Agua Fria River at New Waddell Dam. Where smaller washes intersect the path of the CAP, box or pipe overchutes have been provided to channel the wash over the canal. [Photo 1] Siphons carry CAP water underneath larger streams and rivers, including the Hassayampa, Agua Fria, Salt and Gila rivers. [Photo 2] Elsewhere, dikes have been constructed to prevent surface runoff from flooding the CAP. [Photo 3] Water temporarily impounded by these dikes has created "green-up" areas of vegetation along the upslope side of the aqueduct, improving some existing wildlife habitat.
-4- LAKE MEAD
Q.) NAVAJO Grand Canyon COUNTY
COCONINO COUNTY APACHE COUNTY
Aahfork Flagstaff
Winslow MOHAVE COUNTY YAVAPAI Holbrook 0 COUNTY LAKE HAVASU
N\ HAVASU s RN: _!_ .Joary Parker -It 0 Payson Show Low 0 ROUSE HILLS Springer/1110 0 Wickenburg GILA COUNT( LAKE PLEASANT HAS SAYAMPA ROOSEVELT c? LA PAZ LITTLE /p LAKE er COUNTY HAROU \FIALA
Globe GREENLEE COUNTY GILA 134,._
• GIle Bend Florence
BRADY 0 Safford Yuma PINAL PICACHO MARICOPA COUNTY COUNTY 0- YUMA COUNTY GRAHAM RED ROCK COUNTY
TWIN PEAKS SAN DARIO Tucson PIMA • BRAWLEY SNYDER HILL COUNTY SAN XAVIER BLACK MTN. Benson COCHISE COUNTY
4. Pumping Plant SANTA CRUZ COUNT( • City (9 Naples
CENTRAL ARIZONA PROJECT
Figure 3 Photo 1: Box overchute to channel wash over CAP aqueduct
Photo 2: Construction of CAP siphon under Salt River Photo 3: Flood protection dike adjacent to CAP aqueduct
Photo 4: Wildlife crossing over CAP aqueduct Numerous features were built into the CAP to protect wildlife that might be attracted to the aqueduct for water:
. Eight-foot high fences were constructed on both sides of the CAP for more than 200 miles to keep out desert mule deer, bighorn sheep, javelina and other large animals.
. The top five feet of the concrete lining in the aqueduct has a rough finish so that smaller animals that get through the fence can climb down to the water for a drink and climb back out safely.
. Ramps were constructed in smaller distribution canals so wildlife that fall in the canal while drinking can escape.
Radio telemetry was used to identify important movement paths that intersect the CAP, and specially designed wildlife crossing bridges were constructed at those locations. [Photo 4]
Drainage structures along the CAP were modified for animal use.
Underpasses were built on new roads in key movement areas so that deer and other wildlife would not encounter vehicle traffic.
In addition, Reclamation provided funding for construction of 45 wildlife watering sites located away from the CAP and new roads. It also purchased 4.25 square miles of desert habitat for a deer corridor between the Tucson Mountains and the Schuk Toak District of the Tohono O'odham Nation west of Tucson and buried parts of the CAP aqueduct in this area to preserve the habitat and minimize disruptions to deer, kit fox, desert tortoise, Gila monster and other wildlife.
Reclamation also took steps to protect rare plant life during construction of the CAP. Nurseries, municipalities and individuals were encouraged to salvage cacti and other state-protected native plants before native vegetation was removed for construction. Reclamation also transplanted 400 endangered Tumamoc globe-berry plants that would have been destroyed by CAP construction, purchased 220 acres of existing Tumamoc globe-berry habitat, and funded searches for the plant elsewhere in Arizona and northern Mexico. Those searches revealed that the Tumamoc globe-berry was much more common than previously thought, and the plant subsequently was removed from the endangered species list.
It has been estimated that the United States will spend more than $4 billion to complete the CAP. Much of that cost must be reimbursed by project users. CAWCD was organized as a political subdivision of the State of Arizona for the express purpose of contracting with the United States for repayment of reimbursable CAP costs and
-5- subcontracting with non-Indian water users for delivery of CAP water. To assist in repayment of CAP costs, CAWCD is empowered to levy an ad valorem tax on taxable property within its three-county service area, which includes Maricopa, Pinal and Pima counties. CAWCD's share of CAP costs will approach $2 billion.
CAWCD is also responsible for the operation and maintenance ("O&M") of CAP facilities that have been transferred to CAWCD by Reclamation. With the exception of certain siphons that are still undergoing repair and replacement, Reclamation has transferred O&M responsibility for the entire length of the CAP aqueduct to CAWCD.
B. The Biological Opinion Contradicts Previous Section 7 Consultations on the CAP.
The Biological Opinion represents a fundamental change in the Service's position with respect to the Central Arizona Project. Reclamation consulted with the Service about the potential impact of the CAP on threatened and endangered species more than forty times during the design and construction of the CAP and its associated water delivery systems. In most instances, the Service indicated that the CAP would have no effect on protected species. In the few instances where the Service voiced concern, Reclamation promptly addressed the problem to the Service's expressed satisfaction. The Biological Opinion contradicts the conclusions previously expressed by the Service, without offering any explanation or justification for the contradictions.
Following formal consultation with Reclamation over the proposed construction of New Waddell Dam, which increased the size of Lake Pleasant, the Service issued a biological opinion that found no jeopardy to Gila topminnow populations in the Agua Fria River basin, but recommended construction of a fish barrier as an optional conservation measure.6 Even though Reclamation adopted the conservation measure and constructed the barrier, the Biological Opinion now concludes that CAP water deliveries will jeopardize the Gila topminnow in the Agua Fria River basin above Lake Pleasant. The Biological Opinion did not mention the earlier CAWCS Biological Opinion, much less explain the reason for the different conclusion.
In 1982 the Service was asked "to identify and document problems concerning importation of Colorado River ichthyofauna into central Arizona river systems through CAP facilities," and opined that certain fish transported through the CAP might effect bald eagle forage in the lower Salt and Verde rivers.' To assuage those concerns, the interconnection between the CAP and the Salt River Project ("SRP") was redesigned so that CAP water would be introduced directly into SRP canals below Granite Reef Diversion Dam and electrical fish barriers were installed on the SRP canals between the
6 USFWS, Biological Opinion, Central Arizona Water Control Study, at 8-9 (3/8/83) ("CAWCS Biological Opinion").
Field Supervisor, Ecological Services, USFWS, letter to Reclamation (10/21/82) (only striped bass, blue tilapia and white bass are species of concern); see Preliminary Draft Fish & Wildlife Coordination Act and Habitat Evaluation Report (12/14/82) (same).
-6- CAP/SRP interconnection and the dam, thus insuring that fish from the CAP could not enter the lower Salt and Verde Rivers above Granite Reef Diversion Dam. [Figure 4] The Service reviewed and approved the new design:
"Our major concern relative to the proposed project has been and continues to be the introduction of new aquatic organisms into the Salt and Verde River systems. The decision to deliver water directly into the SRP canals below Granite Reef Diversion Dam, and to construct fish barriers in
the canalsu8 to prevent upstream movement of fish, adequately satisfies this concern.
Ignoring the Service's previous representations, the Biological Opinion concludes (at 18- 19) that delivery of CAP water into SRP canals below Granite Reef Diversion Dam and downstream of the electrical fish barriers will jeopardize native fish and adversely modify critical habitat in the upper Salt and Verde rivers. No new scientific or commercial data was offered to support the Service's changed view.
The Service also conducted Section 7 consultations with irrigation districts during the design and construction of their CAP water delivery systems. With Harquahala Valley Irrigation District ("HVID") and Tonopah Irrigation District ("TID"), for example, the Service stated that the delivery systems being built to deliver CAP water were not expected to have any adverse effects on endangered or threatened species.° The Biological Opinion now concludes (at 17), however, that CAP deliveries to HVID and TID will jeopardize endangered fish in the upper Hassayampa River basin. Again, the Service identified no new scientific or commercial data that might explain the change in its opinion.
The Service did not raise the potential impact of CAP deliveries on endangered fish until the Pima Lateral Feeder Canal was proposed to allow interim CAP deliveries to the Gila River Indian Community ("GRIC") in 1990.10 Although it made no mention of the issue in reviewing a draft environmental assessment in 1986, the Service told Reclamation in 1989 that it was concerned about "the introduction through the Feeder
Field Supervisor, Ecological Services, USFWS, letter to SRP (11/16/84) ("We do not envision any potential environmental impacts associated with the proposed interconnection facilities"); see Field Supervisor, Ecological Services, USFWS, letter to SRP (10/28/87) (same).
9 E.g., Area Manager, USFWS, letter to Franzoy, Corey and Associates (9/14/81) (stating that "no listed or proposed species would be affected by the proposed action in the area of interest [HVID1"); Field Supervisor, Ecological Services, USFWS, memorandum to Reclamation (4/23/84) (Fish and Wildlife Coordination Act Report for HVID, finding "no aquatic resources within the HVID that will be affected by project construction"): Field Supervisor, Ecological Services, USFWS, memorandum to Reclamation (9/13/84) ("no listed, proposed or candidate species would be affected by the proposed action in the area of interest [TID]"); Field Supervisor, Ecological Services, USFWS, memorandum to Reclamation (1/17/85) (Fish and Wildlife Coordination Act Report for TID, finding "no aquatic resources within the TID that will be affected by project construction").
0 1 Ironically, the reason for making these interim CAP deliveries was to protect non-native fish in San Carlos Reservoir on the Gila River upstream of Ashurst-Hayden Diversion Dam. Field Supervisor, Ecological Services, USFWS, memorandum to Reclamation (8/28/89), at 1 (in exchange for receiving 30,000 acre-feet of CAP water, GRIC agreed to leave 30,000 acre-feet of water in San Carlos Reservoir "to maintain a minimum pool for the fishery").
-7- SALT RIVER CAP/SRP A 11120 F P CANAL GRANITE REEF R, INTERCONNECTION DMA X^I r
FORE DAY FISI DARRIER
4, Arlin:MA ARIZONA CANAL - 4. SIPHON DELIVERY '4, (114") VALVE \ 00 CFS) STRUCTURE —
'".■••■•
FISI I DARRIER uSIM FLOW 1 200 CFS / RADIAL \ CO1,11,1011 NIVERDED GATES CANAL DELIVERY (1200 CFS)
t ° .1 1.1 " WEIR .WE1 R
SOUTH 84" PIPE- CANA' -- SOUTH N (1200 CFS) CANAL DELIVERY NN (BOO CFS) USDR TURNOUT EXISTING ... WEIR CAI SIPHON A0111()I/C1-----■ Canal of fish species not native to or not currently found in the Gila River drainage, specifically the stripedfill bass ( Morone saxitilis) [sic], white bass ( Morone chrysops) and tilapia ( Tilapia spp.). Reclamation agreed to study the issue, and the Service specifically limited the scope of that study to species that were "not currently in the San Pedro drainage or in San Carlos Reservoir," stating that it was "not aware of any species outside of striped bass, white bass and tilapia that need to be addressed."12
Despite the Service's repeated statements that the only non-native fish of concern for the CAP were those not already present in the San Pedro River system or San Carlos Reservoir, the Biological Opinion concludes that CAP water deliveries would cause jeopardy if any non-native fish might escape into the Gila River drainage. Thus, the Biological Opinion identifies (at 15) at least thirty-three species of non-native fish that it suggests could present a threat to endangered fish in the upper Gila River basin. Only fourteen of those thirty-three species had been reported in CAP canals, and all but two of those fourteen species (redear sunfish and striped bass) were already found in the San Pedro River or San Carlos Reservoir. As before, the Service did not explain its change in position.
Perhaps the most critical flaw in the Biological Opinion is that it does not even attempt to perform a serious scientific analysis of the issues presented. Instead, the Biological Opinion simply assumes--without regard to available scientific or commercial data--that non-native fish transported through the CAP would escape into local waters, travel great distances through often-dry riverbeds to reach native fish habitat, and negatively impact endangered species. In fact, the Biological Opinion expressly avoids any detailed examination by asserting that a "[s]pecies-specific analysis of the likelihood of non-native species entering and successfully establishing in the Gila River basin, and the potential for impacts to each listed species, would be lengthy, complex, and largely nonproductive."13 Absent such species-specific analysis, the Biological Opinion's jeopardy finding is nothing more than uninformed speculation.
In determining whether an action will jeopardize an endangered species or adversely modify critical habitat, it is imperative that the Service biologist evaluate the particular characteristics of individual species in light of the prevailing conditions that they may expect to face, including water temperature and flow regimes. For example, the Service concluded that rainbow trout to be stocked in the Colorado River would not affect endangered razorback sucker and bonytail chub downstream in Lake Havasu because "the rainbow trout in the Davis Dam to Needles area do not move out of the area due to high temperatures below the Interstate 40 bridge near Topock" and "Lake
Field Supervisor, Ecological Services, USFWS, memorandum to Reclamation (8/28/89).
12 Acting Field Supervisor, Ecological Services, USFWS, memorandum to Reclamation (3/20/90).
1 3 Biological Opinion, at 21 (emphasis supplied).
-8- Havasu is too warm to allow the survival of rainbow trout over the summer."14 Similar analysis here would have shown that non-native species that may be found in the CAP cannot survive in many areas in the Gila River basin long enough to establish self- sustaining populations. Nevertheless, the Biological Opinion did not take into account available data on the hydrology of the Gila River basin and the biology of the fish found in the CAP in concluding that those fish would be able to reach native habitats and jeopardize native species.
The Biological Opinion also uniformly assumed that any non-native fish reaching native habitats would adversely effect endangered species. In other instances, however, the Service has required "direct evidence" of predation by each non-native species of concern before finding jeopardy or adverse modification of critical habitat. For example, the Service concluded that stocking rainbow trout in the Colorado River below Hoover Dam would not adversely modify critical habitat for the bonytail chub because "[r]ainbow trout predation on bonytail chub has not been documented," even though the Service noted elsewhere in the same opinion that rainbow trout are known to prey on humpback chub, razorback sucker and other species:6 Likewise, the Service determined that rainbow trout stocking will not adversely modify critical habitat for the razorback sucker in Lake Mohave because rainbow trout are only known to prey on razorback larvae and that critical habitat designation was intended only to protect adult razorback suckers in Lake Mohave:6 The Biological Opinion, on the other hand, did not require any proof of predation or competition by particular non-native fish on individual endangered species before finding that CAP water deliveries would cause jeopardy and adverse modification of critical habitat.
C. The Endangered Fish.
The Biological Opinion found jeopardy as to four threatened or endangered fish: the spikedace, loach minnow, Gila topminnow, and razorback sucker.
1. Spikedace ( Meda fulgida).
The spikedace is a small, slender fish rarely exceeding 70 millimeters (less than 3 inches) in length. [Photo 5] It has very silvery sides, an olive-gray to brownish back, white underside, and spines in the dorsal and pelvic fins. The spikedace spawns in spring to early summer in response to changes in stream flow and water temperature. Younger females spawn once, while older females may spawn twice each year. The spikedace typically lives only one to two years, feeding primarily on drifting aquatic and terrestrial insects and occasionally the larvae of other fish.
USFWS, Biological Opinion on Fish and Wildlife Stocking of Rainbow Trout and Channel Catfish in the Lower Colorado River (Hoover Dam to the International Border), July 1, 1994 ("FWS Opinion"), at 14.
15 Id. at 15, 13 ("No direct evidence of predation by rainbow trout on bonytail is available").
16 Id. at 15.
-9- Photo 5: Spikedace (Meda fulgida)
Photo 6: Loach minnow (Tiaroga cobitis) CENTRAL ARIZONA PROJECT AQUEDUCT SYSTEM AND POTENTIAL LOCATIONS OF SPIKEDACE
Legend
Spikedace Critical Habitat
Parker
orselthoe em
ett River Sa1t , HOENIX Gila we'
Gi16 Yuma
yo la; *Tucson
Perennial/Regulated Streams Intermittent/Ephemeral Streams Central Arizona Project Aqueduct
Figure 5 Spikedace are found in moderate to large perennial streams, where they inhabit shallow flowing waters, usually less than one meter deep. Adults often congregate in shear zones along gravel-sand bars and broad shallow areas above gravel-sand bars, as well as quiet eddies on the downstream edge of riffles. Smaller, younger fish occupy quiet water along pool margins over soft, fine-grained bottoms. In springtime, spikedace may use shallower waters with swift currents, often over a sand or gravel substrate. Recurrent flooding is very important in maintaining the habitat of the spikedace and also helps it maintain a competitive edge over invading non-native fish species.
The spikedace was identified as a Category 1 candidate species in December 1982 and was listed as a threatened species on July 1, 1986. The Service published a recovery plan for the spikedace in 1991. Presently, the spikedace is found only in Aravaipa Creek, isolated stretches of the upper Gila River in New Mexico, the upper Verde River in Arizona, and Eagle Creek, a tributary of the upper Gila River in eastern Arizona. Critical habitat designated on March 8, 1994, includes 15 miles of Aravaipa Creek, 35 miles of the upper Verde River, and several reaches of the Gila River and its tributaries in New Mexico. [Figure 5]
2. Loach Minnow ( Tiaroga cobitis).
The loach minnow is a small, stream-dwelling member of the minnow family. [Photo 6] This slender, secretive fish is generally less than 3 inches long. The loach minnow is olive colored with blotches of darker pigment. During breeding season, males become brightly pigmented, with colors ranging from blood red to brilliant orange on the lips, fins, and bodies, while females develop yellow pigments on the fins and body. Loach minnow spawn in late winter to early summer, depending on the location. Spawning takes place in riffles where adhesive eggs are deposited on the underside of flattened rocks. Loach minnow feed on insect larvae found among the bottom substrate, particularly mayfly and blackfly larvae. Most loach minnow live no more than 2 years.
The loach minnow is found in small to large perennial streams, using shallow (less than 15 cm deep), turbulent riffles with primarily gravel or cobble substrate and swift currents, sometimes in association with dense beds of coarse, filamentous algae. Because the loach minnow has a reduced gas bladder, it spends most of its life on the stream bottom. Its body density is such that midwater swimming is labored, and the loach minnow swims in swift water only for brief moments while moving from place to place. Recurrent flooding is important to loach minnow survival because it keeps the substrate free of embedding sediments and also helps the fish maintain a competitive edge over invading non-native species.
The loach minnow was identified as a Category 1 candidate species in December 1982 and was listed as a threatened species on October 28, 1986. A recovery plan for the loach minnow was published in 1991. Loach minnow are presently known only in Aravaipa Creek, isolated stretches of the upper Gila River and certain of its tributaries in New Mexico and eastern Arizona, and the east fork and mainstem of the White River
-10- CENTRAL ARIZONA PROJECT AQUEDUCT SYSTEM AND POTENTIAL LOCATIONS OF LOACHMINNOW
Legend
Loachminnow Critical Habitat
Perennial/Regulated Streams Intermittent/Ephemeral Streams - — Central Arizona Project Aqueduct
Figure 6 in the upper Salt River basin. On March 8, 1994, the Service designated as critical habitat for the loach minnow 15 miles of Aravaipa Creek (identical to spikedace critical habitat) and several reaches of the upper Gila River and its tributaries in eastern Arizona and western New Mexico. [Figure 6]
3. GHa topminnow ( Poeciliopsis occidentalis occidentalis).
The Gila topminnow is a very small live-bearing fish. Males are seldom longer than 25 mm (about 1 inch), and females average 30 to 40 mm. It has a tan to olive body, usually with white on the belly. [Photo 7] Gila topminnow generally live near the surface in shallow water. They prefer shallow, warm, fairly quiet waters, but can adjust to moderate currents, greater depths, and fluctuating temperatures. Topminnow may live in springs, cienegas, marshes, and permanent or intermittent streams, and prefer areas with dense mats of algae, vascular plants and debris, usually along stream margins or below riffles, with sandy substrates sometimes covered with organic muds and debris. Gila topminnow feed on small invertebrates such as mosquito larvae, microscopic plants, and organic debris.
The average lifespan of the Gila topminnow is about 1 year, but depends on the breeding cycle. Both fertilization and embryo development occur internally. The female carries two broods simultaneously, one being further developed than the other. Brood size varies from 1 to 20 and the gestation period is 24 to 28 days. Young topminnow can swim, feed and avoid predators a few seconds after birth.
Nine naturally occurring populations of Gila topminnow remain. Eight of these are in the Santa Cruz River basin and one is found in springs near Bylas on the San Carlos Indian Reservation. In addition, Gila topminnow have been stocked in nearly 200 other locations in Arizona, and some of these populations have persisted for several years.17 [Figure 7]
The Gila topminnow was listed as an endangered species on March 11, 1967.18 No critical habitat has been designated for the Gila topminnow. The Service published a recovery plan for the Gila topminnow in 1984, and a revised plan is currently being developed.
4. Razorback sucker ( Xyrauchen texanus).
The razorback sucker was originally proposed for listing as a threatened species on April 24, 1978. That proposal was withdrawn on May 27, 1980, because the Service did not complete the listing process within 2 years as required by the 1978 amendments
17 AGFD, Status of the Sonoran Topminnow ( Poeciliopsis occidentalis) and Desert Pupfish ( Cyprinodon macularius) in Arizona (1991) ("AGFD Status Report"), at 6 & App. A.
18 32 Fed. Reg. 4001 (1967). Photo 7: Gila topminnow (Poeciliopsis occidentalis occidentalis)
Photo 8: Razorback sucker (Xyrauchen texanus) CENTRAL ARIZONA PROJECT AQUEDUCT SYSTEM AND POTENTIAL LOCATIONS OF GILA TOPMINNOW
Legend 400( Gila Topminnow
Q•J; C.)Or
Perennial/Regulated Streams --- Intermittent/Ephemeral Streams - Central Arizona Project Aqueduct
Figure 7 to the Endangered Species Act. The species was subsequently proposed for listing as endangered on May 22, 1990, and the listing became final on October 23, 1991. 56 Fed. Reg. 54957.
The adult razorback sucker is identified by an abrupt sharp-edged dorsal keel behind its head and a large fleshy subterminal mouth that is typical of most suckers. [Photo 8] Adult fish often weigh more than 6 pounds and are more than 2 feet long. Historically, the razorback sucker occupied waters with strong, uniform currents over sandy bottoms. It also lived in eddies and backwaters lateral to river channels, sometimes concentrating in deep places near cut banks or fallen trees. The razorback sucker feeds mostly from the bottom on plankton, algae, insects and decaying organic matter. The species is fairly sedentary except during spawning runs, often moving only a few kilometers over several months.
In the Lower Colorado River basin, razorback suckers have been observed spawning only in Lake Mohave, typically from late January through April in relatively shallow water with a coarse sand, gravel or cobble substrate that is swept free of silt by currents or wave action. The razorback sucker's eggs are adhesive and are deposited in spaces between gravels. The eggs hatch in a few days and the young move to low velocity shoreline habitat for a time. Little is known about the habits of larvae and juveniles because there is little or no recruitment in naturally occurring razorback sucker populations.
The razorback sucker was extirpated from all rivers in the Lower Colorado River basin except the mainstem Colorado River. A sizable adult population remains in Lake Mohave, and the species is sometimes spotted upstream in Lake Mead and the Grand Canyon or downstream on the Colorado River. In the 1980s, the Service and the Arizona Game and Fish Department began to introduce hatchery-raised razorback sucker into the Verde, Salt and Gila rivers in an attempt to reestablish viable populations. Those efforts have been largely unsuccessful, and most of the introduced fry were believed to have been consumed by predatory fish already established in those waters. Flathead catfish were found to be the principal predator of razorback sucker stocked in the Gila River, and flathead catfish are also found in the Salt and Verde river systems. Carp, green sunfish, channel catfish, threadfin shad, smallmouth bass, largemouth bass, bluegill and redear sunfish are also known or suspected to prey on razorback sucker eggs, larvae and fry. All of those non-native species are already established in the Salt and Verde river system.19
Critical habitat for the razorback sucker was designated on March 21, 1994, and includes several reaches of the Colorado River, as well as the Gila River from the Arizona-New Mexico border to Coolidge Dam (including San Carlos Reservoir to full pool
1 9 Hendrickson, Evaluation of the Razorback Sucker ( Xyrauchen texanus) and Colorado Squawfish ( Ptychocheilus lucius) Reintroduction Programs in Central Arizona Based on Surveys of Fish Populations in the Salt and Verde Rivers from 1986 to 1990, at 44 (1993).
-12- CENTRAL ARIZONA PROJECT AQUEDUCT SYSTEM AND POTENTIAL LOCATIONS OF RAZORBACK SUCKER
Legend
410( Razorback Sucker Critical Habitat
— Perennial/Regulated Streams Intermittent/Ephemeral Streams Central Arizona Project Aqueduct
Figure 8 elevation), the Verde River from the boundary of Prescott National Forest near Paulden to Horseshoe Dam (including Horseshoe Reservoir to full pool elevation), and a reach of the Salt River above Roosevelt Lake. [Figure 8]
D. Non-Native Fish in the CAP and Gila River Basin.
The Biological Opinion identified (at 15) 33 species that it claimed have been reported in the CAP and "interconnected" Salt River Project ("SRP") canals. The majority of those species, however, are found only in the SRP system, and contrary to the Biological Opinion's assumption, fish in SRP canals cannot enter the CAP aqueduct. The SRP turnout on the CAP is located just south of the Salt River siphon at an elevation of 1479 feet above sea level. [See Photo 9] At the turnout, CAP water is delivered into a 96-inch diameter concrete pressure pipe that extends for 200 feet downhill to a concrete bulkhead where it joins a 400-foot long, 84-inch concrete pressure pipe built by SRP. The SRP pipe terminates at an elevation of 1330 feet where it empties into a common canal that channels CAP water to the two SRP canals below Granite Reef Diversion Dam. For a fish to travel from one of the SRP canals to the CAP, it would have to swim up through a canal delivery drop structure, pass over a broadcrested weir in the common canal, then climb through 600 feet of concrete pressure pipe having an average slope of 25 percent and a water velocity up to 15.59 feet per second. No fish known to exist in the SRP system is capable of making that journey.
The Biological Opinion also states (at 16) that non-native fish species present in Lake Pleasant will be able to enter the CAP aqueduct when water stored in the lake is released into the canal. That conclusion ignores key facts about the operation of Lake Pleasant and the physical features of New Waddell Dam. CAWCD uses Lake Pleasant as a regulatory storage facility for the CAP by pumping water up from the Colorado River and into the lake during winter months when demand and pumping costs are both relatively low, then releasing water from Lake Pleasant into the CAP during the summer to meet the high demand for water while avoiding higher pumping costs. As a result, water is typically released from Lake Pleasant only from May to October. During those months, Lake Pleasant is highly stratified and dissolved oxygen levels in the hypolimnion make that region anoxic and avoided by fish.2° Because the point at which water is released from Lake Pleasant into the CAP occurs below the upper boundary of the hypolimnion from May to October, there is little likelihood that fish will leave Lake Pleasant and enter the CAP aqueduct. In addition, small impeller clearances in the generators on the outlet works at New Waddell Dam will result in larger fish being crushed or mangled.
The CAP aqueduct system has been surveyed extensively to determine the types and quantities of non-native fish that might be present. As shown in Table 1, fourteen
20 AGED, Phase I: Baseline Limnological and Fisheries Investigation of Lake Pleasant (1990), at 80.
-13- Photo 9: CAP turnout (top) for delivery of water to SRP canals (bottom) species have been reported in the CAP between Lake Havasu and the Brady Pumping Plant, 22 miles south of the Gila River.
Common Name Scientific Name
Threadfin shad Dorosoma petenense I Largemouth bass Micropterus salmoides Bluegill Lepomis macrochirus Green sunfish Lepomis cyanellus I Redear sunfish Lepomis microlophus Black crappie Pomoxis nigromaculatus Striped Bass Morone saxatilis I Channel catfish Ictalurus punctatus Flathead catfish Pylodictis olivaris Ili Yellow bullhead Ameiurus natalis Common carp Cyprinus carpio Goldfish Carassius auratus Mosquitofish Gambusia affinis Red shiner Cyprinella lutrensis
Table 1: Non-Native Fish Reported in CAP
As expected, the non-native species reported in the CAP mirror those found in Lake Havasu, where the CAP intake is located.21
The CAP aqueduct system is not an ideal fish habitat. Because the canal is concrete lined, there are few nutrients available to provide food sources for fish. To move through the CAP, fish must pass through a series of pumping plants without being crushed or mangled along the way. Impeller clearances in CAP pumping plants between Lake Havasu and Lake Pleasant range from 3.5 to 8.1 inches, thus limiting movement by larger fish.22 Larger fish would also have difficulty leaving the CAP because turnout grates have only a 2-inch clearance. Temperatures in the CAP can be uncomfortable or even lethal to some fish. Summertime canal temperatures regularly exceed 25°C in some areas and can often reach 30°C. In short, the CAP system may be incapable of supporting many fish species and certain fish may be physiologically unable to cope with the year-round environment present in some reaches of the CAP aqueduct.
These rather inhospitable conditions may help explain certain trends that have appeared in CAP fish survey data. CAP fish surveys have been conducted since 1987 by electrofishing in the forebays of pumping plants. Electrofishing is an accepted
21 Minckley (1979) USFWS, et al, Lake Havasu Fishery Study (1988).
22 Grabowski, Hiebert & Lieberman, Potential for Introduction of Three Species of Nonnative Fishes into Central Arizona via the Central Arizona Project (1984) ("Grabowski"), at 84.
-14-
Catch per Unit Effort, Electrofishing fish/min in the CAP system, 1987-1992 18
16- CAP Pumping Plants (upstream to downstream) 14- Bouse Hillš 12- Lithe Harquahala 2 1 0- Hassayampa
(-13 8- Salt-Gila
Brady - 6 1 4-
2-
0 1987 1988 1989 1990 1991 Year sampling method used by fisheries biologists to determine species composition and to provide a relative idea of spatial and temporal changes in the fish community. While electrofishing alone does not provide an accurate fish count, it is often used in conjunction with tagging and tag recovery to estimate population size.
Available survey data for 1987 through 1992 shows at least two distinct patterns, neither of which was considered in the Biological Opinion.
First, there has been a significant decline in the number of fish caught at all pumping plants. [Figure 9] The average number of fish per minute (fpm) of electrofishing dropped from approximately 8.56 fpm in 1987 to about 0.7 fpm in 1992. This 92 percent drop in electrofishing catches is indicative of an absolute drop in numbers of fish within the CAP, which translates into a significant reduction in the likelihood of fish being present in water delivered to CAP subcontractors. Moreover, the electrofishing surveys were conducted in the forebays of CAP pumping plants, where fish tend to congregate, so the actual density of fish at CAP turnouts is expected to be even lower. As a result, the probability is small that significant quantities of fish will be diverted from the CAP, and even smaller that significant quantities of a particular species will escape. This makes it highly unlikely that non-native fish from the CAP would be able to establish self-perpetuating populations in native fish habitat in the Gila River basin.
Second, survey results indicate a significant difference in fish composition from one pumping plant to another. Striped bass have been captured in the upper reaches of the CAP aqueduct, but recent surveys have recorded few striped bass downstream of the Hassayampa Pumping Plant. [See Figure 3] Carp are most likely to be found in the middle reaches of the CAP from Hassayampa Pumping Plant to Brady Pumping Plant, but are uncommon above and below that area. In the lower reaches of the CAP, and particularly downstream of Brady Pumping Plant, bluegill is the most common species. These data have a strong bearing on which non-native fish are most likely to be present in CAP water delivered to a given subcontractor or area. In the area of the Salt and Gila rivers, for example, the most likely non-native species to be present in the CAP are carp and bluegill, not striped bass.
Carp and bluegill are common throughout the Gila River basin. In fact, most of the non-native fish reported in the CAP, as well as many other species, were established in the Gila River basin long before any CAP water was delivered, as shown in Table 2.23
23 AGED, Phase I: Baseline Limnological and Fisheries Investigation of Lake Pleasant (1990); Matter, Potential for Transfer of Non-Native Fish in Central Arizona Project Canal Waters to the Gila River System (1991); Jackson, Martinez et al, Assessment of Water Conditions and Management Opportunities in Support of Riparian Values: BLM San Pedro River Properties, Arizona (1987); Jakle, Memorandum to APO-150 Files (2/26/92); Jakle, Memorandum to APO-150 Files (2/10/93); SRP, CAP/SRP Interconnection, Final Environmental Assessment Report (1988); Reclamation, Final Environmental Assessment, 1990 Interim Water Deliveries of CAP Water to San Carlos Irrigation and Drainage District (1990); Kepner, Jakle & Baucom, Middle Gila River Aquatic Study (1983); USFWS Stocking Records; AGFD Stocking Records; Reclamation, Biological Assessment for the Spikedace (Meda fulgida), Central Arizona Water Exchange Project (1988); Hendrickson, Evaluation of the Razorback Sucker ( Xyrauchen texanus) and Colorado Squawfish ( Ptychocheilus lucius) Reintroduction Programs in Central Arizona Based on Surveys of Fish Populations
-15- Species Lake Verde Lower Salt Salt River Middle San Carlos San Pedro Aravaipa (• = in CAP) Pleasant River River Lakes Gila River Reservoir River Creek
• Green X X X X X X X X sunfish
• Threadfin X X X X X X X shad
• Bluegill X X X X X X X
• Common X X X X X X X carp
• Black X X X X X crappie
• Channel X X X X X X X catfish
• Large- X X X X X X X X mouth bass
• Flathead X X X X X catfish
• Redear X X sunfish
Rainbow X X X X trout
• Goldfish X X X
• Red shiner X X X X X X X X
• Yellow X X X X X X X bullhead
• Mosquito- X X X X X X X X fish
Blue tilapia X X X
Fathead X X X X X minnow
Walleye X X
Black X X X bullhead
Table 2: Non-Native Fish in the Gila River Basin (Before CAP)
Many of the non-native fish found in the Gila River basin were intentionally stocked there. Since 1928, the Arizona Game and Fish Department ("AGFD") has stocked nearly 28 million non-native fish into the Gila River basin.24 Much of this stocking activity was supported by federal funds, yet until April 1994, the Service did not conduct Section 7 consultations with respect to non-native stockings by state game and fish departments.
in the Salt and Verde Rivers from 1986 to 1990 (1993); W.E. Barber & W.L. Minckley, Fishes of Aravaipa Creek, Graham and Pinal Counties, Arizona, 11 Sw. NATuRALiST 313 (1966).
24 AGFD Stocking Records.
-16- The Service itself has stocked large numbers of non-native fish into the Gila River basin over the years. Moreover, at the same time it was telling Reclamation that CAP water deliveries jeopardized endangered species because an occasional non-native fish might be able to escape the CAP aqueduct and enter the Gila River basin, the Service continued to stock non-native fish directly into native fish habitat in the region upstream of all CAP delivery points. From 1990 to 1993, the Service stocked 2,546,902 nonnative fish into the Gila and Salt River drainages, apparently without issuing any biological opinion on the effects of that activity.
The Hassayampa River Sub-basin.
The Hassayampa River originates in the Sierra Prieta Mountains in the Prescott National Forest, south of Prescott, Arizona, and flows southward through Wickenburg to the Gila River about 37 miles west of Phoenix. For most of its length, the river is ephemeral, flowing only in response to storm events. The Hassayampa receives runoff from the west slopes of the Bradshaw Mountains and the east slopes of the Weaver Mountains and the Sierra Prieta. From its origin to Morristown, just south of Wickenburg, it flows through mostly rocky terrain where rainfall is often intense, but usually covers only small areas. Because of steep slopes in the area, water concentrates quickly and runs off at a high velocity. In contrast, from Morristown to the Gila River the Hassayampa River traverses a flat sandy plain where flows are infrequent. For 38 miles between Morristown and Arlington, the Hassayampa River is normally dry. [Photo 10]
Centennial Wash is an ephemeral stream that joins the Gila River about six miles downstream from the Gila's confluence with the Hassayampa River. Centennial Wash originates in southwest Yavapai County in the Date Creek Mountain area and flows southwest between the Harcuvar and Harquahala Mountains into La Paz County near Salome, Arizona. From there it turns southeast and flows into Maricopa County to the Gila River. For most of its 110 miles, Centennial Wash runs through a wide flat valley with gentle slopes rising from the valley floor to higher rocky regions and mountains which define the drainage area. In many areas the wash does not show a well-defined channel.
The Gila River in the area of the Hassayampa River and Centennial Wash is largely perennial due to effluent discharges from City of Phoenix wastewater treatment plants. Flows in this reach are also affected by several agricultural users. Upstream of the Hassayampa River, the Buckeye Irrigation Company ("BIC") maintains a diversion dike in the Gila River and diverts water through a headgate structure into the BIC Canal. [Photo 11] Tailwater from the BIC Canal is released into the Hassayampa River just upstream of its confluence with the Gila [Photo 12], causing perennial flow for about 2.8 miles, from just above the USGS gauging station near Arlington to the Gila River confluence [Photo 13].
-17- Photo 10: Hassayampa River near CAP aqueduct, north of Gila River
Photo 11: Buckeye Irrigation Company Canal dike in Gila River channel Photo 12: Tailwater discharge from BIC Canal into Hassayampa River
Photo 13: Hassayampa River at USGS gauge near Arlington The Arlington Canal Company also maintains a diversion dike and canal heading in the Gila River. [Photo 14] The Arlington Canal crosses Centennial Wash and discharges its return flow at Gillespie Dam, about two miles below Centennial Wash on the Gila River. [Photo 15] Constructed in 1921 to divert water for agricultural use, Gillespie Dam was breached by flood waters in January 1993 and has not been rebuilt. [Photo 16] Painted Rock Dam, an earthen flood control structure built by the Corps of Engineers in 1959, is located about thirty-six miles downstream from Gillespie Dam. Although Painted Rock Dam can impound up to 2,492,000 acre-feet of water, it's reservoir is usually dry except after flood events. [Photo 17]
The Biological Opinion contends (at 4, 17) that CAP water deliveries to irrigation districts in west central Arizona could allow non-native fish from the CAP to enter Centennial Wash, from where they could travel downstream to the Gila River, make their way to the Hassayampa River, then swim more than forty miles up the Hassayampa River to jeopardize endangered Gila topminnow that have been stocked in isolated springs and ponds in the upper Hassayampa basin. [See Map 1] That conclusion is contrary to the best available scientific and commercial data.
A. CAP Water Deliveries Cannot Jeopardize Endangered Fish in the Hassayampa River Sub-basin.
The CAP cannot jeopardize endangered species in the upper Hassayampa River basin because the endangered fish in that region are totally isolated from the area where CAP water deliveries are made. In authorizing a nonessential experimental population of woundfin in the upper Hassayampa River, the Service stated that the fish in the Hassayampa River were "totally isolated" from existing populations on the Colorado River and elsewhere because of upstream travel distance, mainstream dams and stretches of dry streambed.25 Indeed, the Service's regulations dictate that experimental populations may only be created where the newly introduced fish are "wholly separate geographically from nonexperimental populations of the same species."26 The same upstream travel distances and dry streambed reaches that are sufficient to maintain "total isolation" for nonessential experimental populations of endangered fish introduced by the Service in the upper Hassayampa River also insure that non-native fish from the CAP cannot invade that region.
According to the Biological Opinion, non-native fish could escape the CAP through water delivered to Harquahala Valley, Tonopah, Buckeye and Roosevelt irrigation districts. However, two of those districts--Buckeye and Roosevelt--have not executed CAP subcontracts and will not receive CAP water in the foreseeable future, if ever. CAWCD has no plans to make water deliveries to either of those entities. Only
25 50 C.F.R. § 17.84(b).
26 50 C.F.R. § 17.80.
-18- Photo 14: Arlington Canal dike in Gila River channel
Photo 15: Tailwater from Arlington Canal collects at Gillespie Dam (bottom) Photo 16: Gillespie Dam (breached in January 1993)
Photo 17: Painted Rock Dam, June 1994 (no water behind dam) Map HWY 93 Hassayampa River Sub-Basin
■
To Irrigation District
Buclke
River Harquahala Valley Irrigation District
• Gila Topminnow • USGS Gauging Stations Gillespie Dam Perennial/Regulated Streams Intermittent/Ephemeral Streams
Plot Size 1:320,000
10
Goa River Miles
*IV e.•11.11 INIW4 Harquahala Valley Irrigation District (HVID) and Tonopah Irrigation District (TID) receive CAP water in this area. Without investigating the structural or geographical features of the HVID and TID systems, the Biological Opinion assumed that non-native fish entering those systems through the CAP could escape into Centennial Wash. A careful examination of those water delivery systems shows that the conclusion of the Biological Opinion is unfounded.
The only portion of the HVID system that crosses Centennial Wash is the Westside Canal. Because that crossing is accomplished by siphon, there is no direct connection between the HVID system and Centennial Wash. The Westside Canal is protected north of Centennial Wash by Reach 1 of the Centennial Levee, a flood control structure built by the U.S. Soil Conservation Service, and south of Centennial Wash by a series of pipe overchutes and box washes that feed storm runoff over the canal and into the wash.27 Standard annual dry-up procedures employed by HVID will result in the total kill of any fish introduced into the HVID system through CAP water deliveries, and there are no permanent ponds in the HVID system that could harbor such fish.28 Thus, any non-native fish that might enter the HVID system will be unable to reach Centennial Wash.
Centennial Wash does not flow through TID lands. Rather, TID is drained by four primary washes--Camp, Winters, Delaney and Fourmile--all of which empty into Winters Wash, which in turn can feed Centennial Wash.29 Thus, any fish that wanted to leave the TID system would have to travel through several smaller washes before they could reach Centennial Wash. Annual dry-up of TID's delivery system will kill any non-native fish that are present, so there is little chance that such fish could enter any natural watercourse.39
Nor is it reasonable to expect that fish could enter Centennial Wash (or any other watercourse) as a result of normal irrigation of agricultural fields. "Delivery canals are highly ephemeral aquatic habitats, being frequently desiccated by canal operators."' Surface flows from irrigation canals are generally consumed by percolation in fields and do not generate a return flow to natural watercourses.32 Indeed, economic and legal
27 Addendum Report to Distribution System Loan Application Report and Feasibility Study for Harquahala Valley Irrigation District (1983).
HVID, Environmental Assessment for Harquahala Valley Irrigation District Irrigation Development Program (1982).
29 Central Arizona Project, Non-Indian Distribution Systems, Status Report (1991).
30 TID, Environmental Assessment for Tonopah Irrigation District irrigation Development Program (1981).
31 Hendrickson, Evaluation of the razorback sucker ( Xyrauchen texanus) and Colorado squawfish ( Ptychocheilus lucius) reintroduction programs in central Arizona based on surveys of fish populations in the Salt and Verde rivers from 1986 to 1990 (1993), at 45.
32 Id. (concluding that "Mew fish which enter diversions are likely to survive to return to the river).
-19- requirements increasingly mandate improved irrigation efficiency and the elimination of return flows. Moreover, the same rainfall event that might cause a normally dry wash to flow briefly would also reduce or eliminate the need for irrigation. It is very unlikely that a farmer would pay to have CAP water delivered, only to allow that water to run off his field and into a wash. CAWCD's experience shows that irrigators reduce their orders for CAP water during rainy periods.
The Biological Opinion acknowledged (at 4) that "[t]he Hassayampa [River] is usually dry throughout the reach of interest." In fact, the Hassayampa can lose 10,000 cfs to subsurface flow between Morristown and the Gila River.33 The Biological Opinion did not point out, however, that Centennial Wash flows even less frequently than the Hassayampa River. Available hydrologic data shows that Centennial Wash flows only about four to five days per year on average.34 Fish would not be able to survive in Centennial Wash belvveen these rare flood events as there are no pools or standing water to support aquatic life.
Flows in the Hassayampa River and Centennial Wash occur almost exclusively due to flood events. As the Service has acknowledged, flood events in the Gila River basin tend to remove or reduce non-native fish populations because the non-natives are washed downstream.35 Native fish, on the other hand, are more adapted to flooding and their physical attributes (e.g., slim bodies, large fins and humps) make them better at holding their position in a stream.36 Flood events also tend to produce high stream velocity and large sediment loads, effectively creating a barrier to upstream fish movement.
Fish swimming speed is a function of species, size, water temperature, oxygen level and motivation. Biologists refer to three categories of swimming speeds: cruising speed, sustained speed and sprint speed. Cruising speed is a speed that can be maintained for long periods of time, such as in migration. A sustained speed is one that can be maintained for several minutes and may be used for passage through a difficult area. For feeding or escaping predators, a fish may achieve a sprinting speed, but such an effort cannot be sustained for more than a minute or two without depleting necessary glycogen reserves."
33 Phillips, USGS, pers. comm.
USGS data, station no. 09517500.
35 Draft Biological Support Document, Colorado River Endangered Fishes Critical Habitat (1993), at 91, 95, 100.
36 Minckley & Meffe, Differential Selection by Flooding in Stream Fish Communities of the Arid American Southwest, in Community and Evolutionary Ecology of North American Stream Fishes, W.J. Mathews and D.C. Heins, eds., 93-104 (1987).
37 Davis, An Investigation of Barriers to Introduction of Fish into Granite Reef Dam Forebay from an Interconnection with the CAP Aqueduct (1984).
-20- Of the non-native fish addressed in the Biological Opinion, the striped bass is considered the strongest swimmer. As the test results in Table 3 show, swimming speeds for striped bass vary according to the size of the fish.38 Based on these data, it is unlikely that a striped bass of 12 inches or less could maintain a cruising speed of more than 3 feet per second. Thus, where the flow velocity in a stream is more than 3 feet per second, striped bass are not likely to move upstream. If a striped bass is not able to overcome stream velocity, a fish of lesser ability surely cannot.
Fish Size Cruising Speed Sustained Speed Sprint Speed
1 inch 1.00 NA NA
5 inches NA 2.75 NA
12 inches NA 4.00 12.00
22 inches NA NA 18.00
Table 3: Swimming Speeds of Striped Bass (feet per second)
During the infrequent flood events that do occur, stream velocity in the vicinity of Morristown is often high because the Hassayampa River channel is highly excised in that area. [Photo 18] Downstream, however, the river enters a flat sandy plain where much of the flow goes subsurface. [Photo 19] Ninety-one percent of the flows at Morristown become subsurface about one mile downstream of the USGS gauging station. [Figure 10] Thus, even when the river is flowing near Wickenburg, it is likely dry further downstream. [Photo 20] A dry river obviously does not permit the movement or survival of any fish attempting to migrate upstream. For the Hassayampa to flow continuously to the Gila River, the flow at Morristown must be approximately 800 cfs or more.39 Only one percent of flows in the Hassayampa exceed that leve1.49 Due to stream channel dimensions, an 800 cfs flow at Morristown creates a stream velocity of at least 5 feet per second, which exceeds the sustained swimming speed of a 12-inch striped bass. In short, a water-to-water connection between the Gila River and the upper Hassayampa River can only be expected a few days each year, and the stream velocity in the Hassayampa on those occasions will prevent even the strongest non-native fish from moving upstream.
m Bell (1973); Freadman (1979) Webb (1971). NA indicates data not available.
39 Phillips, USGS, pers, comm. and USGS data. The minimum flow required at Morristown to ensure a continuous flow to the Gila River when the Hassayampa River channel is saturated from previous flows is estimated to be 800 cfs. Under normal conditions the riverbed is dry, not saturated, and substantially more than 800 cfs would be required before the Hassaympa River would flow to the Gila River.
0 4 USGS data for station no. 09516500.
-21- Photo 18: Hassayampa River channel near USGS gauge at Morristown
Photo 19: Hassayampa River channel about 2 miles south of Morristown Flow Summary of the Hassayampa River cfs near Morristown, Arizona - 9000 •* 4,292 days sampled (USGS 1938-1947, 1991-1994). 8000-
7000-
Volume of flow above which water would be present through the 38-mile stretch 6000- between Morristown and Arlington (1% of flows).
i ja 5000- a3 4000-
3000-
2000- Volume of flow below which discharge becomes subsurface approximately 1 mile below / the USGS gauging station near Morristown (91% of flows). 1000-
- - 0 rrrrrrn rin Frl FITTT I FT1 FT1 EITIFI 11111 0 11 22 33 44 55 66 77 88 99 Percent of flows exceeding Photo 20: Hassayampa River about 15 miles south of Morristown gauge near CAP siphon crossing High water temperatures in the Hassayampa River will also discourage upstream movement by some fish, including striped bass. Near Arlington, the mean monthly temperature of the Hassayampa River from April through October is more than 25°C, exceeding the preferred temperature range for adult striped bass. From May through September, the temperature exceeds 28°C, the avoidance level for adult striped bass. [Figure 11]
Contrary to the Biological Opinion, Painted Rock Reservoir is not "a likely harbor for invading CAP non-natives." Painted Rock Dam, located more than forty miles downstream from the mouth of the Hassayampa River, was designed solely for flood control, not water storage. Thus, the reservoir created by Painted Rock Dam does not maintain water year round. [See Photo 17] Because the reservoir routinely dries up, non-native species cannot "become established in Painted Rock Reservoir [and] serve as a constant source for upstream movement during periods of flow in the rivers," as the Biological Opinion suggests (at 17).
Non-native fish that may be carried downstream by flows in Centennial Wash may have difficulty reaching Painted Rock Reservoir in any event. It is quite possible that those fish would be entrained in the Arlington Canal (which crosses Centennial Wash) and carried to Gillespie Dam where tailwater from that canal is diverted into the Enterprise Canal. [Photo 21] Enterprise Ranch diverts up to 44 cfs from the Gila River at Gillespie Dam through the Enterprise Canal. If non-native fish were able to reach the main channel of the Gila River at Gillespie Dam, they also could be entrained in the Gila Bend Canal, rather than traveling 36 miles further downstream to Painted Rock Dam. Paloma Ranch operates the Gila Bend Canal, which can carry 268 cfs to agricultural lands in the Gila Bend area. Although Gillespie Dam has not been repaired since its breach in January 1993, Paloma Ranch has constructed a diversion channel and installed five pumps to allow it to lift water from the Gila River into the Gila Bend Canal. [Photo 22]
Even if non-native fish from the CAP were able to traverse the long reaches of normally dry streambed in the Hassayampa River, they still would have no impact on Gila topminnow at The Nature Conservancy Hassayampa Preserve (the "Preserve") because, as the Biological Opinion acknowledges (at 17), those fish are located in springs and ponds that are isolated from the river channel. Endangered fish at the Preserve have been planted in a man-made pond (Palm Lake) and fresh water springs all of which lie outside the 100-year floodplain of the Hassayampa River. [Photo 23] During January 1993 flooding (a 25- to 50-year flood event), the peak discharge at Morristown was 26,300 cfs (8,280 cfs mean daily flow).41 Even at that level, the river remained many meters away from and several meters below Palm Lake.42
41 Heaton, pers comm
42 The Preserve has a video tape of the flood waters.
-22- Mean Monthly Temperature of the Hassayampa River and Temperature Requirements of Striped Bass c c 35 ** From USGS data, water years 1990-1995.
30
25 Preferred temperature range of adult striped bass I
15
10 Avoidance level for adult striped bass
Preferred temperature range of Temperature of the adult striped bass Hassayampa River 5 near Arlington.
. " 1 2 3 4 5 6 7 8 9 12 Months (Jan - Dec) Photo 21: Enterprise Canal (center, dry) diverts Arlington Canal tailwater collected at Gillespie Dam (note turbulence in fourth dome from left)
Photo 22: Temporary dike constructed by Paloma Ranch diverts water for pumping into Gila Bend Canal (right) at Gillespie Dam (June 1994) Photo 23: Palm Lake at The Nature Conservancy Hassayampa River Preserve near Wickenburg (not connected to Hassayampa River) The Service has stated that a finding of jeopardy is not warranted where non- native fish are to be introduced into waters that are isolated from the waters occupied by the endangered species.43 Without question, the waters of Palm Lake are isolated from any waters where non-native fish might be introduced by the CAP.
The Service has also indicated that no jeopardy exists where only stocked populations of endangered fish may be affected by the proposed activity. For example, the Service concluded in another biological opinion (the "FWS Opinion") that its own plan to stock 600 rainbow trout per month in 12-Mile Lake on the Colorado River Indian Reservation--the same lake in which the Service stocked razorback sucker in 1993-- would not jeopardize that endangered species because the "[s]tocking [of trout] into 12- Mile Lake would not likely affect the wild population of razorback suckers since it is removed from the river."44 The only Gila topminnow populations in the Hassayampa River basin are stocked.45 Since there are no "wild" populations of Gila topminnow in that area, the FWS Opinion suggests that there can be no jeopardy.
Especially troubling is the fact that the Gila topminnow population at the Preserve was stocked in August 1989, more than four years after the CAP began delivering water to HVID. It is inherently unfair and improper for the Service to sanction the introduction of an endangered species into an area and then assert that the newly created population is jeopardized by an existing federal project that has been operating for several years.46
B. No Critical Habitat Has Been Designated in the Hassayampa River Sub-basin.
CAP water deliveries cannot adversely modify critical habitat in the Hassayampa River basin because no critical habitat has been designated in that area for any of the endangered species identified in the Biological Opinion.
The Endangered Species Act only prohibits the destruction or adverse modification of areas designated as "critical habitat."47 The Act does not extend such
USFWS, Biological Opinion on Fish and Wildlife Stocking of Rainbow Trout and Channel Catfish in the Lower Colorado River (Hoover Dam to the International Border), July 1, 1994 (''FWS Opinion"), at 1 ("the stocking of channel catfish in tanks and other waters with no direct connection to the lower Colorado River is not likely to adversely affect the razorback sucker and bonytail chub or their designated critical habitat"), 14 ("Stockings into isolated waters that do not contain razorback suckers avoid the issue of predation").
44 Id. at 3, 6, 14 (emphasis supplied).
45 Biological Opinion, at 2, 17.
46 The same is true for all razorback sucker populations in the Salt, Verde and Gila river systems. Razorback sucker were extirpated from those areas long before the CAP was built, and any fish that may exist there today were stocked by the Service or AGFD after consultations on the CAP began.
47 16 U.S.C. § 1536(a)(2).
-23- protection to areas that have not been so designated." According to the Biological Opinion (at 17, 24-25), the Hassayampa River is "important recovery habitat," the protection of which is "critical to the recovery of these species." If that were true, then those areas would have been designated as critical habitat because, by definition, "critical habitat" includes all areas that are "essential to the conservation of the species'", and "conservation" is defined to include the recovery of the endangered species.5° Thus, any area that the Service deemed essential to the recovery of an endangered species necessarily would have been included in the critical habitat designation for that species.
III. The Agua Fria River Sub-basin.
The Agua Fria River originates in Arizona's central highlands near Prescott where it drains the east slopes of the Bradshaw Mountains, the south slopes of the Black Hills, and Hutch and Cooks Mesas. The Agua Fria flows southward for about sixty miles to Lake Pleasant, about 35 miles northwest of Phoenix. Above Lake Pleasant, the river experiences extreme fluctuations in flow. During dry periods of the year, the Agua Fria is intermittent and shallow, in some areas less than three inches deep. Under drought conditions, it ceases to flow above ground.
Tule Creek is a small tributary that joins the Agua Fria River at the northern end of Lake Pleasant. Originating from perennial springs, the creek flows intermittently at its upper end, but is an ephemeral wash where it joins the Agua Fria. Under normal flow conditions, there is no water-to-water connection between the Agua Fria River or Lake Pleasant and Tule Creek.
Humbug Creek originates high in the Bradshaw Mountains near Crown King and flows intermittently southward for almost 20 miles to Lake Pleasant. Cow Creek also flows from the Bradshaw Mountains and joins Humbug Creek one to two miles above Lake Pleasant. Both creeks are intermittent or ephemeral in the vicinity of Lake Pleasant.
Gila topminnow were first stocked in Tule Creek around 1968. The site is subject to severe flooding, however, and the population was washed out in 1978. Tule Creek was restocked in 1981 and a June 1994 survey revealed a large population of topminnow.51 The Tule Creek population of Gila topminnow inhabits a source spring and
48 51 Fed. Reg. 19935 (1986) (the limitation of "destruction or adverse modification" to designated critical habitat "is mandated by the strict language of section 7(a)(2) and cannot be altered by the Service").
49 50 C.F.R. § 424.02(d).
0 5 Id. § 424.02(c).
51 USFWS records (8/94).
-24- intermittent flowing reaches of the creek downstream for several hundred meters. Under normal conditions, this population is isolated from the Agua Fria River and Lake Pleasant."
Cow Creek was stocked with Gila topminnow in 1981 and some of those fish subsequently moved downstream into Humbug Creek. The Gila topminnow populations in Cow and Humbug Creek are believed to have been extirpated before CAP water was added to Lake Pleasant. Surveys conducted in 1991, 1992 and 1993 found no Gila topminnow in either location.53 In the past, topminnow had been reported in Cow Creek immediately above a large cement dam which serves as a fish barrier.54 [Photo 24]
The Biological Opinion contends (at 17-18) that storage of CAP water in Lake Pleasant will give non-native fish from the Colorado River access to Tule, Humbug and Cow creeks, thus jeopardizing stocked Gila topminnow in those areas. [See Map 2]
A. The Service Previously Determined that the CAP Does Not Jeopardize Endangered Fish in the Agua Fria River Sub-basin.
In 1983, after formal Section 7 consultations with Reclamation, the Service issued a biological opinion concluding that the construction of New Waddell Dam and the regulatory storage of CAP water in Lake Pleasant did not jeopardize the Gila topminnow.55
The Service's 1983 biological opinion concluded that non-native fish from Lake Pleasant would only be able to reach the Gila topminnow habitat in Tule Creek when an Inflow Design Flood ("IDF") event coincided with flows in Tule Creek." A flood of IDF magnitude "would occur very rarely, if ever."57 Such a flood is not expected to occur within the 100-year lifetime of the CAP or even within 200 years." Because the IDF is so "extremely rare," its "impact to biological resources is essentially hypothetical."59 Nevertheless, as an optional conservation measure, the Service recommended that Reclamation construct a barrier on Tule Creek to preclude migration of non-native fish
52 USFWS, Biological Opinion, Central Arizona Water Control Study, at 8-9 (3/8/83) ("CAWCS Biological Opinion").
USFWS records (8/94).
54 AGFD Status Report, at 46.
CAWCS Biological Opinion.
56 Id. at 9.
57 Reclamation, Draft EIS, Regulatory Storage Division, Central Arizona Project, at 117.
58 Id. (Figure IV-1).
59 Id. at 121.
-25- Photo 24: Concrete dam on Cow Creek above Lake Pleasant
Photo 25: Fish barrier built by Bureau of Reclamation on Tule Creek above Lake Pleasant Map 2 Aqua Fria River Sub-Basin
Dam
L k
Pleasant Gila Topminnow '(Maximum Conservation • Present Location Pool - El. 1702) © Former Location OA Fish Barrier Intermittent/Ephemeral Streams
Plot Scale 1:50,000 1 2
Miles New Waddell Dam ' CAP Canal *R. during such an event.60 Reclamation subsequently built such a barrier. [Photo 25] Even without the barrier, however, the 1983 biological opinion found no jeopardy to the Gila topminnow because "loss of this reintroduced population would not significantly impact the survival of the species."61 Importantly, the 1983 biological opinion did not mention the possibility of any adverse effects on Gila topminnow in Cow Creek or Humbug Creek.
The latest Biological Opinion contradicts the 1983 opinion--without explaining why--and concludes that CAP water deliveries will jeopardize Gila topminnow in intermittent streams such as Tule, Cow and Humbug creeks above the IDF level of Lake Pleasant. However, the Biological Opinion completely ignores key scientific and commercial data showing that non-native fish were already well established in Lake Pleasant before CAP water was introduced in December 1992. From 1987 through 1989, the Arizona Game and Fish Department inventoried Lake Pleasant and the Agua Fria River to determine the types and quantities of fish that were present before the lake became a storage reservoir for the CAP.62 That report was prepared for the Bureau of Reclamation, which also performed studies to determine what types of fish were present in the CAP.63 Table 4 lists the fish identified as being present in the Hayden-Rhodes aqueduct (that portion of the CAP between the Colorado River and the Salt River east of Phoenix) and in Lake Pleasant before CAP water was introduced.64
60 CAWCS Biological Opinion, at 11; USFWS, Draft Fish and Wildlife Coordination Act Report, Central Arizona Project, Regulatory Storage Division, Arizona, at 56 (1983).
61 CAWCS Biological Opinion, at 9.
62 AGFD, Phase I: Baseline Limnological and Fisheries Investigation of Lake Pleasant (1990) ("Lake Pleasant Study").
63 Bureau of Reclamation, Central Arizona Project Canal System, Fishery Investigations Report (1990) ("CAP Fish Survey")
64 Matter, Potential for Transfer of Non-Native Fish in Central Arizona Project Canal Waters to the Gila River System (1991); Lake Pleasant Study.
-26- Hayden-Rhodes Lake Pleasant Aqueduct (CAP) (Before CAP)
Threadfin shad X X
Goldfish X X
Carp X X
Red shiner X X
Yellow bullhead X X
Channel catfish X X
Flathead catfish X
Striped bass X
Green sunfish X X
Bluegill X X
Redear sunfish X X
Largemouth bass X X
Black crappie X X
Blue tilapia X
Golden shiner X
Mosquitofish X
White bass X
White crappie X
Table 4: Comparison of fish in CAP and in Lake Pleasant (before CAP)
Many of these non-native species were stocked in Lake Pleasant. Over the past fifty years, AGED has stocked 60,376 channel catfish, 56,667 largemouth bass, 42,720 rainbow trout, 119,270 bluegill, 25,000 black crappie, 125,000 walleye, 6,300 hybrid sunfish, and 238 white bass into Lake Pleasant.65
As shown in Table 4, only two non-native species found in the CAP--striped bass and flathead catfish--were not already present in Lake Pleasant. Neither of those species can reasonably be expected to pose problems for Gila topminnovv populations
AGED Stocking Records.
-27- in Cow Creek, Humbug Creek, Tule Creek or other small tributaries to the Agua Fria River.
Striped bass are closely related to white bass, which were already established in Lake Pleasant before the CAP was connected.66 AGFD's survey of the Agua Fria River found no white bass in the river above Lake Pleasant.67 Striped bass also are not known to inhabit tributary streams. "Striped bass have been encountered only in the main channel of the Colorado River and at the mouths of major tributaries, but not within any tributaries. For example, no striped bass have been collected in the Bill Williams River between Lake Havasu and Alamo Lake or in the San Juan River flowing into Lake Powell despite repeated surveys."68 Thus, there is no evidence to suggest that striped bass would invade small intermittent creeks such as Cow, Humbug or Tule, where Gila topminnow may be found.
Similarly, the available data suggests that the possible addition of flathead catfish to Lake Pleasant will not jeopardize Gila topminnow in Agua Fria tributaries. Despite the stocking of more than 60,000 channel catfish in Lake Pleasant over the years, AGFD's survey found only one catfish in the Agua Fria River above the lake.69 There is no evidence that catfish now inhabit Cow, Humbug or Tule creeks and, consequently, no evidence to suggest that catfish are likely to inhabit those creeks in the future.
The primary non-native fish threat to the survival of the Gila topminnow is the mosquitofish.79 The mosquitofish was introduced into Arizona waters in 1926 and spread rapidly!' Mosquitofish were well established in the Agua Fria River before CAP water was stored in Lake Pleasant.72 Mosquitofish are also present in the lower reaches of Humbug Creek above Lake Pleasant.73 Yet mosquitofish were not recorded in Tule
66 Grabowski, at 26
67 Lake Pleasant Study, at 59.
68 Matter, at 32 (citations omitted)
69 AGED stocking records; Lake Pleasant Study, at 159.
USFWS, Sonoran Topminnow (Gila and Yaqui) Recovery Plan, at 7-9 (1984); Jackson, Martinez et al, Assessment of Water Conditions and Management Opportunities in Support of Riparian Values: BLM San Pedro River Properties, Arizona, at 43 (1987) ("Mosquitofish depredations appear the major factor in disappearance of this native species"); AGFD, Status of the Sonoran Topminnow ( Poeciliopsis occidentalis) and Desert Pupfish ( Cyprinodon macularius) in Arizona (1991).
T1 Sonoran Topminnow (Gila and Yaqui) Recovery Plan, at 7.
72 Lake Pleasant Study, at 159 (mosquitofish were found in every reach sampled and comprised 15% of all fish collected in the Agua Fria River).
73 AGED, Status of the Sonoran Topminnow ( Poeciliopsis occidentalis) and Desert Pupfish ( Cyprinodon macularius) in Arizona (1991), at 48.
-28- Creek before Reclamation constructed a fish barrier there, and are not present above the barrier today.74
The fish barrier constructed by Reclamation on Tule Creek provides further assurance that the storage of CAP water in Lake Pleasant will not jeopardize the Gila topminnow.75 Indeed, the barrier probably provides greater protection to the Tule Creek topminnow population than existed before New Waddell Dam was built and CAP water introduced. Although the Biological Opinion speculates (at 18) that "the presence of non- native species near the downstream side of the barrier will raise the probability of bait bucket transfer above the barrier," Tule Creek is an ephemeral wash at that point. There is no evidence of fishing in Tule Creek and the physical characteristics of the creek make it highly unlikely that sport fish would persist above the barrier.
In sum, non-native fish are no more likely to affect Gila topminnow in the Agua Fria River basin now than before CAP water was introduced into Lake Pleasant. Thus, there is no reasonable basis to conclude that storage of CAP water in Lake Pleasant will jeopardize any endangered species, and no justification for reversing the Service's 1983 biological opinion.
B. No Critical Habitat Has Been Designated in the Agua Fria River Sub- basin.
As in the Hassayampa River sub-basin, no critical habitat has been designated in the Agua Fria River area for any of the endangered fish identified in the Biological Opinion. Thus, storage of CAP water in Lake Pleasant cannot destroy or adversely modify critical habitat.
IV. The Salt and Verde River Sub-basin.
The Salt River begins at the confluence of the Black and White rivers in eastern Arizona and flows west-southwest to its confluence with the Gila River west of Phoenix. The Black and White rivers drain the Blue Mountains, White Mountains, and the Natanes Plateau before joining to form the Salt River. From there, the Salt River drains the Sierra Ancha and the Mogollon Rim on the north and the Pinal Mountains and Superstition Mountains on the south. Rainfall and snowmelt in the mountains of eastern Arizona provide a perennial flow into reservoirs upstream from Phoenix.
The largest tributary to the Salt River is the Verde River, which originates near Pau!den, Arizona. From its headwaters, Big Chino Wash, the Verde River flows
T4 Id Sonoran Topminnow (Gila and Yaqui) Recovery Plan, at 4.
75 AGFD, Status of the Sonoran Topminnow ( Poeciliopsis occidentalis) and Desert Pupfish (Cyprinodon macular/us) in Arizona (1991), at 46 (the fish barrier on Tule Creek "will prevent invasion of exotics").
-29- southward for 170 miles, draining Big and Little Chino Valleys, the Verde Valley, the western part of the Mogollon Rim and the west side of the Mazatzal Mountains before joining the Salt River about 35 miles northeast of Phoenix.
The Salt River Project ("SRP") operates a series of dams and reservoirs on the Salt and Verde rivers. [Figure 12] Theodore Roosevelt Dam and Reservoir is situated at the confluence of the Salt River and Tonto Creek about 76 miles northeast of Phoenix and was completed in 1911. It is a cyclopean-masonry thick arch structure 280 feet high and 723 feet long at its crest. [Photo 26] Reclamation is in the process of modifying Theodore Roosevelt Dam to provide flood control and improve the safety of the dam. The concrete gravity arch under construction will stabilize the existing masonry structure, add about 77 feet to the height of Theodore Roosevelt Dam and nearly 500 feet to its crest length, and increase the active conservation capacity of the reservoir to 1,591,800 acre-feet. [Photo 27]
Below Theodore Roosevelt Dam on the Salt River are three additional dams: Horse Mesa Dam, Mormon Flat Dam, and Stewart Mountain Dam. Horse Mesa Dam is located about 65 miles northeast of Phoenix and impounds up to 245,138 acre-feet in Apache Lake. It is a 305-foot high concrete thin arch structure completed in 1927. [Photo 28] Mormon Flat Dam is also a concrete thin arch structure. Completed in 1926, the 224-foot high dam is located 51 miles northeast of Phoenix and impounds Canyon Lake, with a total capacity of 57,852 acre-feet. [Photo 29] Stewart Mountain Dam, another concrete thin arch completed in 1930, is 207 feet high and is located 41 miles northeast of Phoenix. [Photo 30] Saguaro Lake, the reservoir created by Stewart Mountain Dam, has a capacity of 69,765 acre-feet.
On the Verde River above its confluence with the Salt are Horseshoe Dam and Bartlett Dam. Horseshoe Dam is a 194-foot high earthfill dam about 58 miles northeast of Phoenix. [Photo 31] Horseshoe Reservoir has a capacity of 139,238 acre-feet. Bartlett Dam is located about 46 miles northeast of Phoenix. It is a 287-foot high concrete multiple arch structure and impounds up to 178,477 acre-feet in Bartlett Reservoir. [Photo 32]
About three miles downstream from the confluence of the Salt and Verde rivers is Granite Reef Diversion Dam, a concrete ogee weir 29 feet high with a crest length of 1,128 feet. [Photo 33] Completed in 1908, Granite Reef Diversion Dam diverts water to SRP canals on both sides of the Salt River for irrigation, municipal and industrial use in the Salt River Valley. Downstream of Granite Reef Diversion Dam, the Salt River flows only in response to storm events or releases from upstream dams. CAP water is delivered to SRP canals below Granite Reef Diversion Dam through the CAP/SRP interconnection. Electrical fish barriers are located on each of the SRP canals above the CAP interconnection and below Granite Reef Diversion Dam. [Photos 34 & 35]
-30- PROFILE VIEW OF SALT AND VERDE RESERVOIR SYSTEMS
Elevation (ft)
ROOSEVELT DAM 2218' 2200 2151' 2136' 2120.2' HORSESHOE DAM 2100'" Roosevelt 2026' p Lake '1,336,734 AF 2000 _ Horseshoe 2000' SPILLWAY 131,427 HORSE MESA DAM 1914' 1891' — BARTLETT DAM Apache Lake 245,138 AF 1 800 1798' 1 MORMON CO ' Bartlett 1748' FLAT DAM C Lake CT \,178,186.,_ AF 1660.5' STEWART SALT TOTAL R; MOUNTAIN 1610.5' — Canyon Lake 1 600 57,852 AF STORAGE: \ DAM 1,709,489 AF VERDE TOTAL 1529' STORAGE: 1506' 309,613 AF Saguaro Lake 69,765 AF 1400
1310' Granite Reef Diversion Dam 1200 Photo 26: Theodore Roosevelt Dam on the Salt River (January 1993) Photo 27: Modified Roosevelt Dam (under construction, August 1994) Photo 28: Horse Mesa Dam on the Salt River Photo 29: Mormon Flat Dam on the Salt River Photo 30: Stewart Mountain Dam on the Salt River Photo 31: Horseshoe Dam on the Verde River Photo 32: Bartlett Dam on the Verde River Photo 33: Granite Reef Diversion Dam on the Salt River Photo 34: Electrical fish barrier on SRP's South Canal below Granite Reef Diversion Dam
Photo 35: Electrical fish barrier on SRP's Arizona Canal below Granite Reef Diversion Dam Critical habitat has been designated for the razorback sucker on the Salt River from Roosevelt Diversion Dam, near the eastern end of Roosevelt Lake, upstream to the US60/SR77 bridge, and on the Verde River from Horseshoe Dam upstream to the U.S. Forest Service boundary near Perkinsville. [See Figure 8] Critical habitat has also been designated for the spikedace on the upper Verde River from Sullivan Lake dam to the confluence with Sycamore Creek. [See Figure 5]
The Biological Opinion contends that CAP water deliveries to SRP will allow non- native fish to enter the Salt and Verde river systems, jeopardizing endangered fish in the upper reaches of those rivers. Specifically, the Biological Opinion asserts (at 18) that non-native fish from the CAP would enter SRP canals through the CAP/SRP interconnection, pass through electrical fish barriers on the upstream side of the interconnection, pass over or through Granite Reef Diversion Dam, be transported around Stewart Mountain Dam, Mormon Flat Dam, Horse Mesa Dam and Roosevelt Dam on the Salt River and Bartlett Dam and Horseshoe Dam on the Verde River, and swim upstream to where the endangered species are located. [See Map 3]
A. CAP Water Deliveries Cannot Jeopardize Endangered Fish in the Salt and Verde River Sub-basin.
With the exception of striped bass, white bass and golden shiner, all of the non- native fish that the Biological Opinion contends might be introduced through CAP deliveries were already present in SRP canals before the CAP interconnection was constructed.76 Moreover, those non-native fish had ample opportunity to migrate upstream since there were no electrical fish barriers on the SRP canals below Granite Reef Diversion Dam before the CAP/SRP interconnection was built. In fact, all of the non-native fish found in SRP canals (with the exception of goldfish) were also present in the Salt and Verde river system above Granite Reef Diversion Dam before any deliveries of CAP water.
The concern expressed in the Biological Opinion seems disingenuous, at best, given the fact that vast quantities of non-native fish have been stocked by AGFD in the Salt and Verde river systems over the years, often with the tacit approval of the Service (which provides significant funding for AGFD's stocking programs). Table 5 shows some of the non-native fish stocked in six large reservoirs on the Salt and Verde rivers.77
76 Matter (1991)
AGFD Stocking Records
-31- Map 3
Hwy 87 Salt and Verde River Sub-Ba
Horseshoe Dam
• Razorback Sucker
Perennial/Regulated Streams 288 Intermittent/Ephemeral Streams
Bartlett Dem Plot Scale 1:320000
Miles
Hwy 88
Granite Reef Diversion Dam er Hwy 88 Black crappie 205,198
Bluegill 1,500,885
Largemouth Bass 1,302,574
Rainbow Trout 1,489,084
Redear Sunfish 183,910
Threadfin Shad 20,500
Channel Catfish 184,975
Walleye 10,854,982
Table 5: Non-native fish stocked in Salt & Verde reservoirs
In addition, AGFD has stocked almost 14 million rainbow trout in the upper Salt and Verde river systems, many of those after the Service wrote its draft biological opinion on CAP deliveries in 1991. For example, as recently as 1993, AGFD stocked rainbow trout in the Salt and Verde rivers, as well as in tributaries such as Black River, Canyon Creek, East Verde River, Oak Creek, Tonto Creek, West Clear Creek and Wet Beaver Creek.78 Stocking of rainbow trout in the Salt River above Granite Reef Diversion Dam continues to this day with the Service's approval.79 Moreover, AGFD's 1995 stocking request includes 240,500 catchable-size trout for the Verde and East Verde rivers, and 157,500 trout for the Salt and Black rivers.
The upper Salt and Verde rivers where spikedace and razorback sucker may be located are "almost completely dominated" by non-native species.8° Smallmouth bass, green sunfish, channel catfish, flathead catfish and carp are found in razorback sucker critical habitat on the Salt River above Roosevelt Diversion Dam and on the Verde River above Horseshoe Dam.81 Carp, channel catfish and flathead catfish are common or abundant in these reaches and have been shown to prey on razorback sucker eggs, larvae and young.82 Mosquitofish, red shiner, fathead minnow, yellow bullhead,
AGFD Stocking Records.
7° See Arizona Republic (12/14/94) (4,000 rainbow trout stocked below Stewart Mountain Dam): Arizona Republic (2/23/95) (13,000 trout stocked or scheduled to be stocked in February 1995).
8° Hendrickson, Evaluation of the razorback sucker ( Xyrauchen texanus) and Colorado squawfish ( Ptychocheilus lucius) reintroduction programs in central Arizona based on surveys of fish populations in the Salt and Verde rivers from 1986 to 1990 (1993).
USFWS, Colorado River Endangered Fishes Critical Habitat, Draft Biological Support Document (1993), at 94, 99; Hendrickson (1993).
82 Id. at 95.
-32- largemouth bass, bluegill, and rainbow trout have also been recorded in the upper Salt and Verde rivers.83
Given the large numbers of non-native fish stocked or established in the Salt and Verde river basins above Granite Reef Diversion Dam, any potential augmentation by fish escaping the CAP would be trivial. In any event, any concern that non-native species could be introduced by CAP deliveries and harm endangered fish in the upper Salt and Verde river systems is eliminated by the electrical fish barriers and large mainstream dams upstream of the CAP/SRP interconnection.
Electrical fish barriers are in operation on both SRP canals below Granite Reef Diversion Dam and upstream of the CAP interconnection. [See Figure 4] Each barrier consists of a series of six pulse generators connected to electrodes located in a weir structure across the bottom of the canal. [Figure 13] The use of six separate pulse generators in series provides redundancy and allows the barrier to continue functioning even if one or two generators fail. In addition, emergency back-up generators are present to supply power to the electrodes in the event that the primary power supply is interrupted.
The electric field created by the barrier invokes a reaction in fish that can range from a slight muscle twitch to complete tetany or even mortality, depending on the strength of the field. A minimum of 0.5 volts per inch is needed to give a fish an unpleasant sensation. Fish are completely tetanized at 1.5 volts per inch, and anything more than 2.0 volts per inch can cause mortality. Testing performed in 1990 on SRP's fish barriers recorded field intensity ranging from 2.1 to 4.0 volts per inch.84
Electrical barriers are most effective when water velocity through the barrier area is sufficient to sweep stunned fish quickly downstream. A velocity of 2 to 10 feet per second is generally recommended. Water velocity through SRP's barriers is maintained by the weir structures housing the electrodes. [Figure 14] With a discharge of only 200 cfs from Granite Reef Diversion Dam, the weir design results in a water velocity of 3.9 feet per second through the electrical fish barrier, as shown in Table 6.85 At higher discharge rates, the water velocity alone through the fish barrier is sufficient to prevent upstream migration by smaller fish.
83 Reclamation, Biological Assessment for the Spikedace (Meda fulgida), Central Arizona Water Exchange Project (1988); Hendrickson (1993).
84 Moorehead, SRP, pers. comm.
85 SRP, CAP/SRP Interconnection, Final Environmental Assessment Report (1988), at 33.
-33-
EXISTING SRP CANAL HEADGATES
EQUIPMENT 34:JILDING
WATER FLOW TOP OF CABLES TO CANAL ELECTRODES
ELECTRODES [ PULSE GENERATOR 1.75 V/cm METER • . .• PULSE NER A TOR G • 1.5 V cm: METER
PULSE GENERATOR 1.25 V/cm METER
PULSE :1.0 V/cm: : : METER GENERATOR
0.75 V,/cm METER PULSE GENERATOR .• .5 V METER PULSE GENERATOR WEIR ON CANAL 80770M PULSE RATE AND DURATION CONTROL MAX WATER LEVEL
,/ CAP NSCHARGE INTO A.C. POWER SOURCE 44(SRP SYSTEM
'EMERGENCY BACK—UP GENERATOR BOTTOM OF CANAL
ELECTRIC FISH BARRIER - PLAN VIEW Figure 13
TOP OF CANAL /
WATER SURFACE
---- / ------S' 4' <
24' 20' 1 2'
WEIR STRUCTURE BOTTOM OF CANAL ELECTRODE LOCATIONS
ELECTRIC FISH BARRIER - PROFILE VIEW Discharge (cfs) Depth (ft) Velocity (fps) 200 0.9 3.9 400 1.3 5.0 800 2.0 6.5 1200 2.6 7.4 1400 2.8 7.8 1700 3.2 8.3 2000 3.5 8.8 2300 3.8 8.8+
Table 6: Water velocity through SRP fish barriers
The Service has stated on several occasions that electrical fish barriers like those in place on the SRP canals below Granite Reef Diversion Dam and elsewhere effectively prevent upstream fish movement. Early designs for the CAP/SRP interconnection had envisioned that CAP water would be introduced into the forebay immediately above Granite Reef Diversion Dam.86 However, due to the Service's concerns about introduction of Colorado River fish into the lower Salt and Verde rivers, the interconnection was redesigned to deliver CAP water directly to SRP canals below Granite Reef Diversion Dam and electrical fish barriers were installed on the SRP canals between the CAP interconnection and Granite Reef Diversion Dam. The Service reviewed the new plans in 1984 and stated:
"Our major concern relative to the proposed project has been and continues to be the introduction of new aquatic organisms into the Salt and Verde River systems. The decision to deliver water directly into the SRP canals below Granite Reef Diversion Dam, and to construct fish barriers in
the canals1187 to prevent upstream movement of fish, adequately satisfies this concern .
The Service reiterated that conclusion when it reviewed the environmental assessment for the CAP/SRP interconnection:
"Our concern with the original design was that it allowed for the introduction of Colorado River ichthyofauna into the Salt and Verde River systems. The change in design [to deliver water directly into SRP canals and construct electrical fish barriers to prevent upstream movement of fish]
86 See Grabowski, at 1
87 Field Supervisor, Ecological Services, USFWS, letter to SRP (11/16/84) ("We do not envision any potential environmental impacts associated with the proposed interconnection facilities")(emphasis supplied).
-34- will effectively prevent this from happening and, therefore, adequately satisfies this concern."88
Ample scientific and commercial data supports the Service's original conclusion that electrical fish barriers of the type used at the CAP/SRP interconnection are effective. The Biological Opinion does not cite any data to the contrary, but instead relies (at 18) on the unspecified "belief' of "some ichthyologists" that "fish can move upstream through the barriers under certain conditions." Interestingly, one of the RPAs rejected in the Biological Opinion (in favor of a transfer of funds from Reclamation to the Service) called for a study of "the effectiveness of the existing electrical fish barriers at preventing upstream fish movement."89 Apparently, the Service was content to base its jeopardy finding on uncited "beliefs" rather than the best available scientific and commercial data. The data strongly suggests that the electrical fish barriers in operation at the CAP/SRP interconnection will prevent upstream migration of non-native fish that might escape the CAP.9°
In addition to the electrical barriers, Granite Reef Diversion Dam provides a physical barrier to upstream fish movement. The average water velocity through the headgate openings of Granite Reef Diversion Dam ranges from 14 to 19 feet per second, precluding upstream movement by all fish less than about 14 inches long.91 Even when the level of the Salt River is low and the headgates are opened wide, the velocity remains 5 to 8.5 feet per second, making it totally impassable for fish smaller than about 6 inches.
Even if non-native fish from the CAP were able to evade the electrical fish barriers on SRP's canals and pass through Granite Reef Diversion Dam, they would still face insurmountable barriers in the form of four large dams on the Salt River and two on the Verde. The Service has stated on numerous occasions that these dams prevent upstream fish movement.92
m Field Supervisor, Ecological Services, FWS, letter to SRP (10/28/87)(emphasis supplied); see SRP, CAP/SRP Interconnection, Final Environmental Assessment Report (1988), at 51 (describing Section 7 review and input by USFWS).
89 Biological Opinion, at 40,
90 On one occasion, the battery to start the back-up generator failed when the main power went out and the barrier was temporarily inoperative. During that brief event, fish from below the barrier moved upstream in the canal but failed to get through Granite Reef Diversion Dam due to the velocity barrier at the dam's headgates. The fish were later captured and moved back below the electrical barrier. After that event, SRP instituted procedures to insure that the situation could not be repeated.
91 SRP, CAP/SRP Interconnection, Final Environmental Assessment Report (1988), at 32.
92 USFWS, Colorado River Endangered Fishes Critical Habitat, Draft Biological Support Document (1993), at 157-59 (listing Horseshoe, Bartlett, Stewart Mountain, Horse Mesa, Mormon Flat, Roosevelt and Coolidge dams as providing "blockage to Colorado River Fish passage," and noting generally that "[m]ajor mainstem dams completely block fish movement"); 50 CFR § 17.84(b)(authorizing experimental, nonessential populations of Colorado squawfish and woundfin in the Salt and Verde rivers and stating that those populations are "totally isolated" from existing populations on the Colorado River and elsewhere because of upstream travel distance, mainstream dams and stretches of dry streambed).
-35- Upstream movement on the Salt River above Roosevelt Dam is further limited by Roosevelt Diversion Dam, which was built in 1906 to provide water for a hydroelectric facility supporting the construction of Theodore Roosevelt Dam. [Photo 36] Several accidents at the dam in the late 1970s and early 1980s led SRP to propose removing it. However, the Service and others argued against removal because, they said, Roosevelt Diversion Dam protected native fish in the upper Salt River by providing a barrier to movement of non-native fish out of Roosevelt Lake. Accordingly, Reclamation agreed to retain the dam and modified it to make it safer and a more effective fish barrier.93 [Photo 37] The dam, located just east of Roosevelt Lake on the Salt River, "restricts nonnative fish in Roosevelt Lake from coming up into the reach [designated as critical habitat for the razorback sucker]."94 Non-native buffalo fish, for example, "though established in Roosevelt Lake now for many decades, and occupying similar rivers within their native range, have clearly been prevented from penetrating the Salt River above Roosevelt Diversion Dam."95 Although the Service was instrumental in retaining Roosevelt Diversion Dam as a fish barrier, there is no mention of the structure in the Biological Opinion.
Numerous smaller dams and diversions on the Verde River above Horseshoe Lake also restrict upstream fish movement. These diversion dams range from "minor, bulldozed cobble dams, up to steel and concrete structures producing up to 3-5 m[eter] vertical, bank-to-bank, drops," many of which "would stop upstream movements."96
CAP deliveries below these mainstream dams are consistent with the Service's recovery plans for both the loach minnow and spikedace.97 Those recovery plans expressly allow stocking of non-native fish in rivers where endangered fish are located, provided such stockings are downstream of "the first absolute barrier to upstream fish movement into spikedace [or loach minnow] habitats."95 The locations from which fish might escape the CAP are downstream of such barriers on the Salt and Verde rivers.
The Biological Opinion acknowledges (at 18) that the six mainstream dams "block fish movement" between Granite Reef Diversion Dam and native fish habitat in the upper Salt and Verde rivers, but concludes that fish from the CAP would inevitably get around each and every one of these dams by "bait bucket transfer." However, human transport
SWCA, Inc., Final Environmental Assessment, Roosevelt Diversion Dam Modifications for Public Safety, Gila County, Arizona (1989).
94 USFWS, Colorado River Endangered Fishes Critical Habitat, Draft Biological Support Document (1993), at 95.
95 Hendrickson (1993), at 48 (citation omitted).
96 Id.
97 The Endangered Species Act requires the Service to develop and implement recovery plans that include specific management actions deemed necessary for the conservation and survival of endangered species. 16 U.S.C. § 1533(f).
95 Spikedace Recovery Plan, at 13; see Loach Minnow Recovery Plan, at 14.
-36- Photo 36: Roosevelt Diversion Dam on the Salt River
Salt River-Project • U.S. Bureau of Iteclamation--* V.S.VorestSerOce U.S. Fisk and Wildlife Service • Arizona tame aid FIIHDEPAILMNEUT
Photo 37: Plaque commemorating efforts to modify Roosevelt Diversion Dam to enhance native fisheries of fish from below a dam to above a dam is neither a direct nor indirect effect of the CAP and, therefore, cannot be considered in determining whether CAP water deliveries will jeopardize endangered species or adversely modify critical habitat.99 The Biological Opinion admits (at 20) that bait bucket transfer "would not result directly from CAP," so it cannot be a "direct" effect. And because bait bucket transfer is not "caused by the proposed action," it also cannot be an "indirect effect" of the CAP.199 Therefore, bait bucket transfer cannot properly be deemed an "effect" of the CAP at all, and it should not be considered here.
In evaluating its own activities, the Service routinely and properly ignores the possible effects of bait bucket transfer. For example, the Service authorized the establishment of an experimental population of woundfin in the upper Gila River after it concluded that the fish would not be able to migrate upstream into New Mexico because certain low-head irrigation dams "form insurmountable barriers to upstream movement of woundfin unless they are purposefully carried over them by man."191 Similarly, the Service concluded that its plan to stock large numbers of rainbow trout and channel catfish in and along the Colorado River would not jeopardize endangered fish located upstream of dams or in waters separated from the stocked fish.192
The only possible conclusion, therefore, is that the electrical fish barriers on SRP canals and the mainstream dams and other barriers on the Salt and Verde rivers will preclude any fish that might escape the CAP from reaching the native habitat of endangered fish.
B. The CAP Will Not Adversely Modify Critical Habitat in the Salt and Verde River Sub-basin.
Because fish from the CAP will be unable to reach native fish habitat in the upper Salt and Verde river basins, CAP water deliveries will have no adverse impact on designated critical habitat for the razorback sucker and spikedace.
V. The Gila River Sub-basin.
The Gila River originates in western New Mexico and flows southwest into Arizona near Duncan. In Arizona, the Gila turns northwest until it is joined by the San Francisco River near Clifton. Much of the flow in the San Francisco River that originates in Arizona
99 50 C.F.R. § 402.02 (definition of "effects of the action").
0 10 50 C.F.R. § 402.02 (emphasis supplied).
50 Fed. Reg. 30191 (1985).
102 FWS Opinion, at 14 ("There are no bonytail in the river below Parker Dam so only effects to razorback sucker will be considered"--ignoring effects on bonytail chub located above Parker Dam; "Stocking into Moovalya Pond likely would not affect the razorback sucker, unless stocked trout are moved from the pond to the river by anglers").
-37- is from the Blue River, which drains the higher elevations along the state's eastern border. From its confluence with the San Francisco River, the Gila River heads southwest toward the Safford Valley where it is joined by the San Simon River, which drains the northeastern slopes of the Chiricahua Mountains and the western slopes of the Peloncillo Mountains. The Gila River runs northwest from Safford to San Carlos Reservoir on the San Carlos Indian Reservation. Along the way, the Gila collects runoff from the Pinaleno and Santa Teresa Mountains on the south and the Gila Mountains on the north. At San Carlos Reservoir, the Gila is joined by the San Carlos River.
San Carlos Reservoir is formed by Coolidge Dam, a multiple dome concrete structure built and operated by the Bureau of Indian Affairs ("BIA"). [Photo 38] Coolidge Dam is 250 feet high and was completed in 1928 to provide water storage for the San Carlos Irrigation Project ("SCIP"), a BIA project that provides irrigation water to Indian and non-Indian lands in Pinal County. Water is released from Coolidge Dam for downstream delivery based on orders received from the Gila River Indian Community ("GRIC") and the San Carlos Irrigation and Drainage District ("SCIDD"). That water is then diverted into the Florence-Casa Grande Canal at Ashurst-Hayden Diversion Dam 8 miles east of Florence and 68 miles downstream from Coolidge Dam. [Photo 39]
Approximately 31 miles downstream from Coolidge Dam, the Gila River is joined by the San Pedro River. The San Pedro originates in desert grasslands of northern Sonora, Mexico, about 30 miles south of the international border. The river enters the United States near Palominas, Arizona, and flows northward for about 125 miles to its confluence with the Gila River near Winkelman. In Arizona, the San Pedro River drains the east slopes of the Huachuca, Whetstone, Rincon and Santa Catalina mountains, and the west slopes of the Mule, Dragoon and Galiuro Mountains.
For most of the reach from Benson north to the Gila River, the San Pedro is intermittent, flowing only in direct response to precipitation. South of Benson, in areas where the streambed intercepts the water table or where it is fed by springs, such as between Hereford and Fairbanks, flows may be continuous. Near Benson, the St. David Ditch and the Pomerene Canal divert water from the San Pedro River. At times of low flows, these diversions take the entire flow of the river.
The highest annual flows along the San Pedro occur from July to September in response to short-duration, high-intensity thunderstorms. These flows are "flashy" and are characterized by extremely rapid rises, high peak flow rates, and rapid declines back to baseflow conditions. Rainfall runoff also produces some flow from December to March, but these winter frontal storms are less flashy and generally produce much lower peak flow rates. From April to June the river is often dry, and low flows also occur in October and November. Near Redington, northeast of Tucson, the San Pedro River is dry almost 50% of the time and carries 1 cfs or less 62% of the time.
In general, the San Pedro River has a sandy bottom that provides little fish habitat in the form of pools, cover or resting space. The shifting sand bottom also does not
-38- Photo 38: Coolidge Dam on the Gila River Photo 39: Ashurst-Hayden Dam on the Gila River produce much algae or benthic invertebrates for fish to consume. Because most of the river is exposed to full sunlight, water temperatures fluctuate radically on a daily and seasonal basis.
The largest tributary of the San Pedro River is Aravaipa Creek, which drains the east slopes of the Galiuro Mountains and the west slopes of the Santa Teresa Mountains before joining the San Pedro north of Mammoth. The upper reaches of Aravaipa Creek are perennial, although flow is small during the dry seasons. The creek is often dry for three to four miles above its confluence with the San Pedro River due to diversions for agricultural103 use. This reach of the creek only flows during storm events of 20 cfs or more.
Critical habitat has been designated for the spikedace, loach minnow and razorback sucker in various reaches of the upper Gila River and its tributaries above Coolidge Dam. In addition, a fifteen mile reach of Aravaipa Creek has been designated as critical habitat for the spikedace and loach minnow. [See Figures 5, 6 and 8]
A. Non-native Fish from the CAP are Not Likely to Reach the Gila River.
The Biological Opinion postulates that non-native fish in CAP water delivered to GRIC and SCIDD would have access to the Gila River by either of two routes. The "direct" route, and the one "of highest concern" to the Service, is through the Florence- Casa Grande Canal, which supplies water diverted from the Gila River by Ashurst- Hayden Dam to the SCIDD and GRIC irrigation systems. The second, "indirect" route is through sumps, ponds or other standing water in the SCIDD or GRIC systems that are in the 100 to 500 year floodplain of the Gila River. Neither of these routes is plausible.
CAP water may be delivered to SCIDD and GRIC through any of four turnouts near Florence. [See Map 4] These turnouts can deliver CAP water to the Florence- Casa Grande Canal, Florence Canal, and Pima Lateral Canal. None of these turnouts is presently in use, as no water has been delivered to SCIDD or GRIC since 1990.
In discussing the potential for fish from the CAP to reach Ashurst-Hayden Diversion Dam through the Florence-Casa Grande Canal in 1989, the Service said: "This problem has already been identified at the Salt River at the CAP-Salt River Project connection and an electrical barrier has been placed to prevent upstream movement into the Salt River at the Granite Reef Diversion Dam."104 Following that advice, an electrical fish barrier like those used at the CAP/SRP interconnection was constructed on the Florence-Casa Grande Canal at China Wash, a short distance below Ashurst-Hayden Dam and above all CAP turnouts feeding the Florence-Casa Grande Canal. [Photo 40]
'°3 Molitor, BLM, pers. comm.
104 Field Supervisor, Ecological Services, USFWS, memorandum to Reclamation (8/28/89).
-39- Map 4 Gila River
Gila Ashurst-Hayden Diversion Dams
Pima N Fish Barrier Lateral A CAP Turnouts — Perennial/Regulated Streams — Intermittent/Ephemeral Streams
Florence
Florence-Casa Grande Canal
Plot Scale 1:180000 Picacho 0 1 2 3 4 s Reservoir I I 1 I I Miles Photo 40: Electrical fish barrier on Florence-Casa Grande Canal at China Wash
Photo 41: Headgate at Ashurst-Hayden Dam delivering water into Florence-Casa Grande Canal Thus, the "direct" route to the Gila River through the Florence-Casa Grande Canal is blocked by an electrical fish barrier.
Even if non-native fish from the CAP somehow made it through the electrical fish barrier at China Wash, they would still have to pass through the canal headgates at Ashurst-Hayden Dam. As at Granite Reef Diversion Dam, the water velocity through the headgates of Ashurst-Hayden Dam will prevent upstream movement by most fish, particularly smaller individuals. [Photo 41]
The Service has speculated that Picacho Reservoir could serve as a "holding facility" for non-native fish from the CAP.105 Non-native species such as channel catfish, black bullhead, carp, tilapia, largemouth bass and mosquitofish were recorded in Picacho Reservoir long before any CAP deliveries were made to the area.106 In addition, northern pike, largemouth bass, channel catfish and tilapia have all been stocked in Picacho Reservoir at one time or another.107 However, several of the species reported in Picacho Reservoir--black bullhead, northern pike and tilapia--are not found in the Gila River above Ashurst-Hayden Dam.108 The fact that these non-native fish have been present in Picacho Reservoir for many years but have shown no tendency to travel up the Florence-Casa Grande Canal, through Ashurst-Hayden Dam and into the Gila River indicates that the invasion route postulated by the Biological Opinion is improbable. With the recent addition of the electrical fish barrier on the Florence-Casa Grande Canal at China Wash, that invasion route has been made even more difficult and unlikely.
The "indirect" route to the Gila River is even less probable. The Service has admitted that the sumps, ponds and pools that might serve as a temporary harbor for non-native fish would only be connected to the Gila River during flood events.109 Those flood events, however, are far more likely to carry any non-native fish downstream, rather than permitting them to swim upstream against the flood."°
Moreover, Reclamation has sampled numerous sumps and ponds in the SCIDD and GRIC systems since the 1990 interim CAP deliveries and found no evidence of
1" Id.
106 Jakle & Baucom, An Inventory of Birds and Fish of Picacho Reservoir (1983).
1 07 AGFD Stocking Records.
108 Matter, at 7-8; Jakle, Memorandum to APO-150 Files (2/26/92); Jakle, Memorandum to APO-150 Files (2/10/93); Kepner, Jakle & Baucom, Middle Gila River Aquatic Study (1983).
1°9 Field Supervisor, Ecological Services, USFWS, memorandum to Reclamation (8/28/89) ("During a flood event when water comes over or around Ashurst-Hayden Dam, the pools and ponds in the river bed can connect with the river allowing fish to move up or downstream").
Draft Biological Support Document, Colorado River Endangered Fishes Critical Habitat (1993), at 91 ("Flood events on the Gila River and its tributaries in this reach results in the periodic removal or reduction in non-native fish populations as they are washed downstream during these events"); Minckley & Meffe (1987).
-40- striped bass, white bass, tilapia, or any other non-native fish that was not already established in the Gila River above Ashurst-Hayden Dam before CAP deliveries began."' Therefore, there is little basis for concluding that those bodies of water will serve as staging areas for upstream migration by non-native fish from the CAP.
Assuming non-native fish from the CAP could reach the Gila River and swim upstream through the flood waters to the base of Ashurst-Hayden Dam, they still would have to get over or around the 9-foot high dam.112 Although the Biological Opinion states (at 19) that "Ashurst-Hayden Dam is not an effective barrier to upstream fish movement," the Service has considered similar dams to be "insurmountable barriers" in other contexts. For example, the Service concluded that an experimental population of woundfin that it planned to establish in the upper Gila River would not be able to migrate upstream into New Mexico due to "three low-head irrigation dams located just inside the New Mexico border."113 "These dams," the Service stated, "form insurmountable barriers to upstream movement of woundfin unless they are purposefully carried over them by man.,014 The Service also considers Roosevelt Diversion Dam an effective barrier to upstream fish movement, even though it is no higher than Ashurst-Hayden Dam.115 Thus, Ashurst-Hayden Dam is a barrier that will limit the upstream movement of non- native fish in the Gila River.
In short, the Biological Opinion's conclusion that non-native fish from the CAP are a threat to invade the Gila River above Ashurst-Hayden Dam is not supported by the best available scientific and commercial data. Even if non-native fish from the CAP did reach the Gila River above Ashurst-Hayden Dam, they would not jeopardize any endangered species.
B. The CAP Cannot Affect Endangered Fish in the Gila River Above Coolidge Dam.
The Biological Opinion asserts (at 20) that CAP water deliveries will jeopardize endangered fish in the Gila River basin above Coolidge Dam. [See Map 5] Spikedace and loach minnow populations may be located in the upper Gila River in New Mexico and in the San Francisco and Blue rivers in eastern Arizona and New Mexico. Gila
Jakle memorandum to APO-150 Files (7/2/91); Jakle memorandum to APO-150 Files (12/12/91) Jakle memorandum to APO-150 Files (2/9/93).
112 The Service has speculated that during high flows fish may be able to "leap the dam," but the Biological Opinion provides no evidence of the leaping ability of any non-native fish that might be present in the CAP. Field Supervisor, Ecological Services, USFWS, memorandum to Reclamation (8/28/89).
113 50 Fed. Reg. 30191 (1985).
114 Id.
115 USFWS, Colorado River Endangered Fishes Critical Habitat, Draft Biological Support Document (1993), at 95 (Roosevelt Diversion Dam "restricts nonnative fish in Roosevelt Lake from coming up into the [upper Salt River]").
-41 -
Map 5 Gila River Above Coolidge Dam
• Spikedace • Loach Minnow • Razorback Sucker • Gila Topminnow Perennial/Ephemeral Streams Intermittent/Ephemeral Streams
Ashurst-Hayden CAP Diversion Dam Coolidge Dam Canal Bylas Ç. Florence-Casa Grande Canal
Florence Canal
VEL
Plot Scale 1:560,000
5 10 16 20 I I MIES topminnow have been reported in springs near Bylas, about 25 miles above Coolidge Dam. Razorback sucker were stocked in the Gila River above Coolidge Dam beginning in 1981, but may not have survived. Minckley did not locate any razorback sucker while surveying the Gila River in the mid-1980s.116
1. Non-native Fish from the CAP Cannot Reach the Gila River Above Coolidge Dam.
Coolidge Dam on the Gila River forms an insurmountable barrier to upstream fish movement."' As noted above, the spikedace and loach minnow recovery plans allow stocking of non-native fish downstream of barriers such as Coolidge Dam.118 No direct or indirect effect of the CAP will cause non-native fish to move over or around Coolidge Dam.
Even if non-native fish were transported around Coolidge Dam by anglers, there are additional barriers to upstream movement on the upper Gila River that were not considered in the Biological Opinion. In 1985, when the Service proposed the introduction of an experimental population of woundfin in the upper Gila, well upstream of Coolidge Dam, some commenters suggested that the woundfin could migrate upstream and adversely affect endangered spikedace populations. The Service ruled out any potential adverse effects, noting that "[t]he spikedace is presently found in the Gila River only as far downstream as 20 miles above the New Mexico/Arizona State line, which is a separation of at least 20 miles, 3 low-head dams and an intermittent section of river from the proposed woundfin reintroduction site."19 Coolidge Dam is at least 60 miles further downstream. Thus, any fish from the CAP that might be carried around Coolidge Dam would still have to travel 80 miles or more and surmount at least 3 dams and one or more intermittent river sections to reach spikedace in the upper Gila River basin.
The potential impact of non-native fish from the CAP is negligible in any event given the current proliferation of non-native fish in San Carlos Reservoir and the upper Gila River. AGFD has stocked more than 230,000 black crappie, bluegill, channel catfish, largemouth bass, rainbow trout and threadfin shad in San Carlos Reservoir, which is touted as one of the top sport fishing locations in the State.126 The Service itself
11 6 Draft Biological Support Document, Colorado River Endangered Fishes Critical Habitat (1993), at 87.
'1' Draft Biological Support Document, Colorado River Endangered Fishes Critical Habitat (1993), at 157-59 (Coolidge Dam "restrict[s] fish movement" and provides "blockage to Colorado River Fish passage"; noting generally that "[m]ajor mainstem dams completely block fish movement").
118 Spikedace Recovery Plan, at 13 Loach Minnow Recovery Plan, at 14.
11 9 50 Fed, Reg. 30191 (1985).
120 AGFD Stocking Records; San Carlos Apache Reservation tourism brochure (noting that the state record 65-pound flathead catfish was taken in San Carlos Reservoir).
-42- stocked more than 340,000 non-native fish on the San Carlos Indian Reservation from 1991 through 1993--at the same time that it was negotiating the draft biological opinion on the CAP with Reclamation.121
Both AGFD and the New Mexico Game and Fish Department ("NMGFD") have stocked rainbow trout and other non-natives throughout the upper Gila basin.122 AGFD has stocked channel catfish, flathead catfish, smallmouth bass and brook trout in the San Francisco River, portions of which have been designated critical habitat for the loach minnow.123 Rainbow trout, brown trout, arctic grayling and largemouth bass have been stocked in the Blue River, also critical habitat for the loach minnow.124 Eagle Creek, once proposed as critical habitat for the razorback sucker, has been stocked with largemouth bass, brown trout and rainbow trout, the latter as recently as 1993.125
As a result of all these stockings, the upper Gila River above San Carlos Reservoir already has an established non-native population of largemouth and smallmouth bass, sunfish, carp, channel catfish and flathead catfish.126 Nevertheless, the main channel of the Gila River does not sustain large populations of most non-native species due to periodic flooding and the lack of backwater habitats.127 Flood events on the Gila River and its tributaries above Coolidge Dam periodically remove or reduce non- native fish populations by washing them downstream.128 Non-native fish from the CAP would be subject to the same natural limitations, so no significant impact on endangered native fish could reasonably be expected.
2. The CAP Will Not Adversely Modify Critical Habitat in the Upper Gila River.
Because Coolidge Dam prevents upstream movement of non-native fish on the Gila River, any fish that are transported by the CAP will be unable to reach critical habitat in the upper Gila River basin. If anything, the CAP may have a positive impact on critical habitat for the razorback sucker. Critical habitat for that fish on the Gila River
121 USFWS Stocking Records.
122 AGFD Stocking Records; Appendix B for Loach Minnow and Spikedace Recovery Plans, at 35, 72 (NMGFD stocking activities).
1 28 AGFD Stocking Records.
1 24 Id.
1" Id.
1 26 USFWS, Colorado River Endangered Fishes Critical Habitat, Draft Biological Support Document (1993), at 89.
1 27 Id.
128 Id. at 91.
-43- includes San Carlos Reservoir to "full pool elevation."129 The water in San Carlos Reservoir is appropriated for downstream agricultural use by SCIDD and GRIC, and the lake rarely remains at full pool elevation. Delivery of CAP water to SCIDD and GRIC would reduce those entities' needs for water from San Carlos Reservoir. Indeed, the very purpose of the 1990 interim CAP deliveries to GRIC was to prevent a dry-up of San Carlos Reservoir, thereby protecting fish in that lake.13°
C. The CAP Will Not Affect Endangered Fish in the San Pedro River.
The Biological Opinion does not assert that the CAP will jeopardize endangered fish in the San Pedro River, but merely contends (at 19) that CAP water deliveries will harm "important recovery habitat" for native fish species in the upper San Pedro River basin. [See Map 6]
1. The Endangered Fish Were Extirpated from the San Pedro River Before CAP Water Deliveries Began.
With the exception of Aravaipa Creek (discussed separately below), the Biological Opinion does not identify any endangered or threatened fish populations in the San Pedro River basin. In fact, studies have shown that none of the four endangered fish the Biological Opinion says are jeopardized by CAP water deliveries currently occupy the San Pedro River.131 According to the BLM study, the last Gila topminnow in the San Pedro River were extirpated by 1979, long before any CAP water deliveries:32 Spikedace and loach minnow disappeared from the San Pedro more than thirty years ago, and the razorback sucker has not been recorded there for more than a hundred years:33 CAP water deliveries cannot jeopardize endangered fish species that are not present.
Even if there were endangered fish in the San Pedro River, fish from the CAP could not be expected to have any appreciable impact. Non-native fish have become firmly established in the San Pedro River without any help from the CAP. Table 7 compares the non-native fish reported in the Salt-Gila reach of the CAP with those already established in the San Pedro River:34
129 59 Fed, Reg. 13400 (1994).
130 Field Supervisor, Ecological Services, USFWS, memorandum to Reclamation (8/28/89), at 1.
1 31 Jackson, Martinez et al, Assessment of Water Conditions and Management Opportunities in Support of Riparian Values: BLM San Pedro River Properties, Arizona, at 38 (1987) ("San Pedro Study").
1 32 San Pedro Study, at 38
13' Id
134 Matter (1991), San Pedro Study.
-44- Map 6 San Pedro River and Aravaipa Creek
Coolidge San Carlos Reservoir Dam Ashurst-Hayden Dam CAP 4'1 Canal"' Florence-Casa Grande Canal Florence Canal vs,z,vaiPa
0
Spikedace Loach Minnow • USGS Gauging Station Perennial/Regulated Streams Intermittent/Ephemeral Stream
Plot Scale 1:700,000
5 10 15 20 26
Mlles United States/Mexico Border
IMMOOOMMI MMMMMM IDMMO
.40•4*INV Camel.. M.P. Salt-Gila San Pedro River Aqueduct (CAP) (Before CAP)
Threadfin shad X X
Carp X X
Flathead catfish X
Channel catfish X
Striped bass X
Green sunfish X X
Bluegill X X
Largemouth bass X X
Black crappie X
Rainbow trout X X
Black bullhead X
Yellow bullhead X
Mosquitofish X
Goldfish X
Brook trout X
Red shiner X
Fathead minnow X
Table 7: Comparison of non-native fish in CAP and San Pedro River (before CAP)
The only non-native species reported in the CAP but not already present in the San Pedro River are striped bass, flathead catfish and black crappie.
History indicates that neither flathead catfish nor black crappie can be expected to move up the San Pedro River from the area of the Gila River. Both species have been present in San Carlos Reservoir for many years and therefore potentially able to move downstream in the Gila River and then upstream into the San Pedro River. The Arizona Game and Fish Department stocked more than 66,000 black crappie in San Carlos Reservoir from 1935 to 1978.135 Flathead catfish have also been recorded there for some time, with the state record 65-pound flathead catfish being caught in San
1 35 AGFD Stocking Records.
-45- Carlos Reservoir in 1951.136 The confluence of the San Pedro and Gila rivers is only about 31 miles downstream from San Carlos Reservoir. By contrast, the mouth of the San Pedro River is 37 miles upstream from Ashurst-Hayden Dam, where the Biological Opinion speculates that fish leaving the CAP might enter the Gila River. Given that black crappie and flathead catfish have had ready access to the San Pedro River for at least forty to sixty years, yet have not been recorded there, it is unreasonable to conclude that those species would escape the CAP and invade the San Pedro River.
Neither can striped bass reasonably be expected to invade the San Pedro River. Contrary to the assertion in the Biological Opinion (at 19), Matter did not conclude that striped bass were "likely to move throughout the Gila and San Pedro Rivers." Rather, Matter indicated that striped bass were unlikely to invade the San Pedro River and Aravaipa Creek: "There are no data to suggest that striped bass have or will select for conditions likely to occur in the San Pedro River."137 Studies have shown that striped bass may approach and enter cool streams when reservoir water becomes too warm.138 However, the Gila River is almost invariably cooler than the San Pedro in summer months.139 Temperatures in the San Pedro River can often reach 25°C or 30°C, making the river potentially lethal for adult striped bass and inhospitable for juveniles.140 In general, the available data indicates that striped bass are not likely to venture into a small stream such as the San Pedro River."' The Biological Opinion cited no data to the contrary.
The intermittent nature of the San Pedro River makes it even less likely that fish from the CAP could move from the Gila River to the upper reaches of the San Pedro. Near Redington, Arizona, northeast of Tucson, the San Pedro River is dry more than half of the time.142 As the Service has acknowledged elsewhere, dry streambeds present a barrier to fish movement.143 Further downstream, at its confluence with the Gila River, the San Pedro River flows at 3 cfs or less 50 percent of the time. When higher flows result from rainfall events, the water in the San Pedro River tends to be heavily silted.
13' Arizona Fish Records (1994) (the state record black crappie is also from San Carlos Reservoir).
137 Matter, at 52-53.
138 Matter, at 28, 32.
1 39 USGS data for station nos. 09474000 & 09473100.
140 USGS data for station nos. 09471000 & 09473100; Matter, at 27-28 (striped bass "will avoid such temperatures if possible")
141 Matter, at 32 (although striped bass are common in the Colorado River, they are not found within major tributaries such as the Bill Williams River and the San Juan River).
142 USGS data for station no. 09472000 (80% of flows are less than 10 cfs).
143 50 C.F.R. § 17.84(b).
-46-
In addition, there are several small dams and diversions that may restrict fish movement into the upper San Pedro River. The Pomerene Water Users Association maintains a concrete diversion dam in the San Pedro River approximately one-half mile downstream of the Highway 80 bridge near St. David. A few miles upstream, the St. David Irrigation District maintains an earthen dam that also diverts water from the San Pedro River for agricultural uses.'"
In short, the best available scientific and commercial data compels the conclusion that CAP water deliveries in the vicinity of the Gila River will not jeopardize endangered fish in the San Pedro River.
2. No Critical Habitat Has Been Designated in the San Pedro River.
No critical habitat has been designated in the San Pedro River for any of the species as to which the Biological Opinion found jeopardy. Although the Biological Opinion refers (at 19, 25) to the upper San Pedro River as "important" or "promising" recovery habitat for endangered fish, as discussed above, the Endangered Species Act affords no protection to habitat unless it has been designated by the Service as "critical habitat"--not merely "important" or "promising" habitat. Accordingly, there is no basis for finding that CAP water deliveries will adversely modify critical habitat in the San Pedro River.
D. The CAP Cannot Reasonably Be Expected to Affect Native Fish or Critical Habitat in Aravaipa Creek.
Although the Biological Opinion postulates that fish escaping the CAP would move up the Gila River to the San Pedro, then up that river to Aravaipa Creek, and up Aravaipa Creek to spikedace and loach minnow habitat, non-native fish already long established in the Gila and San Pedro rivers have had similar (if not better) opportunities to invade Aravaipa Creek but have not done so. [See Map 6]
If the Biological Opinion were correct, all non-native fish found in the middle and upper Gila River should also be found in Aravaipa Creek, since those fish have had ready access to the San Pedro River for many years. However, there are many non- native species resident in the Gila River above Ashurst-Hayden Dam that have not been recorded in Aravaipa Creek, including carp, black crappie, goldfish, threadfin shad, fathead minnow, channel catfish, flathead catfish, bluegill and rainbow trout.145
Likewise, non-native fish found in the upper San Pedro River need only travel downstream to the confluence of Aravaipa Creek and then upstream to critical habitat, a route far easier than that suggested for fish transported by the CAP. Contrary to the
144 Arizona Department of Water Resources, Water Service Organizations in Arizona (1983).
145 AGFD Stocking Records; Matter, at 7-8; Jakle memorandum to APO-150 Files (2/26/92).
-47- Biological Opinion, however, rainbow trout, carp, threadfin shad, channel catfish, bluegill, goldfish and fathead minnow have all been present for a considerable time in the San Pedro River, but have not invaded Aravaipa Creek.146
According to the Service, catfish are the most likely non-natives to co-occur and interact with adult loach minnow and spikedace populations.147 As shown above, both flathead catfish and channel catfish have been present for many years in the Gila and San Pedro rivers, yet neither has shown an ability to invade Aravaipa Creek. Accordingly, catfish that may be transported by the CAP cannot reasonably be expected to affect native fish in Aravaipa Creek.
Red shiner have been identified as the most serious non-native threat to spikedace and, along with mosquitofish, are also considered the most likely non-native to co-occur with small loach minnow.148 The Service has acknowledged that the red shiner is present virtually everywhere that the spikedace once lived and would be impossible to eliminate.149 Red shiner and mosquitofish have both been recorded in Aravaipa Creek.15° CAP deliveries did not cause and will not further this threat to the endangered fish of Aravaipa Creek.151
The Biological Opinion directs (at 27) Reclamation to construct fish barriers on Aravaipa Creek. However, both public and private entities had called for construction of a fish barrier on Aravaipa Creek long before anyone raised the possibility that non- native fish from the CAP might invade the Gila River. AGED, for example, suggested such a barrier in commenting on draft recovery plans for spikedace and loach minnow in September 1988.152 The need for a barrier on Aravaipa Creek was also identified in
146 San Pedro Study, at 38; Matter, at 7-8.
147 Loach Minnow Recovery Plan, at 5 ("Among non-native (introduced) fishes that co-occur in places with adult loach minnow, only ictalurid catfishes are likely to interact strongly with the native"); Spikedace Recovery Plan, at 6 (channel and flathead catfish "frequent riffles occupied by spikedace").
148 Spikedace Recovery Plan, at 6; 59 Fed. Reg. 10910 (1994) ("the red shiner appears to displace the spikedace in some locations and is considered a serious range-wide threat to the spikedace"); Loach Minnow Recovery Plan, at 6.
149 Spikedace Recovery Plan, at 6-7; see 50 Fed. Reg. 30191 (1985) ("Nearly every stream within historic range of the woundfin [which includes the Gila and Salt rivers] has been invaded by the red shiner, and removal of the shiner from selected reintroduction sites would be virtually impossible at worst and temporary at best").
159 Matter, at 7-8.
151 Appendix B for Loach Minnow and Spikedace Recovery Plans, at 33 (AGFD commented in 1988 that spikedace and loach minnow in Aravaipa Creek are "imminently threatened with invasion by Notropis lutrensis [red shiner]"); Sonoran Topminnow Recovery Plan, at 7 (mosquitofish were introduced into Arizona waters in 1926 and spread rapidly throughout the Southwest).
152 Appendix B for Loach Minnow and Spikedace Recovery Plans, at 33 (noting that private property owners also had expressed interest).
-48- the recovery plans for loach minnow and spikedace, but neither plan suggested that the CAP was to blame for any of the factors threatening those species.153
In November 1989--eighteen months before the initial draft of the Biological Opinion was prepared--Service personnel met with interested parties to discuss plans for constructing a fish barrier on Aravaipa Creek. Those present agreed that the chief impediment to constructing such a barrier was a lack of funds. To address that problem, the Service stated that it would request a "contribution" from Reclamation for the barrier project.154 That request has now been presented in the form of reasonable and prudent alternatives in the Biological Opinion.
VI. The Biological Opinion Violated Service Regulations.
A. The Biological Opinion Exceeded the Scope of the Consultation.
A biological opinion is the end result of the formal consultation process outlined in 40 C.F.R. § 402.14. Formal consultation occurs only when requested by a federal action agency. The Service has no authority to issue biological opinions on its own initiative. The request by the federal agency establishes the parameters of the formal consultation by describing the action to be considered and the specific area, listed species and critical habitat that may be affected.155
In this instance, the Biological Opinion states (at 1) that it was prepared in response to a February 1991 request by Reclamation for formal consultation "on the proposal to provide Central Arizona Project (CAP) water to Indian and non-Indian water users in central Arizona." That request for formal consultation concerned CAP deliveries to GRIC and SCIDD only.156 Indeed, the proposal to deliver water to GRIC and SCIDD is what triggered consultation on the introduction of non-native fish through the CAP.157 The biological assessment that accompanied Reclamation's request for formal consultation also was limited to deliveries to GRIC and SCIDD, and the scope of that report had been reviewed and approved by the Service in advance.158 Thus, the only issue properly addressed in the Biological Opinion is the effect of CAP water deliveries to GRIC and SCIDD on listed species or critical habitat.
'53 Spikedace Recovery Plan, at 13 (1991); Loach Minnow Recovery Plan, at 14 (1991).
154 AGED, Memorandum to Interested Parties (12/19/89).
155 40 C.F.R. § 402.14(c).
156 Reclamation, Memorandum to Field Supervisor, Ecological Services, USFWS (2/12/91).
157 Field Supervisor, Ecological Services, USFWS, memorandum to Reclamation (8/28/89).
158 Acting Field Supervisor, Ecological Services, USFWS, memorandum to Reclamation (4/10/90); Matter, at 3.
-49- Significantly, Reclamation's February 1991 memorandum did not request consultation on CAP water deliveries to Harquahala Valley Irrigation District, Tonopah Irrigation District or the Salt River Project--all of which had been the subject of previous consultations.159 Nor did Reclamation request consultation on storage of CAP water in the expanded Lake Pleasant. That issue had already been addressed during formal consultation on the Central Arizona Water Control Study, and the Service had issued a biological opinion finding no jeopardy.169 By addressing deliveries of CAP water to HV1D, TID and SRP and storage of CAP water in Lake Pleasant, the Biological Opinion went beyond the scope of the consultation requested by Reclamation and, therefore, exceeded the Service's authority.
B. The Biological Opinion Improperly Determined the Effects of the CAP.
Service regulations state that the effects of a proposed federal action are to be measured in light of an "environmental baseline" that includes the impacts of past and present federal, state and private actions, as well as proposed federal projects that have already undergone Section 7 consultation.161 That definition presumes that past federal actions, like proposed federal projects, have undergone the requisite Section 7 consultation:
"In issuing its biological opinion on an action, the Service's finding under section 7(a)(2) entails an assessment of the degree of impact that action will have on a listed species. Once evaluated, that degree of impact is factored into all future section 7 consultations conducted in the area.
These impacts062 will continue to be considered as part of the environmental baseline . . . .
Thus, Section 7 "provides a 'first-in-time, first-in-right' process whereby the authorization of federal projects may proceed until it is determined that further actions are likely to jeopardize the continued existence of a listed species or adversely modify its critical habitat."163
159 See Area Manager, USFWS, letter to Franzoy, Corey and Associates (9/14/81) (stating that "no listed or proposed species would be affected by the proposed action in the area of interest [HVIDI"); Field Supervisor, Ecological Services, USFWS, Memorandum to Reclamation (9/13/84) ("no listed, proposed or candidate species would be affected by the proposed action in the area of interest [TID]"); Field Supervisor, Ecological Services, USFWS, Letter to SRP (11/16/84) ("We do not envision any potential environmental impacts associated with the proposed interconnection facilities"); Field Supervisor, Ecological Services, USFWS, Letter to SRP (10/28/87) (same).
160 USFWS, Biological Opinion, Central Arizona Water Control Study (3/8/83).
161 50 C.F.R. § 402.02 (definition of "effects of the action").
162 51 Fed. Reg. 19932 (1986) (emphasis supplied).
163 Solicitor's Opinion, 88 I.D. 903, 905 (1981).
-50- In evaluating the impact of the CAP, the Biological Opinion included in the environmental baseline the effects of all past introductions of non-native fish into the Gila River basin. Most, if not all, of the non-native stocking activities in the Gila River basin have been funded, managed, controlled by or connected with the Service or other federal agencies, yet none of those actions were the subject of Section 7 consultations. For example, both NMGFD and AGFD have stocked millions of non-native fish into the Gila River basin, often upstream of endangered fish populations.164 Those stocking programs are heavily supported by federal funds. In fiscal years 1991 through 1993--while the Service was negotiating RPAs with Reclamation--it provided more than $13 million to AGFD under the Federal Aid in Sport Fish Restoration Program. Almost $8 million of that went to support hatchery and stocking operations designed to augment sport fish populations within the state.166 Until April 1994, however, the Service did not require Section 7 consultations on those stocking programs.
Moreover, the Service itself continues to stock non-native fish throughout the Gila River basin. For fiscal years 1991 through 1993, the Service stocked more than 1,379,000 non-native fish on Indian reservations, military posts and national forests in the Gila River basin.166 Under the reasoning of the Biological Opinion, those non-native fish may affect endangered species. However, there is no indication that the Service conducted Section 7 consultations or issued biological opinions on its stocking activities in the Gila River basin. By including in the environmental baseline activities funded, managed or controlled by the Service that may in themselves cause jeopardy to endangered fish but which were not subjected to Section 7 review, the Biological Opinion violates the "first-in-time, first-in-right" process and effectively shifts the burden of past and ongoing Service programs onto the CAP.
In determining whether CAP water deliveries would jeopardize endangered fish, the Biological Opinion also considered the cumulative impacts of future state and private activities such as non-native fish stocking, including unauthorized private stockings, and the release of aquaculture, aquarium and bait fish. The Biological Opinion concluded (at 26) that these "cumulative effects"--which, in fact and by definition, are unrelated to the CAP--"will have a strong adverse effect on the listed species being addressed here" and, taken as a whole, "may result in severe losses or extirpation of populations of the listed fish."167
The treatment of cumulative effects in the Biological Opinion ignores the fact that most, if not all, non-native stockings by state agencies in the Gila River basin are either federally funded or take place on federal land and, therefore, are subject to the
164 See Appendix B for Loach Minnow and Spikedace Recovery Plans, at 35, 72; AGFD stocking records.
165 AGFD Annual Reports.
1 66 USFWS Stocking Records.
167 See 50 C.F.R. § 402.02 (definition of "cumulative effects").
-51- requirements of Section 7 of the Endangered Species Act. Moreover, events such as "unauthorized stockings," the "release of unwanted aquarium and aquaculture fish," and the "escape" of non-native fish from aquaculture operations are speculative and are not to be considered by the Service in issuing a biological opinion on the CAP.168 Only events that are "reasonably certain to occur" may be considered as cumulative effects under Service regulations.169 This requirement was explained in a solicitor's opinion:
"A non-federal action is 'reasonably certain' to occur if the action requires the approval of a state or local resource or land use control agency and such agencies have approved the action, and the project is ready to proceed. Other indicators which may also support such a determination include whether the project sponsors provide assurance that the action will proceed, whether contracting has been initiated, whether there is obligated venture capital, or whether State or local planning agencies indicate that grant of authority for the action is imminent. These indicators must show more than the possibility that the non-federal project will occur; they must demonstrate with reasonable certainty that it will occur. The more that state or local administrative discretion remains to be exercised before a proposed state or private action can proceed, the less there is reasonable certainty that the project will be authorized. In summary, the consultation team should consider only those state or private projects which satisfy all major land use requirements and which appear to be economically viable."170
By including the cumulative effects of future federal actions as well as non-federal activities that are not "reasonably certain to occur," the Biological Opinion effectively guaranteed a finding of jeopardy, regardless of any actual impact the CAP might have on endangered fish.
By combining past federal activities that were not reviewed under Section 7 (the "environmental baseline"), future activities that are speculative and unrelated to the CAP ("cumulative effects"), and activities such as bait bucket transfer that are not effects of the CAP at all, the Service perverted and utterly foiled the directive of the Endangered Species Act to focus on the effects of the proposed federal action under consideration. Section 7 requires the Service to determine whether a particular federal action, the CAP in this case, will jeopardize an endangered species or result in the destruction or adverse
168 51 Fed. Reg. 19933 (1986) ("Congress did not intend that Federal actions be precluded by such speculative actions," describing speculative actions to include even situations "where proposals have been made, and implementation schedules have been established").
169 50 C.F.R, § 402.02.
° Solicitor's Opinion, 88 I.D. at 908 (emphasis in original).
-52- modification of critical habitat.171 The Biological Opinion ignores that statutory mandate by including the impact of past and future activities that are entirely unrelated to the CAP.
C. The Biological Opinion Did Not Consider the Best Available Scientific and Commercial Data.
As noted at the outset, the Biological Opinion declined (at 21) to undertake any "[s]pecies-specific analysis of the likelihood of non-native species entering and successfully establishing in the Gila River basin, and the potential for impacts to each listed species," stating that such analysis "would be lengthy, complex, and largely nonproductive." Thus, the Biological Opinion did not analyze available hydrologic data showing, for example, that non-native fish could not reasonably be expected to reach upstream areas on the Hassayampa River. Nor did the Biological Opinion evaluate the environmental requirements of non-native fish species to determine, for example, whether striped bass could survive in the San Pedro River or Aravaipa Creek. The Biological Opinion also failed to consider the effects of non-native stocking programs funded by the Service which make any potential impact by the CAP negligible. Likewise, the Biological Opinion did not investigate the design of the CAP or its customers' delivery systems to assess the probability that non-native fish would be able to escape into natural watercourses. In short, the Biological Opinion ignored overwhelming scientific and commercial data indicating that CAP water deliveries will not jeopardize endangered fish or adversely modify critical habitat. [See Appendix 2 for a list of sources not considered in Biological Opinion]
D. The Biological Opinion Used a Double Standard in Evaluating the CAP.
In evaluating the potential impact of CAP water deliveries, the Service applied a much different--and much stricter--standard than it uses in judging its own activities. For example, on July 1, 1994, the Service issued a biological opinion (the "FWS Opinion") which concluded that the agency's stocking nearly 200,000 rainbow trout and channel catfish in the Colorado River in and adjacent to designated critical habitat for the razorback sucker and bonytail chub would not jeopardize the endangered fish or adversely modify critical habitat.172 To reach that conclusion, the Service made a number of assumptions and conclusions that are directly contrary to assumptions and conclusions expressed in its Biological Opinion on the CAP.
The Biological Opinion assumed that dams are essentially irrelevant--i.e., that any non-native fish entering the lower reaches of the Gila River basin would inevitably reach
171 16 U.S.C. § 1536(a)(2).
172 USFWS, Biological Opinion on Fish and Wildlife Stocking of Rainbow Trout and Channel Catfish in the Lower Colorado River (Hoover Dam to the International Border), July 1, 1994 ("FWS Opinion").
-53- areas upstream of dams and negatively impact endangered fish. The FWS Opinion, on the other hand, concluded that dams block fish movement and that areas upstream of dams were not a concern.'"
Similarly, the Biological Opinion's jeopardy finding with respect to endangered fish and critical habitat located upstream of dams on the Salt, Verde and Gila rivers is wholly dependent on bait bucket transfer by fishermen. Absent bait bucket transfer, there is no basis for jeopardy. The FWS Opinion, however, found no jeopardy because stocked trout could not reach endangered fish unless moved by anglers.174 Thus, the Service assumed that bait bucket transfer would occur and was a problem with the CAP, but would not occur and was not a problem for its own project.
The Biological Opinion found jeopardy because non-native fish introduced into irrigation systems (e.g., HVID, TID, GRIC, SCIDD) might reach flowing rivers during flood events. By contrast, the FWS Opinion found no jeopardy where non-native fish were to be stocked in waters "with no direct connection to the lower Colorado River."175
The Biological Opinion assumed that any non-native fish that escape the CAP will move through all occasionally connected waters until they reach endangered fish. The FWS Opinion, however, assumes that movement of non-native fish will be limited and, therefore, not a threat.176 The Biological Opinion did not bother to conduct any species- specific analysis of the ability--or inability--of non-native fish to move from the CAP to waters where endangered fish are located.
The Biological Opinion assumed that any non-native fish will have a negative effect (e.g., predation, competition) on any endangered fish in the vicinity, but expressly avoided any species-specific analysis of such impacts.177 The FWS Opinion, on the other hand, demanded "direct evidence" of predation before finding any adverse impacts. For example, the FWS Opinion concluded that stocking rainbow trout in the Colorado River below Hoover Dam will not adversely modify bonytail chub critical habitat because "Nainbow trout predation on bonytail chub has not been documented," even though the
173 FWS Opinion, at 14 ("There are no bonytail in the river below Parker Dam so only effects to razorback sucker will be considered"; ignoring effects on bonytail chub located above Parker Dam).
1 74 FWS Opinion, at 14 ("Stocking into Moovalya Pond likely would not affect the razorback sucker, unless stocked trout are moved from the pond to the river by anglers").
1 75 FWS Opinion, at 1 ("the stocking of channel catfish in tanks and other waters with no direct connection to the lower Colorado River is not likely to adversely affect the razorback sucker arid bonytail chub or their designated critical habitat"), 14 ("Stockings into isolated waters that do not contain razorback suckers avoid the issue of predation").
176 FWS Opinion, at 14-15 ("Lake Havasu is too warm to allow the survival of rainbow trout over the summer"; "the rainbow trout in the Davis Dam to Needles area do not move out of the area due to high temperatures below the Interstate 40 bridge near Topock"; "There are no razorback suckers or bonytails in the vicinity of these stockings or that could be reached by individual rainbow trout or channel catfish leaving the stocking area").
177 Biological Opinion, at 21 ("Species-specific analysis of... the potential for impacts to each listed species would be lengthy, complex, and largely nonproductive").
-54- opinion notes elsewhere that rainbow trout are known to prey on humpback chub, razorback sucker and other species.178 Likewise, the FWS Opinion states (at 15) that stocking of rainbow trout will not adversely modify critical habitat for the razorback sucker in Lake Mohave because rainbow trout are only known to prey on razorback larvae and Lake Mohave was designated critical habitat because of its population of adult razorbacks. The Service did not perform such a detailed analysis for its opinion on the CAP, much less require "direct evidence" of predation by each non-native species of concern before finding jeopardy.
The Biological Opinion did not differentiate between stocked and naturally occurring populations of endangered fish. In some areas (e.g., Hassayampa, Agua Fria), the only threat identified from CAP was to stocked populations of Gila topminnow. In contrast, the FWS Opinion apparently distinguished stocked from "wild" populations of endangered fish, finding no jeopardy where only a stocked population would be affected. For example, the Service plans to stock 600 rainbow trout per month in 12-Mile Lake, where it also stocked razorback sucker in 1993.178 Nevertheless, the FWS Opinion concluded that "[s]tocking into 12-Mile Lake would not likely affect the wild population of razorback suckers since it is removed from the river."180
Had the Service applied the same standard to the CAP that it applied to its own activities in the FWS Opinion, the Biological Opinion would have found no jeopardy and no adverse modification of critical habitat.
Conclusion
The significant differences between the Biological Opinion on CAP water deliveries and the biological opinion on the Service's plans to stock rainbow trout and channel catfish in the Colorado River highlight the defects of the Biological Opinion. The Biological Opinion adopted a very broad definition of jeopardy in evaluating the effects of the CAP, concluding that the mere possibility of harm--however remote--was sufficient to find jeopardy. In evaluating its own sport fishing projects, however, the Service has taken a narrower, more reasonable view of jeopardy. According to the Service, the FWS Opinion "shows the Endangered Species Act is flexible and can accommodate the needs of endangered native fish and the angling public along the river."181
The Service has also displayed a flexible attitude toward sport fishing in its treatment of state game and fish agencies operating under the Federal Aid in Sport Fish Restoration program. In Arizona, the Service did not perform mandatory Section 7
178 FWS Opinion, at 15, 13 ("No direct evidence of predation by rainbow trout on bonytail is available).
179 FWS Opinion, at 3, 6.
18° FWS Opinion, at 14 (emphasis supplied).
181 USFWS news release (7/1/94) (quoting USFWS Southwest Regional Director).
-55- consultations with respect to federally supported non-native stocking activities by AGFD before April 1994. Significantly, the Service has permitted recent stockings of non-native fish in the lower Salt and Verde rivers above Granite Reef Diversion Dam, even though it concluded in the Biological Opinion that non-native fish in CAP water delivered below Granite Reef Diversion Dam would jeopardize endangered species and adversely modify critical habitat in the upper Salt and Verde river systems.
While flexibility is certainly a commendable trait, the Service should apply the same standard to all federal projects, not merely those that it sponsors or from which it may derive income. Clearly, when CAP activities are judged by the same standard applied to Service projects and state game and fish programs, the jeopardy finding of the Biological Opinion cannot stand.
The aberrations of the Biological Opinion, in conjunction with the specified reasonable and prudent alternatives, create the strong impression that the primary purpose of the jeopardy finding on the CAP was to generate additional funding for Service projects without having to rely on Congressional appropriations. Significantly, nine potential RPAs in the Biological Opinion were rejected in favor of or replaced by transfers of funding to the Service. In total, the Biological Opinion demands the transfer of $500,000 per year for twenty-five years to fund the Service's research and conservation activities, and demands the expenditure by CAP of some $132 million over 100 years to monitor the presence of non-native fish throughout central Arizona. By comparison, the the Service has estimated in its recovery plans for the spikedace and loach minnow that the total cost of recovering those two species would be $230,000 per year for twenty years. Assuming similar figures for the Gila topminnow and razorback sucker, it would appear that the CAP is being required to assume the full financial burden for the recovery of all four endangered species--and then some--even though the Biological Opinion concedes that the CAP is, at most, one small link in the chain of events that has led to the demise of those species and will continue to pressure them in the future.
A similar pattern appears in the Service's consultations regarding the Lake Havasu Fisheries Improvement Program ("LHFIP"). The primary goal of LHFIP is to increase the number of non-native sport fish in Lake Havasu--critical habitat for bonytail chub and also occupied by razorback sucker--by placing 67,482 bass shelters, 54,724 catfish houses, 3,484 bass ambushing cover structures, 1,050 tire towers, and 11,800 brush bundles in 42 different locations around the lake.182 The Service acknowledged that the "project's goal of improving stocks of non-native fish will likely enhance the opportunity for predation on eggs and larvae of bonytail chub and razorback sucker."183 Nevertheless, the Service allowed LHFIP to go forward after its proponents agreed to
182 BLM, Lake Havasu Fisheries Improvement Partnership Program.
183 Field Supervisor, Ecological Services, USFWS, memorandum to BLM (8/28/92).
-56- create up to ten native fish rearing coves for bonytail chub and razorback sucker in Lake Havasu.'"
From these examples, it would seem that the Service is using the threat of a jeopardy opinion to force other agencies to fund the Service's own research and conservation activities. If that is so, and there is every reason to believe that it is, it is an abuse by the Service of the powers granted it under the Endangered Species Act, an abuse that cannot fail to discredit the Service's legitimate efforts to protect endangered species and lead to amendment of the Act itself to prevent such abuses from occurring in the future.
184 USFWS, Biological Opinion for Lake Havasu Fisheries Improvement Partnership Program (2/18/93).
-57- APPENDIX 1
USFWS 1994 BIOLOGICAL OPINION ON CAP United States Department of the Interior
FISH AND WILDLIFE SERVICE P 0. Box 1306 Albuquerque, New Mexico 87103 In Reply Refer To: R2/ES-SE APR 201994 2-21-90-F-119
MEMORANDUM
To: Regional Director, Bureau of Reclamation, Boulder City, Nevada
From: Regional Director, Region 2
Subject: Final Biological Opinion on the Transportation and Delivery of Central Arizona Project Water to the Gila River Basin (Hassayampa, Agua Fria, Salt, Verde, San Pedro, middle and upper Gila Rivers, and associated tributaries) in Arizona and New Mexico
Attached is the Fish and Wildlife Service's final biological opinion on the subject formal consultation under section 7 of the Endangered Species Act of 1973, as amended. This opinion finds that the action would jeopardize the continued existence of the spikedace (Meda fulgida), loach minnow (Tiaro_ga cobitis), Gila topmirinow (Poeciliopsis occidentalis), and razorback sucker (Xyrauchen texanus) and would adversely modify the critical habitat of the spikedace, loach minnow, and razorback sucker. The reasonable and prudent alternative given in this opinion is the product of over 3 years' of negotiation between the Fish and Wildlife Service and the Bureau of Reclamation (BR). We appreciate the efforts of BR to finding a way to conserve the listed species and look forward to working with you on implementation of this opinion. If we can be of further assistance, please contact Sally Stefferud or Tom Gatz in the Arizona Ecological Service State Office, at (602) 379-4720.
Attachment cc: Project Manager, Bureau of Reclamation, Phoenix, AZ Director, U.S. Fish and Wildlife Service, Washington, D.C. (DES) State Supervisors, Ecological Services State Offices, Arizona and New Mexico Project Leader, U.S. Fish and Wildlife Service, Pinetop, AZ Assistant Regional Director - Endangered Species, Region 2 2-21-90-F-119
SUMMARY BIOLOGICAL OPINION ON TRANSPORTATION AND DELIVERY OF CENTRAL ARIZONA PROJECT WATER TO THE GILA RIVER BASIN IN ARIZONA AND NEW MEXICO
Date of the opinion: April 15, 1994
Action agency: Bureau of Reclamation
Project: Transportation and delivery of Colorado River water through the Central Arizona Project (CAP) to various water users in the Gila River basin (excluding the Santa Cruz River subbasin). This biological opinion addresses only the potential of this project to introduce and spread non-native aquatic species. The Santa Cruz subbasin will be the subject of additional formal consultation.
Listed species and critical habitats: Spikedace ( Meda fulaida), loach minnow (TiarOaa cobitis), Gila topminnow ( Ppeciliopsis occidentalis occidentalis), razorback sucker (Xvrauchen texanus), desert pupfish ( Cvprinodon macularius), Colorado River squawfish (Ptvchocheilus lucius), and bald eagle (Haliaeetus lgucocephalus). Critical habitats for spikedace, loach minnow, and razorback sucker.
Biological opinion: Jeopardy for spikedace, loach minnow, Gila topminnow, and razorback sucker. Adverse modification of critical habitat for spikedace, loach minnow, and razorback sucker. (page 1)
Reasonable and prudent alternative (RPA): Implementation of the RPA is necessary to remove the threat of jeopardy from the proposed action. Construction of 4 drop-structure barriers (2 on Aravaipa Creek, 2 on San Pedro River), continued operation of 3 existing electrical barriers on canals, monitoring of non-native fish in specific areas of middle Gila basin and canals, transfer of $500,000 annually to FWS for conservation of Gila basin native fishes and research and non-native fish control, development and implementation of an information and education program about the adverse effects of non-native fish. (pages 26 to 29)
Incidental take statement: Level of take anticipated: Anticipated take is unquantifiable, but will be assumed to have been exceeded if proposed action, as modified by RPA, is altered or not carried out. If the anticipated incidental take is exceeded, consultation must be reinitiated. (pages 29-30)
Reasonable and prudent measures and terms and conditions: Implementation of the RPA. Terms and conditions are mandatory requirements. (page 30)
Conservation recommendations: Implementation of conservation recommendations is discretionary. Construction of 4 drop-structure barriers, encouraging dry-up of CAP connected irrigation canals and other features and management of non-native fishes in those not appropriate for dry-up, organization and facilitation of multi-agency efforts to address conflicts between sport fishing and native fish conservation, and opposition to introduction of additional non-native aquatic species in the lower Colorado River basin. (page 31)
Additional section 7 consultation needs: Further consultation will be required for effects the Santa Cruz River basin (excluding the Santa Rosa Canal system. 2
BACKGROUND INFORMATION
Species Descriptions
Spikedace
The spikedace was listed as a threatened species on July 1, 1986. Critical habitat was designated for the spikedace on March 8, 1994, and includes portions of the Verde and Gila Rivers and Aravaipa Creek. The spikedace is a small silvery fish, with the common name alluding to the well-developed spine on the dorsal fin (Hinckley 1973). Spikedace originally existed throughout much of the Gila River drainage above Phoenix, but is currently known only from Aravaipa Creek (Graham and Pinal Counties, Arizona), the upper Gila River (Grant and Catron Counties, New Mexico), the Verde River (Yavapai County, Arizona), and Eagle Creek (Greenlee County, Arizona). A recent record of the spikedace also exists from the middle Gila River upstream from Ashurst-Hayden Dam (USDI BR 1992). Habitat destruction and competition and predation from introduced non-native fish species are the primary causes of the species' decline ( Propst 21 Al. 1986, Rinne 1991).
Loach Minnow
The loach minnow was listed as threatened on October 28, 1986. Critical habitat was designated for the loach minnow on March 8, 1994, and includes portions of the Gila, San Francisco, Blue and Tularosa Rivers and Aravaipa, Campbell Blue, and Dry Blue Creeks. A small, elongated fish, the loach minnow is olive-colored with small white or orange spots at the base of the dorsal and caudal fins (Minckley 1973). Leach minnow were once common in the Gila River above Phoenix but are now found in only six locations: Aravaipa Creek (Graham and Pinal Counties, Arizona), upper Gila, San Francisco and Tularosa Rivers (Catron and Grant Counties, New Mexico), the White River (Navajo and Gila Counties, Arizona) and the Blue River (Greenlee County, Arizona). Habitat destruction and competition and predation from introduced non-native fish species are the primary causes of the species' decline (Propst 21 Al. 1988, Propst and Bestgen 1991).
Gila Topetiunow
The Gila topminnow was listed as endangered on March 11, 1967. No critical habitat has been designated for this species. The Gila topminnow is a small, livebearing fish found in the Gila, Sonora, and de la Concepcion River drainages in Arizona, New Mexico, and Sonora, Mexico (Minckley 1973, Vrijenhoek II Al. 1985). It was once among the most common species of the Gila River and its tributaries (Hubbs and Miller 1941). Destruction and alteration of its habitat plus competition with and predation by non-native fish species have resulted in extirpation of the Gila topminnow throughout most of its range (USDI FWS 1984, Meffe gt Al. 1983). Nine naturally occurring populations of Gila topminnow remain, all but one located in the Santa Cruz River basin. One naturally occurring population is found in three small adjacent spring systems just off the Gila River on the San Carlos Indian Reservation near Bylas, Graham County, Arizona. Stocked populations of Gila topminnow are found throughout the Gila River basin in Arizona, primarily in isolated springs and spatially intermittent streams.
Desert Pupfish
The desert pupfish was listed as endangered on March 31, 1986. Critical habitat for this species was designated at Quitobaquito Spring, Organ Pipe Cactus National Monument, Pima County, Arizona, and at three locations in Imperial County, California. The desert pupfish is a small fish historically
■ CAP Biological Opinion - April 15, 1994 4
Proiect Description
The CAP was constructed to provide a long term, non-groundwater, water source for municipal, industrial, and non-Indian and Indian agricultural users in Arizona. The water provided through the CAP aqueduct system represents Arizona's allocation of the flow of the Colorado River. The water is taken from the Colorado River at Lake Havasu and is conveyed across the state in a series of large open aqueducts (Figure 1). A storage option for CAP water became available December 1992 following enlargement of Lake Pleasant, an existing reservoir on the Ague Fria River north of Phoenix. Water is pumped into the reservoir when the aqueduct is carrying more than the demand requires and pumped out to make up demands during other times of the year.
The CAP system was declared completed in October of 1993 and its expected project life is 100 years. Water deliveries are currently ongoing to supply agricultural, municipal, and industrial users listed in Tables 1 and 2. Figure 2 shows the general location of the entities receiving water through the CAP.
The issue under discussion in this opinion is the transfer of non-native fish species from the Colorado River and other sources of introduction along the aqueduct system into the waters of the Gila River basin in Arizona and portions of western New Mexico. The CAP aqueduct has been in operation long enough that field collection data support initial hypotheses that fish populations are able to exist in the aqueducts. At present, fish largely originate from Lake Havasu, although reproduction of some species has been documented in the aqueduct and another source of fish became available with storage of CAP water in Lake Pleasant (Grabowski et Al. 1984, USDI BR 1987, USDI BR 1988, Matter 1991).
We have identified several tributaries to the Gila River that, because of the proximity to either the CAP aqueduct or users of CAP water, may have the potential for non-native fish species to be introduced to them via the CAP.
Hassayampa River
The Hassayampa River is crossed by the CAP aqueduct approximately 24 miles upstream of its confluence with the Gila River. Several irrigation districts are in the vicinity of the river; the Harquahala Valley and Tonopah, both of which drain to Centennial Wash, and the Buckeye and Roosevelt which are adjacent to the Gila River at and above its confluence with the Hassayampa. The Gila River at the Hassayampa confluence often has water year round due to treated effluent outflows from the Phoenix metropolitan area and irrigation returns from the agricultural fields. The Hassayampa is usually dry throughout the reach of interest, although permanent water is found upstream near Wickenberg.
Ague Fria River
The CAP aqueduct crosses the Aqua Fria downstream of Lake Pleasant's New Waddell Dam. Beginning in December 1992 CAP water has been stored in the Lake Pleasant. Although the Aqua Fria is seasonally dry above Lake Pleasant, water flows into the lake occurs over a several month period in winter and spring and monsoon generated flash floods are common summer occurrences. Water is occasionally spilled downstream from Lake Pleasant, such as during the flooding in January-February 1993.
■ CAP Biological Opinion - April 15, 1994 •
6
TABLE 2. CENTRAL ARIZONA PROJECT WATER DELIVERIES FOR MUNICIPAL AND INDUSTRIAL USES IN 1993. KC USERS MONTH:* DELIVERIES CALENDER YEAR 1993 Table /2 MC0NTRACT CUSTCMERS JAN FE3 MAR APR MAY JUN JUL AUG SE? CCT NOV DEC TOTAL 113 0311-01...... w■aat.3-SMUS3WMO ors OOOOOOOOO rumemassmamammcamm ...... Az. Water Co. . 0 0 0 o 0 117 137 135 131 135 127 43 82.6 Carefree 0 0 0 11 30 36 41 32 34 18 0 0 202 Cave Lreek : 14 10 10 20 19 25 24 38 19 22 15 15 232 Chandler * .• 117 239 10 12 88 93 53 43 44 135 155 0 1.166 Chaparral .- 83 99 0 0 es 158 159 164 160 155 154 207 1.416 - Elcy 3 5 19 25 25 ES 52 44 31 31 10 28 340. Gilbert 0 0 o 0 o 0 0 0 o 0 4 5 9 Glencale 836 279 25 558 548 926 1,:t8 1,012 991 1,441 569 641 8,918 - X4r1C:44 Countf: A 13 25 47 52 77 72 54 58 35 43 19 493 Mesa 575 570 1,658 1.020 1.5;8 2.074 2,557 2.360 2.136 1.912 1.841 778 19.247 ?hoer!): * : 7.722 134 782 1.326 4.755 7,375 8.394 0,297 4.835 8.199 5,072 5,488 60,969 . Queen Creek n 0 0 0 0 0 0 0 0 0 6 19 25 Rio Veroe ' 0 0 0 250 0 0 0 1 0 0 0 0 250 Scottscale ' : 1.200 1.270 1.496 1.705 1.892 2,179 2.291 2.296 2.222 2.402 1.891 1,719 22.663 Tempe 0 0 0 933 0 0 0 0 0 714 7a0 0 2.427 Tucson : 2,7E6 2,795 4,016 4,765 5.025 5,149 5.228 4,421 4.944 3.423 1.349 1,924 46.315 3133.3111417-33 3.3.3211[33.3333.3131=111.3133213/11war 3331 ...... or MN= 11 =MAWS 31••• ar-• 3-3-33”11143113131MSVINS =SA 3.1116C 31133 SS& *313333.
Sub Total 13.441 5.315 8.118 10,680 14,101 13,Z80 20.936 17.005 15.505 18.530 12,497 10,886 165.495 NON-S12CONT7cACT CUSTOMERS JAN FE3 MAR APR MAY JUN JUL ALM SE? OCT NOV DEC TOTAL . S ...... wimmOMMWOLO ...... 0.731.11.131.533311.-nwlewm=1.3031301MWSUMMWMMSWISSIMWOMoniAmm...... =31MALM Ancala 3 6 18 43 53 AS 49 33 37 44 16 18 353 C.S. McCrossan : 0 5 6 1 4 5 0 0 n 0 0 0 21 Oragasz 0 0 0 0 0 7 4 7 7 0 0 7 32 J.W.J. Con:r Co: 0 1 7 .,'' 24 25 1 1 14 15 9 2 112 mazatzal Farm : 0 1 1 2 3 3 4 4 4 3 1 2 28 ?CL Civil Con:: 0 0 0 ,-,v 0 11 11 1 .4., 4 3 5 46 Picachm Schdol : 0 a 0 • 1- ., s 3 4 0 0 0 0 17 2ed Mt. Rancn : 2 a 25 42 52 SE 68 51 51 75 22 19 491 Sonoran Lnd Grp: 0 o 0 0 0 o 0 0 0 0 0 o o viewoolnt . 2 2 5 12 19 18 18 12 14 18 5 6 131 fl_fln_flSS.taSSflfl.Sfl .SaflS flSflamflflflSSSmaaeWaflflatWSflaaflsSS_flflaS AFLSSNSAFLFLFLFLSTAM
Sub Total 7 22 62 98 18e 190 155 113 138 169 35 59 1.231
TOTAL (ME) 13,448 5.338 8.180 10,778 14.269 18,470 21.094 17.119 15.743 18.789 12.553 10,945 150.729
GRAND TOTAL 12.557 11.575 58.324 68.743 73.223 105,467 131.745 104.554 41.604 25.538 14.079 21.096 669.686
' Customer is role take celiveries tnru SRP intmcoommeeT. 2 Oct - Dec deliveries were mace to incividual farmers uncer excess contracts. - Deliveries may Ineuoe r.emArse/espn.ange ma w-. 2 Oct Nov deliveries were purchased fhdm liehtlkam. ! Doc aell.,tryor were mode to indivicual farmers under e=1.42 contracts.
.CAP Biological Opinion - April 15, 1994 A ▪
8
FIGURE 2. LOCATION OF CENTRAL ARIZONA PROJECT WATER USERS IN CENTRAL ARIZONA. CENTRAL ARIZONA PROJECT* WATER USER ORGANIZATIONS
, PLIA•ANT
WAOCIEU CANAL
TIMICICIRE INN PIAN1.1.7 WYE
••• ••••••■ IMMIIMINI•
Lk.
...Net Yoe
Tampon 1 .0
clouavvcht 1.0. 14orquaNtls VaHey 1.0.
A 721-ni r
. 1 .%1111101ca • al,
•••
n Carios 1.0.0.
Stan! NW 1.0
CAP SUBCONTRACTORS . . .
.. To“at0000 ...O.T Col ■•■•■•• 4., *MI... O ...... , .. ... coo ooro coo■ooT L_i aoT000 T oo b... ■■•■■ Y.. - ••••••■•• Mae, 4 ON al..1 ii yaw. •■••• ••••••• ow. 41 %a Worm ■ 0 i 1I • ,arooma mirol• Ca 1am. I ■•••• Came or Iwo 0... •■■••■•• Colon* am. %mi.. C• ..... N.. ••■•• • ao..,••••■•• bran lay. ma. Co Control Amon. 1.0.13. ...ma *a wawa Ca ...we boom •••■ Ca I 01••■•• Flarwaa I.aam aiaav Ca 1 - - [....aa ...• coo ,..... 0oaa mara.e.* ■••••• 4••• a.. Ca ; it...... pi ..., am. a••■• t 0 • am Mo.., wag, a.. is.... La • Ata-a VaHeyl. mow dee.... ma•ma PI a.m.. wow Ls mme CA a* 10110110 PC04i,11 NM P.m. yew ammo Qom., I... 1 um .0. Ia. 1.4...... row (...... 1 rIowa. KI1A0 0...aw••■ aim. Ca ...... Cs !Mow. Row. taw Pan .•••■• *soma •■• Ca 0 Po.... .01 Own. L.a. I 0 dom. aoma. ma. La I Claommaa ma Con a,. ■ • ■••• Ca IMO as bola 04••••■ MAO , 4 M. ■••••■ .o.... • • . ■••••1 ■••••• C. Piro • C • oakaam Maw 8 P3 •I• II•or mow comoomme, Ca, sow C• i ,...... •...... m.• Cam.. 9 New*. Mep , sway .D•aa Paw GmaNal NI .... C. =...... C....w. ow Nimi • Om (....= ■ • • foam. NW. •■• imago* layloa• anew Ca I . „. , • .2_ lama Ca ON •••■1 ...... am...... __ " Co/m-0am ...... _...... •• lap ••••• ...... „. ammo ■•••• 1..... • ■•• slam i Y.o■ • be.• 1•••• Camme.. • • • la■ •• 111•111 t Clirm Cae go, ••• aaoa, • ,. C• • •••aa ,ao. %Nam ■ • •••■ . ; .18 {••■•••■•• ••••• Cammmoal• ...PI 11•11•• Comma.. KEY MAP Yoe aaa mow Ca do• 11•• Yaw Imffim Csa•■•••• ...... ••■■• *a ••••• 60 am.,am•••• sum alma • ....• ••••• raa••• I •••••••• ma. 1•■■•• Inn. Col MY.= errna ••• aao., 0 ••■• WM. .1 6.•• Tomas ••••• ••• ran.. •■•••• Moo c. 1 LEGEND Maw ICaimmoma awn, Cal mow Cal WA, Now 0.0.. maim Tam Liramo• woo Cal 1.1•014 II •ClISTII•AL WM/ mort-oca•ot LruRC 0001 AA RESIVIVAI 0.1 ••••• C• ona■• • 1 .. away I • ...a • P..... NM. (...... S...... • Cowa• ki Comm. tmor l • • ...P... 4••••••• imam • 4•10 ••••■• • SE. Coommo ama *man Ca 0 Casomm ORR.
• CAP Biological Opinion - April 15, 1994 • 10
FIGURE 3. CENTRAL ARIZONA PROJECT/SALT RIVER PROJECT INTERCONNECTION
CAP/SRP SALT RIVER ARIZONA CANAL INTERCONNECTION GRANITE REEF DAM PROJECT FOR EBAYz FISH BARRIER
ARIZONA ARIZONA CANAL • Atic SIPHON DELIVERY VALVE —I ( 1 1 4") \300 CFS) STRUCTURE +, ....•••••• ••••••■ s..% \ ..., \ % Hsi BARRIER I USE-JR FLOW \ 1 200 CFS METER RADIAL COMMON \ A GATES RIVERBED CANAL ••• DELIVERY (1200 CFS) COMMON WEIR t WEIR \ 7 SOUTH 84" PIPE .------AI \ CAMAI---- ðV SOUTH N (1200 cFS) CANAL DELIVERY J I (800 cFS) List:1R TLIRNOUT EXISTING --- WEIR cAp SIPHON AC)11101/C1 \. CAP Biological Opinion - April 15. 1994 • 12 events may result in flow in the Gila River below Ashurst-Hayden since there would be no diversion of flow into the Florence-Casa Grande Canal.
The delivery of CAP water to SCIP is via a series of turnouts to the Florence- Casa Grande Canal to serve SCIDD lands or via the Pima Lateral Feeder Canal to serve GRIC lands. The first of the turnouts, the Pima Lateral Feeder Canal was constructed in 1989 to deliver CAP water into the existing Pima Lateral Canal (which connects to the Florence-Casa Grande Canal) in exchange for GRIC leaving 30,000 acre-feet of their SCIP water in San Carlos Reservoir to provide a minimum pool for the fishery. In 1990, other turnouts were proposed, and some constructed, from the CAP aqueduct and the Florence-Casa Grande Canal to serve SCIDD. There are, then, existing direct water connections from the CAP aqueduct to the Gila River above Ashurst-Hayden Dam via the Florence-Casa Grande Canal. There may also be instances where other irrigation districts canal systems are connected to the SCIDD system, thus providing an additional introduction point for CAP water and fish to reach the Florence-Casa Grande Canal.
Unlike the other irrigation districts in the area, the first water to SCIP entities was in 1990. The GRIC received 47,548 acre-feet of CAP water, up to 30,000 for their SCIP water left in San Carlos Reservoir, and the rest under an interim contract. The SCIDD received 10,301 acre-feet and SCIP as a separate entity received 24,483 acre-feet. In May 1991 an additional 5,020 acre-feet were delivered and no water has been delivered to these entities since that time. To protect against CAP fish introductions, two electric fish barriers were constructed to deter fish from moving upstream beyond the diversion at Ashurst-Hayden. The first was placed on the Pima Lateral to deter fish from accessing the Florence-Casa Grande Canal and was part of the GRIC exchange program. The second was placed on the Florence-Casa Grande Canal at the U.S. Geological Survey gaging station at China Wash, upstream of any CAP turnout to the Florence-Casa Grande Canal.
Summary
The waters conveyed by the CAP can connect both directly, via existing canals and reservoirs, or indirectly, via return flows, to the rivers of the Gila River basin. This connection opens potential conduits for the transfer of non-native species of fish to the lower Gila River basin tributaries and to the middle and upper Gila River basin above Ashurst-Hayden Dam. The Hassayampa, Aqua Fria, Salt and Verde Rivers, as tributaries of the Gila, will also enable any fish species introduced into their waters from the CAP to reach the Gila River and eventually move upstream to the base of Ashurst- Hayden. Ashurst-Hayden Dam is not an effective fish barrier for several reasons, both structural and operational, and any fish species that reaches the base of the dam would have the opportunity to move beyond it to the middle Gila and San Pedro Rivers. The presence of the CAP connection creates a perennial waterway from the Colorado River to the major streams of the Gila River basin, connecting rivers and canal systems currently more isolated by dams and intermittent stream reaches.
EFFECTS OF THE ACTION
The introduction and spread of non-native fishes and other aquatic organisms have been identified as major factors in the decline of native fishes throughout North America and particularly in the southwest (Miller 1961, Lachner Et. al. 1970, Moyle 1976, Courtenay and Stauffer 1984, Williams et al. 1985). Miller et Al. (1989) report that non-native species were the second most common causal factor in recent extinctions of North American fishes. Although not always the primary factor, non-native species played a part in 68
■ CAP Biological Opinion - April 15, 1994 4 14 more favorable to non-native species than to native species (Minckley 1983, Bestgen and Propst 1986, Bestgen and Propst 1989, Rinne 1991).
Introduction and spread of non-native fish species has also been a major factor in the decline of the Gila River basin fish community (Minckley and Deacon 1968, Meffe 1985, Minckley 1985, Propst et Al. 1986, Propst et al. 1988, Propst and Bestgen 1991, Rinne 1991). The Gila River basin presently supports several non-native fish species. Some of these, such as rainbow trout ( Oncorhvnchus mvkiss) (USDI FWS 1983, Propst It Al. 1992), channel catfish (Ictalurus punctatus)(Bestgen and Propst 1989, Marsh and Brooks 1989), mosquitofish (Gambusia AFFINIS) (Meffe Al. 1983, Meffe 1985), and smallmouth bass ( MICROBTERUS dolomieui) (Propst g Al. 1986), have had substantial adverse impacts to the native fish species. A few, such as fathead minnow ( PIMEVHALES promelas), appear to have had little effect upon native fishes.
Thus, the pre-project baseline is a deteriorated system where native fishes persist in isolated stream sections which have not yet undergone sufficient perturbation to cause extirpation of those species. Non-native species already present exert continuing competitive and predatory pressure on native fishes. Fragmentation of aquatic habitats has disrupted historic patterns of recolonization or augmentation of damaged or destroyed native fish populations from intact up-stream or downstream populations. Additional impacts to these systems and the listed fish from the proposed project will be cumulative to the existing baseline. Therefore, project effects are of more consequence to the species than if the baseline was a pristine condition.
Pinalvsis of Effects
1. Non-native Spocies Which Might b. Introducod Through CAP
In an attempt to define the extent of the potential problem, the striped bass ( Morone saxatilis), white bass (Morone chrysops), blue tilapia (TilaPia aurea), rainbow smelt (Osmerus mordax), and the triploid grass carp ( Ctenopharvngodon IDELLA) were identified in early planning for CAP water deliveries as being non-native fish that were likely be transported by CAP and to adversely impact the native fish species of the Gila drainage.
Matter (1991) evaluated the five species for their likelihood of entry and survival in the Gila River above Ashurst-Hayden Dam. These five species are not currently found in the Gila River basin above Ashurst-Hayden, but all are currently found in the Colorado River system. All five are believed to have adverse effects on native southwestern fishes (Taylor el al. 1984; Evans and Loftus 1987; Miller and Fuiman 1987; W.L. Minckley, Arizona State University, pers. corn., April 1991).
In addition to those five species, we are also concerned about other non- native fish which already inhabit the Colorado and Gila Rivers and the CAP aqueduct system, or which may at some time be introduced into those systems. Such species fall into three categories: those already known and perhaps established in the Colorado River or CAP systems, but not yet in the middle or upper Gila River basin; those not yet found in the Colorado River system, including the Gila River basin; and those already found throughout most of the Colorado and Gila River basins. The CAP and interconnecting canal systems provide enhanced opportunities for such non-natives to colonize, recolonize, or augment existing populations in the Gila River basin. Because of the size and habitat characteristics of the streams of the Gila River basin, it is likely that the primary threat from non-native fish incursion may come from some of the smaller non-native fish species, rather than the larger species included in Matter's study.
■ CAP Biological Opinion - April 15, 1994 • 16 increasingly difficult when the population is continuously augmented from outside.
Although non-native fish are the primary concern in this analysis, the proposed project will also provide an avenue for introduction of other non- native aquatic organisms into the Gila River basin. Various insects, molluscs, crustaceans, plants, and parasitic and disease organisms may be transported into the Gila River basin via the CAP and may have adverse effects on the ecosystem as a whole as well as on the listed fishes specifically. While effects of past non-native invertebrate introductions on native southwestern fishes and invertebrates are poorly documented, such species as the Asian clam (Corbicula panilensis) and the crayfish (Procamberus clarkii) have been implicated in declines of native species (Pieter 1979, Wells gt Al. 1983). Asian tapeworm (Bothriocephalus acheiloanathi) contributed to recent declines of the woundfin (Plaaopterus araentiosimus) in the Virgin River. The tapeworm is thought to have entered the Virgin River with invading red shiner (Heckmann et al. 1986). It is anticipated that other invertebrate non-natives currently expanding their range, such as the zebra mussel (Dreissena volvmorpha) and giant rams-horn snail ( Marisa cornuarietis) would have adverse effects to native fishes and invertebrates (USDI BR 1990, Horne et al. 1992).
2. The Time Frame of Project Effects
It is important to understand that the CAP is not a project whose impact can be measured over a short span of years. Once constructed and operational, the aqueduct system will be bringing water from the Colorado River into the Gila River basin for at least 100 years. Delivery of that water to users gives any species which enters the aqueduct during the period of operation continuous access into the Gila River basin. Once established in the Gila basin, the effects of non-native species on native species will continue far beyond the life of the project. Proper analysis of the potential impacts of the proposed project must include a long-term view since the longer CAP is in operation, the higher the risk of undesirable non-native species entering the system. With the aqueduct and the proposed and ongoing water deliveries, the risks of non-native species reaching the habitat of the listed species are clearly increased and continue long past the construction phase of any feature of the CAP.
3. Potential Sources of CAP-associated Non-native Introductions
There are several avenues by which non-native species may enter the CAP aqueduct. The most direct entry point for non-native species into the CAP aqueduct is from Lake Havasu at the aqueduct head on the Colorado River. This route has been documented as the source of nine species of non-native fish currently present in the CAP aqueduct (Mueller 1989). Non-native species will also be able to enter the aqueduct through Lake Pleasant. Tilapia and white bass are both likely to enter the aqueduct from Lake Pleasant (Matter 1991). Other points of entry are discussed in the project description section above.
Non-native species may also enter the CAP through accidental or intentional release by private citizens without authorization; also known as "bait bucket" introductions. The increasing interconnection of waters of the Colorado River basin through canals and aqueducts adds increasing opportunities for this type of introduction. Increased opportunity for bait bucket introductions was also noted as an important concern in environmental analyses for the Garrison Diversion Unit in North Dakota, an interbasin water transfer project similar to CAP (USDI BR 1990a). CAP provides increased opportunity for spread of bait bucket introductions through much larger portions of the basin than would be possible via the natural stream systems. Bait bucket introductions are made for many reasons: stocking of favorite sport fish, escape or dumping of bait,
.CAP Biological Opinion - April 15, 1994 • 18 populations are present in the basin and a bald eagle nest is located at Lake Pleasant.
Introduction of non-native species into the Ague Fria River basin via CAP is virtually certain and the consequences to listed species will be moderate. The planned fish barrier on Tule Creek will provide some protection to the Tule Creek Gila topminnow population, although the presence of non-native species near the downstream side of the barrier will raise the probability of bait bucket transfer above the barrier. Cow Creek and Humbug Creek, both of which have stocked Gila topminnow populations, already support several non- native species. The Gila topminnow populations in those creeks will probably succumb to increased non-native pressure over the life of the CAP.
C. Salt and Verde Rivers. Because the entry point for CAP-assisted non-native species is bellow the confluence of the Salt and Verde Rivers, the effects on these two river basins are closely related. The potential entry point is at the interconnection of the CAP and SRP systems near Granite Reef Dam on the Salt River (Figure 3). The two major SRP canals head in the pool behind Granite Reef Dam (the forebay). CAP water is transferred from the aqueduct into the canals just downstream from their head. Non-native species could move out of the CAP aqueduct into the SRP canals and from there move upstream into the Salt River above Granite Reef Dam.
The listed species present in the Verde and Salt River basins include bald eagle, spikedace, loach minnow, stocked razorback sucker, stocked Colorado squawfish, and numerous stocked Gila topminnow populations, mostly in isolated springs and small streams. Critical habitat for spikedace is located in the Verde River between Sullivan Dam and Sycamore Creek. Critical habitat for razorback sucker is located in the Verde River between Perkinsville and Horseshoe Dam. In the Salt River, critical habitat for razorback sucker is located between U.S. Highway 60/State Route 77 and the Roosevelt Diversion Dam.
The probability of CAP introduction of non-native species into habitats of the listed species in the Salt and Verde River basins is low, but the consequences of such introductions to the listed species are severe. The probability of upstream movement of fish or other aquatic organisms is limited by several structures. On the Salt River, these structures include the two electric barriers on the SRP canals, which prevent or hinder upstream movement from the canals; Granite Reef Dam, which serves as a barrier to upstream movement in the Salt River bed; and Stewart Mountain, Mormon Flat, Horse Mesa, and Roosevelt Dams, which block fish movement on the Salt River between Granite Reef Dam and the native fish habitat in the upper basin. On the Verde River, fish which surmount Granite Reef Dam are then blocked by Bartlett and Horseshoe Dams from reaching the portions of the river occupied by listed fish.
Despite these barriers, non-native species continue to spread upstream, most likely by bait bucket transport. The presence of CAP introduced non-natives at the base of any given dam increases the risk of those non-natives being transported upstream into the reservoir and river. The electric fish barriers, both on the SRP canals and the Florence-Casa Grande and Pima Lateral Canals (see D. below), are not totally effective at preventing upstream fish movement. Little data exist to support the assumption that these electric barriers totally block upstream fish movement and some ichthyologists have expressed the belief that fish can move upstream through the barriers under certain conditions. In addition, electric barriers are subject to periodic operational failures, such as the one occurring at the SRP barriers on December 23, 1993 (SRP 1994). Following that barrier outage, two grass carp, formerly found only below the barrier, were captured upstream of the barrier.
■ CAP Biological Opinion - April 15, 1994 • 20
Al. 1988). We know of no data on the effects of blue tilapia on native Gila River basin fishes, but redbelly tilapia have been shown to replace desert pupfish in areas of the lower Colorado River basin (Schoenherr 1981).
Hatter (1991) did not consider rainbow smelt likely to spread into the Gila River above Ashurst-Hayden Dam, and therefore not likely to affect native fishes in the middle and upper Gila basin. He believed grass carp may reach the Gila River above Ashurst-Hayden but did not consider them likely to move into the San Pedro or the smaller tributaries. Therefore, he believed them unlikely to pose any significant problems for the listed fish.
With the proposed project, both white and striped bass will likely become a part of the fish fauna in the Gila and San Pedro Rivers and their tributaries (Hatter 1991). Although not likely to reproduce in the Ashurst-Hayden to Coolidge Dam reach, these species are likely to prey on native and non-native fish species in those areas. The effects of these predators, contingent upon their density, could be significant upon the prey populations which would include spikedace, loach minnow, and razorback sucker.
E. Gila River above Coolidoe Dam. There is less likelihood of CAP- mediated non-native introductions into the Gila River and its tributaries above Coolidge Dam. However, the consequences of introductions into that area are the most severe. The largest remaining populations of spikedace and loach minnow are found in the upper Gila, San Francisco, and Blue Rivers along with the largest portions of their critical habitat. Spikedace are also found in Eagle Creek. Several stocked razorback sucker populations are located in this portion of the basin and critical habitat for razorback sucker includes the area between the Arizona-New Mexico border and Coolidge Dam. Natural Gila topminnow populations are found in three springs on the Gila River alluvial plain near the town of Bylas, and stocked populations of Gila topminnow and desert pupfish are scattered throughout the basin. The Gila River between the mouth of the San Francisco River and the Safford Valley is considered to be of high value for recovery of razorback sucker, loach minnow, spikedace, and bonytail chub (Gila elegans) (Hinckley 1985). The Gila River between San Carlos Reservoir and the Arizona-New Mexico border and the San Francisco River from its mouth to the Arizona-New Mexico border have been designated as habitat for establishment of experimental, non-essential populations of woundf in.
Transfer of non-natives over Coolidge Dam would not result directly from CAP, but, in conjunction with bait bucket transfer, will become substantially more likely as a result of the CAP introduction of non-natives into the Gila River below the dam. Coolidge Dam impounds San Carlos Reservoir which supports a warm water sport fishery that is heavily used by local and visiting anglers. A likely scenario for transport of fish from the dam base into the reservoir would be an angler driving to the river just below the dam to collect bait fish for use in the reservoir, then dumping excess bait into the reservoir. Once released into the reservoir, the risk of those non-native species spreading into the upper Gila, San Francisco, and Blue Rivers and their tributaries becomes quite high.
5. Listed Species - Potential Effects
For all non-native introductions, some common principles hold. Aquatic systems can support only a finite biologic resource. With each new species, regardless of the size of its population in the system, there are fewer resources to divide among the other species present. Long-term interactions of introduced and native fish populations are not simple to model or predict, but the record clearly indicates that introduction of non-native fishes into southwestern aquatic habitats coincides with reduction or elimination of
. CAP Biological Opinion - April 15, 1994 A 22 species in the CAP aqueduct or irrigation canals, combined with continuous movement into the river system, may give the non-native species an advantage over native species by insulating the non-native from the effects of limiting natural events, such as drought and floods.
A. Svikedace and Loach Minnow. Analysis of the effects of non-native introductions via CAP on spikedace and loach minnow must take into account the short life span of these fish; both survive only about two years. This makes both of these species particularly vulnerable to short-lived, high-intensity impacts. Loss of two consecutive years of reproduction in either spikedace or loach minnow would result in extirpation of the population. Both could sustain serious adverse impacts from invasion of their habitat by a non-native species even if the non-native fails to become established in the long-term.
Introduction of non-natives via CAP has the potential to affect all remaining populations of spikedace and loach minnow. The threat to the Verde River population of spikedace and the White Mountain population of loach minnow is low, due to protection provided by intervening dams. The Aravaipa Creek population of both loach minnow and spikedace will be at high risk and will probably be extirpated by cumulative impacts of non-native introductions over the life of the project. The upper Gila River and Eagle Creek populations of spikedace and San Francisco, Blue, and upper Gila River populations of loach minnow will probably suffer substantial losses due to CAP-mediated non-native introductions over the life of the project.
Proposed critical habitat for both spikedace and loach minnow includes Aravaipa Creek in Pinal and Graham Counties, Arizona and portions of the upper Gila River in Grant and Catron Counties, New Mexico. For loach minnow, proposed critical habitat also includes the Blue River and the San Francisco River for a stretch downstream from the mouth of the Blue in Greenlee County, Arizona, as well as portions of the San Francisco and Tularosa Rivers in Catron County, New Mexico. For spikedace, proposed critical habitat also includes the upper Verde River in Yavapai County, Arizona. Of these areas, Aravaipa Creek will clearly be affected by non-native introductions resulting from the CAP aqueduct. The Blue, San Francisco, Tularosa, and upper Gila Rivers will be affected by non-native species which may be transported over the dam into San Carlos Reservoir. The upper Verde may be affected by non- natives which eventually are transported above both mainstem dams on that river. Adverse modification of critical habitat is defined as a direct or indirect alteration that appreciably diminishes the value of the critical habitat for both survival and recovery of the listed species. The introduction of additional non-native fish species into the already degraded habitat of the spikedace and loach minnow will appreciably diminish the survival and recovery value of their critical habitats.
B. Gila Topminnow. Gila topminnow are particularly vulnerable to adverse impacts from non-native fish (Meffe gt Al. 1983). The springs near Bylas which support natural populations of Gila topminnow are separated from the Gila River only by short, dry, stream stretches. Mosquitofish have invaded those springs from the Gila River during periods of high flow. Barriers which have been constructed to exclude non-native fish from the springs are only partially successful and are too low to prevent invasion by some species. Introduction of additional non-native species into the Gila River above Coolidge Dam will substantially increase the threat of loss of these Gila topminnow populations.
The Cow Creek and Humbug Creek stocked Gila topminnow populations will probably be extirpated by cumulative non-native introductions over the life of the CAP. The Tube Creek population may survive if the barrier being constructed on that creek successfully prevents upstream movement of non-
■ CAP Biological Opinion - April 15, 1994 • 24 nesting territory at Coolidge Dam is unclear. Matter (1991) believed all five species could live and all but grass carp reproduce in San Carlos Reservoir and from there could move further upstream on the Gila River. No bald eagle nests are known from the Gila River above San Carlos Reservoir. Effects to the bald eagle from introduction of non-native fishes into San Carlos Reservoir will be confined to the Coolidge nest. The degree of effects on the hunting territory will depend on the extent of change in fish species composition.
Introductions and survival of exotic or non-native fish species over the life of the project and beyond will alter fish species compositions in the area of bald eagle nests throughout the Gila River basin. Evidence suggests that bald eagles are not prey specific and will take whatever fish species are available to them, including a large number of non-native species (Biosystems Analysis, Inc. 1991). Unless the newly introduced fish species eliminate significant portions of the forage base presently available for eagles and are themselves not available as prey for some reason, effects to the nesting territories may not be significant.
For the bald eagle, the change in fish fauna over the life of the project may or may not have significant effects to the prey base these birds rely on. Changes to the prey base that still result in abundant, easily catchable prey, albeit of a different species, might not be significant to the maintenance of a nesting territory. Expansion of the newly introduced fish species into a wider range in the Gila River will affect the bald eagle only so far as the individual species posed a concern due to their effects on the prey base.
6. Recovery - Potential Effects
The long-term survival of an endangered or threatened species may require implementation of recovery actions as well as basic protection. Preclusion of recovery opportunities may jeopardize survival. The purposes of Congress in setting forth the Endangered Species Act are very clear. Section 2(h) of the Act states:
"The purposes of this Act are to provide a means whereby the ecosystems upon which endangered species and threatened species depend may be conserved..."
Conserve is defined in section 3(3) to mean: "...to use and the use of all methods and procedures which are necessary to bring any endangered species or threatened species to the point at which the measures provided pursuant to this Act are no longer necessary..."
Thus, the conservation of any threatened or endangered species under the Act clearly requires recovery of that species and protection of ecosystems which would support that recovery. Loss of significant portions of recovery habitat would then be contrary to the purposes of the Act.
A major recovery strategy for endangered and threatened southwestern fishes is their reestablishment within historic range. Since the decline of these species has resulted from the destruction and alteration of their habitat and the introduction of non-native fishes into that habitat, suitable reestablishment sites are rare. Attempts at restoring destroyed or damaged habitat and removing non-native fish have met with limited success (Meffe 1983, Carothers 2t AI. 1989). Therefore, unoccupied streams which appear to still have suitable habitat, or which have a potential for restoration and few or no non-native fishes, are critical to the recovery of these species.
• CAP Biological Opinion - April 15, 1994 • 26
7. Cumulative Impacts
Cumulative effects are those effects of future non-Federal (State, local government, or private) activities on endangered or threatened species or that are reasonably certain to occur during the course of the Federal activity subject to consultation. Future Federal actions are subject to the consultation requirements established in section 7 and therefore, are not considered cumulative in the proposed action.
Various State, private, and local actions will be cumulative to the effects of non-native species introductions via the CAP. The States will continue to introduce sport fish and fish for sport fish prey into waters of the Gila River basin, including species already being stocked and additional species which become desirable for sport fishing. State, private, and local entities will continue to introduce non-native fish and other organisms into waters of the Gila River basin for various purposes, such as insect and weed control. Private individuals will continue to introduce non-native fish through release of live bait fish. Both the States of New Mexico and Arizona permit live bait use in the Gila River basin. Private individuals will continue to make unauthorized stockings of fish and other organisms for a variety of reasons including stocking of sport fish and release of unwanted aquarium and aquaculture fish. Non-native species being raised in aquaculture operations will continue to escape into waters of the Gila River basin.
These State, local, and private introductions of non-native species will have a strong adverse effect on the listed species being addressed here. Cumulatively, these non-Federal impacts may result in severe losses or extirpation of populations of the listed fish. In combination with introductions through CAP, such losses and extirpations pose a severe threat to the survival of all native fishes in the Gila River basin.
REASONABLE AND PRUDENT ALTERNATIVES
Regulations that implement section 7 define reasonable and prudent alternatives (RPA's) as alternative actions, identified during formal consultation, that:
(1) can be implemented in a manner consistent with the intended purpose of the action,
(2) can be implemented consistent with the scope of the Federal agency's legal authority and jurisdiction,
(3) are economically and technologically feasible, and
(4) would, FWS believes, avoid the likelihood of jeopardizing the continued existence of listed species or the destruction or adverse modification of critical habitat.
Many RPA's have been considered in this consultation but were rejected for various reasons during the course of consultation (see Appendix). The following RPA has evolved during the section 7 consultation process.
This RPA contains five primary elements: 1) construction and operation of upstream barriers to fish movement from the CAP aqueduct into native fish habitats, 2) monitoring for non-native fish, 3) transfer of funding to FWS for recovery and protection of listed and candidate Gila basin fishes as mitigation for adverse project effects which cannot feasibly be alleviated below the jeopardy threshold, 4) transfer of funding to the FWS for management
. CAP Biological Opinion - April 15, 1994 • 28
a. A drop structure at the juncture of the Florence-Casa Grande Canal and the Gila River that would form a barrier (with a probability of effectiveness mutually agreeable to FWS, AGFD, and BR) to upstream fish movement from the canal into the river; or
b. A pump system for transferring Gila River water into the Florence-Casa Grande Canal which would preclude upstream movement of fish from the canal into the river.
c. Or other measures agreed upon by BR, FWS, and AGFD.
The effectiveness of the maintenance and operation and the effectiveness of the barriers in preventing upstream fish movement will be reviewed by BR, FWS, and AGFD at intervals not to exceed 10 years. Any changes in operating protocols shall be subject to approval by FWS and AGFD.
If changes in water deliveries, usage, or other factors alter the potential for movement of fish through and between CAP, Florence-Casa Grande, and interconnecting canals, then formal section 7 consultation may be reinitiated and the need for continued operation of the electrical barriers may be reevaluated.
1.4 The existing electrical barrier on the Pima Lateral Canal may be discontinued at the discretion of BR.
2. BR shall, in cooperation with AGFD and FWS, develop and implement a baseline study and long-term monitoring of the presence and distribution of non-native fish in the following areas:
CAP aqueduct SRP canals Florence-Casa Grande Canal Other water delivery canals, as appropriate Salt River between Stewart Mountain Dam and Granite Reef Dam and the electrical barriers Gila River below Coolidge Dam San Pedro River downstream of the U.S./Mexico border
The purpose of this effort shall be to establish baseline data on the presence and distribution of non-native fishes in the target reaches and to detect changes in the species composition or distribution.
Data collection protocols shall be cooperatively formulated by BR, FWS, and AGFD, with the final protocol requiring consensus by all three agencies. This monitoring shall begin no later than 6 months after the date of this biological opinion, and shall continue throughout the life of the project. BR shall submit reports on this sampling to FWS and AGFD on an annual basis throughout the life of the project. Comprehensive reports compiling all previous data collected under this sampling shall be prepared by BR and submitted to FWS and AGFD at the end of every 5 years following initiation of the monitoring. If a non-native fish species is found in an area where they have not previously been found, the FWS and AGFD shall be informed of the finding by telephone within 5 days of the collection.
3. BR shall deposit into an escrow-type account in the name of the FWS the sum of $250,000 annually for 25 years from the date of the first funding transfer. The first funding transfer shall occur no later than three months after the date of this biological opinion and the amount shall be prorated to
p CAP Biological Opinion - April 15, 1994 30 exemption. Harm is further defined to include significant habitat modification or degradation that results in death or injury to listed species by significantly impairing behavioral patterns such as breeding, feeding, or sheltering. Under the terms of section 7(b)(4) and section 7(o)(2), taking that is incidental to, and not intended as part of, the agency action is not considered a prohibited taking provided that such taking is in compliance with the incidental take statement. The measures described below are nondiscretionary, and must be undertaken by the agency or made a binding condition of any grant or permit issued to the applicant, as appropriate.
The FWS anticipates that the proposed transfer of CAP water into the Gila River Basin via irrigation systems under the terms of the reasonable and prudent alternative in this opinion will result in incidental take of spikedace, loath minnow, Gila topminnow, razorback sucker, and Colorado squawfish through habitat modification causing impairment of breeding, feeding, and sheltering.
The anticipated amount of taking due to implementation of this proposed action cannot be quantified. Several species involved are short-lived, have high rates of reproduction, and experience great population fluctuations, thus making population estimates difficult to obtain and interpret. The non-native species which may invade cannot be identified and the timing of the invasions are unknown. Implementation of the RPA is expected to minimize take of the listed fish. Therefore, the level of incidental take anticipated as a result of the proposed action will be assumed to have been exceeded if the proposed action, as modified by the reasonable and prudent alternative, is altered or not carried out as set forth in this biological opinion. If this should occur, the BR must reinitiate consultation with the FWS immediately to avoid violation of section 9. If it is determined that the impact of the additional taking will cause an irreversible and adverse impact on the species, operations must be stopped in the interim period between the initiation and completion of the new consultation.
It is unlikely that any incidental take of desert pupfish or bald eagle will occur as a result of implementation of the proposed action. Accordingly, no incidental take is authorized. Should any take occur, BR must reinitiate formal consultation with the FWS and provide a description of the circumstances surrounding the take.
Reasonable and Prudent Measures
The FWS believes that the RPA given in this biological opinion includes all measures necessary and appropriate to minimize the incidental taking authorized by this biological opinion.
Terms and Conditions for Implementation
In order to be exempt from the prohibitions of section 9 of the Act, the BR is responsible for compliance with the following terms and conditions, which implement the reasonable and prudent measures described above. Implementation of the RPA in the biological opinion will constitute the terms and conditions for this action.
The incidental take statement provided in this opinion satisfies the requirements of the Endangered Species Act, as amended. This statement does not constitute an authorization for take of listed migratory birds under the Migratory Bird Treaty Act, the Bald and Golden Eagle Protection Act or any other Federal statute.
. CAP Biological Opinion - April 15, 1994 • 32
SUMMARY
This concludes formal consultation on the transportation and delivery of CAP water to the Gila River basin (Hassayampa, Aqua Fria, Salt, Verde, San Pedro, middle and upper Gila Rivers and associated tributaries) in Arizona and New Mexico. As required by 50 CFR 402.16, reinitiation of formal consultation is required if: (1) the amount or extent of incidental take is exceeded; (2) new information reveals effects of the agency action that may impact listed species or critical habitats in a manner or to an extent not ^onsidered in this opinion; (3) the agency action is subsequently modified in a manner that causes an effect to the listed species or critical habitat that was not considered in this opinion; or (4) a new species is listed or critical habitat designated that may be affected by this action.
■ CAP Biological Opinion - April 15, 1994 • 34
Courtenay, W.R., Jr., and J.E. Deacon. 1983. Fish introductions in the American southwest: a case history of Rogers Spring, Nevada. The Southwestern Naturalist 28(2):221-224.
Courtenay, W.R., Jr., J.E. Deacon, D.W. Sada, R.C. Allan, and G.L. Vinyard. 1985. Comparative status of fishes along the course of the pluvial White River, Nevada. The Southwestern Naturalist 30(4):503-524
Courtenay, W.R., Jr., D.A. Hensley, J.N. Taylor, and J.A. McCann. 1984. Distribution of exotic fishes in the continental United States. Pp. 41-77 In: Distribution, biology, and management of exotic fishes. W.R. Courtenay, Jr. and J.R. Stauffer, Jr., eds. Johns Hopkins Univ. Press. Baltimore and London.
Courtenay, W.R., Jr., and G.K. Haifa. 1989. Small fishes in strange places: a review of introduced Poseciliids. Pp. 319-331 In: Ecology and evolution of livebearing fishes (Poeciliidae). G.K. Meffe and F.F. Snelson, Jr., eds. Prentice Hall. Englewood Cliffs, New Jersey.
Courtenay, W.R., Jr., and J.R. Stauffer, Jr. 1984. Distribution, biology, and management of exotic fishes. Johns Hopkins Univ. Press. Baltimore and London. 430 pp.
Deacon, J.E., C. Hubbs, and B.J. Zahuranec. 1964. Some effects of introduced fishes on the native fish fauna of southern Nevada. Copeia 1964(2):384- 388.
Drenner, R.W., S.B. Taylor, X. Lazzaro, and D. Kettle. 1984. Particle-grazing and plankton community impact of an omnivorous cichlid. Transactions of the American Fisheries Society 113:397-402.
Eberhardt, S. 1981. San Pedro River basin water quality status report for period 1973-1979. Report of Bureau of Water Quality Control, Arizona Dept. of Health Services, Phoenix.
Echelle, A.A., and P.J. Connor. 1989. Rapid geographically extensive genetic introgression after secondary contact between two pupfish (Cvprinodon, Cyprinodontidae). Evolution 43:717-727.
Evans, D.O. and D.H. Loftus. 1987. Colonization of inland lakes in the Great Lakes region by rainbow smelt, Osmerus mordax: their freshwater niche and effects on indigenous fishes. Canadian Journal of Fisheries and Aquatic Sciences 44(Supp. 2):249-266.
Grabowski, S.J., S.D. Hiebert, and D.M. Lieberman. 1984. Potential for introduction of three species of non-native fishes into Central Arizona via the Central Arizona Project- A literature review and analysis. U.S. Bureau of Reclamation, Denver, Colorado. 225 pp.
Haynes, C.M., R.T. Muth, and L.C. Wycoff. 1982. Range extension for the redside shiner, Riohardsonius balteatus (Richardson), in the upper Colorado River drainage. The Southwestern Naturalist 27(2):223.
Heckmann, R.A., J.E. Deacon, and P.D. Greger. 1986. Parasites of the woundfin minnow, Plaoopterus aroentissimus, and other endemic fishes from the Virgin River, Utah. Great Basin Naturalist 46(4):662-675.
Hendrickson, D.A. and A. Varela-Romero. 1989. Conservation status of desert pupfish, Cvorinodon macularius, in Mexico and Arizona. Copeia 1989(2):478- 483.
CAP Biological Opinion - April 15, 1994 A 36
Hiller, R.R., and L.A. Fuiman. 1987. Description and conservation status of CvPrinodon macularius eremus, a new subspecies of pupfish from Organ Pipe Cactus National Monument, Arizona. Copeia 1987(3):593-609.
Miller, R.R., J.D. Williams, and J.E. Williams. 1989. Extinctions of North American fishes during the past century. Fisheries 14:22-38.
Minckley, W.L. 1973. Fishes of Arizona. Arizona Game and Fish Department. Phoenix, Arizona. 293 pp.
Minckley, W.L. 1983. Status of the razorback sucker, Xvrauchen texanus (Abbott) in the lower Colorado River basin. Southwestern Naturalist 28(2):165-187
Minckley, W.L. 1985. Native fishes and natural aquatic habitats in U.S. Fish and Wildlife Service Region II west of the continental divide. Report to the U.S. Fish and Wildlife Service, Albuquerque, New Mexico. Dept. of Zoology, Arizona State Univ., Tempe, Arizona. 158 pp.
Minckley, W.L. and L.H. Carufel. 1967. The little Colorado River spinedace, Levidomeda vittata, in Arizona. Southwestern Naturalist 13:291-302. Minckley, W.L., and J.E. Deacon. 1968. Southwestern fishes and the enigma of "endangered species". Science 159:1424-1432.
Minckley, W.L., and M.E. Douglas. 1991. Discovery and extinction of western fishes: a blink of the eye in geologic time. Pp. 7-17 in: Battle against extinction: Native fish management in the American west. W.L. Minckley and J.E. Deacon, eds. Univ. of Arizona Press, Tucson.
Minckley, W.L., P.C. Marsh, J.E. Brooks, J.E. Johnson, and B.L. Jensen. 1991. Management toward recovery of razorback sucker (Xyrauchen texanus). PP- 303-357 In: Battle against extinction: Native fish management in the American west. W.L. Hinckley and J.E. Deacon, eds. Univ. of Arizona Press, Tucson.
Moyle, P.B. 1976. Inland fishes of California. Univ. of California Press. Berkeley, California. 405 pp.
Mueller, G. 1989. Fisheries investigations in the Central Arizona Project canal system. Final Report 1986-1989. U.S. Bureau of Reclamation, Boulder City, Nevada. 114 pp.
Pister, E.P. 1979. Report of the Death Valley System Committee of the Desert Fishes Council . Proceedings of the Eleventh Annual Symposium, Desert Fishes Council . Nov. 15-17, 1979.
Platania, S.P. 1990. Reports and verified occurrence of logperches (Percina caprodes and Percina macrolevida) in Colorado. The Southwestern Naturalist 34(1): 87-88.
Propst, D.L. and K.R. Bestgen. 1991. Habitat and biology of the loach minnow, Tiaroaa cobitis, in New Mexico. Copeia 1991(1):29-38. Propst, D.L., K.R. Bestgen, and C.W. Painter. 1986. Distribution, status, biology, and conservation of the spikedace (Heda fuloida) in New Mexico. U.S. Fish and Wildlife Service Endangered Species Report No. 15. 93 pp.
r CAP Biological Opinion - April 15, 1994 • 38
U.S. Department of the Interior, Bureau of Reclamation. 1987. Central Arizona Project Canal System Fishery Investigations Progress Report. Bureau of Reclamation, Boulder City, Nevada. 60 pp.
U.S. Department of the Interior, Bureau of Reclamation. 1988. Central Arizona Project Canal System Fishery Investigations Progress Report. Bureau of Reclamation, Boulder City, Nevada. 59 pp.
U.S. Department of the Interior, Bureau of Reclamation. 1990a. Garrison Diversion Unit joint technical committee report to the United States-Canada consultative group (including the Biology Task Force report). November 1990. Bureau of Reclamation, Billings, Montana. 57 + 73 pp.
U.S. Department of the Interior, Bureau of Reclamation. 1990b. Memorandum to San Carlos Irrigation Project re electric barrier outage. July 9, 1990. Bureau of Reclamation, Phoenix, AZ. 2 pp.
U.S. Department of the Interior, Bureau of Reclamation. 1992. Summary of fish and water quality sampling along the San Pedro River from Dudleyville to Hughes Ranch near Cascabel, Arizona, October 24 and 25, 1991, and the Gila River from Coolidge Dam to Ashurst/Hayden Diversion Dam, October 28-31, 1991. February 26, 1992. Bureau of Reclamation, Phoenix, Arizona. 11 pp.
U.S. Department of the Interior, Fish and Wildlife Service. 1982. Southwestern bald eagle recovery plan. U.S. Fish and Wildlife Service, Albuquerque, New Mexico. 65 pp.
U.S. Department of the Interior, Fish and Wildlife Service. 1983. Arizona trout recovery plan. U.S. Fish and Wildlife Service, Albuquerque, New Mexico. 37 pp.
U.S. Department of the Interior, Fish and Wildlife Service. 1984. Gila and Yaqui topminnow recovery plan. U.S. Fish and Wildlife Service, Albuquerque, New Mexico 56 pp.
U.S. Department of the Interior, Fish and Wildlife Service. 1991a. Endangered and threatened wildlife; the razorback sucker ( Xyrauchen texanus) determined to be an endangered species. Federal Register 56(205):54957- 54967.
U.S. Department of the Interior, Fish and Wildlife Service. 1991b. Colorado squawfish revised recovery plan. U.S. Fish and Wildlife Service, Denver, Colorado. 56 pp.
Vrijenhoek, R.C., M.E. Douglas, and G.K. Meffe. 1985. Conservation genetics of endangered fish populations in Arizona. Science 229:400-402.
Wells, S.M., R.M. Pyle, and N.M. Collins. 1983. The IUCN invertebrate red data book. International Union for the Conservation of Nature and Natural Resources, Gland, Switzerland. 301 pp.
Williams, J.E., D.B. Bowman, J.E. Brooks, A.A. Echelle, R.J. Edwards, D.A. Hendrickson, and J.J. Landye. 1985. Endangered aquatic ecosystems in North American deserts with a list of vanishing fishes of the region. Journal of the Arizona-Nevada Academy of Science 20(1):1-62.
Williams, J.E., and D.W. Sada. 1985. Status of two endangered fishes, Cyprinodon neyadensis mionectes and Rhinichthys osculus nevadensis, from two springs in Ash Meadows, Nevada. The Southwestern Naturalist 30(4):475- 484.
. CAP Biological Opinion - April 15, 1994 4 40
9. Assist in treatment of Aravaipa Creek with rotenone; rejected due to need for extensive evaluation of environmental consequences including take and possible jeopardy of the listed fish.
10. Request all agencies to stop stocking non-native fish in the Gila River drainage; rejected as unenforceable and therefore ineffective in relieving jeopardy. Incorporated into a more comprehensive Conservation Recommendation.
11. Request the Game and Fish Departments of Arizona and New Mexico to stop live bait use in the Gila River basin; rejected due to high probability of ineffectiveness. Incorporated into a more comprehensive Conservation Recommendation.
12. Continuation of funding for the bald eagle nest watch program; rejected because the finding for bald eagle is non-jeopardy and this element is not effective in relieving jeopardy for listed fish.
13. Various locations and combinations of physical and electrical barriers, particularly on the Gila River; rejected as parts of concepts 1(a) and (b).
14. Construction of sets of barriers on the Gila and lower San Pedro Rivers to form management zones between barriers that could be managed to control or remove non-native species; rejected as part of concept 1(b).
15. Require annual dryup or non-native fish management in all canals, ditches, siphons, sumps, and other open water sources and conveyances of CAP and entities receiving CAP water; this was part of a an approach which included preventing access of non-natives to the CAP aqueduct which was rejected as part of concept 1(a). Specific actions were moved into the Conservation Recommendations.
16. Modification of points of discharge into the Gila River downstream from canals and ditches to prevent pooled year-round water; rejected as part of concept 1(a).
17. Conduct a program of test applications of piscicide to the Gila River between Coolidge and Ashurst-Hayden Dams and possibly the San Pedro River to refine techniques for non-native fish removal and native fish reestablishment; rejected because existing techniques and expertise are adequately developed for contingency purposes and due to the need to conduct extensive analysis of the potential need for and impacts of such an effort.
18. Maintenance of the existing Pima Lateral electrical barrier; rejected as unnecessary due to other actions in RPA.
19. Study the effectiveness of the existing electrical fish barriers at preventing upstream fish movement; replaced by transfer of funding to FWS for research to support non-native fish management.
20. Prohibit fishing and public access in the CAP; rejected as part of concepts 1(a and (b).
21. Oppose the introduction of additional non-native aquatic species into waters of the Colorado River basin; rejected as part of concepts 1(a) and (b) and placed into the Conservation Recommendations.
22. Preparation and implementation by BR of management plans for controlling or removing Invading non-native fish species; replaced by transfer of funding to FWS for these actions.
■ CAP Biological Opinion - April 15, 1994 • APPENDIX 2
SOURCES NOT CONSIDERED BY USFINS BIOLOGICAL OPINION Arizona Bureau of Mines. 1969. Mineral and Water Resources of Arizona. University of Arizona; Tucson, Arizona.
Arizona Department of Water Resources. 1994. Arizona Riparian Protection Program Legislative Report. Phoenix, Arizona.
Arizona Department of Water Resources and University of Arizona. 1983. Water service organizations in Arizona. Arizona Department of Water Resources; Phoenix, Arizona.
Arizona Game and Fish Department. 1928-1993. Stocking records of native and non-native fish, locations included.
Arizona Game and Fish Department. 1980-1993. Stocking records of native fish, locations included.
Arizona Game and Fish Department. 1992. Hatchery annual report-1992. Phoenix, Arizona.
Arizona Game and Fish Department. 1994. Stocking requests for fiscal year 1995.
Arizona Water Commission. 1975. Phase I, Arizona State Water Plan: Inventory of Resources and Uses. Phoenix, Arizona.
Barber, W.E. and W.L. Minckley. 1966. Fishes of Aravaipa Creek, Graham and Pinal Counties, Arizona. Southwestern Naturalist 11:315-324.
Brooks, J.E. 1985. Factors affecting the success of Gila topminnow ( Poeciliopsis o. occidentalis) introductions on four Arizona National Forests. Prepared for Office of Endangered Species, U.S. Fish and Wildlife Service.
Burbidge, R.G. 1969. Age, growth, length-weight relationship, sex ratio, and food habits of American smelt, Osmerus mordax (Mitchill), from Gull Lake, Michigan. Trans. Am. Fish Soc. 98:631-640.
Cole, G.A. 1981. Habitats of North American desert fishes. Pp. 477-492 In: Fishes in North American deserts. R.J. Naiman and D.L. Soltz, eds. Wiley-Interscience Pub.; New York, New York.
Constantz, G.D. 1981. Life history patterns of desert fishes. Pp. 237-290 In: Fishes in North American deserts. R.J. Naiman and D.L. Soltz, eds. Wiley-Interscience Pub.; New York, New York.
Courtenay, W.R., Jr. and J.E. Deacon. 1982. Status of introduced fishes in certain spring systems in southern Nevada. Great Basin Nat. 42:361-366. Meffe, G.K., D.A. Hendrickson, and J.N. Rinne. 1982. Description of a new topminnow population in Arizona, with observations on topminnow/mosquitofish co-occurrence. Southwestern Naturalist 27:226-228.
Meffe, G.K. and W.L. Minckley. 1987. Persistence and stability of fish and invertebrate assemblages in a repeatedly disturbed Sonoran desert stream. American Midland Nat. 117(1):177-191.
Miller, A.S. and W.A. Hubert. 1990. Compendium of existing knowledge for use in making habitat management recommendations for the upper Colorado River Basin. United States Department of Interior, Fish and Wildlife Service, Region 6; Denver, Colorado.
Minckley, W.L. 1968. The common fishes of the Salt River reservoirs. Wildlife Views 15:4-13.
Minckley, W.L. and J.E. Brooks. 1979. Fishes. Pp. 510-531 In: W.L. Minckley and M.P. Sommerfeld, Resource inventory for the Gila River complex, eastern Arizona, Final report. United States Department of Interior, Bureau of Land Management, Safford, Arizona and Arizona State University, Tempe, Arizona.
Minckley, W.L. and J.E. Brooks. 1980. Transplantations of native Arizona fishes: records through 1980. Journal of Arizona-Nevada Academy of Science 20(2):73-89.
Minckley, W.L. and S.W. Carothers. 1979. Recent collections of the Colorado River squawfish and razorback sucker from the San Juan and Colorado Rivers in New Mexico and Arizona. Southwestern Naturalist 24:686-687.
Minckley, W.L., and G.K. Meffe. 1987. Differential selection by flooding in stream fish communities of the arid American Southwest. Pp. 93-104 In: Community and evolutionary ecology of North American stream fishes. W.J. Mathews and D.C. Heins, eds. University of Oklahoma Press; Norman, Oklahoma.
Minckley, W.L., J.N. Rinne and J.E. Johnson. 1977. Status of the Gila topminnow and co-occurrence with mosquitofish. United States Department of Agriculture, Forest Service Research Paper RM-198; Denver, Colorado.
Morgensen, S.A. 1990. Phase I: Baseline limnological and fisheries investigation of Lake Pleasant. Arizona Game and Fish Department, Research Branch; Phoenix, Arizona.
New Mexico Game and Fish Department. 1901-1991. Stocking records of native and non-native fish, locations included.
Osmundson, D.B. and L.R. Kaeding. 1991. Recommendations for flows in the 15-mile reach during October-June for maintenance and enhancement of endangered