Klamath River Basin: Reclamation Met Its Water Bank Obligations, 
but Information Provided to Water Bank Stakeholders Could Be	 
Improved (28-MAR-05, GAO-05-283).				 
                                                                 
Drought conditions along the Oregon and California border since  
2000 have made it difficult for the Bureau of Reclamation	 
(Reclamation) to meet Klamath Project irrigation demands and	 
Klamath River flow requirements for threatened salmon. To augment
river flows and avoid jeopardizing the salmon's existence,	 
Reclamation established a multiyear water bank as part of its	 
Klamath Project operations for 2002 through 2011. Water banks	 
facilitate the transfer of water entitlements between users. This
report addresses (1) how Reclamation operated the water bank and 
its cost from 2002 through 2004, (2) whether Reclamation met its 
annual water bank obligations each year, (3) the water bank's	 
impact on water availability and use in the Klamath River Basin, 
and (4) alternative approaches for achieving the water bank's	 
objectives.							 
-------------------------Indexing Terms------------------------- 
REPORTNUM:   GAO-05-283 					        
    ACCNO:   A20140						        
  TITLE:     Klamath River Basin: Reclamation Met Its Water Bank      
Obligations, but Information Provided to Water Bank Stakeholders 
Could Be Improved						 
     DATE:   03/28/2005 
  SUBJECT:   Fishes						 
	     Land reclamation					 
	     Program management 				 
	     Water conservation 				 
	     Water resources conservation			 
	     Water resources development			 
	     Wildlife conservation				 
	     Performance measures				 
	     Land management					 
	     Water supply management				 
	     Farmland						 
	     Irrigation 					 
	     Bureau of Reclamation Klamath Project		 
	     (CA/OR)						 
                                                                 

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Product.                                                 **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
******************************************************************
GAO-05-283

                 United States Government Accountability Office

                     GAO Report to Congressional Requesters

March 2005

KLAMATH RIVER BASIN

 Reclamation Met Its Water Bank Obligations, but Information Provided to Water
                      Bank Stakeholders Could Be Improved

                                       a

GAO-05-283

[IMG]

March 2005

KLAMATH RIVER BASIN

Reclamation Met Its Water Bank Obligations, but Information Provided to Water
Bank Stakeholders Could Be Improved

  What GAO Found

Reclamation has changed how it operates the Klamath Project water bank, as
it has gained more experience, to help it meet its growing obligations and
mitigate costs. For example, Reclamation initially obtained most of the
water for the water bank by contracting with irrigators to either forego
irrigation altogether (crop idling), or use only well water (groundwater
substitution). It later added the option to pump well water into the
irrigation canals for others to use (groundwater pumping). For the period
2002 through 2004, Reclamation's water bank expenditures totaled over $12
million, and the cumulative cost could exceed $65 million through 2011.

GAO's analysis of water bank contracts and river flow records found that
Reclamation met its water bank obligations by acquiring and delivering the
required amount of water for 2002 through 2004. However, Reclamation has
not provided stakeholders with systematic and clear information concerning
the water bank's management and status and its decision to use river flow
data that are not publicly available limited stakeholders' ability to
monitor water bank activities. This has led to confusion and doubt among
stakeholders on whether Reclamation met its water bank obligations.

The water bank appears to have increased the availability of water to
enhance river flows by reducing the amount of water diverted for
irrigation, but the actual impacts are difficult to quantify because
Reclamation lacks flow measurement equipment and monitoring data for the
Klamath Project. Reviews by external experts of the impacts of the 2002
and 2003 crop idling contracts indicate that significantly less water may
have been obtained from these contracts than Reclamation estimated. Given
the uncertainty surrounding how much water can be obtained from crop
idling, in 2004 Reclamation officials decided to rely primarily upon
metered groundwater wells for the water bank. However, Reclamation has
since learned that groundwater aquifers under the Klamath Project, already
stressed by drought conditions, have shown significant declines in water
levels and are refilling at a slower than normal rate in recent years. As
a result, Reclamation is considering lessening its reliance on groundwater
for the 2005 water bank but is uncertain if it can meet its water bank
obligations, particularly for spring flows, while increasing its reliance
on crop idling.

Although several alternative approaches for achieving the water bank's
objectives have been identified by Reclamation and other stakeholders,
limited information is available regarding their feasibility or costs.
Some alternatives to the water bank include permanently retiring Klamath
Project land from irrigation or adding new short-term or long-term
storage. Each alternative has been considered to varying degrees, but
significant analysis is still needed on most alternatives before any
implementation decisions can be made. Meanwhile, Reclamation and the
National Marine Fisheries Service have an ongoing dialogue regarding the
water bank and will likely reconsult on Klamath Project operations,
including the water bank, in 2006.

United States Government Accountability Office

Contents

  Letter

Results in Brief
Background
Reclamation Modified Its Water Bank Operations from Year to Year

as Its Obligations and Costs Increased Reclamation Met Its Water Bank
Obligations, but Information Provided to Stakeholders Could Be Improved

The Water Bank Appears to Have Increased the Availability of Water for
River Flows by Reducing Irrigation Use, but the Extent of Its Impacts is
Unclear

Limited Information Is Available Regarding Alternative Approaches

for Achieving Water Bank Objectives Conclusions Recommendation for
Executive Action Agency Comments and Our Evaluation

1 5 8

14

17

25

30 40 41 41

Appendixes

              Appendix I: Appendix II: Appendix III: Appendix IV: Appendix V:

Scope and Methodology
Information on Water Bank Applications and Contracts
Comments from the Department of the Interior
Comments from the Department of Commerce
GAO Contacts and Staff Acknowledgments

GAO Contacts
Staff Acknowledgments

43

45

50

52

55 55 55

Tables   Table 1: Reclamation's Water Bank Expenditures, Fiscal Years 
                                                                    2002 
                                      to 2004                              16 
             Table 2: Number of Water Bank Applications and Contracts by 
                                                                    Type 
                                    and by Year                            45 
           Table 3: Volume Represented by Water Bank Applications and    
                            Contracts by Type and Year                     47 
             Table 4: Acres of Land Offered in Applications and Accepted 
                                                                   under 
                            Contracts by Type and Year                     49 
Figures   Figure 1: The Klamath River Basin Figure 2: Reclamation's        
                                  Klamath Project                        2 10
              Figure 3: 2002 Augmented Klamath River Flows and Key Dates   20 
              Figure 4: 2003 Augmented Klamath River Flows and Key Dates   21 

Contents

Figure 5: 2004 Augmented Klamath River Flows and Key Dates 22 Figure 6:
Groundwater Pumping for the Klamath Water Bank 28 Figure 7: Upper Klamath
Lake Expansion Options 34 Figure 8: Long Lake Storage Option 38 Figure 9:
Proportion and Number of Water Bank Contracts by Type,

2002 to 2004 46 Figure 10: Proportion and Volume of Water Acquired for the
Water Bank by Contract Type, 2002 to 2004 48

Abbreviations

BLM Bureau of Land Management
FWS Fish and Wildlife Service
NMFS National Marine Fisheries Service
NOAA National Oceanic and Atmospheric Administration
USGS U.S. Geological Survey

This is a work of the U.S. government and is not subject to copyright
protection in the United States. It may be reproduced and distributed in
its entirety without further permission from GAO. However, because this
work may contain copyrighted images or other material, permission from the
copyright holder may be necessary if you wish to reproduce this material
separately.

A

United States Government Accountability Office Washington, D.C. 20548

March 28, 2005

The Honorable Henry A. Waxman Ranking Member Committee on Government
Reform House of Representatives

The Honorable Mike Thompson House of Representatives

Located in the Upper Klamath Basin, the Bureau of Reclamation's
(Reclamation) Klamath Project (Project) is a federal water project
spanning the borders of southern Oregon and northern California. Initiated
in 1905, the Project was designed to dam the Upper Klamath Lake to manage
Klamath River flows, drain nearby lakes and marshlands to create
approximately 200,000 acres of farmland, and provide farmers with
irrigation water through an elaborate system of canals and drains. As a
result, Project operations largely determine the amount of water flowing
in the Klamath River, which subsequently passes through several
hydroelectric generating dams before running freely into the Lower Klamath
Basin and emptying into the Pacific Ocean, as shown in figure 1. On
average, about 1.5 million acre-feet-nearly 500 billion gallons-of water
pass from the Upper Basin to the Lower Basin annually.

Figure 1: The Klamath River Basin

                              Source: Reclamation.

Drought conditions since 2000 have made it difficult for Reclamation to
balance the demands for irrigation water by farmers on the Project with
the requirements for specific river flows and lake levels for threatened
and endangered species. The southern Oregon/northern California coho
(coho), a species of salmon native to the river, was listed as threatened
under the Endangered Species Act in 1997, and two species of sucker in
Upper Klamath Lake were listed as endangered in 1988. Project operations
were severely impacted in April 2001 when Reclamation cut off water
deliveries to the majority of Project irrigators in order to meet river
flow and lake level requirements to protect the coho and suckers under the
act. As a result, agricultural production was impaired or eliminated on
much of the Project, and some individuals engaged in acts of civil
disobedience to protest Reclamation's actions.

To avert future crises similar to 2001, Reclamation proposed a new 10-year
Project operations plan for 2002 through 2011. As required under the act
and applicable regulations, Reclamation consulted with the National Marine
Fisheries Service (NMFS) on its biological assessment of the plan to
determine its effect on listed species.1 NMFS issued a final biological
opinion finding that the plan would jeopardize the continued existence of
the coho and likely adversely modify its critical habitat.2 In its
opinion, NMFS recommended an alternative plan to protect the coho, which
included the establishment of a multiyear water bank to provide additional
river flows that would better protect critical habitat. Reclamation
incorporated the water bank program into its Project operations plan
through 2011. The biological opinion also provides for reinitiation of
consultation to modify the plan if, for example, new scientific
information on river flow requirements for coho becomes available.

1NMFS, within the Department of Commerce's National Oceanic and
Atmospheric Administration, is responsible for administering the act with
regard to ocean dwelling and anadromous species, such as salmon, which
live part of their lives in freshwater and part in saltwater. See 50
C.F.R. S: 402.01(b)(2004); See GAO, Endangered Species: Federal Agencies
Have Worked to Improve the Consultation Process, but More Management
Attention Is Needed, GAO-04-93 (Washington, D.C.: Mar. 19, 2003).

2The biological opinion was the subject of litigation in 2003 when it was
challenged as arbitrary and capricious by several environmental groups.
The U.S. District Court in California agreed, in part, and remanded the
opinion to NMFS with instructions to amend it by addressing certain
deficiencies. The court added that the biological opinion would remain in
place until NMFS' amendment was issued. Pacific Coast Federation of
Fishermen's Associations v. Bureau of Reclamation, U.S. Dist. LEXIS 13745
(N.D. Cal. 2003).

To comply with the biological opinion, Reclamation must meet certain river
flow requirements and also provide a water bank of 30,000 acre-feet in
2002, 50,000 acre-feet in 2003, 75,000 acre-feet in 2004, and 100,000
acre-feet from 2005 through 2011 to supplement river flows.3  In broad
terms, a water bank is an institutional mechanism that facilitates the
transfer of water entitlements between users and/or uses. As such,
Reclamation's water bank is not a physical reservoir where water can be
deposited and withdrawn as needed but is an administrative process under
which Project irrigators who volunteer to participate are paid by
Reclamation to forego their contractual entitlement for one irrigation
season in order to make more water available for release downstream. Water
accrues to the water bank over the course of the year when participants do
not use irrigation water for their crops as they normally would. By March
31 each year NMFS and Reclamation will meet to determine a flow schedule,
including water bank deliveries. According to Reclamation, the water bank
provides a temporary means to augment river flows for threatened species
and also allows it to meet its contractual responsibility to deliver water
to Project irrigators while longterm, basinwide solutions for balancing
water demands are evaluated. To collaboratively develop potential
long-term solutions to improve habitat conditions, some of which could
increase river flows, the biological opinion also requires Reclamation to
initiate a Conservation Implementation Program to bring together
stakeholders, including federal agencies, tribes, and the states.

This report addresses (1) how Reclamation operated the water bank and how
much it cost from 2002 through 2004, (2) whether Reclamation met its
annual water bank obligations each year, (3) the water bank's impact on
water availability and use in the basin, and (4) alternative approaches
for achieving the water bank's objectives.

To address the objectives of this report, we visited the Klamath Project
and met with and collected documentation from Reclamation and NMFS
officials, as well as representatives from other stakeholder federal
agencies, tribes, irrigators, commercial fishermen, academics, and
conservationists. To determine how Reclamation operated the water bank and
its costs, we analyzed water bank planning, contracting, and expenditure
documentation. To determine whether Reclamation met its annual water bank
obligations, we analyzed water bank contracts and

3Reclamation identifies the water bank as a pilot program that the agency
contends is not subject to the National Environmental Policy Act.

documentation of Klamath River flows. To describe the water bank's impact,
we analyzed relevant land, surface water, and groundwater use data;
reviewed relevant studies; and met with stakeholders. We did not review
the water bank's impact on fish species because the short history of the
water bank makes it difficult to obtain reliable information. To describe
alternative approaches to the water bank, we met with potential land
sellers, reviewed studies of water storage options, and reviewed the
status of basinwide efforts to increase flows. We performed our work
between May 2004 and February 2005 in accordance with generally accepted
government auditing standards.

Results in  Brief	Reclamation has modified its water bank operations from
year to year as its obligations and costs increased. Reclamation obtained
water for the water bank by contracting with irrigators to either (1)
forego irrigation altogether (crop idling), (2) irrigate using only well
water (groundwater substitution), or (3) pump well water into the
irrigation canals for others to use (groundwater pumping). Based on each
year's experience, Reclamation modified its water bank operations to
better meet its increasing obligations and to mitigate costs. For example,
in 2003, Reclamation solicited applications for water bank participation
from irrigators to either forego irrigation water or substitute
groundwater at fixed rates, while in 2004, Reclamation broadened the
program's selection criteria to include contingency contracts for
groundwater pumping that could be activated "as needed" to deliver
additional water and sought to reduce costs by competitively bidding rates
with irrigators. However, as Reclamation's water bank obligation increased
each year, the water bank's expenditures also increased. Reclamation's
water bank expenditures through 2004 totaled more than $12 million. Based
on Reclamation's projected annual costs of about $7.6 million for fiscal
years 2005 through 2011, the cumulative cost of the water bank could
exceed $65 million through fiscal year 2011.

While Reclamation has met its water bank obligations each year since 2002,
its management and accounting practices have created confusion for
stakeholders. Our analysis of water bank contracts and river flow records
found that Reclamation acquired and delivered the required amount of water
for 2002 through 2004. However, the manner in which the agency has managed
and accounted for the water bank has caused confusion for stakeholders,
such as tribes and irrigators. For example, on issues where the biological
opinion is silent-such as how to count any water spilled from dams to
prevent flooding and regarding when, or if, Reclamation can

reclassify baseline river flow requirements-Reclamation has not been clear
in communicating what actions it took and why it took those actions,
resulting in a lack of transparency for stakeholders regarding the
operation of the water bank. Furthermore, Reclamation has not provided
stakeholders with systematic and clear information concerning the water
bank's status or operations and its decision to use river flow data that
are not publicly available has limited stakeholders' ability to
independently monitor water bank activities. This has led to confusion and
doubt among stakeholders on whether Reclamation actually met its water
bank obligations. We are recommending that Reclamation take steps to
improve its communications regarding the operation of the water bank.

The water bank appears to have increased the availability of water to
enhance river flows by reducing the amount of water diverted for
irrigation in the Project, but there is uncertainty regarding the extent
of its impacts on river diversions and groundwater use. In 2003, when the
water bank primarily relied on crop idling to obtain water, 20,335 Project
acres were unirrigated, about 60 percent more than 2002. However, because
of annual variations in irrigation demand and because Reclamation does not
have reliable water flow measurement equipment on the Project and
monitoring data for the Project, assessing the precise impact of the water
bank on river flows has been an ongoing issue. Moreover, throughout the
life of the water bank, Reclamation has used varying assumptions regarding
the amount of water that can be saved by crop idling as more research and
information has become available about this practice. Because of the
uncertainty about how much water crop idling provided in 2003, in 2004,
Reclamation officials decided to rely primarily upon metered groundwater
sources for the water bank. However, the U.S. Geological Survey (USGS) and
Oregon Water Resources Department have found evidence that groundwater
aquifers under the Project, already stressed by drought conditions, are
refilling at a slower than normal rate in recent years. Many wells have
shown significant declines in water levels, and an increasing number of
wells have been deepened to reach groundwater in Klamath County in recent
years. Reclamation is considering lessening its reliance on groundwater
pumping and substitution for the 2005 water bank, but is uncertain whether
it can meet its water bank obligations, particularly for spring flows,
while increasing its reliance on crop idling.

Although several alternative approaches for achieving the water bank's
objectives have been identified by Reclamation and other stakeholders,
limited information is available regarding their feasibility or costs.
Possible alternatives to the water bank have been studied to various
levels of detail,

including permanently retiring Project land from irrigation or adding new
short-term or long-term water storage capacity. For example, under the
land retirement alternative, at least 50,000 acres of irrigated land would
need to be permanently removed from agricultural production to achieve an
estimated 100,000 acre-foot reduction in irrigation. However, the
feasibility and costs of land purchases and the impacts of this
alternative on river flows and the agricultural economy have not been
fully assessed. Similarly, there are several options for increasing water
storage, either by expanding storage on Upper Klamath Lake or by building
a separate reservoir. However, there is little reliable information
available regarding the total costs, environmental impacts, and certainty
of water availability for storage under these alternatives. Although NMFS'
2002 biological opinion required the collaborative study of the
feasibility of alternatives to increase river flows, Reclamation and other
stakeholders are still developing the framework for this process. In the
interim, Reclamation and NMFS have an ongoing dialogue regarding water
bank management and will have the opportunity to consider alternative ways
to more effectively manage the water bank when they meet for a planned
reconsultation on the biological opinion in 2006. For example, Reclamation
officials may consider proposing more flexibility to manage water bank
volumes in wet or above average water years, thus preserving funding and
resources for dry years.

We are recommending that Reclamation take steps to improve the information
provided to stakeholders regarding water bank management and accounting by
regularly and systematically providing-through media such as a water bank
Web-link or a monthly or biweekly press release- public information on the
rationale and effects of management decisions related to forecasted water
availability, unexpected spill conditions, or other significant events, as
well as regularly updated information regarding the water bank's status,
including the amount of water bank deliveries to date. In commenting on a
draft of this report, the Departments of Commerce and the Interior
generally agreed with our findings. Reclamation concurred with our
recommendation, agreeing to add a water bank page to its Internet Web site
that will include background information on the water bank, current
information that is regularly updated, such as the status of water bank
deliveries, and links to other relevant Web resources. Commerce and
Interior provided written technical comments which we incorporated as
appropriate. We requested comments from the Department of Agriculture but
none were provided. Interior's comments appear in appendix III and
Commerce's comments appear in appendix IV.

Background	The Klamath River Basin, spanning the southern Oregon and
northern California borders, covers over 15,000 square miles. The Klamath
River originates in the Upper Basin, fed by Oregon's Upper Klamath Lake, a
large, shallow body of water composed of flows from the Sprague,
Williamson, and Wood Rivers. The river subsequently flows into the Lower
Basin in California, fed by tributaries including the Shasta, Scott,
Salmon, and Trinity Rivers, and empties into the Pacific Ocean. River
flows and lake levels depend primarily upon snowpack that develops during
the winter months, melts in the spring, and flows into the river basin.
Rainfall and groundwater from natural springs also contribute to flows. On
average, about 1.5 million acre-feet of water pass from the Upper Basin to
the Lower Basin annually at Iron Gate Dam.

The Secretary of the Interior authorized construction of the Klamath
Project in 1905.4 Reclamation dammed Upper Klamath Lake, drained and
reclaimed Lower Klamath and Tule Lakes, stored the Klamath and Lost
Rivers' flows, and provided irrigation diversion and flood control on the
reclaimed land. About 85 percent of the Project lands obtain irrigation
water from Upper Klamath Lake and the Klamath River, while Gerber
Reservoir, Clear Lake, and the Lost River supply the remainder of the
Project. Water is delivered to Project lands using an elaborate system of
canals, channels, and drains, including diversions directly from the
Klamath River. The distribution system is considered highly efficient,
ensuring that water that is diverted for use within the Project is reused
several times before it returns to the Klamath River. Homesteading of the
reclaimed lands began in 1917 and continued through 1948.

As shown in figure 2, the Project is currently composed of about 207,000
acres of irrigable lands. Historically, about 200,000 acres of Project
lands have been in agricultural use annually.5 For example, in 2003, the
most recent year for which data is available, about 202,000 acres were
considered to be in agricultural use, of which about 180,000 acres were
irrigated and harvested. Crops grown and harvested on the Project include
alfalfa, barley, oats, wheat, onions, potatoes, and peppermint, and cattle

4Interior authorized the Project under provisions of the Reclamation Act
of 1902, 32 Stat. 388.

5Includes harvested, unharvested, and fallowed land. Fallowing is the
practice of not seeding land for one or more seasons, for example, to
destroy weeds or conserve soil moisture.

graze on more than 40,000 acres of irrigated pastureland. In addition to
farm and pastureland, four national wildlife refuges were set aside by
executive orders in conjunction with the construction of the Project.6 The
refuges, managed by Interior's Fish and Wildlife Service (FWS) support
many fish and wildlife species and provide suitable habitat and resources
for migratory birds of the Pacific Flyway. About 23,000 acres of the two
refuges within the Project water delivery area-Tule Lake and Lower Klamath
National Wildlife Refuges-are leased for agricultural purposes.7

6The Lower Klamath Lake Wildlife Refuge was established in 1908; the Clear
Lake Wildlife Refuge was established in 1911; the Upper Klamath Lake and
Tule Lake Wildlife Refuges were established in 1928.

7The Kuchel Act (Pub. L. No. 88-567 (1964)) specifies that refuge lands be
leased for agricultural use to the extent consistent with their primary
purposes, waterfowl management. Contracts are issued for 5 to 8 years but
require annual renewal. These lands are the most productive lands in the
Klamath Basin and represent about 10 percent of the land area receiving
Project water.

Figure 2: Reclamation's Klamath Project

                              Source: Reclamation.

Note: Due to space limitations, Upper Klamath Lake is not shown to scale
with the Klamath Project.

Reclamation, through contracts, provides water for irrigation and
hydropower production and must also provide water for the national
wildlife refuges. Reclamation has entered into contracts with numerous
irrigation districts and individual irrigators on the Project to provide
for the repayment of Project costs and the right to receive Project water.
The contracts most commonly specify a land acreage amount to be covered by
the contract-not a specific water amount to be delivered. Also by contract
with Reclamation, California-Oregon Power Company (now PacifiCorp)
obtained the right to use certain amounts of water, after requirements of
the Klamath Project are satisfied, for hydropower generation at its
privately owned and independently operated dams on the Klamath River
downstream of the Project.8 PacifiCorp's southernmost hydropower dam, Iron
Gate Dam, located about 20 miles downriver of the Oregon-California
border, is the last control point before Klamath River flows run freely to
the Pacific Ocean. Finally, the national wildlife refuges have federally
reserved rights for the water necessary to satisfy the refuges' primary
purposes, and Reclamation must satisfy refuge water needs after its other
obligations are met.

Reclamation is also obligated to protect tribal trust resources, such as
water and coho salmon. The Klamath River Basin is home to four federally
recognized tribes, identified by Reclamation as the Klamath Tribes in the
Upper Basin area of Oregon, and the Hoopa Valley Tribe, Yurok Tribe, and
Karuk Tribe in the Lower Basin area of California. Each tribe has
longstanding cultural ties to the Klamath River, its tributaries, and
native fish species. Furthermore, the Klamath, Hoopa, and Yurok tribes
have, either by treaty or executive order, reserved rights to sufficient
water quality and flows to support all life stages of fish life in
protection of tribal fishing rights.9 As with all federal agencies,
Reclamation has a trust responsibility to protect these tribal resources
and to consult with the tribes regarding its actions in a
government-to-government relationship.

Reclamation must comply with the Endangered Species Act to ensure that any
action it authorizes, funds, or carries out is not likely to jeopardize
the continued existence of any listed species of plant or animal or
adversely modify or destroy designated critical habitat. Interior's FWS
and

8PacifiCorp's hydroelectric dams are operated under a Federal Energy
Regulatory Commission license; relicensing of the power project is
scheduled for 2006.

9The Karuk Tribe of California did not obtain federal tribal recognition
until 1979, and is seeking recognition of a fishing right.

Commerce's NMFS are responsible for administering the act.10 If FWS or
NMFS finds that an agency's proposed activity is likely to jeopardize a
threatened or endangered species or destroy or adversely modify its
critical habitat, then a "reasonable and prudent alternative" that would
avoid such harm must be identified. Three species of fish that are of
particular importance to the cultures of the tribes-the threatened
southern Oregon/northern California coho, and the endangered Lost River
sucker and shortnose sucker-are affected by Project operations. NMFS
listed the coho as threatened in 1997, and FWS listed the two species of
suckers, which populate Upper Klamath Lake and rivers other than the
Klamath, as endangered in 1988.

Drought conditions since 2000 have complicated Reclamation's efforts to
balance the irrigation water demands on the Project with the requirements
for specific river flows and lake levels for threatened and endangered
species. Reclamation operates the Project according to an annual
operations plan that helps the agency to meet its various obligations and
responsibilities, given varying hydrological conditions. In 2001,
responding to Reclamation's biological assessment of its proposed Project
operations plan, FWS and NMFS issued biological opinions that suggested11
Reclamation take numerous actions, including maintaining higher water
levels in Upper Klamath Lake and two reservoirs on the Lost River and
higher flows of the Klamath River below Iron Gate Dam. Because of the new
biological opinions and drought conditions, Reclamation was prohibited
from releasing normal amounts of water to most Project irrigators, which
impaired or eliminated agricultural production on much of the Project.

Subsequently, Reclamation proposed a new 10-year Project operations plan
for 2002 through 2011. NMFS reviewed Reclamation's biological assessment
of the plan to determine its effect on listed species, and issued a final
biological opinion on May 31, 2002, directing Reclamation to establish a
multiyear water bank to provide additional river flows.

10Along with NMFS, the FWS administers the act and is responsible for
protecting terrestrial, or land-dwelling species, and freshwater animal
and plant species, including suckers.

11Although the opinion "suggested" that Reclamation implement certain
measures, such as a water bank, to prevent harm to the threatened or
endangered species, such recommendations were required to be implemented
if Reclamation was to be protected under the act from enforcement actions
for prohibited "takings," for example, actions that result in harm,
harassment, or the killing of protected species. See 16 U.S.C. S:S:
1532(19), 1536(b)(4), 1536(o)(2), 1538(a)(1)(B).

Reclamation incorporated this water bank into its Project operations plan
through 2011. NMFS and Reclamation can reconsult on the requirements of
the biological opinion as warranted, for example, if new scientific
information on river flow requirements for fish is developed.
Reconsultation is likely during 2006, when ongoing studies of Klamath
River flows are expected to be completed.

Although NMFS' biological opinion recommended a water bank as an
alternative and specified the amounts of water to be provided each year,
it provided Reclamation little specific guidance regarding the structure,
management, or operation of the water bank. Water banking is broadly
defined as an institutional mechanism that facilitates the legal transfer
and market exchange of various types of surface water, groundwater, and
water storage entitlements. Water banks have been proposed or are
operating in almost every western state. However, significant differences
exist in the way that each bank operates with respect to market structure,
degree of participation, pricing, regulatory oversight, environmental
objectives, and other factors.12

Under Reclamation's water bank program, participating irrigators would be
paid to forego their contractual entitlement to water for one irrigation
season in order to make more water available for release into the river.
Water acquired by Reclamation would accrue to the water bank over the
course of the year as participants did not divert water for irrigation
purposes as they normally would. A schedule for delivery of additional
flows is determined by NMFS and Reclamation by March 31 each year, with
the majority of the water bank provided in the spring and early summer
when the water is most needed by the coho. According to Reclamation, the
water bank would enable the agency to augment river flows for threatened
species and also meet its contractual responsibility to deliver water to
Project irrigators until other solutions for balancing water demands were
identified. Reclamation was also required to initiate a Conservation
Implementation Program that would bring together basin stakeholders,
including federal agencies, tribes, and the states, to collaboratively
develop

12California's state administered Drought Water Bank, operating since
1991, is a 1-year leasing program to reallocate water between users during
drought conditions and acts as a clearinghouse that pools water and
allocates supplies to critical demands in the state. In contrast, Oregon's
Deschutes Water Exchange Groundwater Mitigation Bank, administered by the
Deschutes Resources Conservancy, facilitates surface water leases and
time-limited transfers to create mitigation credits for groundwater
pumping in the Deschutes Basin.

long-term solutions, some of which would increase flows, such as surface
water storage and groundwater resource development.

Reclamation Modified Its Water Bank Operations from Year to Year as Its
Obligations and Costs Increased

Reclamation modified its water bank operations from year to year as its
obligations and costs increased. Reclamation acquired water for the water
bank by contracting with irrigators for the water needed to augment
Klamath River flows as required by the biological opinion. As it gained
more experience each year, Reclamation modified its water bank operations
to better meet the increasing obligations and to mitigate costs. As its
annual obligations increased, Reclamation's annual water bank expenditures
also increased, totaling more than $12 million through 2004. Based on
Reclamation's estimated annual cost of about $7.6 million for fiscal year
2005 and onward, the cumulative cost of the water bank could exceed $65
million through fiscal year 2011.

Reclamation Modified Its Water Bank Operations from Year to Year to Meet
Increasing Obligations

Reclamation initiated the Klamath Project water bank program in 2002, as
recommended under NMFS' biological opinion, with the objective of
purchasing irrigators' water entitlement for one irrigation season so that
this water could be used to provide additional Klamath River flows for
threatened coho salmon. The water bank is not a physical reservoir of
stored water but an administrative mechanism through which Reclamation
contracts with irrigators both on and off the Klamath Project. Through
these contracts irrigators agreed to either (1) forego irrigation
altogether (crop idling), (2) irrigate using only well water (groundwater
substitution), or (3) pump well water into the irrigation canals for
others to use (groundwater pumping), thus making water available to
augment river flows. Water accrues to the water bank over the course of
the irrigation season as water bank contractors forego irrigating their
land by crop idling or groundwater substitution and as groundwater is
pumped into canals under water bank contracts. However, because
Reclamation is required to provide large amounts of water in spring and
early summer before sufficient water has accrued to the water bank, it
actually "borrows" water for the bank from short-term storage supplies.
This water is later replaced by foregone irrigation water over the course
of the year.

Reclamation modified its water bank operations each year, changing its
composition, selection process, contracting process, and program rules as
it gained experience to meet its increasing obligations. In 2002 when the
obligation was 30,000 acre-feet, the water bank sources included crop

idling off-Project and groundwater pumping; in 2003 when the obligation
was 50,000 acre-feet, sources included crop idling on-Project and
groundwater substitution; and in 2004 when the obligation was 75,000
acrefeet, all three sources of water were included in the water bank.
Reclamation modified the selection process from relying on only two
irrigators in 2002-without a public application process13-to soliciting
applications from any qualified irrigator in both 2003 and 2004. In 2004,
Reclamation solicited water bank applicants earlier in the year than it
had in 2003, in part, to allow successful applicants more lead time in
planning their irrigation.14 Reclamation also modified the contracting
process to obtain more flexibility by competitively bidding contract rates
in 2004 rather than paying a fixed rate as in 2003 and entering into
contingency contracts for groundwater pumping that could be activated "as
needed" to deliver additional water to meet its increasing water bank
obligation and uncertain delivery schedule. These contingency contracts
allowed Reclamation to acquire only the amount of water it needed to meet
the agreed upon delivery schedule. Finally, Reclamation expanded its
program rules to make participation in the water bank more practical and
attractive to potential applicants. For example, Reclamation changed the
rules for the 2004 water bank to allow harvesting of crops on land under
crop idling contracts, reflecting the fact that some crops such as alfalfa
can grow with water from subsurface moisture alone.

Similarly, Reclamation modified its monitoring process for the water bank
over time. For example, in 2003, Reclamation monitored every participant
for compliance with the program rules. Enforcement staff examined and
tested each crop idling parcel of land at least once over the course of
2003's water bank to ensure that no intentional irrigation occurred. In
addition, Reclamation relied on self-policing by irrigators who called in
tips identifying potential cheaters. A Reclamation official estimated a
greater than 95 percent compliance rate in 2003 and only terminated the
contract of one participant who intentionally irrigated fields after
deciding to withdraw from water bank participation without notifying
Reclamation. In

13While NMFS wanted Reclamation to operate a water bank in the spring of
2002, the final biological opinion was not released until May 31, 2002.
Nevertheless, NMFS and Reclamation agreed to operate the water bank from
May 1 to May 31, before the biological opinion was finalized and released.

14Those participating in the water bank do not have to forego irrigating
all of the land they own in a given year, but they must forego irrigation
entirely for the year on those lands accepted into the water bank.

contrast, during 2004 Reclamation sought to reduce enforcement costs and
increase efficiency by examining and testing crop idling parcels of land
only toward the end of the year while following up on tips identifying
potential cheaters throughout the year. In 2004, Reclamation found no
intentional violations.

Water Bank Costs Could Reclamation's water bank expenditures through
fiscal year 2004 exceeded Exceed $65 Million through $12 million and could
total more than $65 million through 2011. As shown Fiscal Year 2011 in
table 1, Reclamation's total expenditures have increased annually as the

water bank obligation has grown from 30,000 acre-feet in 2002 to 75,000

acre-feet in 2004.

Table 1: Reclamation's Water Bank Expenditures, Fiscal Years 2002 to 2004

                                  Fiscal year

                    Expenditures       2002       2003       2004       Total 
        Groundwater substitution                                  
            or pumping contracts $1,000,000 $1,788,711 $4,009,451  $6,798,162 
           Crop idling contracts          0 2,700,789     637,258   3,338,047 
         Klamath Basin Rangeland                                  
        Trust contracts for off-                                  
             Project crop idling    948,300          0    690,221   1,638,521 
                  Administration      2,479    175,233    255,119     432,831 
                           Other     10,215     22,213    144,785     177,213 
                           Total $1,960,994 $4,686,946 $5,736,834 $12,384,774 

Source: Reclamation, Klamath Basin Area Office.

Note: Dollar amounts are not adjusted for inflation.

Reclamation attributes the increasing costs of the water bank to the
increasing annual volume of water purchases, as well as increasing
administrative costs due to the large increase from 2002 to 2003 in the
number of contracts to manage and the addition of the groundwater pumping
program in 2004 and its associated contract negotiations. Reclamation
estimates that the 100,000 acre-foot water bank requirements for fiscal
years 2005 through 2011 will cost at least $7.6 million annually, bringing
the total water bank costs to more than $65 million. For 2005 and onward,
according to Reclamation, the water bank will be a specific budget

item in its budget request.15 Accordingly, Reclamation requested $7.626
million for fiscal year 2005 and plans to gradually increase annual budget
requests to about $7.660 million by 2011.

Reclamation's expenditures fall into five categories: groundwater contract
costs, crop idling contract costs, Klamath Basin Rangeland Trust contract
costs, administrative costs, and other costs. Reclamation's largest water
bank expenditures were for groundwater contracts with irrigators-for both
substitution and pumping-totaling nearly $7 million, or 55 percent of
total expenditures from 2002 through 2004. Reclamation's second largest
water bank expenditures were for crop idling contracts with Project
irrigators, totaling about $3.3 million, or 27 percent of total
expenditures. Reclamation's contracts with the Klamath Basin Rangeland
Trust to forego irrigation of pastureland outside of the Klamath Project
totaled more than $1.6 million, or about 13 percent of total water bank
expenditures through 2004. Reclamation's administrative costs-mainly
payroll and overhead- for planning and implementing the water bank
comprised about 3 percent of total water bank expenditures. Reclamation
also incurred other costs related to the operation of the water bank, such
as water quality analysis, contract compliance monitoring, and a contract
for assistance from the Oregon Water Resources Department.

Reclamation Met Its Water Bank Obligations, but Information Provided to
Stakeholders Could Be Improved

Reclamation met its water bank obligations to provide additional water to
supplement Klamath River flows each year since 2002. However, the manner
in which the agency has managed and accounted for the water bank has
caused confusion for some stakeholders, such as tribes and irrigators, and
has reduced the transparency of the water bank's status and operation.

15The 2002 through 2004 water banks were unanticipated and thus not
included in Reclamation's budget requests nor specifically provided for in
agency appropriations. Instead, funds appropriated for feasibility studies
under the Klamath Basin Water Supply Enhancement Act of 2000 were used, as
well as other budget sources such as the Central Valley Project and the El
Paso Water Reclamation and Reuse Program.

Reclamation Met Its Water Bank Obligations Each Year

According to NMFS and Reclamation officials, Reclamation's obligation is
to both acquire the amount of water required in the biological opinion
each year and deliver the water-some of it or all of it-in accordance with
the schedule mutually agreed to by both agencies. Regarding the
acquisition of water, NMFS concluded, and our analysis of Reclamation
contract records verified, that Reclamation met its obligation to acquire
30,000 acre-feet in 2002, 50,000 acre-feet in 2003, and 75,000 acre-feet
in 2004, by contracting for about 47,000; 59,000; and 111,000 acre-feet,
respectively. Appendix II provides detailed information on water bank
applications and contracts. According to Reclamation officials, they
contracted to acquire more water than required, in part to serve as a
buffer against unexpected changes in water conditions and as insurance
against uncertainty about how much water is actually obtained from crop
idling.

Regarding the delivery of water to augment flows, NMFS concluded, and our
analysis of USGS river flow records verified, that Reclamation met its
obligation each year as established in the schedule agreed upon with NMFS.
We found that, in total, Reclamation augmented Klamath River flows by
approximately 30,000 acre-feet within the brief 2002 water bank time
frame-meeting its 30,000 acre-feet schedule requirement; by more than
71,000 acre-feet in 2003-surpassing its 50,000 acre-feet schedule
requirement; and by more than 95,000 acre-feet in 2004-surpassing its
74,373 acre-feet schedule requirement. According to Reclamation officials,
these augmented flows represent water provided per water bank requirements
plus additional releases of water purchased and stored to meet tribal
trust obligations. Because the water bank is not a physical pool of water
allowing the constant measurement and monitoring of deposits and
withdrawals, estimating the status of water bank accruals or deliveries
and differentiating water bank deliveries from tribal trust deliveries
during the year is neither precise nor easy. Reclamation views water bank
deliveries as simultaneously meeting both its requirement to augment river
flows under the biological opinion and its tribal trust responsibilities.
However, to account for its annual deliveries, Reclamation officials have
generally counted augmented flows as first satisfying the water bank
requirement and consider excess flows, such as the approximately 20,000
acre-feet delivered above the water bank requirement in 2003 and 2004, as
tribal trust deliveries.

Augmented flow is defined as the volume of water in excess of base flows
measured at Iron Gate Dam.16 Klamath River base flows are determined
according to "water-year types." Based on an April 1 forecast of snowpack
and runoff, Reclamation initially classifies each year as Wet, Above
Average, Average, Below Average, and Dry in accordance with the biological
opinion.17 Each classification requires a specific base flow of water at
Iron Gate Dam. Forecasts are updated at least monthly, incorporating
actual water conditions as the year progresses, and providing Reclamation
with increasingly accurate data with which to determine if the water-year
type needs to be reclassified during the year.

In 2002, the water bank operated from May 1 to May 31, during which
Reclamation met its water bank delivery obligation by augmenting flows by
approximately 30,000 acre-feet, as shown in figure 3. NMFS released its
final biological opinion on May 31 directing Reclamation to operate a
water bank through 2011.

16Klamath River flows-actual and base-are calculated by taking the average
daily flow rates measured in cubic feet per second (1 cubic foot per
second per day equals 1.9835 acrefeet).

17The Department of Agriculture's Natural Resources Conservation Service
provides the forecasts.

Figure 3: 2002 Augmented Klamath River Flows and Key Dates

Cubic feet per second 2,500

2,000

1,500

1,000

o  2002 water  o  2002 water bank  o  Water-year  o  Fish die-off  o  Fish
die-off

bank begins	ends type changed begins ends from Below

o  Final 2002 Average toNMFS biological Dryopinion released

Average daily Klamath River flow

Base river flow required for a Below Average water-year

Base river flow required for a Dry water-year

Augmented flow equals 29,965 acre-feeta

Sources: GAO analysis of USGS river flow data and Reclamation water bank
and water-year type data.

aThe augmented flow during May 2002 consists of water bank deliveries
only.

Note: Klamath River flows are measured at Iron Gate Dam.

In 2003, Reclamation met its water bank delivery obligation by augmenting
flows by at least 50,000 acre-feet as agreed in its schedule with NMFS.
Heavier than expected rainfall in early spring prompted Reclamation to
move the official start of the water bank from April 1 to May 21, as shown
in figure 4. The water bank operated between May 21 and October 31,

during which time Reclamation reclassified the water-year type from Dry to
Below Average due to better than expected water conditions.

           Figure 4: 2003 Augmented Klamath River Flows and Key Dates

                             Cubic feet per second

3,500

3,000

2,500

2,000

1,500

1,000

4/1/03: 5/21/03: 6/13/03: 7/9/03: 10/31/03:

o  Initial start date  o  Spill condition  o  Water-year  o  Reclamation 
o  2003 water for 2003 water ends type changes declares water bank ends
bank  o  2003 water from Dry to bank exhausted,

o  Klamath Project bank begins Below subsequently

irrigation season (revised start Average reopened

begins date)  o  "Spill condition" exists

                        Average daily Klamath River flow

                 Base river flow required for a Dry water-year

            Base river flow required for a Below Average water-year

                    Augmented flow equals 71,428 acre-feeta

                             Spill condition exists

Sources: GAO analysis of USGS river flow data and Reclamation water bank
and water-year type data.

aThe augmented flow between May 21 and October 31 consists of both water
bank and tribal trust deliveries.

Note: Klamath River flows are measured at Iron Gate Dam.

In 2004, Reclamation met its water bank delivery obligation by augmenting
flows by at least 74,373 acre-feet as agreed in its schedule with NMFS. As
shown in figure 5, the 2004 water bank started on April 1 and ended on
October 31, as planned. The water-year type was reclassified from Below
Average to Dry on May 7, shortly after the water bank began due to a
smaller than expected water supply.

           Figure 5: 2004 Augmented Klamath River Flows and Key Dates

                             Cubic feet per second

4/1/04: 5/7/04: 10/31/04:

o  2004 water  o  Water-year type  o  2004 water bank begins changed from
Below bank ends

o  Klamath Project Average to Dry irrigation season begins

                        Average daily Klamath River flow

            Base river flow required for a Below Average water-year

                 Base river flow required for a Dry water-year

Augmented flow equals 95,176 acre-feeta Sources: GAO analysis of USGS
river flow data and Reclamation Water Bank and water-year type data.

aThe augmented flow during 2004 consists of both water bank and tribal
trust deliveries.

Note: Klamath River flows are measured at Iron Gate Dam. Average daily
flow data for October 11, 2004, is missing. Therefore, the 2004 augmented
flow calculation does not include data for that date.

Reclamation's Management and Accounting Practices Confused Stakeholders

Although Reclamation met its annual water bank obligations each year, the
manner in which the agency managed and accounted for the water bank
confused stakeholders. Specifically, for two issues where the biological
opinion is silent-how to count spill water released to prevent flooding,
and whether Reclamation can reclassify the water-year type designation
midyear due to changing water conditions-Reclamation has not been clear in
communicating related management and accounting decisions. Furthermore,
Reclamation has not provided stakeholders with systematic and clear
information concerning the water bank's status or operations, and its
decision to use river flow data unavailable to stakeholders limited
stakeholders' ability to independently monitor water bank activities. This
has led to confusion and doubts among stakeholders on whether Reclamation
actually met its water bank obligations.

Reclamation's management of the water bank during 2003's spill
condition-when water was released by dams to prevent overflow or
flooding-and its lack of clear communication with stakeholders, caused
significant confusion. Heavier than expected rainfall in early spring of
2003 caused a "spill condition" to exist on April 1 when the water bank
was set to begin. However, the biological opinion does not specify how
much, if any, spill water can be counted as a water bank delivery. In the
absence of specific guidance from NMFS, Reclamation could have counted
spill toward water bank deliveries in one of three ways: (1) up to the
amount already scheduled for delivery during the spill; (2) in its
entirety, including water above the scheduled amount; or (3) not at all.
According to Reclamation officials, they eventually decided to reset the
water bank's start to May 21-when the spill condition ended-and started
counting augmented flows as of that date. However, Reclamation did not
clearly communicate its decisions to stakeholders leading to confusion
among stakeholders on how, or if, Reclamation was meeting its water bank
obligations. For example, according to some tribal representatives,
Reclamation provided a preliminary status report in July stating that over
20 percent of the water bank-over 11,000 acre-feet-was delivered during
the spill. This contradicted a Reclamation official's statement that the
agency had retroactively reset the water bank's start date to after the
spill conditions ceased. Reclamation officials concede that stakeholder
confusion as a result of these actions was understandable. Subsequently,

NMFS and Reclamation agreed that, beginning in 2004, spill water will be
counted only up to the amount already scheduled for delivery by the water
bank.

Similarly, Reclamation's reclassification of the water-year type-which
determines the base river flow requirement-has also caused confusion for
stakeholders. Like spill conditions, the biological opinion is silent on
whether the water-year type can be reclassified midyear after its initial
determination in April. NMFS and Reclamation officials contend that
reclassifying the water year type to reflect changing water conditions is
necessary to reflect the most current and accurate data available.
However, some stakeholders, such as the tribes, contend that midyear
reclassification is not allowed under the biological opinion and could
lead to the improper manipulation of water bank delivery schedules. While
Reclamation issued a press release informing the public of its
reclassification of the water-year type in 2003, the impact of such
changes on the water bank was not clearly articulated. For example,
Reclamation did not mention that it would also change its estimate of
year-to-date water bank deliveries as a result of the midyear
reclassification in water-year type, leading to stakeholder concerns that
water deliveries were being manipulated to benefit irrigators at the
expense of fish. Reclamation officials believe that reclassifying
water-year types is in compliance with the biological opinion and that the
confusion related to reclassifying the water-year type stems from
Reclamation's attempt to incorporate the most recent and accurate data on
water conditions in their water bank delivery schedules.

Reclamation also has not clearly or systematically communicated the water
bank's status and operations, further increasing stakeholder confusion.
Specifically, Reclamation does not have a systematic mechanism to
communicate information regarding the water bank to all stakeholders.
Rather than regularly providing updated calculations of year-to-date
deliveries to all stakeholders simultaneously through a single mechanism,
such as a Web site or regularly scheduled press releases, Reclamation
provides information on the water bank and its status "upon request" and
through occasional press releases. Consequently, different stakeholders
receive different information at different times. According to Reclamation
officials, they meet regularly with the tribes and discuss the water
bank's status. However, Reclamation does not systematically seek feedback
on the operation of the water bank from all stakeholders, limiting the
opportunities to clarify misunderstandings. As a result, after several
years of operation, questions continue to persist among stakeholders,
including

some Project irrigators and tribes, on basic topics such as the purpose of
the water bank. Reclamation placed some information about the water bank
application process on its Web site; however, Reclamation has not made
other water bank information-such as the year-to-date status- available
since that time, in part, because Reclamation has been reluctant to
release status information that will almost certainly require revision
later in the year.

Finally, Reclamation's use of river flow data generated by PacifiCorp to
estimate the water bank's river flow augmentation has reduced the
transparency of the water bank and limited the ability of stakeholders to
independently monitor the operation of the water bank. The PacifiCorp data
used by Reclamation to calculate actual Klamath River flows is not
available to the public. Therefore, interested stakeholders must use a
different source-the publicly available USGS data on actual Klamath River
flows-to calculate year-to-date water bank deliveries. The PacifiCorp and
USGS flow data differ because each uses a different formula to calculate
the average daily flow. Thus, Reclamation and stakeholders will arrive at
different augmented flow calculations, depending upon which data source
they use. For example, we found that, in 2003, augmented flows appeared to
be about 2,500 acre-feet greater when using USGS data than when using
PacifiCorp data. Furthermore, Reclamation, using PacifiCorp data, would
calculate that it had met its water bank obligation on a different date
than a stakeholder would using USGS data, creating the potential for
stakeholder confusion and doubt regarding the status of water bank
deliveries. Reclamation officials told us that as of October 2004 they
began using the publicly available USGS data to calculate and communicate
the water bank's status.

The Water Bank Appears to Have Increased the Availability of Water for
River Flows by Reducing Irrigation Use, but the Extent of Its Impacts is
Unclear

Reclamation's water bank appears to have increased the availability of
water to enhance river flows by reducing irrigation water use on the
Project, but there is uncertainty regarding the extent of its impacts on
river diversions and groundwater resources. In 2003, when the water bank
primarily relied on crop idling to obtain water, there was a significant
increase in the amount of land not using irrigation water compared with
recent years. While it was likely that a reduction in river and lake
diversions for Project irrigation resulted, a university study funded by
Reclamation found that the reduction attributable to the water bank alone
was highly uncertain due to the lack of effective flow measurement
equipment and monitoring data for the Project. Because Reclamation was
uncertain about how much water crop idling actually provided to the water

bank, Reclamation shifted to groundwater substitution and pumping as the
primary sources for the 2004 water bank. However, USGS and Oregon state
officials have since found evidence that groundwater aquifers under the
Project, already stressed by drought conditions, are being pumped by an
increasing number of wells and refilling at a slower than normal rate,
prompting Reclamation to consider lessening its future reliance on
groundwater substitution and pumping.

Crop Idling under the Water Bank Has Reduced the Amount of Irrigated Land,
but the Extent of Its Impacts on River Diversions Is Unclear

In 2003, Reclamation obtained about 60 percent of its water bank
acquisitions by contracting with irrigators for crop idling on nearly
14,500 acres of land, based on the assumption that water foregone from
irrigation on those lands would be available to enhance river flows. Crop
idling contributed to a significant increase in the amount of land not
irrigated in 2003, compared with recent years. For example, according to
Reclamation's 2003 crop report, a total of 20,335 Project acres were not
irrigated, which is about a 60 percent increase over 2002 when 12,546
acres of land were not irrigated, and well exceeds the average of 7,665
acres of Project land not irrigated due to agricultural fallowing
practices from 1998 through 2000-the three years preceding Reclamation's
restriction of irrigation water in 2001.

Although the number of acres of crop land idled is a useful indication of
the water bank's impacts, it does not provide a reliable estimate of the
true extent to which irrigation water has been made available for river
flows. According to Reclamation officials, the precise impact of the water
bank cannot be determined because of year-to-year variation in irrigation
demand and its determining factors such as temperature, precipitation, and
crop types. Moreover, throughout the life of the water bank, Reclamation
has used varying assumptions about the amount of water that can be saved
by crop idling as more research and information has become available about
this practice. Specifically, in 2002, Reclamation assumed that it could
obtain about 5 acre-feet of irrigation water per acre of crop idling, in
2003 and 2004 assumed 2.5 acre-feet, and is currently assuming that it can
obtain 2 acre-feet per acre through crop idling. To help it quantify the
actual results of the water bank, Reclamation has turned to other
organizations for assistance. For example, after the 2002 water bank was
completed, Reclamation engaged USGS to review the assumptions and results
for its off-Project crop idling. In February 2004, USGS reported to
Reclamation that, based on the available data, the amount of water
actually obtained per acre of crop land idled during the 2002 water bank
was most likely in the range of .9 to 1.3 acre-feet of water per acre.

Similarly, in 2003, Reclamation was again unable to obtain precise
information on the measurable impacts of the water bank for the year, so
it contracted with California Polytechnic State University to study this
issue. This study concluded that without effective flow measurement
equipment and monitoring data for the Project it could not precisely
estimate the impact of the water bank in reducing Upper Klamath Lake and
Klamath River diversions to the Project. According to the study, in 2003
the reduction in diversions compared with 2000 may have ranged from 11,000
to 71,000 acre-feet and, moreover, this reduction may have been
attributable to numerous other factors in addition to the water bank, such
as heavy rainfall, a large amount of groundwater pumping, changes in
irrigation district operations, and awareness among Project irrigators of
the need to reduce water use. Based on subsequent university analysis,
Reclamation now estimates that it actually obtained about 2 acre-feet of
water per acre from crop idling in 2003 and 2004.

Despite the ongoing uncertainty regarding the impact that reducing the
amount of irrigated land has on the availability of water for river flows,
Reclamation officials told us they must continue to rely on crop idling
for a significant portion of the water bank. While some stakeholders favor
taking farmland out of irrigation, they are also uncertain of the extent
to which crop idling reduces diversions for irrigation. For example, both
tribal and fishing industry representatives told us that they doubt that
Reclamation can accurately estimate how much additional water is actually
made available to the river. Some irrigators question the effectiveness
and accountability of crop idling as a strategy for the water bank, and
also are concerned about the economic impacts to taking farmland out of
production.

Groundwater Pumping for the Water Bank under Drought Conditions Raises
Concerns about the Impacts on Aquifers

Because of the uncertainty regarding the measurable impact of crop idling,
Reclamation shifted to groundwater for most of its water bank acquisitions
in 2004; however, the impact of groundwater pumping on basin aquifers
during ongoing drought conditions is largely unknown and continued
reliance may not be sustainable. Reclamation obtained over 70 percent of
the 2004 water bank deliveries by pumping nearly 60,000 acre-feet of
water, either to substitute for irrigation water or to fill canals for use
by others. Figure 6 below shows a groundwater pump delivering water into a
canal for the water bank. According to Reclamation officials, in the
absence of stored water, groundwater pumping is the only way to meet
required flows in the spring and early summer because land idling provides
little water in the April through June time period. An advantage of
groundwater pumping

for the water bank is that, unlike crop idling, flow meters on pumps and
wells allow the exact measurement of the amount of groundwater being used
in place of river diversions for irrigation.

Figure 6: Groundwater Pumping for the Klamath Water Bank

Source: GAO.

The impact of groundwater pumping on Upper Basin aquifers, however, is not
well understood, and its use during drought conditions is a matter of
growing concern for Reclamation and others. The basin has suffered drought
conditions since 2000, resulting in less rain and snowmelt to fill lakes,
rivers, and aquifers. Recognizing that water demand would cause more users
to turn to groundwater but that there is little reliable information on
the groundwater hydrology of the Upper Klamath Basin, USGS and the Oregon
Water Resources Department initiated a cooperative study in 1998 to study
and quantify the Upper Basin's previously unknown groundwater flow system.
The study, funded in part by Reclamation, is expected to be substantially
completed in 2005.

Nevertheless, USGS and Oregon Water Resources Department officials have
found evidence that groundwater aquifers in the Upper Basin, already

stressed by drought conditions, are being pumped by an increasing number
of newly drilled wells and refilling at slower than normal rates in recent
years. According to state officials, well drilling sharply increased after
2000, and an increasing number of domestic wells have needed to be
deepened-a symptom of dropping water levels-in Klamath County during that
same time frame. According to state records for Klamath County, Oregon,
from 1998 to 2000, 14 irrigation wells were drilled; from 2001-when
Project deliveries were restricted-through 2003, 124 irrigation wells were
drilled. From 1998 to 2000, 21 domestic wells were deepened; from 2001 to
2003, 30 domestic wells were deepened; and in 2004, another 13 were
deepened. Furthermore, USGS officials have identified wells in various
parts of the Upper Basin, within and outside the Project boundaries, which
have shown significant water level declines. For example, wells outside
the Project have shown declines of up to 10 feet since 2000, thought to be
primarily attributable to climatic conditions. Wells within the Project
have shown a variety of responses to pumping- some wells seem to decline
during irrigation season and then recover substantially during winter
months, while other wells have shown steady year-to-year declines, some
dropping more than 15 feet.

Reclamation engaged USGS in May 2004 to conduct an assessment of their
current water bank strategies and any potential strategies that could help
the agency meet its obligations. Specifically, Reclamation asked USGS to
(1) document current and planned water bank activities, (2) assess the
effectiveness of the 2003 and 2004 water banks, (3) determine if
sufficient information is available to assess the impact of the water bank
on Klamath River flows, and (4) develop a matrix of water bank management
options, including their potential positive or negative consequences. In
December 2004, USGS officials briefed Reclamation officials on their
assessment, presenting the pros and cons of various management options to
assist Reclamation's 2005 water bank planning. Reclamation officials are
considering lessening their reliance on groundwater pumping and
substitution for the 2005 water bank but are uncertain whether they can
meet their water bank obligations, particularly for spring flows, while
significantly increasing their reliance on crop idling.

Limited Information Is Available Regarding Alternative Approaches for
Achieving Water Bank Objectives

While several alternative approaches for achieving the water bank's
objectives have been identified by Reclamation and other stakeholders,
limited information is available with which to reliably judge the
feasibility or costs of these alternatives. Possible alternatives to the
water bank include permanently retiring Project land from irrigation,
expanding Upper Klamath Lake storage, or building a new reservoir separate
from the lake. A large amount of Project land was offered for retirement
by willing sellers in 2001, and a number of storage options have been
evaluated to some extent, but implementation is not imminent for any of
these alternatives. Although one of the objectives of the Conservation
Implementation Program, required under NMFS' 2002 biological opinion, is
the collaborative study of the feasibility of water storage and
groundwater development alternatives, Reclamation and other stakeholders
are still developing the framework for that process. In the interim,
Reclamation and NMFS have an ongoing dialogue regarding water bank
management and will likely reconsult on Klamath Project operations,
including the water bank, in 2006.

A Significant Amount of Irrigated Land Would Need to Be Retired to
Adequately Enhance River Flows, with Unknown Costs and Impacts

As an alternative to the water bank, permanently retiring a large area of
irrigated Project land could provide 100,000 acre-feet of water to enhance
Klamath River flows, but little reliable information is available to
comprehensively assess this option. It is not known with any certainty the
amount of irrigated land that would need to be retired to replace the
water bank, how much irrigated land for retirement could actually be
obtained from sellers, or the price at which it could be obtained.
Furthermore, while this option is viewed positively by some Klamath River
stakeholders, the potential impacts on the agricultural economy from
retiring a large portion of Project lands is cause for concern in the
farming community.

The amount of irrigated land that would need to be retired to reduce
irrigation and enhance river flows by 100,000 acre-feet can be roughly
estimated at about 50,000 acres but is not precisely known. As discussed
earlier in this report, estimates of forgone irrigation water can prove to
be much less than expected, and the lack of reliable water flow
information on the Project makes it difficult to accurately determine the
specific effects of crop idling-which is the short-term equivalent of
permanent land retirement-and other strategies for reducing river
diversions. Reclamation, irrigators, and tribal representatives told us
that they believe that retiring irrigated land would reduce river
diversions, but none are certain as to precisely by how much.
Nevertheless, Reclamation, based on its most recent estimate of the amount
of irrigation water obtained from

crop idling for the water bank, assumes that irrigation is reduced by
about 2 acre-feet of water per acre idled. Using this assumption,
Reclamation estimates that at least 50,000 irrigated acres-about 30
percent of the acreage currently irrigated by water from Upper Klamath
Lake and the Klamath River-would need to be retired to reduce irrigation
by 100,000 acre-feet. However, according to Reclamation officials, because
crop idling provides little water from April to June, such land retirement
by itself will not provide sufficient water to meet spring river flow
requirements under the biological opinion. Furthermore, the actual
reduction in irrigation would depend upon factors such as the extent of
irrigation on the land before it was retired and how it is used after
retirement.

Although there may be a fairly large number of potential willing land
sellers on the Project, the amount of irrigated land actually available
for purchase and permanent retirement is not known. In 2001, the American
Land Conservancy (Conservancy)-a national, nonprofit organization involved
in land conservation efforts-obtained 1-year agreements with 78 different
landowners to purchase over 25,000 acres of irrigated land for the purpose
of land retirement. The Conservancy made agreements with willing
sellers-who, according to Conservancy officials, were generally aging and
fearful of future drops in property values-expecting that the federal
government would purchase the land for retirement. However, according to
the Conservancy, Reclamation was not interested because the land was not
in a single block. Moreover, according to Reclamation officials, the
federal government is not interested in acquiring more land in the Klamath
Basin. Subsequently, the Conservancy's agreements with the sellers lapsed.
Whether a coalition of willing sellers could be put together again is
unknown. An incentive to potential sellers could come from an expected
increase in power rates in 2006. According to a recent Oregon State
University economic study, an increase in power rates could raise
agricultural production costs by an average of $40 per sprinkler-irrigated
acre, potentially making agriculture unprofitable on as much as 90,000
acres of Project land. This scenario could potentially make more land
available for sale and might even result in some voluntary land retirement
due to lack of profitability, thus increasing river flows.

Additionally, the price at which land might be obtained for retirement is
unknown. According to the Conservancy, the appraised value of the
potential willing sellers' land in 2001 was $3,000 per acre. However,
based on 2001 estimates from an Oregon State University and University of
California economic study, the market value for Project irrigated land can
range from $300 per acre for Class V soils-the lowest quality for

agricultural purposes-to $2,600 for Class II soils- some of the better
agricultural soil on the Project. In addition, Project landowners are
concerned that property values may have decreased due to the uncertainty
of water deliveries for irrigation after the 2001 water restriction. Using
the 2001 price estimates from the universities' study, the total cost to
retire 50,000 acres, assuming the land is available from willing sellers,
could range from $15 to $130 million, depending upon the mixture of low
and high valued land offered for sale.

Finally, while tribal representatives and others favor significant
irrigated land retirement as a means to reduce demands on the river, the
extent of impacts on the agricultural economy is cause for concern in the
Project farming community. Tribal representatives and downstream fishing
representatives told us that irrigated land retirement is essential to
restoring the balance between the supply of and demand for water in the
basin. However, according to Klamath irrigators, the Klamath agricultural
economy is fragile and must maintain close to current levels of
agricultural acreage in production to sustain its infrastructure.
Irrigators argue that retiring large amounts of irrigated farmland on the
Project could eliminate or adversely impact key aspects of agricultural
infrastructure, such as fuel, transportation, equipment and fertilizer
suppliers, and affect a whole host of other dynamics of the agricultural
community. However, retiring land with the lowest agricultural value could
help minimize the potential negative effect on the region's agricultural
economy. According to the study by Oregon State University and the
University of California, retiring lands with the least productive soils
and, therefore, lowest agricultural value, would have the smallest
potential negative effect on the region's agricultural economy.

Expanding Upper Klamath Lake or Building a Separate Reservoir Are under
Consideration, but Costs and Impacts Have Not Been Extensively Evaluated

Adding water storage capacity in the Klamath River Basin could provide an
alternative to the water bank for river flow augmentation, and several
options to either expand Upper Klamath Lake or build a separate reservoir
have been considered or pursued to various extents. In general, Klamath
River stakeholders-irrigators, tribes, federal entities, and others-view
either option favorably as a potential solution to help balance competing
water demands. However, the extent and reliability of information
regarding the total cost for each water storage option, the amount of
water potentially provided, the certainty and sustainability of water
storage, and the environmental impacts are largely unknown.

Upper Klamath Lake Expansion	Upper Klamath Lake (including adjoining
Agency Lake) is the primary source of water for the Project and the
Klamath River. To satisfy the water contracts of irrigators, as well as
river flow and lake level requirements, Upper Klamath Lake must be full at
the start of the irrigation season. However, when the lake exceeds its
maximum storage capacity-generally due to heavy runoff before the
irrigation season begins-the lake goes into "spill condition," releasing
water into the Klamath River (and eventually into the Pacific Ocean) to
avoid flooding the surrounding area. Expanding the lake's capacity by
purchasing and flooding adjoining properties with water that would
otherwise be spilled would enable Reclamation to preserve this water for
peak demand periods-late spring to early fall-for both fish and
irrigators. It would also reduce irrigation demand from these lands,
leaving more water in the lake for Project and other uses.

In 1998, Reclamation prepared a report identifying numerous options for
expanding the lake, but only six options were evaluated regarding their
feasibility for water storage development. Collectively the six options
have the potential to provide approximately 100,000 gross acre-feet of
water, however, according to Reclamation officials, evaporation losses
would reduce the net usable water storage to about half that amount. As
shown in figure 7, the six water storage options-listed roughly from north
to south and by proximity to each other-include Agency Lake Ranch, Barnes
Ranch, Wood River Ranch, the Williamson River Delta Preserve, Caledonia
Marsh, and Running Y Marsh.

                 Figure 7: Upper Klamath Lake Expansion Options

                              Source: Reclamation.

o 	Agency Lake Ranch is a 7,125-acre, Reclamation-owned marshland located
on the west side of Agency Lake. Reclamation purchased the land in 1998 to
store spill water during periods of high inflow to the lake. Agency Lake
Ranch currently has the capacity to store about 13,000 gross acre-feet
without flooding neighboring properties. However, it has the potential to
store up to 35,000 gross acre-feet of water if the existing levees
surrounding the land are raised, at an unknown cost.

o 	Barnes Ranch is a privately owned 2,671-acre pasture bordering the west
side of Agency Lake Ranch with the capacity to store 15,000 gross
acre-feet of water if the levees surrounding the property are improved. If
the Barnes Ranch is acquired and Reclamation removed the levees bordering
Agency Lake Ranch and Agency Lake, a combined total of approximately
40,000 gross acre-feet of water could be stored and would potentially fill
to this capacity in most years. In January 2004, Reclamation had Barnes
Ranch appraised for $5.9 million, but the owners and Reclamation have not
yet agreed on a purchase price.

o 	Wood River Ranch is an approximately 3,000-acre site on the north end
of Agency Lake, adjacent to Agency Lake Ranch. The Bureau of Land
Management (BLM) purchased Wood River Ranch in 1994 to restore as a
wetland, among other objectives. Because of its proximity to Agency Lake
Ranch and Barnes Ranch, Reclamation officials would like to convert the
land to store approximately 7,500 gross acre-feet of water. However, local
BLM managers feel that this would not be compatible with the existing
goals and objectives of the Klamath Resource Management Plan, telling us
that converting the land to water storage would destroy wildlife habitat
and reverse a 10-year, multimillion dollar restoration effort accomplished
with many private contributors.

o 	The Williamson River Delta Preserve is a 7,440-acre site, located at
the southern end of Agency Lake, that was converted from wetland to
farmland in the 1930s and 1940s. The Nature Conservancy purchased two
properties-Tulana Farms in 1996 and Goose Bay Farms in 1999- and is
developing a restoration plan for the combined site. With the
encouragement and financial support of Reclamation, the Nature Conservancy
has considered the option of returning the properties to Upper Klamath
Lake. Reclamation estimates that the preserve would add 35,000 gross
acre-feet of water storage capacity, at relatively low cost with the
Nature Conservancy's collaboration.

o 	Caledonia Marsh is a privately owned 794-acre farm on the southern end
of Upper Klamath Lake with the potential capacity to store nearly 5,000
gross acre-feet of water. According to Reclamation, the owner has

expressed interest in selling; however, the surrounding levees would need
to be improved and the Highway 140 road bed raised to protect the
neighboring property, Running Y Marsh. The cost of these improvements has
not been determined.

o 	Running Y Marsh is a privately owned 1,674-acre farm and wetland area
adjacent to Caledonia Marsh with the potential to store about 10,000 gross
acre-feet of water if converted to lake storage. However, because of the
high value crops grown there, the owner is not currently interested in
selling the property to Reclamation.

For all of these options, while it would be relatively easy to determine
the amount of additional water storage provided by measuring changes in
the lake surface area, there are a number of associated uncertainties and
constraints. For example, since these storage areas are essentially
extensions of the lake itself, filling the additional capacity is
dependent upon adequate flows into the lake-if the lake does not fill to
capacity, the storage areas would not be filled to their capacity. In
addition, use of the additional stored water in these areas would be
constrained by the minimum lake level requirements set out by the FWS
biological opinion for Upper Klamath Lake to protect the two species of
sucker. As an extension of the lake, the new storage areas could not be
drained below these minimum levels. Finally, the environmental impacts of
developing water storage areas vary and would need to be addressed by
Reclamation as part of the water storage development process.

Separate Reservoir Development	The development of a separate reservoir
would create a long-term storage area in the Klamath Basin that could far
surpass the capacity of the water bank as a source of flows for the river,
potentially benefiting all Klamath River stakeholders and protected
species. Evaluation of such potential water storage areas has focused on
Long Lake Valley, located southwest of Upper Klamath Lake. Developing Long
Lake Valley into a reservoir would enable water to be stored that would
otherwise be spilled into the Klamath River when Upper Klamath Lake's
water level exceeds the maximum lake elevation. Reclamation, irrigators,
and others generally agree that Long Lake Valley is the most viable option
currently available for new reservoir development.

According to Reclamation, converting Long Lake Valley into a reservoir
could yield up to 250,000 acre-feet of water, with a depth of 250 to 300
feet when full. Thus, Long Lake represents "deep" water storage, which
generally contains colder water-beneficial to fish-than shallow Upper

Klamath Lake can provide. Reclamation indicated that the reservoir's
250,000 acre-foot capacity would be filled by pumping water from Upper
Klamath Lake to Long Lake between March and June, using the piping system
shown in figure 8. However, much like the Upper Klamath Lake expansion
options, the certainty of Long Lake's water supply depends entirely upon
the availability of spill water to fill it and, according to NMFS
officials, the impacts on the river of diverting these flows to a
reservoir need to be studied. Once filled, Long Lake could provide a
sustainable supply of water to supplement river flows. In addition, the
amount of water stored by Long Lake and delivered to enhance river flows
could easily be measured by metering water flow in the pipeline to and
from the lake or, potentially, in a pipeline emptying directly into the
Klamath River.

Figure 8: Long Lake Storage Option

                              Source: Reclamation.

Reclamation completed an initial study of the geology of Long Lake Valley
in March 2004, which determined that Long Lake Valley's floor would

provide a good barrier to prevent water leakage. Geologic investigations
of Long Lake Valley are continuing in 2005. To date, Reclamation has not
conducted a full feasibility study for Long Lake development, and it will
not do so until a funding plan has been established. Reclamation estimates
that a feasibility study would take three years to complete and would cost
approximately $12 million. Subsequently, reservoir construction funds
would need to be obtained. There are no reliable estimates available, but
Reclamation's most recent projection of construction costs is about $350
million, not including real estate acquisition costs. The Long Lake
development project would take at least 10 years to complete, which means
that Long Lake would not address any immediate water demand issues in the
Klamath Basin. Based on Reclamation's initial study, if Reclamation can
address funding, technical, and environmental impact requirements, Long
Lake may offer a promising long-term storage option for the Klamath Basin.

Reclamation's Conservation Implementation Program Is Still Being Developed

Storage options and other potential long-term solutions to water quantity,
quality, and wildlife resource issues are expected to receive greater
attention in coming years under Reclamation's Conservation Implementation
Program. In addition to the water bank, NMFS' 2002 biological opinion
required Reclamation to establish such a program, and Reclamation and
other stakeholders began developing the framework for future collaboration
in 2003. One of the objectives of the program is the development and
implementation of feasibility studies to identify opportunities for
increased water storage and groundwater development alternatives. The
Governors of the states of California and Oregon and heads of the
Departments of the Interior, Agriculture, and Commerce, as well as the
Environmental Protection Agency, signed an agreement in October 2004 to
coordinate their efforts to achieve program objectives, and Reclamation is
currently preparing a third draft program document for stakeholder review.

The Water Bank Could Be Reclamation and NMFS will have the opportunity to
discuss revising some

Modified in 2006	elements of the biological opinion, including the water
bank, when they meet for an expected reconsultation in 2006.
Reconsultation could address the following potential changes to the
biological opinion, affecting Reclamation's responsibility for river
flows, its water bank obligation, and how it operates the water bank:

o 	Adjusting Reclamation's level of responsibility for ensuring Klamath
River flows to reflect information currently being developed regarding the
water quality and quantity requirements of Klamath River fish, as well as
historic natural flows of the Klamath River. Based on a recent USGS study
of irrigated acreage in the Upper Basin, Reclamation- currently held
responsible for ensuring 57 percent of needed flows- may suggest reducing
that number to about 40 percent. Such an adjustment would not directly
alter Reclamation's water bank obligations; however, it would decrease
Reclamation's overall responsibility for ensuring Klamath River base flows
by increasing the responsibilities of other basin stakeholders, such as
the states and other federal agencies. According to NMFS, such a change
would need to be considered within the context of the U.S. District
Court's 2003 criticism of the allocation of responsibility for providing
flows.

o 	Not requiring a water bank in Above Average or Wet water years, thus
eliminating the cost and effort of obtaining and managing the water bank
when natural flows are abundant.

o 	Changing the method for determining water-year types from a five-tier
system to a more incrementally adjustable method that would cause less
dramatic changes in flow requirements, thus addressing one of the concerns
raised by stakeholders. Currently being piloted by Reclamation with FWS
for managing Upper Klamath Lake levels, this method would reduce the
magnitude of changes and the need for significant water bank delivery
recalculations.

Conclusions	Water shortages in the Klamath River Basin have created
serious conflicts and placed Reclamation in the difficult position of
balancing competing demands for water among numerous stakeholders. Over
the last three years, Reclamation has demonstrated commitment and
resourcefulness in this task, particularly under drought conditions, by
implementing and meeting the obligations of the temporary water bank.
However, whether Reclamation can continue meeting its water bank
obligation using current methods is unclear, given the uncertain results
of crop idling and the unknown sustainability of groundwater pumping. This
uncertainty adds urgency to Reclamation and stakeholder efforts to
collaboratively identify and evaluate long-term solutions. In the mean
time, because the water bank acts as the primary mechanism for balancing
competing demands for water, Reclamation must be able to clearly
communicate to stakeholders how the water bank is managed and how water is
accounted for. This

information will make the management and accountability for this public
resource more transparent to all those that rely on and are affected by
the water bank.

Recommendation for Executive Action

We are recommending that Reclamation take steps to improve the information
provided to stakeholders regarding water bank management and accounting by
regularly and systematically providing-through media such as a water bank
Web-link or a monthly or biweekly press release- public information on the
rationale and effects of management decisions related to forecasted water
availability, unexpected spill conditions, or other significant events, as
well as regularly updated information regarding the water bank's status,
including the amount of water bank deliveries to date.

Agency Comments and Our Evaluation

We provided copies of our draft report to the Departments of Agriculture,
Commerce, and the Interior for their review and comment. We received a
written response from the Under Secretary of Commerce for Oceans and
Atmosphere that includes comments from the National Oceanic and
Atmospheric Administration (NOAA) and from Interior's Assistant Secretary,
Policy, Management and Budget that includes comments from Reclamation and
BLM. Overall, NOAA stated that the report accurately reflects the history
of the water bank, and Reclamation expressed appreciation for GAO's
efforts to report on the complex Klamath River Basin situation. We
requested comments from Agriculture, but none were provided.

Reclamation agreed with our recommendation to improve the information
provided to stakeholders regarding water bank management and accounting.
Reclamation agreed to implement steps to enhance water bank communications
through systematic feedback to stakeholders with information regarding the
water bank. Reclamation said that it would add a new page to its Web site
exclusively for the water bank, which will include background information,
new information as it becomes available, links to relevant Web resources
such as USGS' Klamath River gauge at Iron Gate Dam, and graphics showing
the status of water bank flow augmentation. This information will be
updated at least biweekly, with notices posted to direct stakeholders to
updated information. Reclamation plans to complete these changes to its
Web site by June 30, 2005.

NOAA, Reclamation, and BLM provided comments of a factual and technical
nature, which we have incorporated throughout the report as appropriate.
Because of the length of the technical comments provided by Reclamation
and BLM, we did not reproduce them in the report. Interior's transmittal
letter and response to our recommendation are presented in appendix III,
and NOAA's comments are presented in appendix IV.

We are sending copies of this report to the Secretaries of Agriculture,
Commerce and the Interior, appropriate congressional committees, and other
interested Members of Congress. We also will make copies available to
others upon request. In addition, the report will be available at no
charge on the GAO Web site at http://www.gao.gov.

If you or your staffs have any questions about this report, please contact
me at (202) 512-3841 or [email protected]. Key contributors to this report
are listed in appendix V.

Anu K. Mittal Director, Natural Resources and Environment

Appendix I

Scope and Methodology

To determine how the Bureau of Reclamation (Reclamation) operated the
water bank and how much it cost, we analyzed Reclamation's water bank
planning, contracting, and expenditure documentation. We researched and
analyzed laws, regulations, the National Marine Fisheries Service's (NMFS)
biological opinion, and related court cases pertinent to the water bank
and how it operates. For each year of the water bank, we reviewed and
analyzed data on applications and contracts in comparison with the
biological opinion requirements. We reviewed and analyzed expenditures for
contracts and program administration, as well as future budget request
estimates, for total costs incurred to date and expected future costs of
the water bank. Finally, we interviewed staff from Reclamation, NMFS, and
other relevant agencies, as well as stakeholders-including representatives
from tribal, commercial fisheries, and irrigator groups-on water bank
program obligations, operations, and monitoring.

For each year of the water bank program, we reviewed and analyzed data on
water bank contracts to determine whether Reclamation met its water bank
acquisition obligations, and we reviewed and analyzed scheduled base
Klamath River flows, as well as the daily average Klamath River flows,
using both U.S. Geological Survey (USGS) and PacifiCorp-generated data to
calculate the augmented flows to determine whether Reclamation met its
water bank delivery obligations. We interviewed staff from Reclamation and
other relevant agencies, as well as stakeholders- including
representatives from tribal, commercial fisheries, and irrigator groups-on
water bank program obligations, operations, and monitoring.

To describe the water bank's impact on water availability and use in the
Klamath River Basin, we interviewed staff from Reclamation, USGS, the
Oregon Water Resources Department, California Polytechnic State
University, and the Klamath Basin Rangeland Trust. We gathered and
analyzed Reclamation crop reports, a USGS study of irrigation water use,
and a California Polytechnic State University study of the 2003 water bank
to describe the impact of crop idling on river flows. To describe the
impacts of groundwater use, we collected and analyzed Oregon Water
Resources Department information on groundwater pumping, well drilling,
and well deepening in Klamath County, Oregon, and USGS information on well
levels in the Upper Basin. We also collected descriptions of the joint
USGS/Oregon Water Resources Department study of Upper Basin groundwater
and the USGS study of Reclamation's water bank. In addition, we
interviewed and obtained relevant documentation from stakeholders
including irrigators, tribes, and commercial fisheries. We did not review
the

Appendix I Scope and Methodology

water bank's impact on fish species because the short history of the water
bank makes it difficult to obtain reliable information.

To describe alternative approaches to the water bank, we collected
information and interviewed staff from Reclamation and the Bureau of Land
Management, as well as potential land sellers, irrigators, irrigation
experts, economists, and conservationists. We also toured the Klamath
Project area by plane and car to visit and observe potential irrigated
land retirement options and water storage areas. In addition, we collected
and analyzed documentation of potential water storage locations, a study
of options for increasing water storage, as well as a Reclamation study of
a potential new reservoir. Finally, we reviewed the requirements for
coordinated efforts among stakeholders in NMFS' biological opinion and the
status of basinwide planning to increase river flows.

To assess the reliability of the noncomputerized data we received, we
interviewed officials most knowledgeable about the collection and
management of each data set. We assessed the relevant general and
application controls and found them adequate. In addition, we reviewed the
methodology of the economic and water use studies and interviewed the
authors to discuss their scope, data quality, and results. Finally, we
conducted tests of the reliability of computerized data. On the basis of
these interviews, tests, and reviews, we concluded that the data from the
various sources and studies were sufficiently reliable for the purposes of
this report.

We performed our work between May 2004 and February 2005 in accordance
with generally accepted government auditing standards.

Appendix II

Information on Water Bank Applications and Contracts

As shown in table 2, the total numbers of applications from irrigators
seeking participation in the water bank decreased from 2003 to 2004;
Reclamation did not solicit applications in 2002. The total number of
contracts for participation has fluctuated up and down since the inception
of the water bank.

Table 2: Number of Water Bank Applications and Contracts by Type and by
Year

                                                   2002    2003          2004 
               Total water bank applications       a, b    521            449 
                                 Crop idling          a    335     
                    Groundwater substitution          a    186     
                         Groundwater pumping          b          b          b 
                  Total water bank contracts          2    315     
                                 Crop idling          1    223     
                    Groundwater substitution          0     92     
                         Groundwater pumping          1          0 

Source: GAO analysis of Reclamation data.

aDue to the timing of the water bank in 2002, Reclamation negotiated
contracts without a formal application process.

bReclamation negotiated contracts for groundwater pumping outside of the
formal application process.

Reclamation shifted its contracting emphasis from primarily crop idling in
2003 to primarily groundwater contracts in 2004. As such, the number of
groundwater contracts (groundwater pumping plus groundwater substitution)
has grown to represent a larger proportion of all contracts as
Reclamation's water bank obligation increased, as shown in figure 9.

Appendix II Information on Water Bank Applications and Contracts

Figure 9: Proportion and Number of Water Bank Contracts by Type, 2002 to
2004 Number of water bank contracts

350

300

250

200

150

100

50

0

2002 2003 2004

Groundwater pumping Groundwater substitution Crop idling Source: GAO
analysis of Reclamation data.

Note: In 2002, there were two water bank contracts-one for crop idling and
one for groundwater pumping.

The volume of water (acre-feet) offered in water bank applications
increased by almost 50 percent from 2003 to 2004. The volume of water
Reclamation acquired through contracts more than doubled since the water
bank's inception, as shown in table 3.

Appendix II Information on Water Bank Applications and Contracts

Table 3: Volume Represented by Water Bank Applications and Contracts by
Type and Year

                                                      2002      2003     2004 
          Total volume represented by applicationsa   b, c  104,151   154,908 
                                        Crop idling      b  49,274     75,637 
                           Groundwater substitution      b  54,877     79,271 

cc

Groundwater pumping

         Total volume acquired through contractsa  47,072  59,332d   110,877e 
                                      Crop idling  27,072 35,389 d     22,582 
                         Groundwater substitution       0 23,943 d     16,656 
                              Groundwater pumping  20,000         0   71,639e 

Source: GAO analysis of Reclamation data.

aAcre-feet of water.

bDue to the need to establish the water bank expeditiously in 2002,
Reclamation negotiated contracts without a formal application process.

cReclamation negotiated contracts for groundwater pumping outside of the
formal application process.

dReclamation voided two of these 2003 water bank contracts-one crop idling
contract for 64.45 acrefeet and one groundwater substitution contract for
254.29 acre-feet.

eReclamation only purchased 82,257 of the 110,877 acre-feet it acquired
through contracts in 2004, as it did not need to exercise the options on
all of its contingency groundwater pumping contracts in order to meet the
delivery schedule.

As shown in figure 10, from 2002 to 2004, Reclamation has increased the
volume of groundwater as a proportion of the total water bank acquired by
contract.

Appendix II Information on Water Bank Applications and Contracts

Acre-feet (in thousands)

120

105

90

75

60

45

30

15

0 2002 2003 2004

Groundwater pumping

Groundwater substitution

Crop idling

Source: GAO analysis of Reclamation data.

Note: Reclamation was able to meet the water delivery schedule without
exercising the options on all of its contingency groundwater pumping
contracts, purchasing only 82,257 of the 110,877 acre-feet it acquired in
2004. Of this 82,257 acre-feet of water actually purchased, 43,019
acre-feet (52 percent) is attributable to groundwater pumping alone, while
73 percent of actual purchases is attributable to groundwater sources
overall.

As shown in table 4, the total irrigated land acreage offered in water
bank applications and accepted under contracts has increased since the
inception of the water bank.

Appendix II Information on Water Bank Applications and Contracts

Table 4: Acres of Land Offered in Applications and Accepted under
Contracts by Type and Year

                                                      2002     2003      2004 
            Total acreage offered in applications      a,b  47,215     67,508 
                                      Crop idling        a  23,093     33,841 
                         Groundwater substitution        a  24,122     33,667 
                              Groundwater pumping     b, d     b, d      b, d 
           Total acreage accepted under contracts    3,161  25,469     22,371 
                                      Crop idling   3,161c  14,430    15,497c 
                         Groundwater substitution        0  11,039      6,874 
                              Groundwater pumping        b        b         b 

Source: GAO analysis of Reclamation data.

aDue to the need to establish the water bank expeditiously in 2002,
Reclamation negotiated contracts without a formal application process.

bBecause groundwater pumping does not represent water foregone from a
particular area of land, the amount of land represented by the pumping is
not applicable.

cCrop idling included an off-Project contract with the Klamath Basin
Rangeland Trust that accounted for all 3,161 acres of crop idling in 2002
and 11,133 acres in 2004.

dReclamation negotiated contracts for groundwater pumping outside of the
formal application process.

Appendix III

Comments from the Department of the Interior

Note: Because of the length of the technical comments provided by
Interior, we are not including them here. We have incorporated suggested
changes into the report as appropriate.

Appendix III
Comments from the Department of the
Interior

Appendix IV

Comments from the Department of Commerce

Appendix IV
Comments from the Department of
Commerce

Appendix IV
Comments from the Department of
Commerce

Appendix V

                     GAO Contacts and Staff Acknowledgments

GAO Contacts	Anu K. Mittal (202) 512-3841 Edward M. Zadjura (202) 512-9914

Staff 	In addition to those individuals named above, Brad C. Dobbins,
David A. Noguera, and Tama R. Weinberg made key contributions to this
report. Also

Acknowledgments	contributing to the report were John W. Delicath, Philip
G. Farah, Curtis L. Groves, Julian P. Klazkin, Kim M. Raheb, and Monica L.
Wolford.

The Government Accountability Office, the audit, evaluation and

GAO's Mission EURinvestigative arm of Congress, exists to support Congress
in meeting its constitutional responsibilities and to help improve the
performance and accountability of the federal government for the American
people. GAO examines the use of public funds; evaluates federal programs
and policies; and provides analyses, recommendations, and other assistance
to help Congress make informed oversight, policy, and funding decisions.
GAO's commitment to good government is reflected in its core values of
accountability, integrity, and reliability.

The fastest and easiest way to obtain copies of GAO documents at no
costObtaining Copies of is through GAO's Web site (www.gao.gov). Each
weekday, GAO postsGAO Reports and newly released reports, testimony, and
correspondence on its Web site. To

have GAO e-mail you a list of newly posted products every afternoon,
goTestimony to www.gao.gov and select "Subscribe to Updates."

Order by Mail or Phone The first copy of each printed report is free.
Additional copies are $2 each.

A check or money order should be made out to the Superintendent of
Documents. GAO also accepts VISA and Mastercard. Orders for 100 or more
copies mailed to a single address are discounted 25 percent. Orders should
be sent to:

U.S. Government Accountability Office 441 G Street NW, Room LM Washington,
D.C. 20548

To order by Phone: EURVoice: (202) 512-6000 TDD: (202) 512-2537 Fax: (202)
512-6061

To Report Fraud, Contact:

Waste, and Abuse in Web site: www.gao.gov/fraudnet/fraudnet.htm

E-mail: [email protected] Programs Automated answering system: (800)
424-5454 or (202) 512-7470

Gloria Jarmon, Managing Director, [email protected] (202)
512-4400Congressional U.S. Government Accountability Office, 441 G Street
NW, Room 7125 Relations Washington, D.C. 20548

Public Affairs EURPaul Anderson, Managing Director, [email protected]
(202) 512-4800 U.S. Government Accountability Office, 441 G Street NW,
Room 7149 Washington, D.C. 20548

                           PRINTED ON RECYCLED PAPER
*** End of document. ***