Integrated Watershed

Management
Integrated Watershed
Management
Principles and Practice
Isobel W. Heathcote

School of Engineering

University of Guelph
 Contents
Preface ix

Chapter 1 Introduction 1

1.1 Current Issues in Water Management / 5

1.2 Characteristics of Effective Watershed Management / 7

1.3 Why "Integrated" Management? / 9

1.4 A Recommended Planning and Management Approach / 12

Chapter 2 The Watershed Inventory 15

2.1 Physical Features and Landforms / 15

2.2 Climate / 17

2.3 Soils, Infiltration, and Runoff? / 24

2.4 Streamflow / 31

2.5 Groundwater / 34

2.6 Water Quality / 38

2.7 Pland and Animal Communities / 46

2.8 Land Use / 52

2.9 Social and Economic Systems / 56

2.10 Valued Features and Activities / 59

Chapter 3 Problem Definition and Scoping 61

3.1 Identifying Current Water Uses and Use Impairments / 61

3.2 Identifying Current Water Users (Stakeholders) / 74

3.3 Setting Targets for Future Use / 78

3.4 Scoping the Plan / 88

Chapter 4 The Consultation Process 99

4.1 The Need for Public Involvement / 99

4.2 Principles of Consultation / 101

4.3 Identifying Interested Publics / 106

4.4 Public Involvement Techniques and Processes / 112

Chapter 5 Developing Workable Management Options 135

5.1 Identifying the Sources / 135

5.2 Creating a Long List of Management Options / 138

5.3 Types of Options / 139

5.4 Developing Mutually Exclusive Management Alternatives / 143

5.5 Evaluation Constraints and Criteria / 146

Chapter 6 Simple Assessment Methods 156

6.1 The Watershed Inventory / 157

6.2 Scoping / 161

6.3 Developing and Screening Management Alternatives / 163

 6.4 Outputs of a Simple Assessment Process / 182

6.5 An Example of the Application of Simple Assessment Procedures / 186

Chapter 7 Detailed Assessment Methods 195

7.1 The Detailed Watershed Inventory / 196

7.2 Scoping / 213

7.3 Developing and Screening Management Alternatives / 225

Chapter 8 Costing and Financing 244

8.1 Scope and Measures / 244

8.2 Costing Major Public Works / 257

8.3 Benefit-Cost Analysis / 260

8.4 Allocation of Costs Among Multipurpose Projects / 267

8.5 Quantifying Intangibles / 268

8.6 Incorporating Risk and Uncertainty in Economic Analyses / 271

8.7 Capital Financing (Sourcing) / 283

Chapter 9 Legal, Institutional, and Administrative Concerns 287

9.1 The Evolution of Modern Environmental Protection Legislation / 288

9.2 Common-Law Causes of Action / 296

9.3 The Making of Laws / 300

9.4 Existing Legal Frameworks for Water and Environmental Management

/ 307
 9.5 Administrative and Institutional Systems / 323

9.6 Transboundary Water Issues and Free Trade / 325

Chapter 10 Environmental and Social Impact Assessment 328

10.1 The History of Environmental Assessment Policy / 328

10.2 Overview of the EA Process / 330

10.3 Challenges in Environmental Assessment / 334

10.4 Environmental Assessment Methodologies / 339

10.5 Social Impact Assessment / 350

10.6 Strategic Environmental Assessment / 354

10.7 Monitoring and Follow-up / 356

10.8 The Shortcomings of Environmental Assessment / 357

Chapter 11 Choosing the Best Plan 362

11.1 Evaluating the Effectiveness of Options and Strategies / 362

11.2 Implementation Considerations / 370

11.3 Case Study: Watershed Management for Gander Lake, Newfoundland

/ 371

Chapter 12 Implementing the Plan 375

12.1 Principles of Water Resources Administration / 376

12.2 Planning for Successful Implementation / 381

12.3 Why Implementation Sometimes Fails / 391

 12.4 Case Studies / 394

12.5 Lessons Learned / 403

Index 407
 Preface
Conservation is a state of harmony between men and land.

Aldo Leopold, A Sand County Almanac

To waste, to destroy, our natural resources, to skin and exhaust

the land instead of using it so as to increase its usefulness, will
result in undermining in the days of our children the very
prosperity which we ought by right to hand down to them
amplified and developed.

Theodore Roosevelt, Message to Congress, December 3, 1907

Water is possibly our most precious natural resource. The abundance and
quality of water drives all human systems and those of most other
organisms as well. Yet until the 1970s, most water management practices
sought to solve single, localized problems without taking account of the
impacts of those actions on the biophysical, economic, and social elements
of the larger watershed system. Over the past twenty years, a strong global
consensus has begun to develop around the notion that the watershed is, in
fact, the best unit for the management of water resources. Now, countries
in every part of the world try to place water management actions in the
context of natural and human systems: watersheds, and the human
communities in them.

This book builds on the experience of many individuals and agencies

around the world as they have grappled with the challenge of integrated
watershed management. Almost without exception, meeting that challenge
has required collaboration among specialists in widely varying disciplines-
engineering, biological sciences, economics, sociology, law, and ethics-and
among government agencies, private industries, nongovernment
organizations, and the pub lic. This level of collaboration marks a dramatic
change from the technocratic approaches to water management that were
typical of earlier generations. It also marks the beginning of a new, and
integrated, style of water management, embracing the contribution of a
variety of disciplines and viewpoints in the development of strong water
management strategies.

This book aims to present an integrated approach to watershed

management to those who are now, or in future will be, responsible for the
management of precious water resources. It is intended for use by senior
undergraduate and graduate students and by water management
professionals. It has several key objectives:

1. To present a rational framework for the development of water

resources management strategies.

2. To provide an introduction to the technical elements and tools of

water management.

3. To illustrate the interplay of disciplines in the management of water

resources problems.

4. To demonstrate that water resources planning is not a clear-cut

scientific or engineering technique with a single "right" answer, but
rather a continuous social process intended to move a community
closer to its goals for environmental quality.

Many agencies and individuals have contributed to the development of

this book. In particular, I would like to recognize the advice, comments,
and technical support provided by past and present staff of the Ontario
Ministry of Environment and Energy's watershed management group; the
Grand River Conservation Authority (Cambridge, Ontario); the Institute
for Water Resources, U.S. Army Corps of Engineers; the U.S.
Environmental Protection Agency water management and hydrologic
simulation groups; and the Canadian Institute for Environmental Law and
Policy (Toronto). I would also like to acknowledge the special support and
advice I have received from many individuals, in particular Walter Lyon,
Harold "Jack" Day, David Eaton, Mark Killgore, Tony Smith, Bill James,
Trevor Dickinson, Ramesh Rudra, Hugh Whiteley, Doug Joy, Don
Weatherbe, George Zukovs, and Mike Fortin. On a personal level, I would
like to thank my husband, Alan Belk, and my children, Elspeth Evans, Zoe
Belk, and Edward Belk, and my parents, Blake and Barbara Heathcote, for
their continued support and encouragement through the long process of
research and writing. Finally, warmest thanks go to the editorial and
production staff at John Wiley and Sons, particularly Dan Sayre, Neil
Levine, Ira Brodsky, and Millie Torres-Matias, who have cheerfully and
enthusiastically seen this hook through its development and publication.

ISOBEL HEAFHCOTE

University of Guelph

November /997
1
 Introduction
Water is one of our most precious resources. In moist, temperate regions,
water is the fundamental mechanism in chemical flux and cycling. In and
regions, access to water lies at the heart of much conflict. Every living
organism on this planet requires water in some form. Water, therefore,
regulates population growth, influences world health and living conditions,
and determines biodiversity (Newson 1992).

For thousands of years people have tried to control the flow and quality
of water. McDonald and Kay (1988) document water disputes of 4,500
years ago in the Mesopotamian cities of Lagash and Umma. Engineering
works related to military and urban development, drainage works,
irrigation projects, and water diversions can all be documented over
thousands of years. Bonnin (1988) notes that the year 1989 was the
2,000th anniversary of a Roman decree (senatus- consultus) to the effect
that:

Ne quis ayuam oletato clo/o 1nalo ubi publice saliet si quis oletarit
sestertiorum X mi/a multa esto.

That is,

It is forbidden to pollute the public water supply: any deliberate

offender shall be punished by a fine of 10,000 sesterces.

Water provided resources and a means of transportation for

development in North America-and placed limits on that development in
some areas. Even today the presence or absence of water is critical in
determining the uses to which land can be put.

Yet despite this long experience in water use and water management,
humans have failed to manage water well. Through the nineteenth century
and much of the twentieth century, economic development in many
countries was rapid, and often at the expense of sound water management.
Frequently, optimism about the applications of technology-whether dam-
building, wastewater treatment, or irrigation measures-vastly exceeded
concerns or even interest in their environmental shortcomings. Pollution
was viewed as the inevitable consequence of development, the price that
must be paid if economic progress was to be achieved.

Rachel Carson's publication of Silent Spring in 1962 was a turning point

in public views about the environment in general and about water in
particular. The hook drew attention to the rapid deterioration of water
quality and the role of industrial polluters in that decline. Over the next
decade governments around the world strove first to understand, and then
to limit, misuse of water, establishing stronger environmental-protection
legislation, more efficient administrative structures, and better oversight of
public and private water users.

In March 1977, the United Nations (UN) sponsored a conference on

water at Mar del Plata, Argentina. The conference is viewed by many
researchers (e.g., Lee 1992; Koudstaal et al. 1992; Biswas 1992) as a
landmark event in water management. The conference resulted in an
"action plan," including recommendations targeted at meeting the goal of
safe drinking water and sanitation for all human settlements by 1990. The
Mar del Plata conference made specific reference to the problem that water
resources would be increasingly under siege as the need for economic
development came in conflict with the desire for protection of the
environment. The Mar del Plata recommendations for water management
policy can be summarized as follows:

1. Each country should formulate and keep under review a general

statement of policy relating to the use, management, and conservation
of water as a framework for planning and implementation. National
development plans and policies should specify the main objectives of
water-use policy, which in turn should be translated into guidelines,
strategies, and programs.

2. Institutional arrangements adopted by each country should ensure

that the development and management of water resources take place
within the context of national planning, and that there be real
 coordination among all bodies responsible for the investigation,
development, and management of water resources.

3. Each country should examine and keep under review existing

legislative and administrative structures concerning water
management and, where appropriate, should enact comprehensive
legislation for a coordinated approach to water planning. It may be
desirable that provisions concerning water resources management,
conservation, and protection against pollution be combined in a
unitary legal instrument. Legislation should define the rules of public
ownership of water and of large water engineer ing works, as well as
the provisions governing land ownership problems and any litigation
that may result from them. This legislation should be flexible enough
to accommodate future changes in priorities and perspectives.

4. Countries should make necessary efforts to adopt measures for

obtaining effective participation in the planning and decision-making
process involving users and public authorities. This participation can
constructively influence choices between alternative plans and
policies. If necessary, legislation should provide for such participation
as an integral part of the planning, programming, implementation, and
evaluation process.

This "action plan" emphasizes a strong, centralized, and national

commitment to water management. Yet even 20 years later, the problems it
was intended to solve remain significant. Lee (1992) lists the following
difficulties as continuing to exist, despite a growing global consensus,
confirmed at Mar del Plata, about the need for careful, strategic water
management:

1. The dominance of unregulated water uses

2. Inadequate and ineffective water resources management

3. A high degree of inefficiency in many water-related public utilities

4. A failure to retain trained staff of all types

 5. Overcentralization and bureaucratization of decision-making
authority

6. Inappropriate and inadequate water legislation

The significance of Mar del Plata probably lies in the fact that it
recognized, formally and globally, that existing water management policy
was failing to reach its goals. The disappointing progress in the years since
the conference has encouraged many authors to reexamine the Mar del
Plata action plan and the reasons for continued inaction. Many of these
papers were written between 1990 and 1992, possibly in preparation for
the UN Conference on Environment and Development held in Rio de
Janeiro in June 1992.

The Rio meeting provided an important forum for the discussion of

global environmental issues and reinforced the need for continued action,
including the protection of biological diversity. At the meeting, 156 nations
signed a Convention on Biological Diversity, which aims to protect
biodiversity and restore damaged ecosystems. Many countries have begun
to develop formal policy and programmatic responses to the Convention,
and a Global Environment Facility is under development by the UN and
the World Bank to provide loans for projects that have environmental
benefits in preserving biodiversity and maintaining natural habitats, which
reduce the emission of greenhouse gases, stop pollution of international
waters, and protect the ozone layer.

These actions clearly speak to the need for management of systems, not
system components. The level of consensus on this notion is now almost
unprecedented. Lee (1992) argued that overcentralization of water
management, like overcentralization of social and economic systems, has
failed and must be replaced with locally responsive systems at the
watershed level. Koudstaal et al. (1992) reaffirm this idea, noting that there
is no single, clear water management "problem," so it is difficult to focus
public attention on water and to develop a single centralized approach to
water management. Increasingly, authors are calling for an emphasis on
"achieving rational, efficient use of water locally" (Lee 1992), including
water management institutions that are "appropriate to local conditions and
not centrally, inflexibly, imposed."

This perspective, widely endorsed, clearly supports the notion of water

management on a watershed, not state or national, basis. The O.vfrd
English Dic- tionarv defines watershed as "a narrow elevated tract of land
separating two drainage basins," or "the thin line dividing the waters
flowing into two different rivers." A watershed, therefore, is the boundary
of a drainage basin. In the two decades since Mar del Plata, however, the
term watershed has come to mean also the drainage basin itself, or
"catchment" of the river system. This book employs the common usage of
watershed as a drainage basin: an area of land within which all waters flow
to a single river system.

As early as 1980 authors such as Schramm were observing that a

watershed is an integrated system, "holistic in nature ... [with the] whole ...
greater than the sum of its parts" (Schramm 1980). Today there is a clear
global consensus that the watershed is the most appropriate unit for water
management (cf. Newson 1992; Lee 1992; Koudstaal et al. 1992;
Goodman and Edwards 1992; Nickum and Easter 1990; McDonald and
Kay 1988). Figure 1.1 illustrates the interplay of forces affecting integrated
watershed management.
 Figure 1.1 Forces affecting integrated watershed management (after
Koudstaal et al. 1992).

This book responds to that call for water management on a watershed

basis, providing a range of techniques and approaches that can be used to
investigate the biophysical, social, and economic forces affecting water
and its use. This chapter examines some of the fundamental issues
currently facing water resources and proposes a general framework for
integrated water management.

1.1 CURRENT ISSUES IN WATER MANAGEMENT

In the wide debate leading up to the Rio meeting, a number of authors

analyzed the forces affecting water management. There is remarkable
consensus among these authors-who come from countries around the
world-about the current issues confronted by water managers (Viessmann
1990; Goodman and Edwards 1992; Nickum and Easter 1990):

Water Availability, Requirements, and Use

 • Protection of aquatic and wetland habitat

• Management of extreme events (droughts, floods, etc.)

• Excessive extractions from surface and ground waters

• Global climate change

• Safe drinking water supply

• Waterborne commerce

Water Quality

• Coastal and ocean water quality

• Lake and reservoir protection and restoration

• Water quality protection, including effective enforcement of

legislation

• Management of point- and nonpoint-source pollution

• Impacts on land/water/air relationships

• Health risks

Water Management and Institutions

• Coordination and consistency

• Capturing a regional perspective

• The respective roles of federal and state/provincial agencies

• The respective roles of projects and programs

• The economic development philosophy that should guide planning

 • Financing and cost sharing

• Information and education

• Appropriate levels of regulation and deregulation

• Water rights and permits

• Infrastructure

• Population growth

• Water resources planning, including

Consideration of the watershed as an integrated system

Planning as a foundation for, not a reaction to, decision making

Establishment of dynamic planning processes incorporating

periodic review and redirection

Sustainability of projects beyond construction and early operation

A more interactive interface between planners and the public

Identification of sources of conflict as an integral part of planning

Fairness, equity, and reciprocity between affected parties

There is a marked trend from the early 1980s to the present time in
several aspects of water management. Lee (1992) suggests that these
trends may derive from shifts in economic forces, from aggressive
economic growth in the early period to more fragmented, less successful
growth in recent years. Certainly, a global economic recession through the
early to mid- I 990s has made governments more cost conscious and
businesses more willing to evaluate the impacts of their actions in advance
of implementation. The 1980s saw significant growth in public concerns
about the environment, and these have not abated even in the face of'
economic downturns. The trends that can be observed include:

1. A move from end-of-pipe (reactive) pollution control measures

toward pollution prevention.

2. Increasing concern about chronic effects and "invisible" threats-for

instance, to human health-as compared with acute effects and visible
problems.

3. Increasing awareness that point-source controls are generally well in

hand, and that urban and agricultural nonpoint sources of pollution are
now major contributors to surface and groundwater impairment.

4. A shift from local action (e.g., abatement of a single point source) to

global management strategies (e.g., the Rio Convention on
Biodiversity). This trend is also reflected in the growing consensus
about the value of watershed management, as compared with
management by political boundaries such as those of municipalities.

5. Growing mistrust of, or perhaps understanding of the limits of,

technology, and increased reliance on education and extension
activities to change consumer behavior.

6. Increasing consensus that the user or exploiter of resources should

pay for any damage done by that use: the "user pay" or "polluter pay"
principle.

These principles are inherent in an effective integrated water

management strategy, as discussed in Section 1.2.

1.2 CHARACTERISTICS OF EFFECTIVE WATERSHED

MANAGEMENT

Generally speaking, water management can be considered effective when

it:
 1. Allows an adequate supply of water that is sustainable over many
years

2. Maintains water quality at levels that meet government standards and

other societal water quality objectives

3. Allows sustainable economic development over the short and long

term

We may, in fact, have reached a point-perhaps signaled by recent

environmental disasters like Love Canal and the Cuyahoga River and by
water supply crises in many communities-at which it is clear that future
water use must be sustainable or development in some regions will halt.
Sustainability implies closer cooperation between water users than has
typically been experienced in the past. It also implies consideration of the
needs of the community, not just the individual-a difficult proposition for
many water users.

Goodman and Edwards (1992) state that the in this context word plan
can mean any one of the following:

• A single-purpose, single-unit plan to serve a specific need, such as a

demand for water or abatement of a water-related problem

• A multipurpose and multiproject plan

• A regional plan for water resources development, preservation, or

enhancement, staged over a period of time with one or more planning
horizons

• A national plan for water resources development, preservation, or

enhancement

Planning may proceed in many ways. Bishop (1970) notes that clear
goals may be set or a process may simply proceed without goals. One
agency may lead and control the process, perhaps even forbidding
participation by other groups or agencies. Or the process may be clearly a
multiparty and multiper- spective one, with community consensus
established at every step. Some planning processes consist of a rigid
schedule of meetings, formally chaired and run; others employ a flexible
workshop or "kitchen table" approach in which discussion is open and
unstructured. There are advantages and disadvantages to each method, but
the point that must be stressed is that the choice of a planning process is
often highly context-dependent; that is, its success will depend very much
on the characteristics of the planning area, the water management issues in
the area, and the interests and needs of the community of water users. It is
increasingly clear, however, that unilateral planning processes that seek to
exclude the public will fail-if not in the planning stage, then in
implementation. The rapid rise of public interest in, and knowledge about,
environmental issues through the 1980s has created a climate in which
public participation is expected and, indeed, required in almost every
planning situation.

Schramm (1980) offers the following general guidelines for successful

river basin planning:

1. The institutional framework for the project must allow consideration

of a wide range of alternatives to solve observed problems, including
those that may be outside the specific responsibilities of the planning
bodies.

2. The planning agencies must have the expertise needed for

multipleobjective planning and evaluation procedures, especially in
economic, social, and environmental areas.

3. The institutional framework must facilitate adaptation of the plan to

meet changing national, regional, and local priorities.

4. The institutional framework must seek representation of all parties

affected by the specific development plans and management.

5. The institutional framework must reward initiative and innovation

among the members of the technical team and within cooperating
agencies.
 6. The technical team must be sufficiently free from day-to-day
responsibilities so that they can concentrate on long-range planning
and anticipation of future problems.

7. The institutions must have the capacity for learning and improving
over time, including sufficient continuity over time and the ability to
evaluate past programs.

8. There must be sufficient authority within the institutional framework

to enforce conformity of execution with construction and operating
plans.

9. The institutional framework must be capable of guaranteeing an

acceptable minimum level of professional performance by the
technical team.

10. The plan implementation stage must include provisions for the
timely and qualitatively and quantitatively sufficient supply of needed
services by other agencies, as well as provisions to assure continued
functioning-i.e., operation, repair, and maintenance of the facilities
and services provided.

Schramm emphasizes the need for coordination and cooperation at

local, regional, and national levels, noting that:

Planning and plan implementation do not proceed in a rarified vacuum

derived from lofty, immutable principles that are a law unto themselves.
Planning is done for people and people have different and often
competing wants, desires, and hopes; political institutions should be
designed to meet those wants. One of the best ways to condemn
planning efforts to oblivion or failure is to turn the task over to a self-
contained, isolated team of experts who fail to communicate with one
another, the people their plans are to serve, and those with political,
decisionmaking authority. Within this dynamic, competing world of
human wants and values there is no ultimate reality or single-
dimensioned optimum that can be determined by scientific methods
alone.
1.3 WHY "INTEGRATED" MANAGEMENT?

The idea of trans-media environmental management-management using

the "ecosystem" concept-is a relatively new one. In large part, it was born
of experience showing that single-medium or single-source management
was not successful in meeting short- or long-term goals. Until the mid-
1970s, for instance, almost all pollution control effort was directed at
controlling point sources like sewage treatment plants and industrial
discharges. The International Joint Commission's Pollution from Land Use
Activities Reference Group (PLUARG) (PLUARG 1983) examined the
reasons that phosphorus-reduction efforts in the Great Lakes Basin had
stalled. Its research showed that remediation of the lakes would require
integrated management plans for both point and nonpoint sources
throughout the entire Great Lakes Basin. In some areas, point-source
controls would be most cost-effective; in others, the focus would have to
be on nonpointsource controls. Without the overview provided by an
integrated strategy, costly management efforts would continue to fail.

Sometimes water management efforts have been unsuccessful because

they have focused on a single medium (water) rather than on other
environmental components such as sediment, air, or biological tissue.
Mercury poisoning at Minamata, Japan, and the Wabigoon-English River
system, northwestern Ontario, Canada, are excellent examples. In each
case, an industrial facility had discharged large volumes of wastewater
containing inorganic mercury into receiving waters. In each case the
inorganic mercury (which is relatively nontoxic to humans and other
organisms) was converted in the water column and sediments to methyl
mercury, which is highly bioaccumulative and persistent in body tissues.
The methyl mercury was readily taken up by invertebrates in the waters,
which in turn were eaten by larger species such as fish, and these larger
animals were consumed by humans. The humans, at the top of this
particular food chain, received concentrated doses of methyl mercury,
which accumulated in their own body tissues, causing a wide range of
nervous system impacts. Subsequent abatement efforts aimed at
eliminating mercury-using technologies in the Wabigoon facility (a pulp
and paper mill) were successful. Nevertheless, more than 20 years after the
technology change, mercury continues to be released into the water from
river and reservoir sediments and, possibly, from residual deposits within
the mill. Consideration only of effluent quality from the mill might suggest
that the problem was "solved" 20 years ago; in fact, methyl mercury
continues to cycle through the Wabigoon-English River system as a result
of trans-media environmental phenomena.

Very often, water management strategies have failed because they

neglected to incorporate the full range of values and perspectives present
among water users or agencies with an interest in water management.
Wilkes (1975) noted that the provision of adequate water supplies in the
Rhine River watershed is hampered because different agencies are
responsible for water supply and for water quality, and the two are not
always effectively coordinated. Wilkes also observed that watershed
management requires the involvement of regional, state, national, and
international agencies-a measure that was unnecessary at the level of local
water management and pollution control. The transition from local to
watershed management can be difficult, because interested agencies may
not have the necessary authority to take on new management roles, may
encounter varying political influences, or may simply not work very well
together in managing water resources. In multilingual countries or
watersheds, or in less developed countries where external agencies like the
World Bank may be involved in planning initiatives, cooperation across
agencies and disciplines may he more difficult still.

"Integrated" watershed management, although a strategy that is

increasingly advocated in the literature, is therefore still a relatively new
concept. McDonald and Kay (1988) observe that there have been "few real
attempts to provide integrated management information and even fewer
evaluative studies of the policy and management of integration within the
water resources field." More and more agencies are now establishing
administrative frameworks that permit and even encourage management of
water on a watershed basis. Less frequently are water management
activities integrated with other resource management activities affecting or
affected by water. Heathcote (1993) notes that these may include, at
minimum, the intensity and nature of agricultural activities, forestry, and
commercial fisheries. Although integration at the watershed level is
increasingly possible, integration at larger scales is, in the words of
McDonald and Kay (1988). "conspicuously absent," although there are
clear advantages to integrated water management at the international scale
(especially in international river basins) and even at the global scale.

In the mid-1980s, the Canadian federal government established a formal

inquiry on federal water policy, in response to a growing awareness that
Canadian water resources were potentially at risk of overuse and
undeiprotec- tion. The inquiry called for "visionary policies" for the
management of water resources in Canada (Pearse et al. 1985).

The inquiry drew attention to rising water consumption rates in Canada,

conflicting water uses in many areas, and deteriorating water quality,
especially in the heavily populated Canada-U.S. border region. Throughout
their reports, the inquiry panel stressed the need for caution and
prevention, rather than careless use and reaction. Among their
recommendations were several relating to the administrative structures of
water management and the need for what they termed "comprehensive
management." In particular, they called for (Pearse et al. 1985):

• A watershed plan sufficiently comprehensive to take into account all

uses of the water system and other activities that affect water flow and
quality

• Information about the watershed's full hydrological regime

• An analytical system, or model, capable of revealing the full range of

impacts that would be produced by particular uses and developments
in the watershed

• Specified management objectives for the watershed, with criteria for

assessing management alternatives in an objective and unbiased way

• Participation of all relevant regulatory agencies

 • Provisions for public participation in determining objectives and in
management decisions

These recommendations are particularly notable because they come

from a seasoned group of experts in a country that has long considered
itself to have infinite water resources. In the decade following the release
of the inquiry's report, many of the panel's admonitions about excessive
water use and deteriorating water quality have been proven correct, and the
need for integrated watershed management is now seen as urgent. In less-
water-rich countries, including the United States and many European
countries, population density and a limited resource base make integrated
watershed management essential for sustainable water use. As global
consensus about the need for integrated management grows, it may now be
social and economic forces, rather than technical considerations, that
determine the success of an integrated watershed planning effort. In this
regard, Thompson (1982) has remarked:

Without systematic methods for taking account of uncertainty, the

tendency of the regulator is to obscure the fact that scientific
controversy exists. This tendency in turn reinforces the public's
unrealistic expectations that science and technology can supply the
answers. Closely related to the uncertainty problem is the fact that
scientific conclusions, when they are applied to solving human
problems, invariably incorporate a range of value judgments. If these
are acknowledged, the tendency of the pragmatic professional is to say
that, since the issue involves value questions, it might as well be
confronted as a political choice without the need for an expensive and
time-consuming scientific analysis. Alternatively, the regulator may
ignore the value question and carry forward the pretense that his
decision is a purely technical one. In this case it is better not to pursue
the scientific inquiry too far!

1.4 A RECOMMENDED PLANNING AND MANAGEMENT

APPROACH
This book presents a planning approach that is rooted in rational decision
making-that is, systematic development and comparison of management
alternatives. But this general approach is placed within the context of an
informed public and a rapidly changing external environment. It endorses
Bishop's (1970) observation that water management planning is a process
of achieving social change. In that sense, it is a consensus-building
process, not a unidimensional scientific exercise. This theme is stressed
throughout the hook. Chapter 4 offers a range of techniques that have been
proven to he effective in developing social consensus about water
management planning.

The general approach presented in this book can he summarized as

follows:

1. Develop an understanding of watershed components and processes,

and of water uses, water users, and their needs (Chapter 2).

2. Identify and rank problems to be solved, or beneficial uses to be

restored (Chapter 3).

3. Set clear and specific goals (Chapter 3).

4. Develop a set of planning constraints and decision criteria, including

any weights that may he assigned to criteria (Chapter 5).

5. Identify an appropriate method of comparing management

alternatives (Chapters 6 and 7).

6. Develop a list of management options (Chapters 5, 6, and 7).

7. Eliminate options that arc not feasible because of time, cost, space, or
other constraints (Chapters 5, 6, and 7).

8. Test the effectiveness of remaining feasible options using the method

identified in (5) and the decision criteria and weights identified in (4)
(Chapters 6 and 7).
 9. Determine the economic impacts and legal implications of' the
various feasible management options (Chapters 8 and 9) and their
environmental impacts (Chapter 10).

10. Develop several good management strategies, each encompassing

one or more options, for the consideration of decision makers (Chapter
I I).

11. Develop clear and comprehensive implementation procedures for

the plan that is preferred by decision makers (Chapter 12).

Planners (and authors) find it helpful to divide the planning process into
these discrete steps. In reality, however, the planning process is dynamic
and continuous. Several tasks or steps may be under way simultaneously.
Planning direction may change radically if new information comes to light,
if political forces change dramatically, or if community consensus is
redirected for other reasons. Above all, integrated watershed planning and
management must be responsive and adaptive to changing conditions. This
means that a good watershed plan is not a single product, such as a
document that sits on an agency bookshelf. Instead, it is a framework for
continued dialogue about water and the watershed. Ideas endorsed by the
planning team must be revisited and reviewed periodically to determine
whether they are still acceptable or could be improved. New technologies
and management thinking must be incorporated into the evolving plan.
Most of all, the watershed management plan must reflect the current
societal consensus about the value of water as a resource, about
responsibilities and social attitudes, and about the community's vision of
an ideal watershed state. Integrated watershed management is, therefore, a
journey, not a destination.

REFERENCES

Bishop, Bruce. 1970. Public Participation in Planning: A Multi-Media

Course. IWR Report 70-7. Fort Belvoir, Va.: U.S. Army Engineers
Institute for Water Resources.
Biswas, Asit K. 1992. Sustainable water development: A global
perspective. Water International 17 (1992): 68-80.

Bonnin, Jacques. 1988. Were urban water systems improved over the last
20 centuries? Water International 13 (1988): I0-16.

Goodman, A. S., and K. A. Edwards. 1992. Integrated water resources

planning. Natural Resources Forum 16(1): 65-70.

Heathcote, Isobel W. 1993. An integrated water management strategy for

Ontario: Conservation and protection for sustainable use. In
Environmental Pollution: Science, Policy and Engineering , edited by B.
Nath, L. Candela, L. Hens, and J. P. Robinson. London: European Centre
for Pollution Research, University of London.

Koudstaal, Rob, Frank R. Rijsberman, and Hubert Savenije. 1992. Water

and sustainable development. Natural Resources Forum 16(4): 277-290.

Lee, Terence. 1992. Water management since the adoption of the Mar del
Plata Action Plan: Lessons for the 1990s. Natural Resources Forum
16(3): 202-211.

McDonald, Adrian T., and David Kay. 1988. Water Resources Issues and
Strategies. New York: Longman Scientific and Technical and John Wiley
& Sons.

Newson, Malcolm. 1992. Water and sustainable development: The "turn-

around decade"? J. Envir. Planning and Management 25(2): 175-183.

Nickum, James E., and K. William Easter. 1990. Institutional arrangements

for managing water conflicts in lake basins. Natural Resources Forum
14(3): 2 10-220.

Pearse, P. H., F. Bertrand, and J. W. MacLaren. 1985. Currents of Change:

Final Report of the Inquiry on Federal Water Policy. Ottawa: Queen's
Printer.
Pollution from Land Use Activities Reference Group (PLUARG). 1983.
Nonpoint Source Pollution Abatement in the Great Lakes Basin: An
overview of Post-PLUARG Developments. Report to the Great Lakes
Water Quality Board of the International Joint Commission. Windsor,
Ont.: International Joint Commission.

Schramm, Gunter. 1980. Integrated river basin planning in a holistic

universe. Natural Resources Journal 20: 787-805.

Thompson, Andrew R. 1982. Water law-The limits of the management

concept. In Environmental Law in the 1980s: A New Beginning, edited
by P. Z. R. Finkle and A. R. Lucas. Proceedings of a colloquium
convened by the Canadian Institution of' Resources Law, Faculty of Law,
University of Calgary, Calgary, Alberta, November 27-29, 1981.

Viessman, Warren, Jr. 1990. Water management issues for the nineties.
Water Resources Bulletin 26(6): 883-891.

Wilkes, Daniel. 1975. Water supply regulation. In Regional Management

of the Rhine, edited by Chatham House Study Group. London: Chatham
House.
2
 The Watershed Inventory
The lifeblood of a watershed ecosystem is water. Water movement in the
system is affected by many physical, chemical, and biological features and
processes. An understanding of these features and processes is an essential
first step in assessing the condition of a watershed ecosystem and the
impacts of management actions on it. A watershed inventory also provides
the building blocks with which to create predictive models of the system
for application in evaluating the impacts of proposed management actions.

The following sections provide an overview of the elements contained

in a typical watershed inventory.

2.1 PHYSICAL FEATURES AND LANDFORMS

2.1.1 Bedrock Geology

Rock underlies all watershed ecosystems. The nature of this rock

determines the character of overlying soils and influences the ultimate
movement of water draining through those soils.

Rocks are formed in one of three ways: by an igneous, sedimentary, or

metamorphic process. Each type has its own special physical qualities and
chemistry.

Igneous rocks are formed when molten rock (magma) is pushed up

through the cooler regions of the earth's crust from deeper zones. As the
magma cools, the minerals within crystallize and bond in ways that create
characteristic patterns. Igneous rocks typically show a texture of finely to
coarsely interlocking grains, sometimes containing larger crystals (usually
of quartz or feldspar) in a finer-grained matrix. The chemical properties of
these rocks become important through the process of rock weathering, in
which the quality of water flowing over the rock surface is affected by the
chemistry of the rock itself. Rock weathering can occur either through
physical processes, such as fracturing and freezethaw forces, or through
chemical processes such as dissolution and absorption.

Sedimentary rocks are composed either of fragments of mineral and

rock eroded from preexisting rocks and then deposited by natural forces, or
of' materials precipitated from aqueous solutions. Thus, sandstones come
from sandy sediments, limestones and dolomites come from finer
sediments with a high proportion of shell and carbonate, and shales come
from the finest clay sediments. Salt and gypsum are sedimentary rocks
formed from the precipitation of materials in solution. As with igneous
rocks, the chemistry of sedimentary rock becomes a significant influence
on watershed soil and water chemistry. This is particularly true for rocks
containing high concentrations of calcium carbonate, an important
buffering agent especially in freshwater systems. Sedimentary rocks
usually exhibit a layered or stratified structure, reflecting the depositional
processes involved in their creation.

Metamorphic rocks are created when igneous or sedimentary rocks are

subjected to further heat and pressure. thus modifying their crystal
structure (but not their chemistry) and their physical properties. Slate, for
example, is the metamorphosed form of shale; marble is the
metamorphosed form of limestone. Metamorphic rocks show a more
pronounced alignment of crystals than is seen in igneous or sedimentary
rocks, a result of the action of heat and directional pressure on preexisting
crystal structures.

In the context of watershed management, the geological processes that

create these minerals are probably of less interest than their physical
structure and chemistry. Important aspects of physical structure include
fractures, folds, and faults, all of which (among other processes) are
important in groundwater flow and, thus, in the distribution of water within
the watershed ecosystem.

The United States and Canada, and many other countries worldwide,
maintain extensive geological databases. Often, this information is
available in map or report form specific to a given area. The agencies
involved may also have electronic bulletin board facilities and/or World
Wide Web sites summarizing investigative activities and available
publications. The following are the primary Web sites currently operated
by these agencies:

2.1.2 Surficial Geology and Landforms

Overlying bedrock in most regions is a layer of softer sediments, or soils.

These materials develop gradually from the action of plants and physical
forces on bedrock and contain a mixture of eroded rock and mineral
fragments and organic debris; the precise mixture determines the soil
chemistry, physical properties such as drainage characteristics, and
fertility.

These soft sediments can be shaped and realigned in many ways by the
action of ice, water, and wind. In northern regions the force of advancing
and retreating glaciers tens of thousands of years ago had an important and
lasting impact on landforms. A variety of glacial features can be seen in
these areas, including moraines, created by the pushing effect of massive
ice sheets; till plains, mixtures of soil and rock left behind by the plowing
glacier; and sorted beds of sand and gravel formed by meltwater drainage.

Glacial action was also very important in carving the landscape in these
northern regions, including, for instance, the shaping of lake and river
basins. It is often possible to see vestiges of ancient beaches and shorelines
near the margins of modern water bodies. These structures can provide
clues about the water levels and flow patterns that prevailed in glacial
times and thus give insight into modern drainage phenomena rooted in
glacial structures.

In warmer regions the action of wind and water can be as important in

sculpting landforms as glacial action in the north. Arroyos, for example,
reveal where water flows seasonally or less often, or perhaps where it
flowed in the distant past. In the prairie, wind forces can be important in
sorting and redistributing soils, as was the case during the dust storms of
the 1930s.

State and national geological surveys should also be able to provide

information on watershed topography and the historical forces that led to
the development of present landforms. Detailed topographic maps,
available for most regions, are an excellent source of information about the
degree, shape, and length of slopes (important in controlling runoff and
erosion). This information, in conjunction with a knowledge of bedrock
geology, lays the foundation for a more detailed inventory of the natural
and built features of a watershed. It can give a general indication of the
historical and modern forces affecting water flow and, thus, the
distribution of surface waters, indigenous and domesticated flora and
fauna, and the human systems based on these factors.

2.2 CLIMATE

Climate, including temperature, wind force and direction, and

precipitation, is an extremely important influence on water resources and
biological processes in a watershed.

2.2.1 Temperature, Evaporation, and Transpiration

The national weather services in most countries collect information about

ambient solar radiation, maximum and minimum air temperature, and
related measurements across a wide network of monitoring stations. These
data are usually available to the researcher through public and university
libraries or directly from the agencies involved. Increasingly, such
information is available in elec tronic format, vastly simplifying its use in
predictive computer models or other analytical methods. As with
geological data, the World Wide Web is an excellent starting point for
information about the availability of meteorological data. Current Web
sites for key U.S. and Canadian agencies involved in meteorological data
collection are:
 National Weather Service (NWS) (links to NWS offices
throughout the United States) vortex.atms.unca.edu/-
staylor/nws

Atmospheric Environment Branch, Environment Canada

www.ns.doe.ca/ aeb/aebhome

As with precipitation measurements and other climatological data, it is

important to remember that climate can vary dramatically with the local
environment, so a general rule of thumb is that climatological data must he
collected as close as possible to the watershed area(s) of interest. Where
available data are inadequate or too distant to be relevant, it may be
appropriate to collect additional measurements using recording gauges at
key locations in the watershed.

For water managers, air temperature is of interest because of its

influence on evaporation and transpiration, and on the growth of plants in
the watershed. Evaporation is the transfer of water from the liquid state to
the vapor form. It occurs over the surface of open water bodies through the
escape of water molecules (which are always in motion) from the liquid
into the air. Evaporation is said to occur when the transfer from liquid to
air exceeds the transfer from air to liquid; when the reverse is true,
precipitation is said to occur. Dalton's law gives the formula for
evaporation from free-water surfaces:

E :C(e.,-ed) (2.1)

C = a constant

e, = the saturated vapor pressure at the temperature of the water

surface in mm Hg

e,j - actual vapor pressure of the air (e, times relative humidity) in
mm Hg
 The value of the constant has been estimated by Rohwer (1931) as:

C- (0.44+0.073W)(1.465-0.00073p) (2.2)

p = atmospheric pressure in mm Hg at OoC

The value of E calculated using this constant has units of mm/day. For
reser voirs, the calculated value of E should be multiplied by 0.77 (Schwab
et al. 1993). Meyer (1942) calculated the value of C for pans and shallow
ponds using the relationship:

C= 15+0.93W (2.3)

(Note that the value of E calculated in this fashion has units of mm/month,
that vapor pressure should also be measured at 7.6 m height, and that air
temperature is calculated as the average of the daily minimum and daily
maximum temperatures.) Meyer's calculation of C for small lakes and
reservoirs is similar:

C- 11+0.68W (2.4)

Evaporation also occurs from land surfaces. Where soils are saturated
and the water table is high, evaporation rates may approach those from
free-water surfaces. As saturation decreases, evaporation rates will
diminish rapidly. Artificial cover on a soil surface (for instance, by
mulching) or a natural cover (for instance, by densely growing vegetation)
further restricts evaporation from the land.

Transpiration is the transfer of water molecules into the air through the
tissues of living plants. So in an area such as a rice paddy, evaporation
(from the water surface) and transpiration (from the growing plants) may
occur simultaneously. Transpiration rates are usually measured in the field-
a complex process that can involve "bagging" plants or trees to capture all
moisture. Experimental data of this type is sometimes subject to criticism
in that the process of enveloping a growing plant with plastic or other film
may alter tissue temperatures and thus affect the rate of natural processes.

Transpiration is sometimes expressed as a ratio of the weight of water

transpired to the weight of dry plant tissue. Schwab et al. (1993) quote
typical transpiration ratios for common plants, as shown in Table 2. I.
Figure 2.1 Hypothetical patterns of precipitation and evapotranspiration over
a watershed.

Evaporation and transpiration rates are often combined into a single

"evapotranspiration" rate, which can be determined from field
measurements using either controlled evaporation techniques and/or mass
balance approaches (calculation of evapotranspiration losses by difference,
given known inputs and outputs from a system).

Across most of eastern Canada and the United States, approximately to

the Mississippi River valley, average annual precipitation exceeds
evaporation. This regime also applies in the Pacific Northwest. For the rest
of the country west of the Mississippi, evaporation rates generally exceed
precipitation rates, with the result that there is a net annual loss of water
from the regions of this area. Figure 2.1 illustrates these concepts.
2.2.2 Precipitation

Ultimately, all the water in a watershed cones from precipitation-rain,

snow, or dew. Water vapor is always present in the atmosphere, even over
the driest regions. The deposition of this moisture as precipitation depends
on complex physical phenomena, both at the local and at the regional-or
larger-scale. The most casual observer knows that rainfall is seldom
uniformly distributed over a given area: it may be raining in one part of a
city, for example, but dry in another. It is virtually impossible to model
atmospheric processes, and thus to predict rainfall intensity, with certainty.
A sound understanding of precipitation phenomena must therefore be
based on observation, preferably over a period of time long enough to
capture the full range of natural variability.

The following discussion focuses on the collection of rainfall data as the

most general case; a brief discussion of snowfall/snowmelt data is included
at the end of this section.

Rainfall measurement aims to collect information about the duration of

a rainfall event, as well as its intensity, or rate. There are two challenges in
collecting this information. The first is to ensure that the rain gauge
collects a representative sample (in other words, accurately reflects the
quantity of rain that would have fallen on open ground, as if the gauge had
not been present). The second is to ensure that the number and placement
of rain gauges is sufficient to give an accurate picture of rainfall conditions
over the entire area of interest-for instance, a watershed.

The first challenge, that of collecting a representative rainfall sample,

relates primarily to the design of the rain gauge and its placement in the
study area. Careless placement of gauges (for instance, under trees or close
to sheltering buildings), can reduce the amount of rain reaching the gauges
and give an unrepresentative sample. Less obvious errors include problems
arising from water splashing into, or out of, the gauge container, errors
related to improperly leveled gauges, and instances in which the gauge
itself absorbs water and measured rain is less than actual rain entering the
gauge. A correctly placed gauge is located close to the ground but
sufficiently elevated to exclude water splashed in from the surrounding
land surface. Usually, this position is somewhere between 30 cm and I m
from the ground; the decision may relate to routine practice in the agency
responsible for rain gauging, to topographic features, to expected snow
cover, or to some combination of these factors.

The simplest rain gauge is simply a graduated cylinder. A funnel placed

over the opening can allow collection of rain over a larger area (thus
improving collection efficiency) but requires that the gauge be calibrated
to reflect the larger area of capture. Simple gauges of this type are
generally limited to sites with good access, where regular visits for reading
and maintenance are possible. In more remote areas, gauges of larger
capacity ("storage" or "totalizer" gauges) are often used instead.

Most modern rain gauges incorporate capability for continuous

recording of data, whether by mechanical (e.g., clockwork) or digital
means. Three types of gauges are commonly used: tipping bucket gauges,
weighing gauges, and gauges that record water level by means of a float
device. Tipping bucket gauges use a small two-compartment tipping
container that, when full, tips to one side, closing an electric circuit and
recording a specified (calibrated) quantity of rain. As the container
empties, it brings the other side under the filling funnel and the process
begins again. Tipping bucket gauges are durable and reliable and are
widely used. In heavy rains they may, however, underrepresent actual rain
fall through spillage during filling/tipping. They also generate rainfall
records that are stepped rather than continuous (smooth) and are usually
considered unsuitable for measuring snowfall.

Weighing devices, as the name suggests, measure precipitation by

weight. Rain or snow falling into the container is continuously weighed on
a balance, and the result is transmitted to a recording device. This type of
gauge is better than tipping bucket devices in areas with frequent snow, but
its capacity is, of course, limited, so it cannot be left unattended for long
periods in areas with heavy precipitation. Weighing devices are also
sensitive to wind effects, which can cause the balance to tremble and
create an error in the recorded data.
 Float-type gauges employ a float that rises with accumulating rainwater.
The position of the float is then continuously recorded on a strip chart or
other recording device. Without heating devices (which have sometimes
been employed in cold climates), these gauges are unsuitable for use in
freezing weather.

Newer gauges employ digital recording techniques in place of older

strip chart recording and are capable of transmitting data in electronic form
over dedicated telephone lines or by similar means.

The second challenge in monitoring precipitation is that of collecting a

reasonable sample across a diverse area. Weather radar technology is now
widely used for the prediction of rainfall events and for assessment of the
spatial distribution and intensity of rainfall over the area of interest.
(Increasingly, this technology is also used in flood prediction and other
hydrological forecasting requiring distributed rainfall data as input; see, for
example, Kouwen and Soulis (1994).

Radar data may be available more or less continuously over the period
preceding, during, and following a storm, so the analyst can piece together
a time series of precipitation phenomena. Although methods are available
to convert radar data into quantitative estimates of precipitation volume,
these methods can be difficult to apply accurately. Radar depends on echo
intensity from rain droplets, and because storms vary tremendously in the
size and distribution of these droplets, it has been difficult to develop a
consistent relationship between echo response and actual rainfall rate.

The simplest method of estimating precipitation over an area (as

opposed to a single point) is probably to calculate the arithmetic average of
several point measurements throughout the area of interest. Although
useful for small, topographically homogeneous watersheds, this method is
flawed in that precipitation seldom falls uniformly over a given area (as a
result of topography and other factors) and gauges are seldom uniformly
distributed throughout the area of interest. A simple arithmetic average
may therefore yield an unrepresentative estimate of "average"
precipitation. More complex methods, such as the isohyetal method and
the Thiessen method, attempt to interpolate precipitation between gauges,
the former method using weighted average precipitation between isohyets
(lines of equal rainfall depth) and the latter using polygonal subareas, the
center of each being it rainfall gauge. The Thiessen method is rel atively
straightforward to employ but is not suitable for areas of high elevation.
The isohyetal method is considered the most accurate but requires
considerable skill and careful application in areas of varied topography.
Figure 2.2 illustrates the disposition of precipitation falling over a drainage
basin.

In colder regions snow may comprise 50% or more of total

precipitation. It is therefore an important element in the watershed water
budget, and essential to include in a watershed inventory.
Figure 2.2 Disposition of precipitation falling over a drainage basin (adapted
from Viessman and Lewis 1996).

Snow sampling presents special challenges in comparison with

measurement of rainfall or streamflow. The most important difference, of
course, arises from the need to accommodate freezing conditions in the
collection, recording, and transmission of data.

A simple and obvious means of measuring snow is to insert a ruler in

the snow pack to the full depth of the snow. Yet measurement of snow
depth may be of little use in estimating water content, which differs
depending on the age and consistency of the snow. To obtain information
about water equivalency, snow packs may be cored and the cored material
either weighed or melted and measured in a graduated cylinder.
Radioisotope methods involve burial of a gamma emitter (e.g., cobalt 60)
under the ice pack. The thickness of the accumulating snow pack is then
proportional to the decrease in gamma particle emissions. Weighing snow
gauges have also been developed, but they are likely to be less accurate
because of the tendency of adjacent snow to support the weighing plate
and thus cause an underestimate of snow accumulation.

Snow cover (i.e., proportion of area covered by snow) is also of interest,

particularly during snowmelt periods. Even aerial photographs of snow
cover can give an indication of watershed areas that are probably still
frozen and storing water as snow, versus areas that are melting and thus in
a position to release their stored moisture.

2.2.3 Wind

Wind conditions affect many land and water phenomena in a watershed,

particularly evapotranspiration and wind erosion. Wind, like precipitation
and ambient temperature, is commonly measured by regulatory agencies
through a network of meteorologic stations, in part because of its primary
importance in aviation. Also like those parameters, wind conditions are
typically highly variable, even over small areas, as a result of natural
topography and the influence of structures. Wind speed and direction are
usually measured with an anemometer, which may be either mechanical
(e.g., rotating cup type) or electronic (e.g., "hot wire" or "hot film"
anemometers, which measure wind speed as a function of heat loss over a
surface). Wind records are widely available in printed and electronic
formats from regulatory agencies such as the United States National
Weather Service and Environment Canada's Atmospheric Environment
Branch.

2.3 SOILS, INFILTRATION, AND RUNOFF

2.3.1 Soil Classification and Permeability

In effect, soil is a porous medium that blankets bedrock. Its ability to

transport water depends on the size and condition of channels through the
porous medium. These factors in turn depend on the size of soil particles,
the degree to which individual soil particles are aggregated into larger
masses, and the arrangement of individual particles and aggregates. Where
soil channels are large and relatively permanent, the soil provides an
effective transport mechanism for water-in other words, it is highly
permeable.

Soil permeability can change over time, particularly through soil

compaction or the development of a thin, impermeable layer at the surface,
such as may be caused by repeated action of raindrops on soil particles.
The energy in a raindrop dislodges soil particles and moves them, with the
flowing water, over the soil surface. Smaller particles eventually settle into
the gaps between larger fragments, so over time the porous surface can
become clogged and compact.

Soils are classified into various groups depending on their permeability,

infiltration rate, and surface runoff potential. Permeability refers to the rate
of water movement through a saturated soil (when gravitational forces
dominate) and is defined by Darcy's law:

q = KhA/L (2.5)

K = hydraulic conductivity of the flow medium (distance/time)

h = head or potential causing flow (distance)

 A = cross-sectional area of flow (distance2)

L = length of the flow path (distance)

Soil permeability is affected by a range of factors, including particle

size distribution (texture), degree of compaction, and chemical
composition, especially salt concentration. Infiltration rate is a broader
concept, referring to the ability of a soil to "draw up" water through a
combination of gravitational forces and capillary action. Infiltration rate is
expressed in units of volume per unit time per unit area or, roughly, depth
per unit time. Infiltration is the mechanism by which plants receive water
and dissolved nutrients and by which groundwater supplies are maintained
and replenished. It is affected by the physical structure of the soil and any
cover on the soil surface (such as mulch or crops) and by soil moisture
content, temperature, and rainfall intensity. Figure 2.3 illustrates the
relationship between soil type, and soil porosity and moisture content.

The United States Soil Conservation Service uses four categories to

define soil infiltration characteristics (see Table 2.2). Infiltration refers to
the entry of water into the soil surface (Schwab et al. 1993) as opposed to
other drainage phenomena such as overland flow. Soil classification maps
are available from the SCS for most regions of the U.S.

2.3.2 Infiltration

Infiltration rate (that is, water transport through an unsaturated soil) can be
predicted by Horton's (essentially empirical) formula (Horton 1939):
Figure 2.3 Typical drainage characteristics of soils (adapted from Novotny
and Olem 1994).

l = .lI + (.fo -.I`,)e k ` (2.6)

f,. = the constant infiltration capacity as t approaches infinity

f~ = infiltration capacity at time of onset of infiltration

k = a positive constant for a given soil and initial condition

t = time
 Other methods for predicting infiltration, such as that proposed by
Green and Ampt (1911), are essentially variations on Darcy's equation
using assumptions about uniformity of vertical flow, water content, and
soil hydraulic conductivity. Holtan (1961) proposed a method that reflects
the gradual exhaustion of soil moisture storage as a soil becomes saturated.
Simplified versions of more complex models, such as that proposed by
Philip (1983), appear in several major hydrological models. Philip's model
calculates infiltration rate from soil sorptivity and conductivity of the
wetting front and takes the following form:

[(t) = I St 1 /2 + K (2.7)

S = sorptivity of the soil (a function of soil "suction")

K = conductivity of the wetting front

t = time

In the field, infiltration is frequently measured using a device called an

infiltrometer. An infiltrometer is essentially an enclosed plot of land to
which moisture is applied using sprinklers. Infiltration is simply computed
as the difference between the volume of water applied and the volume
collected as surface runoff. Infiltration rates can vary significantly even
over small areas of soil (tens of square meters or less), so this method can
be expected to yield only an approximate estimate. Infiltrometer results
should therefore be evaluated relative to other infiltrometer experiments,
and should not be considered to be accurate representations of natural
phenomena.

Infiltration data can also be obtained by comparing direct measurements

of rainfall and measured runoff from gauged streams. These "infiltration
indices," as they are called, give a general picture of infiltration over a
watershed (or subwatershed) area rather than the infiltration characteristics
of a particular plot of land. See Figure 2.4.

2.3.3 Runoff

Precipitation falling over land that does not evaporate or infiltrate into the
soil must, by default, run off the land surface. Runoff estimation is
important, for instance, in the design of urban stormwater structures such
as storm sewers and treatment wetlands. Runoff rates can be calculated in
several ways. An older method, now often regarded as unacceptably
simplistic by many authors, is termed the rational method or the rational
formula. It is expressed as:

q = 0.002CiA (2.8)
Figure 2.4 Typical infiltration-runoff curves developed from infiltrometer
data (adapted from Schwab et al. 1993).

C == runoff coefficient (dimensionless ratio of peak runoff rate to

rainfall intensity)

i = rainfall intensity

A = watershed area in hectares (ha)

This formula requires that a design return period (a period of time over
which performance will he judged) be selected. For example, large
structures may he designed to contain major runoff events that may occur
only once every 20 to 50 years, or more, whereas constructed wetlands or
nonstructural measures may be designed to contain runoff that occurs once
every 10 years, or even less. Clearly, the more extreme the event (the
"once-in-200-year storm," for example), the larger the structure must be to
contain it.
 The formula also requires selection of it rainfall intensity (for the design
return period), which is equal to the "time of concentration" of the
watershed-the time required for water to flow from the most remote parts
of the watershed to the outlet. Time of concentration assumes saturated soil
conditions and all minor depressions filled with water. When the duration
of a storm exceeds the time of concentration, the analyst can assume that
all parts of the watershed (not just those closest to the outlet) are
contributing to observed flow patterns. Time of concentration can also be
estimated in various ways, and indeed there is a long-standing debate in
the literature about the best way to calculate this parameter. One simple, if
crude, approach is to divide the length of flow by the estimated velocity of
flow to obtain travel time. The sung of channel travel time(s) and those for
overland flow equals the time of concentration.

Figure 2.5 Measured hydrograph for the Grand River at Galt, Ontario,
January 3-10, 19?? (adapted from Kouwen and Soulis 1994).

The rational method makes several important simplifying assumptions:

1. That the frequencies of rainfall and runoff are similar;

 2. That rainfall occurs at uniform intensity over the time of
concentration; and

3. That rainfall intensity is uniform over the watershed area.

These assumptions are usually acceptable in smaller, homogeneous

basins but quickly break down in larger systems. The rational method
should therefore be used cautiously or replaced with another runoff
estimation approach such as the Soil Conservation Service method (United
States Soil Conservation Service 1990).

A hydrograph plots runoff rate versus time and is a useful representation

of watershed response to a given storm (see, for example, Figure 2.5). The
SCS method of estimating runoff employs a simplified triangular
hydrograph approach and was developed from hundreds of measurements
compiled over a variety of real systems and storms. In the SCS method, the
peak runoff rate (the top of the hydrograph) is calculated as:

q = q,,AQ (2.9)

q„ = unit peak flow rate (m3/sec per ha/mm of rainfall)

A -- watershed area in ha

Q -- runoff depth in mm

Q is calculated separately as:

(1-0.2S )' I+0.8S (2.10)

S = maximum potential difference between rainfall and runoff in

mm, starting when the storm begins (usually obtained
empirically)
 In the SCS method, the time to peak flow (top of the hydrograph) is:

TI, = D 2 + Tr. = D 2 + 0.6T,. (2.1 1)

TI, - time of lag

T( = time of concentration

I) -- duration of excess rainfall (units of time)

T, is the longest travel time in the basin and can he calculated as:

T Lo.s((I000/N) - 9]0.7 ~ 44075 0.5 (2.12)

L = longest flow length

N runoff curve number

S~ _ average watershed slope in m/m

Like the rational method, the SCS method is best used on small,
homogeneous watersheds without steep slopes or major areas of ponding.

Chow (1988) discusses a variety of other methods for the prediction of

runoff. Typically, these relate peak runoff to watershed area and surface
characteristics using an exponential relationship of the general form:

q = KA' (2.13)

K coefficient intended to represent the physical characteristics of

the watershed
 A -= watershed area

x = a constant for a given location

2.4 STREAMFLOW

Streamflow is an important component of water resources in a watershed

ecosystem. Certainly, it is among the most visible and perhaps also the
most easily measured. For these reasons, most jurisdictions maintain
networks for the sampling, often continuous, of streamflow. This section
describes some basic approaches to these measurements, along with their
advantages and disadvantages.

Streamflow is the rate of flow of water and is expressed in units of

volume per unit time (e.g., m3/sec). To determine streamflow, the analyst
must obtain two pieces of information: the cross-sectional area of the
stream at the location of interest and the velocity of that flow. The first
involves units of area, for instance, square meters; the second involves
units of distance per unit time-for instance, meters per second. Multiplying
the two yields units of volume (e.g., cubic meters) per unit time.

Measurement of streamflow therefore comprises two activities:

estimation of stream cross-sectional area and measurement of flow
velocity. There is often a large quantity of historical data available for
major rivers and streams. As a result, it is often possible to discern for each
location in a given system a characteristic relationship between water level
("stage") and the flow (or "discharge") passing that location. This
relationship is of considerable value in the collection of streamflow data,
because if the stage-discharge relationship is well understood, simple
measurements of water level can yield reasonably accurate information
about flow.

Water level, or stage, is measured by reference to a fixed point of known

position. The simplest water level measurement is obtained with a vertical
rod or stick clearly marked with appropriate units (inches, feet, or
centimeters). This gauge is attached to some permanent vertical structure,
like a bridge or piling, in such a way that it is visible at all possible water
levels. Water level is then simply read off the gauge by an observer
standing at the side of the river.

Another simple device for measuring water level is the wire-weight

gauge. With this method, a reel of wire or metal tape is attached to a fixed
vertical structure (again, a bridge or piling is suitable) and a weight
attached to the end of the wire. Either the wire (or tape) can be marked
with appropriate graduated units or the gauge can be calibrated so that
each revolution of the reel allows a known length of wire to be unwound.
Water level is then measured by lowering the weight from the reel until it
touches the surface of the water.

These manual gauges require regular observation by trained personnel

and thus are subject to human error both in measurement and in recording.
The need for on-site staff usually reduces the frequency at which manual
measurements are taken; most manual gauges are read only once a day, yet
the stage may change dramatically over a period of an hour or less. It is
therefore both costly and onerous to collect detailed streamflow data using
manual gauges.

More convenient, particularly in remote locations, are automated

recording stream gauges. These may be of several types. Float-type
gauges, as the name implies, use a float on the water surface to track water
level changes. Stage is then automatically recorded on a strip chart or
digital device and can also he telemetered to remote locations. Float
gauges are easily upset by river-borne debris and ice, and by vandalism, so
they are usually housed in an enclosure with a stilling basin linked to the
stream by pipes. Usually, more than one pipe is used to ensure continuous
recording even when one pipe becomes clogged-a not infrequent
occurrence at periods of high flow and high sediment transport. Pressure-
activated gauges do not require a stilling well; however, they are also
usually enclosed in a housing to protect against damage. Pressure-activated
gauges employ a sensor that tracks changes in a pressure diaphragm
submerged in the river. Pressure is transmitted to a manometer located in
the gauge house, where it is recorded on a continuous recorder and
possibly also telemetered to a remote location.

Recording gauges are typically accurate to within a few millimeters and

may be operated for periods of up to it year or more without major
maintenance. More usually, these types of gauges are serviced on a
monthly basis to ensure smooth functioning and guard against major gaps
in the streamflow record that could be caused by malfunction.

As with precipitation gauges, the placement of water level gauges is

critical to their accuracy. Because the stage-discharge relationship is often
used to generate streamflow data, water level gauges should he located
where water level and flow are strongly correlated (e.g., scatter less than
5% around a plotted linear relationship) and where the stage-discharge
relationship is relatively independent of flow conditions. This condition
may exist upstream of a small dam or weir, for example. provided that the
location is far enough upstream to avoid backwater effects from the control
structure. If downstream channel characteristics change through significant
erosion or deposition, the stage-discharge relationship will probably also
change. In such a case, it will be necessary to redefine the stage-discharge
relationship by collecting additional data on stream cross sections and flow
velocity. Figure 2.6 illustrates a typical stage-discharge relationship.

Discharge, or flow, measurement requires an understanding of both

stream cross-sectional area and average flow velocity across the stream.
Estimation of average velocity requires a number of velocity
measurements across the cross section, which are then averaged to give an
estimate for the full channel width. Channel depth is relatively easily
measured using graduated lines or wires with a weighted end. Depth
measurements are typically obtained at regular intervals across the
channel. Twenty or more measurements may be taken across a single
stream cross section. At each measurement point, stream velocity is also
measured, usually with the aid of a mechanical current meter. In shallow
streams these measurements can be obtained on foot using hip- or chest-
waders. In deeper waters measurements are usually taken from a bridge or
a moored boat. Velocity is measured just above the stream bed (at 80% of
total depth) and just below the surface (at 20% of total depth) in each
segment, and the two values are averaged. This approach produces a more
accurate velocity reading because it avoids the effects of drag at the
sediment-water and water-air interfaces. In very shallow water velocity is
measured at mid-depth. This approach is illustrated in Figure 2.7. Under
some circumstances, especially where flow dynamics are complex, it may
be appropriate to collect a more complete velocity profile at certain
stations, with measurements at 10 cm intervals from surface to bottom.

When all necessary data have been collected for a given segment, the
flow for that segment is calculated as:

Figure 2.6 Computed stage-discharge curve for Little Sixmile Creek,

Jacksonville, Florida (from Schmidt et al. 1996, reproduced with permission
from Computational Hydraulics International).
Figure 2.7 Hypothetical stream-gauging scheme showing location of cross-
sectional segments and velocity measurements.

a(n0 2 rk).x) q + - 2 (2.14)

a = area of segment ( measured depth x width, expressed in

length2)

rk).2 = measured velocity at 20%- of total depth

ry).s = measured velocity at 80% of total depth

Total flow through the full cross section is then calculated as the sum of
all individual segment flows, as follows:

Q= qi+q2+q3...+q„ (2.15)

qi, q2, q3 ... q„= individual segment flows

 Viessnian and Lewis (1996) and other basic hydrology texts provide a
full discussion of current streamflow measurement practice.

From these basic measurements, it is possible to develop a plot relating

water level (stage) to flow (discharge). This plot is called by several
names, most commonly "stage-discharge curve" or simply "rating curve."
Once the descriptive data have been collected, the stage-discharge curve
should remain applicable unless major changes occur in the stream cross
section or in the stability of downstream controls. If these changes occur, a
new stage-discharge curve must be developed. In some rivers this will
occur several times a year; in others, perhaps only once every few years.

Strcamflow data are collected and published by a variety of government

agencies. Reported values usually are mean daily flows, but average
annual and average monthly flows, as well as annual maximum and
minimum values, may also be reported. If the data have been collected on
a continuous basis (as is likely the case for most major river systems),
continuous flow data may also be available, often in electronic format.
Increasingly, such data (like those for the well-known Hubbard Brook
research program) are available freely over the World Wide Web or on
Internet bulletin boards.

2.5 GROUNDWATER

Water that has passed through the land surface into underlying soils and
rocks is called groundwater. This subsurface region contains both saturated
and unsaturated zones. In the very deepest regions of the earth, water in
liquid form is unavailable and all water molecules are chemically
combined with other sub stances. Above this layer lies a deep, saturated
zone through which liquid water flows, confined by impervious rock
strata. Novotny and Olem (1994) define an aquifer as a saturated
permeable geologic underground stratum that can transmit significant
quantities of water, and an aquitard as a less permeable layer that may be
significant in the regional transport of water but which has a permeability
insufficient to permit economical development. (The term aquiclude refers
to a geologic formation that is entirely impermeable-a rare occurrence, and
consequently a rarely used term.) Most aquifers are formed of
unconsolidated materials including soil and gravel. Some sedimentary
rocks also permit the movement of large quantities of water, but igneous
and metamorphic rocks, which are essentially impermeable, form aquifers
only when they are fractured or otherwise mechanically rendered
permeable.

The top of the saturated zone forms a "water table," which can be
located with the aid of drilling apparatus. Closer to the land surface,
smaller basinshaped rock formations may capture water, creating locally
saturated zones and "perched" water tables. The transition from saturated
to unsaturated zones is not an abrupt one. Rather, capillary action at the
interface acts to draw water up from saturated zones into less saturated
areas, creating a "fringe" effect at the saturated-unsaturated boundary.

Groundwater is measured by injecting and tracking tracer materials in

the groundwater, by the use of field permeameters, or by drilling into the
aquifer and determining the pressure of the flowing water. Normally, water
in the drill hole will rise to the surface of the water table. When the water
surface in the well is higher than the groundwater table, the aquifer is said
to have artesian conditions.

When groundwater is extracted-for instance, pumped out for well water

extraction-the water surface in and near the drill hole or well will be lower
than the original groundwater table. The distance between the well and the
point where no significant influence on water level is observed is called
the "zone of influence" of the well; the zone of lowered water level is
referred to as the "drawdown," which is located in a "cone of depression."
These terms are illustrated graphically in Figure 2.8.

The one-dimensional flow of water in a saturated zone or aquifer is

described by Darcy's law:

q = KhA/L (2.16)
 K = hydraulic conductivity of the flow medium (distance/unit
time)

h = hydraulic head or potential causing flow (distance)

A = cross-sectional area of flow (distance2)

L = length of the flow path (distance)

This equation is valid for most conditions of one-dimensional flow,

whether vertical (as in downward infiltration), horizontal, or upward.
Groundwater is discharged into surface waters (lakes, streams, and
wetlands) at the land-water interface. Groundwater is therefore a major
(perhaps the major) contributor to the base flow for most surface waters.
Table 2.3 gives typical hydraulic conductivities for a variety of soil and
rock materials.

The chemistry of water moving through an aquifer is altered by reaction

with chemicals in the soil, adsorption onto solids in the aquifer, bacterial
action, and similar processes. (Note that this is, nevertheless, a less
complicated system than in surface waters, which are also subject to the
influence of light and volatilization.) Important in these chemical
transformations arc acid-base reactions-for example, those that occur when
acidic precipitation falls on neutral or alkaline soils. As rainwater passes
through soils and over rocks, it dissolves minerals (which minerals, and
how much is dissolved, depends on the relative characteristics of the
rainwater and the soil or rock). Carbonate chemistry is often important in
such a system. Bicarbonate (HCO1) is created when limestone and
dolomite are dissolved. The dissolution of limestone is expressed as:
 Figure 2.8 Influence of well water extraction on ground water table (after
Viessman and Lewis 1996).

That of dolomite is expressed as:

These sedimentary rocks are common throughout much of the world, and
thus their influence on surface water and groundwater chemistry is
significant and well documented. When acidic precipitation (rich in the
hydrogen ion, H+) flows through limestone or dolomite zones, the
bicarbonate content of the rock buffers, or absorbs, excess hydrogen ion
and thus reduces the acidity of the flowing water, yielding water and
carbon dioxide:

This reaction has the effect of increasing the content of polyvalent ions
in the water, particularly calcium (Ca2+) and magnesium (Mg2+). Water
supplies in limestone/dolomite areas are typically high in these ions and
are termed "hard."

A variety of other minerals are also dissolved in groundwater flow.

Among these are sulfides of zinc, lead, and other heavy metals, which,
when dissolved, release those metals into solution and thus into the domain
of drinking water, irrigation water, and livestock water supplies. For
example:

and

Other chemical phenomena that can affect the quality of groundwater

include redox reactions, in which oxidized forms such as sulfates are
converted to reduced forms such as sulfites (and vice versa). Adsorption-
desorption phenomena are also important in transportation of certain
nutrients and priority pollutants such as heavy metals and trace organic
compounds. In this process, adsorption sites are provided by the
electrically charged surfaces of clay minerals and organic particles and
certain other chemical forms. Phosphorus, ammonia, many heavy metals,
and hydrophobic organic compounds, such as some pesticides, are
attracted by these charged sites and bind to the particle. When the particle
is transported-for instance, in spring runoff-the attached pol lutant is
transported with it. (This phenomenon is partly responsible for the thorny
problem of remediating large volumes of polluted sediments in urbanized
streams, lakes, and harbors. To remove the target pollutants, it is often
necessary to remove the entire polluted sediment mass. Removal by
dredging or suction can disturb settled sediments and cause desorption to
occur, with the result that older, buried pollutants are once more released
into surface waters.)

Groundwater quality is also strongly affected by local sources such as

malfunctioning septic systems, leaking below-ground manure storage
systems, land application of sewage sludges and manures, and similar
processes. These activities can convey bacteria, viruses, parasites, and
cysts into groundwater along with nutrients and solids. Aging or ruptured
underground storage tanks formerly used for volatile or nonvolatile wastes
are often sources of various industrial chemicals introduced into
groundwater. Indeed, the range of wastes, coupled with the difficulty of
reversing groundwater contamination, make this one of the most serious
environmental challenges currently faced by our society. Even sanitary
(municipal) landfill sites, if not properly designed and operated, can be a
significant source of numerous materials including solids, phenols,
nitrogen compounds, and various toxins into groundwater.

A common soil quality impact in warmer climates occurs when land is

overirrigated. Water applied to fields contains some level of dissolved
minerals, but water evaporated from those fields is, of course, mineral-
free. Thus, with time, there is an accumulation of minerals, or salts, in
irrigated land. This problem can be addressed by applying excess irrigation
water or by growing salt-tolerant crops. Excess application of water can in
turn leach other materials from the soil and into underlying groundwaters.

A direct impact on human health can occur when high levels of nitrate
are leached into groundwater from agricultural fields where nitrogen
fertilizers or manures have been applied. The disease methemoglobinemia
is particularly troublesome in infants and older adults, who tend to produce
less gastric acid and have higher digestive system pH levels than the rest of
the population. At these higher pH levels, nitrate ions (N03 ). which are
common in food and drinking water, are readily converted into nitrite ions
(NO2 ). The nitrite ions then react with hemoglobin to create an inactive
methemoglobin form, reducing its oxygen-carrying capacity and causing
oxygen deficits in the affected individual. (Healthy adults are usually able
to convert methemoglobin hack to oxyhemoglobin and thus retain
sufficient blood oxygen, even at relatively high levels of nitrate/nitrite.)

2.6 WATEFI QUALITY

2.6.1 Influences on Water Quality

Water quality directly affects virtually all water uses. Fish survival,
diversity, and growth; recreational activities such as swimming and
boating; municipal, industrial, and private water supplies; agricultural uses
such as irrigation and livestock watering; waste disposal; and general
aesthetics-all are affected by the physical, chemical, biological, and
microbiological conditions that exist in watercourses and in subsurface
aquifers.

Many factors influence water quality; some of these have been

discussed in previous sections. The chemistry of bedrock and surficial
geology and the drainage characteristics of soils can determine whether
natural waters are acid or alkaline, sediment-laden or clear, high in heavy
metals and dissolved salts or relatively free of those constituents. Physical
processes like erosion can add large quantities of suspended sediment to
surface waters.

Biological processes draw nutrients from surface waters and soil

moisture, and decaying tissues release nutrients into the watershed
ecosystem. Photosynthesis uses sunlight and carbon dioxide to create plant
sugars and oxygen, so where aquatic plant growth is dense, daytime
dissolved oxygen levels are likely to he high. At night, however,
respiration-which uses oxygen for tissue maintenance and repair-
dominates over photosynthesis, so dense aquatic plant growth can also
mean nighttime dissolved oxygen depletion in a stream.

Although pristine waters are suitable for most purposes, the minimum
acceptable quality of water depends very much on the water use. Water for
irrigation, for example, should be low in dissolved salts, but water
intended for livestock should be low in bacteria. Water for industrial
processes should usually be of much higher quality than water for
industrial cooling. Water for municipal drinking water supply must not
only be safe to drink, but should ideally contain low concentrations of
materials such as calcium carbonate, iron, and similar materials that can
cause costly infrastructure damage or add unpleasant characteristics to the
finished water. Human health considerations demand low levels of a range
of contaminants, yet salmon and trout populations may be even more
sensitive than humans to many water quality constituents.

Water quality impairment is often a trigger for conflict in a watershed,

simply because degraded water quality means that desired uses are not
possible or are not safe. Chapter 3 discusses approaches to "scoping" a
watershed management plan, including compilation of a list of desired uses
and the water quality targets appropriate for each. This simple matching
process provides an early evaluation of where use impairments may exist
and, thus, where remedial action may be warranted.

Water quality indicators are of several main types. The following

sections give a brief overview of the principal indicators in current use. A
more complete discussion is available in Tchobanoglous and Schroeder
(1987).

2.6.2 Physical Indicators of Water Quality

Water Clarity The most basic indicator of water quality, and an excellent
measure of aesthetic potential, is water clarity. Clarity is measured in
different ways. One method often used for lakes, and occasionally for
rivers, is Secchi disk depth. The Secchi disk is a flat disk about 30 cm in
diameter, with the top surface painted in quarters alternating black and
white. The disk is lowered on a line until it disappears from view, and a
note is made of the depth of disappearance. The disk is then lowered
farther and raised again, with a note made of the depth at which it is once
again visible. The average of the two depths-ofvisibility is recorded as the
"Secchi disk depth." This measure is inexpensive and easy to use and
provides results that are intuitive for the lay person. It does not, however,
yield information as to the causes of reduced clarity, which may be of
diverse origin.

Water clarity is also measured with mechanical devices, including the

turbidity meter and the transmissometer. These devices, which measure
light transmission through the water column, are more costly and more
difficult to operate (and calibrate) than a Secchi disk, and the data they
generate is harder for the lay person to understand. They are, however, less
subject to human error and interpretation than the Secchi disk and have
largely replaced that device in modern water quality monitoring.

Suspended Sediment Suspended sediments (particulates) arise from eroded

silts and clays, organic detritus and plankton. Waters high in suspended
sediment have a turbid or "muddy" appearance that makes them
unattractive for swimming and bathing. High solids content can also
demand additional treatment for municipal water supplies and can limit the
utility of water for industrial purposes. Sediment settled out of turbid water
can foul fish spawning habitat, clog fish gills, and otherwise disrupt a
habitat for aquatic organisms. Although natural processes like erosion
contribute a high proportion of suspended sediment in some systems,
human activities such as agriculture can increase the rate of these
processes and can add new sources of sediments such as municipal and
industrial wastewaters. Urban storm drainage is a major source of
suspended solids in many cities with storm sewer systems discharging to
surface waters. Suspended sediments can also indicate the presence of
other pollutants such as phosphorus, heavy metals, and some pesticides,
which readily adsorb to the surface of clay and silt particles. Table 2.4
gives particle size ranges for a variety of inorganic and organic solids.
Conductivity Conductivity is a measure of the electrical conductance of
water-its ability to conduct an electric current. Electricity is conducted by
ions in a solution, so conductivity tends to increase with increased ionic
strength (quantity of dissolved salts). Although a gross measure of water
quality, conductivity can be a valuable surrogate for other water quality
constituents that may be both more costly and more time-consuming to
measure. Continuous conductivity measurement is common in industrial
operations (as is continuous pH measurement) to provide an early warning
of problems ("upset") in the process or other nonroutine circumstances.

Hardness Dissolution of calcium and magnesium minerals can add high

levels of these multivalent cations to natural waters. This "hard" water can
cause problems with scaling and clogging in pipes and other infrastructure,
reduce the life of water-using appliances such as water heaters and electric
kettles, and reduce the effectiveness of household washing agents. Pipe
clogging due to hard-water "scale" is a major contributor to reduced water
pressure in residences and institutions.

Water Temperature Water temperature is often a useful measure of water

quality, although it is understandably influenced by many factors,
including local vegetative cover (which may shade a stream from incident
sunlight and thus reduce its temperature), meteorological conditions such
as cloud cover and wind, and industrial or other effluents-for instance,
cooling waters (which by definition have received excess heat from heat-
generating processes such as thermal power generation). Temperature is a
critical factor in habitat choice for aquatic organisms. Too high or too low
an ambient temperature will deter certain organisms from entering an area;
extreme high or low temperatures can be lethal. Temperature is often a
central factor in the decline of desired "pan fish" species such as salmon
and trout, especially in urban areas. These species, sometimes termed
"cold-water" species, prefer cool temperatures and, although they may pass
through warm water zones, will not reproduce successfully in conditions
where ambient temperatures are too high. In heat-polluted surface waters,
it is not uncommon to see less desirable species such as carp and alewife
come to dominate the fish community. Thermal pollution-elevated
temperature in receiving waters-is common near the effluents of thermal
generating stations and where urban stormwater runoff has picked up heat
from sun-warmed roofs and parking lots. The role of temperature in
determining species distribution and reproductive success probably relates
to the fact that temperature affects the rate of many biochemical processes.
Temperature can, of course, also affect the rate of chemical reactions in the
abiotic environment and is a central factor in evaporative processes.

Aesthetics The term aesthetics encompasses a broad, if subjective, range of

concerns about water quality. In many systems, surface waters are valued
not just for their utility in drinking, recreation, agriculture, and industrial
use, but also simply for their beauty. Degraded aesthetics may not have
direct health effects but can create major economic impacts-for instance,
through diminished revenues from tourism and recreation. Waters of high
aesthetic quality are typically free of materials that will settle and form
objectionable deposits, contain little or no floating debris, oil, or scum, and
do not have an objectionable color, odor, taste, or turbidity. Aesthetics may
also be taken to include the presence of desirable species, such as trout and
salmon, or the absence of undesirable species, such as carp or excessive
insect fauna.

2.6.3 Chemical Indicators of Water Quality

Dissolved Oxygen Of all chemical indicators of water quality, dissolved

oxygen may, in many ways, be the most important. Most aquatic animal
species require abundant oxygen for life and reproduction; some will
tolerate lower oxygen conditions, and it is often the presence of these in
the aquatic community that signals water quality impairment. High in-
stream oxygen levels are also helpful in promoting the biological
assimilation of organic and nitrogenous wastes--an important consideration
downstream of effluent discharges from sewage treatment plants and some
industrial facilities. High levels of dissolved oxygen are not usually of
concern, but low levels, highly variable levels, or a complete absence of
oxygen can create conditions that are inhospitable for aquatic life. See
Figure 2.9.

Nutrients Several common water pollutants are often grouped together as

"nutrients" because they are essential for the growth and maintenance of
bio logical tissue. The three pollutants most commonly described as
nutrients are phosphorus, nitrogen, and silica; the latter is in such abundant
supply relative to demand that it seldom controls the growth of nuisance
plants or algae, and it is not discussed further in this section. Phosphorus
and nitrogen are, however, considered to be important parameters for
pollution control and are therefore often the focus of concerted efforts in
watershed cleanup. Both of these elements are commonly found in nature.
Phosphorus is used in many biological systems, most fundamentally in
adenosine diphosphate and triphosphate, which are key players in glucose
metabolism and energy conversion. Nitrogen is probably most familiar for
its role in building protein, and in the decay products of protein such as
ammonia and urea. In most temperate freshwater systems nitrogen is
abundant, and it is phosphorus that is the so-called limiting nutrient-the
nutrient whose concentration determines whether additional biomass can
or cannot be produced. In marine systems, the reverse is often true:
phosphorus is abundant and nitrogen is the limiting nutrient.
Figure 2.9 Influence of organic wastes on oxygen concentrations in a stream
(after Ontario Ministry of Environment 1987).

There are many sources of nutrients in most watersheds, including

atmospheric processes and rock weathering, but in most developed
watersheds the most significant sources are often municipal and industrial
waste discharges and agricultural activities. Phosphorus and nitrogen
compounds are considered "nonconservative" in nature, because they are
likely to react with other substances and with biological tissue. They are
unlike suspended sediment (a "conservative" pollutant) in that it can be
difficult to develop a mass balance for phosphorus and nitrogen
compounds in a watershed-that is, to account for all the sources and sinks
of these materials within the system and fully explain their behavior. These
nutrients are therefore challenging to simulate accurately, and it can be
very difficult to develop accurate predictions of ecosystem response to
nutrient management actions.

In converting from one form to another, nitrogen compounds can cause

impacts on other water quality parameters, especially dissolved oxygen.
For example, the chemical conversion of ammonia to nitrite, and then
nitrate, uses oxygen. High ammonia levels, as may be found downstream
of a sewage treatment plant outfall or a poorly managed manure storage
facility, can therefore also have a significant impact on instream dissolved
oxygen resources (the unionized fraction of ammonia is also acutely toxic
to aquatic life). High concentrations of nitrate in drinking water can cause
serious health effects in human infants through methemoglobinemia (see
Section 2.5).

Heavy Metals Our society uses many metals in many forms, and with time
these have found their way into soils and receiving waters and, in some
cases, into biological tissue as well. Heavy metals such as copper, nickel,
zinc, cadmium, chromium, lead, and mercury are common in industrial
use. In urban centers most small industries and many larger ones have for
decades used the municipal sewer system to convey liquid effluents to a
central sewage treatment plant. This has been a good arrangement for
many public utilities, which can charge sewer-use fees and generate
revenues while providing a convenient treat ment option for industry. The
difficulty has been that sewage treatment plants are not designed to remove
metals. In some cases wastewater concentrations of a given metal have
been high enough to affect treatment processes at the plant. In many,
perhaps most, others, metals simply attach to particulate matter in the
wastewater stream and are removed with "biosolids," the mixture of dead
bacterial cells, inorganic particles, and other materials that comprise
sludge. Up to 70% "removal" of metals may occur in publicly owned
treatment plants. These metals are, however, not truly removed but simply
transferred to the solid phase. Some metals do remain in solution and are
discharged with treated effluents into receiving waters. Metals may
therefore reenter the ecosystem in several ways: in aqueous solution, in
leachate from landfills or agricultural lands where sewage sludges have
been spread, or in air emissions from incinerators used to dispose of
sewage sludges.

Metals are of concern as a group because they can impair enzyme

function. The degree of impairment depends very much on the metal and
the organism affected. In humans, heavy metal poisoning characteristically
involves symptoms of gastric upset and central nervous system
impairment, including altered perception (e.g., peripheral vision) and gait.
Several metals, including lead and mercury, have been linked to smaller
birth weights and delayed cognitive development in young children. High
concentrations of heavy metals also cause acute and chronic toxicity in
nonhuman organisms and (depending on the metal) may also have
mutagenic, teratogenic, or carcinogenic effects. Some metals (mercury,
lead, and tin, for example) are known to convert from relatively nontoxic
inorganic forms to highly toxic, bioaccumulative, and persistent organic
(methyl or ethyl) forms, probably through bacterial mediation.

Trace Organic Compounds Almost unknown 20 years ago, industrial

organic compounds, pesticides, and herbicides are rapidly gaining
prominence in the field of water management. Until recently laboratory
analytical protocols were unable to measure these substances at very low
levels, and it was assumed that levels so low could not be important to
human or ecosystem health. In recent years, however, improved analytical
techniques and better knowledge of the toxicology of these agents have led
to increased concern about them.

There are at least 60,000 chemicals in regular use in the Great Lakes
Basin alone, and many of them have been developed in the last 50 years.
Some substances, which may be almost undetectable in water, have been
labeled "hydrophobic" because of their affinity for particulate matter or, in
some cases, for fats and oils. Many are toxic (that is, cause impaired
function in one or more organs), and some are highly persistent and
bioaccumulative. A substance is considered bioaccumulative when plant or
animal enzyme systems cannot effectively detoxify and excrete it. Over
time, an organism's "body burden" of the substance increases, simply
because intake rates vastly exceed elimination rates. A corollary of this
finding is that animals higher in the food chain-secondor third-level
carnivores (animals that eat other carnivores)-will themselves receive a
highly concentrated "dose" of the substance with each meal. When humans
consume a top carnivore species like lake trout, they too ingest all of that
organism's accumulated body burden of the contaminant-and cannot
themselves readily excrete it.
 (Many of the most toxic and persistent organic (carbon-based)
substances, including dioxin, DDT, and PCBs, contain chlorine and are
thus termed "chlorinated organics." Residual chlorine from water and
wastewater disinfection or bleaching processes is not itself in this
category-in fact, it is highly volatile-but it is acutely toxic to fish and other
aquatic life forms even at relatively low concentrations.)

2.6.4 Biological Indicators of Water Quality

Bacteria Thousands of bacteria strains can be detected in even the most

routine household items; few of these will actually cause disease in
humans or other organisms. For decades, fecal coliform and fecal
streptococcus bacteria (common in the feces of warm-blooded animals)
have been used as indicators of the presence of fecal pollution. These
groups are easy to identify, primarily by appearance; laboratory analyses
for their presence are quick and inexpensive. The groups do, however,
encompass a wide range of bacteria, both pathogenic and nonpathogenic.
There is now a growing body of epidemiological evidence showing that
fecal coliform and fecal strep counts do not correlate well with the
incidence of gastrointestinal illness, so these indicators are gradually being
replaced by more specific measures. Assays for individual organisms, or
smaller groups, has become more routine with the development of
improved laboratory techniques. Sometimes fecal coliform, total coliform,
or fecal streptococcus assays will be used with measurements of one or
more specific strains to obtain a more precise picture of the likelihood of
illness occurring.

One organism now commonly used in water quality assessment is

Escherischia coli, a common gut bacteria in warm-blooded animals,
including humans. Enterococci (species of the fecal streptococcal group,
which includes Streptococcus faecium and S. faecalis, which are always
found in the human gut) and Pseudomonas aeruginosa, which causes eye,
ear, and skin infections in humans, give a more reliable indication of health
risk than traditional indicator groups. Other pathogens, for instance
Salmonella, Shigella, Staphylococcus aureus, Campylobacter jejuni,
Legionella spp., and viruses can also be measured in water, but their
analysis is much more costly and time-consuming than tests for indicator
groups, so they are seldom used unless a specific problem is under
investigation.

The sources of bacteria and viruses primarily involve the feces of

warmblooded animals, including humans. In a watershed, this can mean
sewage treatment plant effluent that is incompletely disinfected,
stormwater runoff (carrying animal feces), combined sewer overflows
(carrying a mixture of human sewage and stormwater and diverted to a
watercourse before treatment), and animal wastes such as manure storage
and feedlots. These sources are directly affected by precipitation events. As
a result, levels of bacteria and viruses in receiving waters are typically
highly variable, difficult to correlate with other biological indicators or
bacteriological species, and almost impossible to model with accuracy.
Simulation is even more difficult because these organisms, as living
entities, grow, reproduce, and die and can even overwinter in lake and river
sediments if conditions are suitable.

Parasites A number of parasitic organisms occur in freshwater systems;

some of them have serious health implications for humans and other
animals. Among these are Schistosoma sp., parasitic worms causing
schistosomiasis (bilharziasis). In temperate climates some species of
schistosomes do not actually live in the human body but are carried by
freshwater molluscs and birds. These species cause dermatitis, or
"swimmer's itch," in humans exposed to them. In warmer regions most
schistosome species live in the human body and can cause serious
gastrointestinal illness and even death.

Giardia lamblia, a waterborne parasitic protozoan carried by the beaver,

is of increasing concern in recreational areas, where it may be ingested in
drinking water or in water used for swimming or bathing. Giardia causes
mild to serious gastrointestinal illness in the affected person and is readily
transmitted by oral contact with contaminated materials. About 7% of the
parasitologic samples submitted in the United States contain this organism.
Giardia is more resistant to chlorination than are bacteria and, in some
cases, may also be more resistant to medical intervention than other
parasitic infections.

Crvptosporidium is a more recently identified pathogenic protozoan that

may be as important as Giardia in causing diarrhea] illness. Outbreaks of
cryptosporidiosis are associated primarily with ingestion of inadequately
treated drinking water. It is now believed that Crvptosporidium is widely
present in the feces of warm-blooded animals, including domestic farm
animals, so major sources of this organism are thought to be improperly
stored manure and, in some cases, faulty septic systems.

2.7 PLANT AND ANIMAL COMMUNITIES

In addition to geologic and meteorologic forces, a watershed is affected by

a variety of biological processes. An inventory of watershed biota therefore
aims to do three things:

1. Determine the number and types of plant and animal species present
in the watershed area;

2. Estimate the number of individuals of each species; and

3. Investigate the interrelationships between the species and their abiotic

environment.

Depending on the desired detail of the watershed inventory and the com
plexity of uses and issues in the basin, these objectives may be satisfied by
a review of existing data (for instance, in the scientific or government
literature) or new data may be required. The latter is a costly and time-
consuming prospect and one that may be unnecessary except at a research
level. In either case, the analyst is usually dealing with problems of
sampling from a vastly diverse and complex natural system, with all the
pitfalls such sampling implies.

At a cursory assessment level it is usually possible to obtain relative or

qualitative assessments of the plant and animal species present in the
watershed, along with their probable responses to management actions of
various types. It is, however, very difficult to obtain a detailed and
comprehensive understanding of these systems and, thus, to predict
ecosystemic responses accurately and quantitatively. For major
management initiatives likely to have impact on these systems, there may
be value in pilot scale implementation to test outcomes "in nature" before
full, and costly, implementation.

2.7.1 Using Existing Data

The literature on field ecology provides extensive guidance on estimating

the number of species present in an area and the number of individuals per
species. Most state and provincial natural resources agencies have
available detailed information about regional biota, including the biotic
region or ecotone within which the watershed falls, the common and
uncommon species in the area of interest, and the nature and location of
fragile or valued habitats. Information about rare or endangered species
can also be obtained from these agencies or from regional or national
biodiversity protection organizations.

Detailed information about the distribution of individual species can be

found in one of the many readily available field guides such as those
produced by Peterson or Audubon." These guides may be most helpful for
larger species of mammals, birds, and reptiles and for trees and shrubs. For
these organisms, field guides usually provide maps of geographic
distribution accompanied by detailed commentary-for instance, about
species habitat preference and diet.

Information of this kind is, however, seldom as available for

invertebrate species such as insects and their larvae. Yet these organisms
can be very important in assessing the health of aquatic systems or soil
quality, especially at a micro-habitat level. Surveys of the nature and
diversity of macroinvertebrate populations have been used for many years
to identify areas where water quality impairment has had an impact on
stream ecology. These surveys are relatively inexpensive (as compared
with laboratory analysis of individual chemical substances) and can be
helpful in identifying areas requiring more detailed assessment. State and
provincial agencies responsible for natural resources or envi ronmental
protection may have historical data on invertebrate fauna. University
departments of biology, agriculture, or soil science are also often engaged
in research in these areas. Theses reporting this research are usually
available on microfilm or microfiche through university or public
reference libraries.

Occasionally it may be necessary to conduct field surveys within the

watershed to update existing information or fill data gaps. The following
sections provide a brief overview of techniques for this purpose.

2.7.2 Indices of Biodiversity and Similarity

Classical ecological wisdom holds that ecosystems that are diverse and
complex are also likely to be more stable and more able to recover
following disruption. Although there is not universal agreement on these
concepts, it is often true that a healthy, robust ecosystem is characterized
by a large number of species but with no species heavily dominant or
overrepresented in terms of individuals.

Environmental biologists use a variety of indices to describe the species

diversity of a habitat and the similarity of that habitat to another. Some of
these indices are quite old and now well entrenched in the ecological
literature, few, if any, of them are wholly objective or accurate. They may,
however, provide useful tools with which to assess the condition of 'a
watershed system of interest.

Diversity indices usually employ measures of the number of species

present and the number of individuals per species. Since these measures
also imply the probability of encountering an individual of a given species
(or reencountering one, in the case of mark-recapture methods, as
discussed in Section 2.7.3), some indices express number of species
present in terms of probability.

Among the oldest and most famous of diversity indices is Simpson's

Index of Diversity (Simpson 1949), which is expressed as:
 D= I -E(p,)2 (2.17)

/r proportion of individuals of species i in the community

Simpson's Index reflects the simple principle that the probability of

randomly picking two organisms of different species is equal to I minus
the probability of picking two organisms of' the same species. The
probability of picking an organism of any given species is simply its
proportion in the community. p;. So the probability of picking to organisms
of the same species is simply the joint probability (p; x pt), or (p;)-. The
difficulty with Simpson's Index is that it tends to give little weight to rare
species and more weight to common species.

The Shannon-Weiner Index attempts to add information about the

evenness of individuals among species. This function was developed
independently by Shannon and Weiner and bears both their names. It is
sometimes mistakenly referred to as simply "Shannon's Index" or as the
"Shannon-Weaver Index." It is expressed as:

H = E(p,)(In p,) (2.18)

pi = proportion of individuals in species i

It is possible to calculate a theoretical maximum equitability (evenness)

of individuals among species using the Shannon-Weiner approach, as
follows:

Hmax = S S In, S) - In S (2.19)

 Equitability, sometimes written as .1, can then be calculated as the ratio
of observed diversity (H) to H,,,;,x:

H J = Hmax (2.20)

Similarly, indices typically employ measurements of the number of

species occurring in each of two communities to be compared, and the
number of species occurring in bulb communities. A good example is
Jaccard's Similarity Coefficient, 1:

1 = J x 100 a+h -j (2.21)

j = the number of species occurring in both communities

a = the number of species occurring in community a

b = the number of species occurring in community b

2.7.3 Field Data Collection Techniques

Field data collection is an expensive prospect, so if the effort is made to

undertake a field survey, it is worth taking care to make sure that the
results are both quantitative and representative. Usually, this means careful
attention to the data requirements of desired analytical techniques, such as
statistical hypothesis tests. Ideally, the analyst should have a clear question
in mind in designing the field survey, such as "How many species are
present in the sample plot?" or "What is the average number of individuals
species in the sample plot?" Clear formulation of the hypothesis to he
tested simplifies selection of an analytical technique and provides valuable
guidance in sampling design.

Frequently, sampling of biological populations and communities

(whether aquatic or terrestrial) employs random placement of plots of
known area and a complete inventory of the biota within the test plot. Plots
of a square-meter area are typically used for sampling small organisms
such as insects and aquatic invertebrates. Larger plots are necessary for
larger organisms, and plots of several hundred square meters or even
kilometers may he required to survey very large or uncommon species or
those that routinely move across large distances.

Field sampling of animal populations is usually much more difficult

than sampling of plant populations, for the simple reason that animals are
mobile. Sampling of animal populations can employ mark-recapture
methods, which involve capturing and "marking" (for instance, tagging) a
number of individuals of the target species, releasing those individuals, and
then recapturing animals of the same species after some period of time.
The recaptured sample will include both marked and unmarked
individuals, the proportions of which give an estimate of total population
size.

Aerial photography can be helpful in assessing large tracts of

vegetation, such as forest cover or crop cover, and in tracking certain kinds
of large animals, especially those that are very mobile but quite visible.
The harp seal population in the Gulf of St. Lawrence has been estimated in
this way with some success. The disadvantage of aerial methods is that it is
not always possible to distinguish the sex or age of an animal from the air,
yet these characteristics can be vital in establishing the health and breeding
status of a population. Similarly, it can he difficult to identify plants to the
species or subspecies level, especially if the canopy is dense or the
vegetation diverse. Although sometimes less labor-intensive to mount,
aerial techniques may therefore have less value than traditional on-ground
methods in locations where there is heavy forest cover, where access by air
is dangerous, or where the presence of aircraft can interfere with normal
behavior patterns and, thus, population distribution.

Air photo techniques and satellite imagery have been used successfully
in inventories of managed forests and other terrestrial plant systems and
may have potential for the remote evaluation of some water quality
characteristics. The advent of Global Positioning System technology has
made it possible to collect field data with considerable accuracy as to
geographic location and thus simplifies the problem of locating a precise
sampling station when sampling is to be repeated.

All field sampling techniques, whether on-ground or remote, involve

some error in estimation. Depending on the sampling method and the way
it is applied, this error can be very large and thus difficult to interpret in the
formation of public policy. It is therefore essential that estimates of
population size and distribution (including sex and age distribution as well
as density) he reported with an accompanying estimate of sampling error
or uncertainty.

A more complete description of current ecological field data collection

techniques is provided in Magurran (1988).

2.7.4 Assessing Interrelationships

Although counting the number of species and the number of individuals

per species is a relatively straightforward matter, it is much more difficult
to evaluate the interrelationships between species and (perhaps of most
concern to watershed managers) the implications of management actions
that might change those relationships.

A common method of representing the interrelationships between

species is the food web diagram. The food chain is a related, but simpler,
concept. A food chain shows only a single species at each tropic level: one
species of primary producer (plant), one species of primary consumer
(herbivore) that feeds on that plant, one species of secondary consumer
(carnivore) that feeds on the herbivore, and so on to the top of the food
chain, arriving at the species that has no major predators, the "top
carnivore." For the purposes of watershed management, the food web
offers a more useful approach, showing the interlinkages of individual
food chains and, thus, usually a range of habitats from aquatic to
terrestrial, a range of species encompassing herbaceous and woody plants,
and a range of vertebrate and invertebrate animals. Figure 2. 10 illustrates
a typical food web.
 Conceptual representations like a food web diagram are helpful in
elucidating relationships that may not be apparent on first glance but that
may be vital to key watershed functions. For example, heavy grazing
pressure in rangeland tends to favor low-growing plants like Polvgonum
and discourage reproduction of taller grasses and herbs. Release of that
grazing pressure on a grassland (by fencing to exclude cattle access) could
(depending on other climatic and soil conditions) allow grass species to
recover and even dominate the system. As a result, surface roughness is
significantly increased, runoff is diminished, and changes in local stream
hydrology may be observed. It is likely that these changes would also be
accompanied by shifts in the composition of small mammal communities,
especially granivores like rats and mice, and larger animals who feed on
those species. These changes may in turn affect recreational opportunities,
aesthetics, and agricultural activities.

Although schematics like the food web diagram can help identify
relationships that may he in place, they are not as helpful in evaluating-or
predicting-the rates of mass- or energy-transfer phenomena. These are
strongly affected by local physical, chemical, and biological conditions; an
accurate evaluation therefore requires detailed investigation. There is a
steadily expanding body of literature on ecosystem processes and the
factors that affect them. This literature may provide sufficient information
for an initial estimate of system rates and insight into what may control
them in the watershed of interest. An example of qualitative food web
analysis is provided in Bodini et al. (1994).
 Figure 2.10 Example of a simple food web.

2.8 LAND USE

Up to this point, we have primarily been concerned with taking inventory

of natural features in a watershed and (to a certain extent) understanding
their behavior in space and time. Equally important in most systems are
what may be termed the "built" components of the environment-the
structures, machines, and systems constructed by humans and the social
and economic systems that have encouraged those changes in the natural
watershed system.

A critical first step in assessing this built environment is to understand

current and anticipated land use within the watershed. The term laid use
implies use by humans: the ways in which humans formerly, presently, or
may in the future change the landscape for the purposes of resource
extraction and processing, housing, and transportation.

Assessment of land use demands three elements: an understanding of

the nature of human activities currently practiced in the basin, an estimate
of the areal extent of each activity (and perhaps its proximity to valued
watershed features), and the ways in which these aspects are likely to
change in the near and far future. An analysis of land use yields insights
into many facets of the watershed system, including altered drainage
regimes, pollutant sources, valued natural and built features, and,
indirectly, regional economic forces and community priorities. These in
turn become important in anticipating the impact of possible management
actions within the basin.

Figure 2.11 gives a conceptual representation of the interrelationships

between major systems in a watershed. Even at this simple scale, it is easy
to see that minor changes in one system (for instance, climate) can have
important implications for the structure and function of other systems (for
instance, soils, hydrologic processes, human and nonhuman animal
communities, plant communities, and so on). Similarly, a change in human
activities, especially land use, can have a dramatic impact on natural
processes.

Figure 2.12 illustrates the effect of increasing urbanization on a

hypothetical stream system. With the building of more roads, parking lots,
housing, and major industrial, commercial, and institutional structures,
more of the land surface becomes impermeable. Rain falling over the area
no longer infiltrates as easily or as quickly, so much more is "excess" flow,
or runoff, over the land surface. This runoff reenters streams much more
quickly now than in the preurban state, so streamflow response to rainfall,
and to the cessation of rain, is much faster in an urbanized watershed.

The element of time is also important in understanding land use. Past

land use practices have an influence on the present watershed and the ways
in which current land uses may be constrained. For example, past open-pit
mining activities may have created physical changes (pits, altered
drainage, and so on) that limit the usefulness of the site for residential
development. The same will clearly hold true for the influence of current
land use on future development. For this reason, many urban centers have
devoted time and thought to the development of long-term land use
forecasts to guide present development. These forecasts may include
estimates of projected population growth and reflect community consensus
about intended areas for future residential, industrial, and commercial
growth. Long-term planning for transportation corridors will likely also be
a part of the plan. Many communities employ a system of land use zoning,
so that individual parcels of land are designated for use only for certain
activities.

Figure 2.11 Interaction of natural processes in a watershed ecosystem (after

Lang and Armour 1980).
 Compiling an inventory of current land use need not he a tedious task.
An obvious place to start is local and regional governments with an
interest in planning and zoning. Many such governments have official
land-use plans incorporating current and projected land use maps. If maps
for a particular watershed are not readily available, topographic maps or
even road maps can provide a starting point for a land use inventory.
Agricultural extension professionals usually have quite detailed
information about regional farming practices and even individual farm
operations. Information held by public agencies is increasingly available in
digital format, such as Geographic Information Systems, and thus readily
extracted for use in simulation models or for other analysis.

Remote sensing technology is also well developed in this area. Satellite

or high-altitude aerial photography is an excellent source of accurate
information and, indeed, forms the basis for most modern watershed
cartography. The frequency of imagery may not be sufficient to identify
small changes in the watershed with time (changes that may, for example,
be important in assessment of biological communities) but will likely be
more than adequate for a general inventory of land use activities in the
basin.

Land use is often divided into categories, such as open space (rural, non
agricultural, undeveloped land), forested/woodlot, parkland, agriculture
(which may be subdivided into rangeland and cropland or even into more
specific categories reflecting production type), residential, commercial,
and industrial. Again, depending on the interests of the analyst, these broad
categories may be further subdivided. Table 2.5 gives an overview of
typical land use categories and the pollutant generation potential of each.
Figure 2.12 Effect of urbanization on rate and volume of runoff (after Lang
and Armour 1980).

These different land uses alter the land in different ways. In terms of the
movement of water in the basin, land use often changes the permeability of
the land surface, thus affecting surface drainage systems and natural
hydrology. The most dramatic example of this is the widespread paving of
roads, highways, and parking lots, and the high proportion of roof area to
total watershed area, in heavily urbanized watersheds. Water diverted off
these impermeable surfaces is routed overland or to stormwater collection
systems and thence to receiving waters, moving faster and carrying more
heat and pollutants than would be the case in an undeveloped watershed.

Land use also affects pollutant yield, both in terms of the type of
pollutant generated by a particular land use and in terms of the total mass
of pollutant released. This phenomenon has been recognized for many
years, and there is now an excellent literature on the unit-area pollutant
loads that can be expected from various land use types (see Table 2.6).

Land use may also be important in determining the applicable

legislation (Chapter 9), oversight agency, and interested nongovernment
organizations. Agricultural activities, for example, are subject to different
legislation than are manufacturing facilities, have available different loan
and grant opportunities, are overseen by government agencies with an
interest in protecting and promoting agriculture and its products, and have
well-developed farmer and soil conservation outreach networks. Different
land uses clearly look different to an observer, implying landscape
aesthetics that are altered to a greater or lesser degree. Although
appearance may be a minor concern in some communities or landscapes, it
can also be an important contributor to community economics, for
instance, through tourism and recreational opportunities available, or
unavailable, in the basin.

Systematic decision making regarding appropriate land uses for

different parts of a watershed system has a central role in protecting or
impacting valued landscapes. Often this entails zoning-a system of land
use categories with allowable uses designated for every part of the basin.
Without a clear land use planning system, development can proceed both
unattractively and, arguably, unsustainably. Bangkok, Thailand,
experiences staggering urban traffic congestion, forcing most commuters
to spend four to six hours a day in their cars. This in turn creates serious
urban air pollution problems and directly affects the health of hundreds of
thousands of Bangkok residents every year. Yet Bangkok's traffic problems
may have their roots in the simple absence of a comprehensive land use
planning framework. Without such a plan, landowners have been able to
build houses, dig roads and laneways, and subdivide property without
regard for the cumulative result on the community. Some parcels of land
have now been subdivided to the point where they are too small to build on
and cannot be adequately serviced, frequently deteriorating into slums. An
impressive system of urban highways has been built and more are under
construction, but the city lacks an adequate system of "feeder" roads to
service these highways. Drivers exiting a freeway may enter directly into a
small laneway system-and grind to a halt. In Bangkok the jurisdictional
authority for transportation and land use planning is not at all clear, and it
will require a major effort on the part of Thailand's federal and state
governments to clarify these roles before the problem of land use, and
transportation, planning can be resolved.

2.9 SOCIAL AND ECONOMIC SYSTEMS

The natural and built environments of a watershed are usually overlain by

human social systems. Beyond the obvious systems of urban centers and
farm steads, human systems can include the social and economic
infrastructures necessary to support thousands or millions of people.

(Source: Waller and Novak 1981; Wanielista 1978)

 In the past these social and economic systems were often ignored in
water management, viewed, perhaps, as separate from and irrelevant to
surface or ground water hydrology and the structures that might be built to
manage those systems. Today, however, it is widely recognized that social
and economic systems are an integral part of the watershed ecosystem,
affecting not only the physics and chemistry of natural waters but also the
attitudinal and economic forces so central to successful implementation of
water management actions. Indeed, under many environmental assessment
laws, "the environment" is defined as including human social and
economic systems. Yet these systems can be difficult to describe-and to
inventory.

What is meant by "social and economic systems"? The term can

encompass the obvious components of commercial and industrial activity,
major institutions such as hospitals and universities, and residential
development. Less obviously, it can be assumed to include religious
systems, because they entail socially significant structures (e.g., churches,
temples, synagogues; graveyards; sites of aboriginal religious value) and
the belief systems of the religion(s), which have at least the potential to
conflict with watershed activities. Considerations such as social and
economic stability may also be important elements within a community, as
well as what might be termed "quality of life"-that intangible quality of a
safe, satisfactory living environment, with all essential services available
and affordable.

Social and economic systems have the potential to influence a

watershed ecosystem in three main ways beyond those described earlier in
regard to land use:

1. By influencing the attitudes and priorities of watershed residents and

decision makers

2. By affecting the value that may be placed on individual watershed

features and activities, and thus the importance they are given in
watershed planning
 3. By constraining the financial resources available to resolve watershed
issues

These forces must not be underestimated in watershed management. A

company that is profitable and progressive can provide important
leadership in the basin through its use of effective waste management
practices, water reuse and recycling, and responsible ultimate disposal of
wastes. On the other hand, a major employer that is struggling financially
may be hard pressed to pay its routine bills, let alone install costly new
pollution control devices. In the latter case, community residents may
support the company even when it is a known polluter, simply because of
their desire to retain jobs and economic stability in the community.
Community attitudes are especially influential in actions relating to water
conservation. Figure 2.13 shows the influence of demand management
strategies (measures to encourage consumers to use less water) on
projected water consumption patterns in the city of Providence, Rhode
Island. The figure illustrates four possible water consumption scenarios,
depending on whether demand is unconstrained (no water conservation
measures employed), or "baseline," "moderate," or "aggressive" demand
management is undertaken. It is not necessary to understand the details of
these schemes to realize that community support is an essential factor in
successful implementation. That support will depend on community
attitudes and social priorities: whether a green lawn is highly prized, for
example, or whether residents would favor a brown lawn to conserve
limited community water supplies.
 Figure 2.13 Projected influence of demand management strategies on
community water consumption patterns for the City of Providence, Rhode
Island (after Deb et al. 1995).

Information about social and economic systems is often qualitative

rather than quantitative, although census surveys can provide a good
overview of average household income, age and sex distribution, spending
patterns, religious preference, and similar information related to these
concerns. Most valuable of all are sources within the basin, including local
Chambers of Commerce, environmental or other (for instance, labor)
public interest groups, and service clubs such as Rotary or Kiwanis Clubs.
Community leaders usually have an excellent understanding of the
economic and attitudinal forces at play in their areas and can confirm
impressions gleaned from published sources. Even community "yellow
pages" listings can give a sense of the major industrial, commercial, and
institutional activities in the area and may yield insights into community
attitudes and priorities.
2.10 VALUED FEATURES AND ACTIVITIES

Each watershed ecosystem contains certain valued features or activities.

Often these are unremarkable to the casual observer, yet fiercely defended
by local stakeholders. It is almost impossible to give a guide to such
features, except by example.

The city of Wellington, New Zealand, is located on the South Coast of

that country's North Island. Just east of the city lies a rock formation called
Red Rocks-of little interest to European settlers but prominent in local
Maori myths and legends. In recent years quarrying activities near the site
have been closed down and relocated farther inland to protect the
appearance and cultural importance of this site.

Another example illustrates the very local nature and small scale of
some valued features. Asked to suggest an example of a valued feature, a
young student from a farming background cited his grandfather's
horseradish patch, which had been cultivated by the family for more than a
hundred years. The family was adamantly opposed to any management
intervention such as tile drainage or grassed waterways that might affect
the quality and yield of horseradish from this plot.

Other examples of valued features may be historical sites such as a

standing rock circle or a battleground, a vista or feature of great natural
beauty, or simply a geological oddity such as a rock shaped like a man's
head. These features may have little obvious importance to the casual
observer, and indeed their value may be invisible (as in the case of the
battleground), but to ignore them is to invite antagonism and public
displays of concern.

An inventory of valued features is best compiled with the assistance of

local residents, including interest groups, historical societies, community
leaders, and similar individuals. Widespread advertising in local media can
invite submissions by the wider population and ensure a good coverage of
issues of concern.
REFERENCES

Bodini, Antonio, Giovanni Giavelli, and Orazio Rossi. 1994. The

qualitative analysis of community food webs: Implications for wildlife
management and conservation. J. Envir. Management 41: 49-65.

Bruce, J. P., and R. H. Clark. 1966. Introduction to Hvdrometeorologv.

Oxford: Perga- mon Press.

Chow, Ven Te, David R. Maidment, and Larry W. Mays. 1988. Applied
Hydrology. New York: McGraw-Hill.

Deb, A. K., F. Grablutz, and P. Gadoury. 1995. Demand management

strategies for Providence Water Supply Board. In Integrated Water
Resources planning for the 21st Century, edited by M. F. Domenica.
Proceedings of the 22nd Annual Conference, Cambridge Massachusetts,
May 7-11, 1995. New York: American Society of Civil Engineers, Water
Resources Planning and Management Division.

Green, W. H., and G. A. Ampt. 1911. Studies in soil physics. The flow of
air and water through soils. J. Agr. Sci. 4: 1-24.

Holtan, H. N. 1961. A Concept for Infiltration Estimates in Watershed

Engineering. Washington, D.C.: USDA Agricultural Research Service.

Horton, R. E. 1939. Approach toward a physical interpretation of

infiltration capacity. Proc..Soil S'ci. Soc. Am. 5: 399-417.

Kouwen, N., and E. D. Soulis. 1994. Weather radar and flood forecasting.
In Current Practices in Modelin,~ the Management ofStormwater
Impacts, edited by W. James. Boca Raton, Fla.: Lewis Publishers.

Lang, R., and A. Armour. 1980. Environmental Planning Resourcehook.

Ottawa: Envi- ronnment Canada, Lands Directorate.

Maguran. A. E. 1988. Ecological Diversity and Its Measurement. London:

Croon Helm.
Meyer, A. F. 1942. Evaporation .from Lakes and Reservoirs. St. Paul:
Minnesota Resources Commission.

Novotny, V., and H. Olem. 1994. Water Quality: Prevention, Identification,

and Management of Diffuse Pollution. New York: Van Nostrand
Reinhold.

Ontario Ministry of the Environment. 1987. Technical Guidelines /or

Preparing a Pollution Control Plan. Report from Urban Drainage Policy
Implementation Committee, Technical Sub-Committee No. 2. Toronto:
Ontario Ministry of the Environment.

Philip, J. R. 1957. The theory of infiltration: I: The infiltration equation and

its solution. Soil Sci. 83(5): 345-357.

. 1969. Theory of infiltration. In Advances in Hvdroscience, Vol. 5, edited

by V. T. Chow. New York: Academic Press.

. 1993. Infiltration in one, two, and three dimensions. In Advances in

Infiltration. ASAE Publi. No. 11-83. St. Joseph, Mich.: American Society
of Agricultural Engineers.

Rohwer, C. 1931. Evaporation from Free Water Surfaces. USDA Tech.

Bull. 271. Washington, D.C.: U.S. Government Printing Office.

Schmidt, M. F., M. J. Bergman, D. R. Smith, and B. A. Cunningham. 1996.

Calibration of SWMM-EXTRAN using short-term continuous
simulation. In Advances in Modeling the Managcmenl of* Stornnvater
Impacts, edited by W. lames. Guelph, Ont.: Computational Hydraulics
International.

Schwab, Glenn O., Delmar D. Fangmeier, William J. Elliot, and Richard K.

Frevert. 1993. Soil and Water Conservation Engineering. 4th ed. New
York: John Wiley & Sons.

Simpson, E. H. 1949. Measurement of diversity. Nature 163: 688.

Tchobanoglous, G., and E. D. Schroeder. 1987. Water Qualit
Characteristics. Modeling, Modification. Reading, Mass.: Addison-
Wesley.

United States Soil Conservation Service. 1990. Engineering Field Manual.

Chapter 2. Washington. D.C.: US SCS.

Viessman, W., Jr., and G. L. Lewis. 1996. Introduction to Hydrology. 4th

ed. New York: HarperCollins College Publishers.

Waller, D. H., and Z. Novak. 1981. Pollution loading to the Great Lakes
from municipal sources in Ontario. .1. Water Pollution Control Fed.
53(3): 387.

Wanielista, M. P. 1978. Stormwater- Management Quontitm' and Quality.

Ann Arbor, Mich.: Ann Arbor Science.
3
 Problem Definition
and Scoping
The watershed inventory discussed in Chapter 2 gives an overview of
physical, chemical, and biological processes operating in a watershed
ecosystem. It seldom, however, points the way to a clearly defined
"problem" to be solved by management actions. Yet disagreements about
what problem is to be solved can create significant obstacles to effective
watershed management, even if stakeholders agree on most issues and
conditions in the system.

The process of problem definition begins with a vision of what the

watershed should be like, in terms of beneficial water uses and the quantity
and quality of water required for them. It then continues to an evaluation
of the disparity between existing and ideal conditions. The vision of an
"ideal" watershed will vary according to the basin and the people living
and working in it. But whatever the ultimate goals, this process is
fundamental in determining where conditions are less than ideal and where
restorative actions should be begun.

Often, watershed residents are acutely aware of degraded conditions,

simply because degradation often constrains beneficial uses. Section 3.1
examines some approaches to identifying water uses and use impairments
in a watershed system.

3.1 IDENTIFYING CURRENT WATER USES AND USE

IMPAIRMENTS

Virtually every watershed is of some use, or value, to human communities,

even if that use is simply one of aesthetic enjoyment or preservation of the
natural condition. A helpful list of potential water uses developed by Great
Britain's National Rivers Authority (NRA) is presented in Table 3.1.
(Source: NRA 1933a)
 Although not obvious "water users," a variety of land-based activities
are also important in affecting water resources in a watershed. Most
important among these are probably forestry, including the logging and
replanting of trees, agricultural practices such as tillage, planting,
harvesting, and drainage works, and construction activities. Each of these
activity classes has potential to affect local hydrologic processes and soil
quality and permeability. As a result, each can be important in altering
sedimentation regimes and changing the ability of upland areas to retain
moisture, thereby affecting the volume and patterns of surface and
groundwater flows. These processes, if altered, can in turn have an impact
on valued watershed features and activities such as potable water quality,
fisheries, and wildlife habitat. Their influence cannot, therefore, be ignored
in an evaluation of water uses and water-using interests in the basin.

3.1.1 Estimating Population Size

A preliminary list of water uses is an essential foundation for more detailed

analysis. The next step, that of calculating current and future demand for
those uses, is more complex.

Most estimates of water demand require an estimation of population

size and potential growth over the planning horizon. Population size is a
key variable in determining not only water demand, but demand for other
water-using activities such as electric power generation, manufacturing,
agriculture, and recreation, and the intensity of these activities within the
watershed. Forecasted population size is also a useful measure of the
potential market for water-related goods and services.

Since many water structures such as sewer systems and dams have long
useful lives, it is appropriate that estimates of population size be made for
the present, the near future, and the distant future-the time at which major
structures are likely to require replacement. Goodman (1984) observes that
our record of accuracy in long-term forecasting is not good, in part because
of changing societal patterns (such as an increasing number of women in
the workplace, leading to a higher frequency of delayed parenting and
smaller average family size). Goodman reviews several authors' methods
of predicting population growth for the city of Kingston, Jamaica (see
Figure 3.1), concluding that projections of population growth within two to
five years can vary by as much as 10%. Ten years into the future, estimates
differ by 20% to 25c1c. Fifteen years into the future, the spread in
estimates can be 50% or more, depending on the estimation method used.
(Simple extension of historical growth trends yielded an estimate that was
in the low to moderate range: this projection was ultimately adopted as
"most likely.")

Despite difficulties in achieving an accurate projection, long-term

population forecasting is an important element in watershed management
and one that can be continuously updated as more current information
becomes available.

The size of the current population is most easily estimated from census
records, usually compiled every 10 years, and/or municipal (e.g., tax)
records and similar documents. The simplest population growth model
takes the form:

P,=PO~+B-D+I-O (3.1)

PO = population at present time

 Figure 3.1 Population projections for Kingston and St. Andrew, Jamaica
(adapted from Goodman 1984).

B = number of births in planning period

D = number of deaths in planning period

I = number of individuals immigrating into area during planning

period

O = number of individuals emigrating from area during planning

period

Census information, often available for many decades, is an excellent

source of data for the individual terms in this equation. Generally
speaking, projections are more accurate for larger areas (e.g., a nation)
than for smaller ones (e.g., a county) and for overall population size rather
than for the specific characteristics (e.g., average income) of a given
population. Forecast accuracy is also affected by the age of the data used,
so it is always best to begin with the most recent data available.

Although complex models incorporating detailed county-level data on

age, sex, fertility rates, and similar details are available, these are difficult
to apply and increase the number of possible error sources in the final
population projection. This type of model may have application in some
specialized watershed planning situations but is likely to be less accurate,
and, hence, less useful to the watershed manager, than aggregated
(lumped) models of total population size.

Whether based on aggregated or disaggregated models of population

growth, estimates of population size can be used to generate forecasts
about work force availability, markets for goods and services, demand for
schools and hospitals, and similar elements important in watershed
planning. Information about the existing distribution of industry and other
activities in the basin can be used with population forecasts to predict the
intensity of land-based activities such as agriculture and construction in the
future. These predictions will, however, be only as accurate as the
information on which they are based, and they will be most useful if
regularly updated to reflect changing watershed conditions.

Despite all precautions, population forecasts are always fraught with

error. This problem may be dealt with in two ways. First, the data used in
developing population growth forecasts should, where possible, be
corroborated by other sources or methods. For example, data obtained
from census information should be checked using municipal tax records or
similar sources. And forecasts can often be compared with national or
regional trends prepared by government agencies.

A second way of guarding against error in population forecasting is to

avoid using a single, probably uncertain, estimate. Instead, the analyst can
employ a sort of sensitivity analysis, in which high, moderate, and low
estimates are made to generate a range of projected population sizes.
Although this renders the analysis somewhat more complex, there is
immense value in these best case/typical case/worst case estimates,
particularly in forecasting long-term (and therefore very uncertain)
population growth and associated water demand. As an example, the full
range of forecasts presented in Figure 3.1 could be used to generate a
range of domestic water demand estimates. A "best case" (lowest demand)
estimate could be compiled using the lowest population growth estimates.
A "worst case" estimate, using the highest population growth estimates,
would give an idea of the highest likely demand. The two together bracket
a range of possibilities within which the true condition will almost
certainly fall.

3.1.2 Estimating Water Demand

Estimates of population size and the intensity of various human activities

form the basis for estimates of water demand. Water demand is usually
expressed in two categories: water withdrawals, the total amount of water
that is removed from natural systems for human use, and consumptive
demand, the volume of water that is removed by humans but which is
"consumed" in the removal and thus unavailable for return to natural
systems.

Pearse et al. (1985) estimated that, of 38 billion m3 of water withdrawn

from Canadian sources in 1981, 54 billion m3 were used (recirculation
allowed these supplies to be "stretched" 1.4 times) and 4 billion ni 3 (about
10% of the total) were consumed. For consumptive uses, thermoelectric
power generation and agriculture were by far the largest consumers.

Water demand varies widely, depending on geographic location and

activity (see, for example, Van der Leeden et al. 1990, World Water 1989).
In general, Canada and the United States rank among the heaviest
domestic water users in the world (see Table 3.2). By contrast, some water-
poor countries such as Botswana, with its Kalahari Desert, may exhibit
domestic water consumption rates of only 8 L/person/day.

(Source: World Resources Institute 1993)

 Table 3.3 illustrates the fact that different countries use water
differently. Certainly, climate has much to do with these differences.
Warm, dry countries such as Egypt and India use a high proportion of
extracted water for irrigation to replace or supplement natural rainfall. By
contrast, cool, wet countries like Canada and Poland use a higher
proportion of water for electric power generation (especially cooling
waters) and for industrial purposes.
 Water use also implies nonwithdrawal uses such as fishing, swimming,
and navigation. Although these uses do not extract water, they may require
a certain volume of water to be present and thus must be included in any
consideration of present or future water demands.

Western European countries such as Belgium, France, Germany, and the

Netherlands have a limited resource base and a much higher population
density than much of Canada or the United States. These factors may have
influenced their modest water use (consumption rates in those countries
are typically 150-200 L/person/day) and high water prices. (Note that not
all water-poor countries are careful in their water use. In some wealthy but
water-poor Gulf nations, high water use is considered a sign of affluence.
These cultural attitudes have been serious obstacles to water conservation
efforts.)

Generally speaking, water demand is influenced by population size and

density, annual per capita income, quality of supply, and annual rainfall.
These variables can be used to make reasonably accurate predictions of
water demand for a given area. Clearly, the larger the population, the more
water consumed within the area. Warmer climates usually require higher
water use, for drinking, irrigation, livestock watering, building foundation
watering (to avoid cracking), and similar uses. Annual rainfall affects the
volume of water available for withdrawal and thus may impose limits on
allowable water use.
 Much water resources planning is, however, still based on assumptions
of steadily increasing per capita demand. Certainly, as growth proceeds in
a watershed area, total water demand may also be expected to increase.
But per capita demand may also increase through social forces and life-
style expectations. A recent residential subdivision application in the city
of Guelph, Ontario, included a proposal to require all new houses to install
in-ground sprinkler systems. The houses were to be costly "executive"
homes, and the developer's vision included uniformly green lawns across
the development. In response to citizen outcry against this requirement to
use water, Guelph City Council required the deletion of this provision in
approving the plan of subdivision. By contrast, the neighboring regional
municipality of Kitchener-Waterloo, which faces increasingly urgent water
supply shortages over the next decade, has significantly reduced water
demand by aggressive public education programs and municipal subsidies
of water-conserving devices. In Kitchener-Waterloo, a brown lawn is seen
as prestigious, a symbol of responsible citizenship and concern for the
environment.

Water consumption also varies widely in industrial, commercial, and

institutional settings. Even within the food processing industry, for
example, production of canned fruit may use only one-fifth to one-tenth of
the water required for the same quantity of beer or soft drinks. Pulp and
paper manufacturing and mineral mining and refining are among the
heaviest industrial water users, with withdrawal rates approaching 2,000
times those of some food processing industries. Table 3.4 gives some
typical wastewater discharge rates (indirectly reflecting water withdrawal
rates) for industries in the United States and Canada.

Estimating current water demand, like current population size, is a

relatively straightforward task. The simplest approach is to estimate daily
per-capita water demand (municipal water supply records are a good
source of this information) and multiply per-capita use by the population
served. This estimate can be extended into the future by using projected
population figures and assuming the same usage rate as currently exists.
 This crude estimate can be improved if separate information is available
for different user types, especially residential versus industrial,
commercial, and institutional users. Many municipalities employ water
metering to track water use by individual residences and industries so that
users can be billed for water use on a unit basis. (Both increasing block
pricing-the more you use, the more you pay per unit used-and decreasing
block pricing-the more you use, the less you pay per unit-are used in
charging metered consumers for water service; the former is clearly
preferable in terms of encouraging efficient water use.) These metered
records provide the most current and accurate basis on which to calculate
current basin water uses.

In some areas (many older residential areas in Canada, for instance)

residential water use is unlimited and unmetered, and billed on a flat-rate
basis. In England less than half of I% of the residential population has
water meters (Heathcote et al. 1996). Water metering is much more
common in water-poor areas, in newer development, and in industrial
facilities.
(Source: Eckentelder 1970; Metcalf and Eddy 1991)

Metcalf and Eddy (1991) offer an example of a water demand

calculation based on the following water supply data:

Solution: Assuming a serviced area of 147 households and an average

household ("service") size of 2.43 people based on local planning data,
daily water consumption can be calculed from the given water supply data
above:

Metcalf and Eddy's hypothetical data illustrate another important point

about water supply and demand calculations. In almost every system a
certain proportion of water is "lost" through pipe leakage and other
unaccounted system losses. In a well-maintained system, this amount will
be a small proportion of the total, perhaps less than 15%. In most older
cities the proportion is somewhat higher, 30% or more. A very high loss
rate points to an urgent need for infrastructure repair. In the preceding
example unaccounted-for system losses amount to about 24% of the total,
calculated on a daily basis as follows:

or
 It is important to use data on water sales (consumption) rather than
water production because of the problem of unaccounted-for system
losses. In other words, water production almost always exceeds water
demand, sometimes by a considerable margin. Using water production
records to forecast demand may lead to serious error in computed
estimates.

It is somewhat more difficult to estimate future water demand under

circumstances of changing usage, especially under water conservation
programs or, possibly, under an assumption of increased per-capita use.
Like population forecasting, estimation of future water demand is an
inexact science. The members of Canada's Inquiry on Federal Water Policy
(Pearse et at. 1985) calculated that total municipal withdrawals in Canada
(estimated at 4,263 million m3 in 1981) could rise to 8.458 million m3 or
fall to 3,984 million m3 by 2011, depending on whether conservation
policies (and pricing) are or are not adopted. These figures do not include
estimates of municipally drawn industrial demand, which the Inquiry
estimated at an additional 811 million m3 in 1981, nor rural residential use,
which was estimated at 347 million m3.

In the early 1970s, the United States National Academy of Sciences

examined a range of technological advances with the potential to increase
or decrease future water demand. Among the latter group were increasing
reliance on advanced nuclear, wind, and water power, improvements in
industrial cooling technology (ponds and towers), altered plant genetics to
improve drought tolerance, and movement away from surface irrigation to
subsurface irrigation and "xeriscaping"-landscaping for water
conservation. Most of these predictions are now implicit or explicit in
water management plans. For instance, recently tightened industrial
discharge regulations have prompted the adoption of newer, more water-
efficient technologies (one example is improved water reuse in the metal
mining/refining sector), and this has had a spin-off effect in reducing
overall water consumption patterns.
 Efforts to promote water conservation have increased steadily over the
past decade. To some degree these efforts have resulted in decreased water
demand, particularly among residential customers. Industrial, commercial,
and institutional water use patterns have been more difficult to change,
although some reduction in demand is also apparent in these uses. Beard
(1996), Heathcote (1993, 1995), and many others have suggested that the
abundance of water in most of Canada and the eastern United States has
made water managers complacent about water supplies. As the Ontario
MISA Advisory Committee (1991) wrote:

[In Canada, t]he traditional assumption of water is there is always

enough, it's always clean and it's always free. Today's reality is that
there is not enough, it is not always clean, and it will never again be
free.

In many watersheds, it takes a water supply crisis to focus public and

institutional attention on water demand and possible mechanisms for
demand reduction; few water-rich jurisdictions currently undertake
aggressive demand reduction programs voluntarily.

To address the problem of demand forecasting, water managers must

therefore make a number of critical planning assumptions; often these will
be specific to a particular region or even a particular watershed. These
assumptions include:

1. Estimated population growth rate over the planning horizon

2. Estimated precipitation entering the area

3. Estimated volume of surface and/or groundwater available for future

extraction

4. Nature of water-using industries, commercial establishments, and

institutions in the basin
 5. Attitudes toward water use versus conservation in the basin; i.e.,
likelihood of uncontrolled versus controlled water use

When the majority of water withdrawals in a basin are for domestic

potable water supply, forecasts can he made on a uniform per-capita basis.
If industrial, commercial, or institutional uses are important in the basin,
however, each major water demand category-often each individual major
user-must be predicted separately.

Water demand for nonwithdrawal uses such as navigation and fishing

can he expressed in similar fashion using estimates of population and
industrial growth. Table 3.5 shows some of the primary considerations in
estimating water demand for nonwithdrawal uses.

As with population forecasts, it is often wise to make several estimates

of future demand to bracket a range of possible outcomes.

3.2 IDENTIFYING CURRENT WATER USERS

(STAKEHOLDERS)
The water uses present in a basin are simply the physical expression of
human behavior patterns. Implicit in an understanding of water uses,
therefore, is the need to understand water users-their attitudes, their needs
and wants, and the priorities they set on different features and activities.
Water managers ignore users at their peril: community support is one of
the most important elements in successful implementation of management
schemes, and lack of that support can be one of the most formidable
obstacles.

Water users include a broad array of individuals and groups with an

interest in basin water resources. The particular mixture of users differs
from watershed to watershed, but in most cases includes some or all of the
entities discussed in the following paragraphs.

3.2.1 Government Agencies

Some government agencies have a direct role in water use and must
usually be included in decision making about watershed management.
These often include municipal governments and their public utilities
(because of their role in water supply and wastewater treatment and
disposal), natural resources agencies (through their interest in protection of
fish and wildlife species and habitat), public health agencies, and agencies
involved with shipping and navigation. These agencies may have a
statutory or regulatory responsibility to oversee water use within their
jurisdiction and may thus be in a position to forbid or place constraints on
certain management approaches.

A second category of government agencies includes those that oversee

and regulate water-using activities, such as the United States
Environmental Protection Agency (EPA), Environment Canada, and state
and provincial pollution control agencies. Others in this category are
electric power generation agencies and utilities and those engaged in
overseeing agriculture and urban development.

Still other agencies have a less direct, but nevertheless important,

interest in water use. These may include agencies overseeing tourism
(recreational water use and aesthetic considerations) and industrial
development (which could be limited by scarce or low-quality water
supplies).

It is important to realize that different arms of government may have

different agendas in water management, and that interagency conflict in
making decisions about water resources is therefore not uncommon. For
instance, government departments concerned with timber management and
extraction may be much less concerned with the impact of logging on
streamflows than they are with issues of safety and efficiency. Similarly, it
is not uncommon for public health agencies to recommend that domestic
water users flush their taps for several minutes before drawing water in
order to reduce concentrations of lead and other contaminants that may
have accumulated in standing water. Such a recommendation, however, is
in direct conflict with local programs to reduce domestic water use and
conserve scarce water supplies. These differences are rooted in agency
mandates and are entirely understandable. Yet they have the potential to
create major obstacles in allocation of funds-and funding responsibilities-
for watershed management activities.

3.2.2 Industrial Water Users

In some ways industrial water users are among the most straightforward to
understand. There are usually a limited number of major industries in a
basin, and these are relatively easy to list and characterize. Although the
concerns of one industry may differ from those of another, in general
industries are interested in obtaining water of adequate quality and
quantity for process and cooling water purposes, and in their regulatory or
voluntary responsibilities in discharges of wastes to receiving water.

Industrial water users are often accessible through sector-based or other

interest groups. These include sectoral associations, Chambers of
Commerce, and regional industrial societies. As a result, it is usually easy
to make a first contact with these users and to sustain that contact through
decision-making and implementation phases.

3.2.3 Commercial Shipping and Fishing Interests

In areas with larger river or lake systems, commercial shipping and fishing
interests may be important. Like land-based industries, these are often
accessible through sector associations and are in any case likely to be well
known in the watershed area. A critical consideration for shipping interests
is water levels, because if these are not maintained at adequate depth for
navigation, commercial interests can be severely impacted. In major
systems such as the Great Lakes, commercial fishing interests employing
larger vessels may also be concerned about water levels. More frequently,
however, fishing interests relate to the preservation of fish stocks and, thus,
to other water conditions, particularly water quality.

There is strong potential for commercial shipping and fishing interests

to come into conflict with those of private landowners, particularly around
the issue of water levels. High water levels may, for instance, be very
advantageous for shipping and navigation but can increase property
damage through flooding and erosion of waterfront lots. Commercial
fisheries may also come into conflict with recreational fisheries or wildlife
habitat preservation efforts. The most prominent example of this may be
the collapse of the Atlantic groundfish fisheries. Commercial fishers blame
what they believe to be a large harp seal population for preying on scarce
cod (or cod prey species) and thus preventing a faster recovery. Detailed
food web analysis demonstrates, however, that Atlantic cod comprise only
a small part of the seals' diet, and that fish populations may be more
strongly affected by physical conditions such as temperature and salinity
than by predation. The scientific evidence underlying this debate is not
clear-cut, allowing the conflict around fishing rights, sealing rights (and
protections), and employment issues to escalate in Atlantic Canada
(Hutchings and Meyers 1994).

3.2.4 Residential (Private) Water Users

By far the most numerous water users in most areas are private residents.
As a group, their primary concern is usually the quality of water for
consumption: whether the water they buy or draw privately is safe to drink
and tastes and looks good. A second concern, almost as important, is that
of water supply: whether they will be able to use the quantity of water they
believe they require, not only for drinking, washing, and cooking, but for
activities like lawn and garden watering, car washing, and foundation
watering. Many also have an interest in recreational water-based activities
such as swimming, boating, and nature enjoyment.

Residential water users are difficult to consult, in that their interests are
diverse and, as a group, they are seldom well organized or easily
accessible. Yet their property interests and strong links to valued local
features make residential users very sensitive to changes in water
management practice. As its members become better informed and more
involved in environmental issues, the public has also become less trusting
of government and more willing to challenge public policy in consultation
forums and the media. Increased opportunity for public participation over
the past 10 years has also been instrumental in teaching private citizens
that there is a difference between public information sessions and
meaningful involvement in decision making. In most areas citizens now
expect to have a substantive role in the development of water management
policy for their region. Meaningful involvement of the public in watershed
planning has tremendous advantages for the water manager as well. First,
it is an opportunity to learn from the vast range of knowledge and expertise
in the community. Second, it is helpful in identifying local issues and
values, such as those related to valued features and social systems (see
Sections 2.8 and 2.9). And it provides a forum in which water managers
can clarify and resolve conflicts before the implementation stage.

3.2.5 Public Interest Groups

Virtually all communities and watersheds have some representation by

organized "public interest" groups, sometimes termed nongovernment
organizations (NGOs). (Occasionally the acronym ENGO is used, for
environmental nongovernment organization.) It is important to note that
these groups do not necessarily reflect all or even a majority of public
opinion on a given issue or watershed system. Often, they have strong and
clearly stated agendas of their own. The World Wildlife Fund, for example,
is dedicated to the preservation of wildlife and, by extension, wildlife
habitat. This laudable goal has, however, sometimes placed the
organization in conflict with watershed residents where the goals of
preservation conflict with the needs of human communities (see, for
example, Alpert 1993; Richardson 1993).

Public interest groups thus, as the name implies, represent specific,

often narrow, interests within a community. Some have as their primary
goal a heightened public awareness of issues central to their cause.
Although not always backed by in-depth research, these organizations have
value in the watershed planning exercise simply because they highlight
sensitive or controversial issues. Other groups may have access to
impressive research resources and be in a position to produce
comprehensive reports on key themes. The expertise in these groups is,
again, a valuable adjunct to the planning process and may include
perspectives and insights that have largely been ignored by conventional
planning approaches.

It is essential to remember that however valuable in watershed planning,

representation by public interest groups does not constitute representation
by the public. A meaningful consultation regime, therefore, includes both
public interest groups and members of the general public. Chapter 4
discusses the consultation process in more detail.

3.2.6 Aboriginal Communities

Absent in some watersheds, at least as an organized group, aboriginal

communities are extremely important in others. These groups have all of
the same concerns and interests as the nonaboriginal residential water user,
but may have others as well. Critical to many aboriginal societies is the
relationship between humans and the land-in a sense, a landscape and its
resources are respected not so much for what they can yield for human use,
as for what they mean in terms of human history and values. Aboriginal
communities may also be subject to different laws and agreements,
especially treaty agreements, than would be the case for nonaboriginal
groups. Reservations, for instance, may be under the jurisdiction of the
federal government or may be self-governed. By contrast, neighboring
communities outside the reservation are more likely to be answerable first
to state and provincial regulators. Historical nonwater issues, such as
unresolved land claims, may also influence aboriginal attitudes and
decision making in a watershed.

Native self-government is an issue of increasing importance in water

management institutions. Some aboriginal groups have expressed interest
in, and jurisdiction over, the setting of water quality standards on
aboriginal lands. Although these issues are rarely fully resolved, they raise
questions about jurisdiction and harmonization of standards that may
become contentious and divisive in watershed management.

3.3 SETTING TARGETS FOR FUTURE USE

The next step in identifying water management issues within a basin is to

develop targets for each water use. A comparison between existing
conditions (as revealed by the watershed inventory; see Chapter 2) and
ideal conditions (discussed in this section) can indicate areas where
beneficial uses are currently impaired and restoration may be desirable.

This process should involve the setting of targets not only for current
uses but also for any envisioned over the planning horizon, which may be
10, 20, or 50 years into the future. An area may, for example, hope to
encourage the establishment of a particular kind of industry in the future;
the likelihood of the industry's actually settling in the basin may depend
very much on the quality and quantity of water available for its use.

In some cases the ideal conditions for two water uses will differ and
may even be incompatible, as in the conflict between navigation
authorities and private landowners over water levels. These conflicts, if not
already apparent, will likely emerge in public consultation and may be
resolved through conflict resolution processes, by mediation, or even by
unilateral government decision (see Chapter 4). At the scoping stage (see
Section 3.4), it is more important to understand where each beneficial
water use is currently, or may in the future be, constrained by existing
water conditions.
 The following sections provide an overview of how targets are set for
major water uses.

3.3.1 Water Quantity

Water quantity targets are closely linked to estimated water demand,

whether for withdrawal, consumptive, or nonwithdrawal uses. This topic
has already been covered in some detail in Section 3.1.2. The question of
whether or not the watershed has, or will have, enough water for all
intended uses is fundamental to watershed planning. Indeed, it is often a
water supply crisis that prompts concerted water management planning. In
setting targets for water quantity, the water manager must be able to
answer several key questions:

1. What proportion of reliable annual flows is currently required to meet

withdrawal demands and service nonwithdrawal uses?

2. What proportion of estimated withdrawals is consumed, that is, not

returned to the hydrologic system for reuse?

3. How are these supply-to-demand ratios likely to change in the future

in response to population growth, consumption rates (which may be
related to various factors including per capita income and conservation
policies), or climate change?

The answers to these questions will provide a preliminary assessment of

whether available flows are adequate to service existing and projected
needs, and the estimated excess (if any) of supply over demand.

3.3.2 Water Quality

As discussed in Chapter 2, the term water quality is necessarily a

subjective one, interpreted differently by different agencies and user
groups. Over the past quarter-century, the United States and Canada have
often worked together in setting water quality targets, most commonly
through the Canada-U.S. Interna tional Joint Commission (IJC), a
binational body set up under the Canada-U.S. Boundary Waters Treaty of
1909. (There is also a U.S.-Mexico Border Environment Cooperation
Commission (BECC), established in 1993, whose goals relate to the
coordination and implementation of environmental infrastructure projects
under the North American Free Trade Agreement (NAFTA). BECC is not
currently involved in the development of binational water quality
standards.)

The IJC is an organization whose mandate and structure make it almost

unique. Charged with oversight of the boundary waters, the IJC has for
decades provided a forum in which scientists and water managers from
both countries can meet and discuss technical issues of joint interest. One
such has been the setting of water quality objectives, most recently
contained in the 1987 revision of the Canada-U.S. Great Lakes Water
Quality Agreement (IJC 1987). Table 3.6 gives some of these objectives,
which were formally ratified by the two countries on November 18, 1987.

Over the past decade, however, there has been less agreement between
the two countries, and between these countries and others in the world, on
what water quality objectives are appropriate and manageable. In the
United States, for example, the Great Lakes Initiative strove to set a full
range of water quality targets, some (but not all) more stringent than those
in the Great Lakes Water Quality Agreement.

Possibly the greatest disagreement between jurisdictions exists in the

area of targets for trace substances such as heavy metals and industrial
organic compounds. Different jurisdictions have very different approaches
to this task, in part because ambient levels of the substances are so low and
toxicity is so difficult to define. One jurisdiction may define loxic as
causing an observable effect such as a tumor or lesion, while another may
include more subtle effects, such as reduced reproductive success and
changes in perception and behavior, in the definition of toxicity.

A further complication arises in that different jurisdictions employ

different bases for their target setting. Some set "maximum allowable"
levels, others set targets for ideal conditions, some for drinking, some for
the protection of aquatic biota, and so on. As a result, few standards or
targets are directly comparable; rather, each must be interpreted through its
rationale and intended application.

Finally, there is a difference between water quality standards, which are

legally binding requirements (for instance, for drinking water supply in
some systems), and water quality targets, objecla,cs, criteria, or guidelines,
which carry no weight under the law and are intended to reflect an ideal
condition. Where standards are in place, they must he met or those
responsible for meeting them will suffer sanctions. If only guidelines are
available, water managers are expected to work toward them with the
understanding that achieving targets may take some years. In some cases
multitier systems exist, with a minimum water quality level specified as a
standard and more protective (hut harder to achieve) levels designated as
objectives. Some health-related criteria specify one level for children and
another for adults; others give a short-term exposure limit and a longer, or
chronic, exposure limit.
 Table 3.7 gives some examples of the variations in water quality
objectives used by different jurisdictions.

The wide variation in water quality targets, their rationales, and the
manner in which they are applied makes it imperative that water managers
determine applicable standards and objectives for their regions of interest.
This information is most easily obtained from federal, state, or provincial
pollution control agencies.

3.3.3 Fisheries

Fisheries activity in a watershed may be commercial (in which case it is

likely an important basin employer and economic force as well),
recreational, or both. Fish stocks are also a good measure of overall
aquatic ecosystem health. Generally speaking, fisheries are assessed
through measurements of fish community composition, both in terms of
species and in terms of age and sex distribution (see also Section 2.7). This
information may then be compared against some ideal condition, including
both species and numbers of individuals. Often (but not always), the ideal
will be that which was believed to have existed before widespread human
settlement and disturbance of the natural environment. But because fish are
living organisms not simply responsive to chemical and physical change,
the challenge in fisheries management and fisheries restoration is to
achieve not only species presence (which can be maintained by artificial
stocking), but a self-sustaining fishery.

Setting targets for watershed fisheries therefore encompasses two

primary elements-desired species composition and desired stocks for each
species-but also secondary considerations in terms of the physical,
chemical, and biological conditions required to support a self-sustaining
population of each desired species. Physical conditions include aspects
such as water depth, temperature, and the availability of appropriate
spawning and nursery habitat (often degraded by excessive siltation in
areas with dense urbanization or intensive agriculture). Good chemical
condition implies an absence of compounds that are acutely or chronically
toxic to fish; biological requirements include an abundance of healthy prey
species.

Federal and state/provincial natural resource agencies are excellent

sources of information about the current status of fisheries in the region of
interest, and previously identified issues in fisheries management. The
British National Rivers Authority has published a series of "strategy"
documents (NRA 1993a), including one on fisheries throughout England
and Wales. Table 3.8 illustrates the major components of that strategy.

In addition to setting targets for fisheries restoration, the NRA Fisheries

Strategy illustrates several features essential to watershed planning. First, it
has a general goal or vision toward which to strive. (In some systems, this
goal might be more explicit-for instance, itemizing ideal species
composition and popula tion sizes.) Second, it identifies areas of current
fisheries impairment (problem definition). Third, it identifies a number of
causes of impairment, a step that will help in developing effective
management actions. And finally, it lists "key indicators," some very
specific, that it will use to monitor progress in maintaining, improving, and
developing fisheries. (The strategy also includes a discussion of the
legislative, financial, and institutional framework within which fisheries
management is practiced; such factors are discussed in more detail in
Chapters 8 and 9.)
(Source: Ontario Ministry of the Environment and Energy Accessible
Standards Information System, 1994)
(Source: NRA 1993a)
 Together, these elements provide the NRA with a management
framework that is implicitly dynamic. That is, the organization is able to
measure the status of fisheries, on a continuous basis if desirable, and
compare progress against targets. Regular feedback also allows it to
modify or refine management actions in response to changing conditions.

3.3.4 Ecology and Conservation

Virtually every watershed subject to management scrutiny has experienced

a long history of alteration through human activities. Yet many, perhaps
most, watersheds contain features that are rare, of particular scientific
interest, sensitive to disturbance, or simply valued as relics of
predisturbance times. Sometimes an area is valued simply because it is
beautiful. These elements properly form part of a management strategy, but
setting targets for them is a more challenging task than setting water
quality objectives or desirable water depths.

One commonly expressed target for conservation uses is simply "more"-

that is, to protect more watershed area from further disturbance. The
difficulty with such a goal is that its vagueness may allow it to be
overridden by more explicit, and thus apparently more urgent, watershed
goals. Yet even this simple approach is useful in establishing a benchmark
against which progress can be measured: any improvement is a step toward
the goal of "more."

An alternative approach is to divide the broad concept "conservation"

according to separate and distinct uses. These are likely to include some or
all of the following (some may overlap):

• Rare or endangered species and/or habitats

• Valued historical or archeological remains

• Areas of great natural beauty

• Habitats especially susceptible to disturbance, such as wetlands or

flood plains
 • Designated park lands

• Nature reserves

• Scientific research stations

Having established which of these uses is present and valued in the

basin, targets can be set by deciding, perhaps based on community
consensus, whether existing resources are adequate or additional resources
are needed. For example, the community may decide that an existing park
land area of 50 km2 is adequate for current uses but should ideally be
expanded to 75 km2 over the next 20 years. The target of 75 km2 then
becomes a measurable goal against which progress can he judged over
time. Similarly, a manager could set explicit targets, for instance, to protect
a given area of sensitive habitat in the future, or a certain list of historical
monuments. It is important to remember that the definition of these various
uses may change with time. What today appears an innocuous wood
structure may tomorrow be revealed as an early example of pioneer
architecture, and the definition of "wetland" may change according to
prevailing scientific views.

Several major environmental management agencies, including the

United States Fish and Wildlife Service, Environment Canada, and the
English National Rivers Authority, have implemented management
programs based on inventories (e.g., of various classes of wetland) and
have set targets to protect or enhance specified additional resources in each
category. An advantage of this approach is that resource-specific targets
lend themselves more easily to public education and outreach programs,
including the land stewardship programs currently promoted throughout
North America and Europe.

3.3.5 Navigation

Shipping and boating are essential uses in many watersheds. Indeed,

recreational boating may be one of the most widely practiced water sports
in some areas, with demand steadily increasing. Yet the terms "shipping"
and "boating" cover a vast range of activities. Setting targets for these uses
requires careful consideration of the water craft using local waterways and
the uses to which they will be put. Lund et al. (1995) discuss the problem
of developing dredging schedules for the lower Mississippi River, because
that river has "uncertain and varying channel bottom elevations (due to
sedimentation and scour), river stages, and vessel traffic of various drafts."
Commercial navigation may be constrained at a given depth-limited
crossing by:

• Uncertain draft

• Cargo

• Destination

• Each vessel's light-loading, delay, lightering, and diversion options

Any or all of these factors may contribute to the costs of delayed or

avoided navigation. For commercial shipping and navigation, the issue of
certainty of draft is as important as estimated draft itself. Water managers
must therefore set goals for commercial navigation that reflect the
expected maximum draft of relevant vessel classes, the required duration
of the navigation season, and the required certainty of draft and related
considerations such as turning area.

Typical categories of recreational use include:

• Open-water boating, including motor boating, sailing, and windsurfing

• Still-water boating, including canoeing

• White-water boating, including rafting and kayaking

Each of these uses is also constrained in some way by water conditions

such as channel depth, channel width, turbulence, availability and
condition of locks, and similar features. Here again, the requirements of
individual categories of use can form the basis for specific use-restoration
targets.
 Finally, some navigable waterways are administered by navigation
authorities whose requirements must be met in any watershed plan. Other
lakes and streams are navigable in some sense but are not overseen by a
special authority.

3.3.6 Hydroelectric Power Generation

In many systems electric power generation is one of the largest water

users, both in terms of volume of water withdrawn and in terms of the
economic value of the products of that water use. Hydroelectric power
imposes special requirements on water management: as McDonald and
Kay (1988) have said, geography may, in fact, be the most serious
limitation on the development of hydropower.

Until very recently the limitations of hydroelectric power generation

technology required a minimum head of about 7 in for a viable installation.
Over the past 15 years technologies employing heads as low as 5.2 m have
been used in the Mississippi River basin (Keevin 1982), and newer,
"micro-head" lowcapacity (18 to 125 kW) stations in China use only about
2 m as the smallest head (McDonald and Kay 1988). According to these
authors, only a few nations in the world possess the right combination of
flow and topography for hydroelectric power, but for these nations this
water use is of primary importance. In order of importance, based on
production capacity under at'eragc flows (socalled Ga„), they are China
(with 13.5% of global hydroelectric power capacity), the countries of the
former USSR (together 1 I.1%), the United States (7.2%), Zaire (6.7%r~),
Canada (5.5%), and Brazil (5.3%). If capacity is calculated based on flows
that obtain 95% of the time (so-called G95), the order of importance and
capacities are somewhat altered: Zaire (13.97%), China (10.84%),
countries of the former USSR (together 9.02%), Brazil (8.58%), Canada
(5.45%), and the United States (4.89%).y

Geography is not the only consideration in siting hydroelectric

installations. Upstream development can affect river sedimentation and
flow regimes, with detrimental effects to river hydrology and reservoir
capacity. Even without examination of the many environmental
implications of dams and inundation, it is clear that there is a need to
determine the desired capacity of any hydroelectric facility in the planning
area, the necessary average or 95% frequency flows to maintain that
capacity, and the present and future risks of changes to flows that may
result from upstream activities. These factors will he important
performance targets for any watershed plan incorporating hydroelectric
power generation.

3.3.7 Recreation

The vast range of potential recreational activities makes it difficult to

develop a generalized list of targets for ideal recreational use. Furthermore,
the requirements of one use, such as bird-watching, will differ from those
of other uses, such as hiking and picnicking.

Generally speaking, recreational water uses can he broken into several

categories, which can he helpful in determining targets for restoration.
Even though these distinctions may seem arbitrary, they provide useful
insight into the ways that water-related resources are, or could be, used.

Distance-Related Activities Activities such as hiking, mountain hiking, off-

road vehicle driving, and bicycling usually require some sort of prepared
trail or path. The costs of building and maintaining such trails is usually
proportional to their length. Similarly, the extent of opportunity for these
activities-whether an existing condition or some future ideal-can be
measured in terms of kilometers of available trail.

Area-Related Activities Other activities, such as habitat preservation or

simple nature enjoyment, require a certain area in order for them to be
effective. (The term area may be qualified by some characteristic related,
for instance, to habitat type: e.g., so many square kilometers of native dune
habitat.) Land area can therefore be used as a direct measure of the
"quantity" of such activities currently available or planned for the future.

User-Days The term recreation implies infrequent leisure or hobby use for
activities such as bird-watching or recreational fishing. The number of
available user-days can therefore he a good measure of the current state of
the resource or of some desired future conditions.

3.4 SCOPING THE PLAN

3.4.1 The Meaning of Scoping

This chapter has developed a simple framework for identifying ideal water
conditions in a watershed:

1. Identify the categories of water users in the basin.

2. Determine desired water uses.

3. Set short- and long-term goals (targets) for each use.

The challenge now is to determine which uses and watershed areas

require management intervention and which do not. This process is often
called "scoping" the watershed plan.

Scoping has two important components: boundary setting and focusing.

Boundary setting refers to the conscious limitation of the plan to a
specified geographic area, a particular time period, and, often, a small
number of species or populations of particular interest. Focusing is the
identification of issues that have overriding importance in the basin and
should therefore be treated in depth in the planning exercise.

Scoping can be a more challenging process than may be apparent, partly

because of differences in perception and priority among participants. It is
not usually difficult, for example, to generate a comprehensive list of use
impairments (areas where beneficial uses are currently constrained by
inadequate quality or quantity of water). Much more difficult is setting
priorities among the various use impairments. Which urgently demand
attention'? Which can be deferred until a later, perhaps more affluent,
time'? These are questions that must be resolved at the community level,
through appropriate consultation processes, to avoid acrimony and reduced
implementation success.
 There are three keys to successful scoping:

1. A clear definition of each problem to be solved

2. Specific goals for restoration

3. Community consensus on the importance attached to each element of

the plan.

Defining the "problem" may seem the simplest clement of these three,
but it may, in fact, be the most difficult. The simplest and perhaps best
approach may be to state problems in terms of use impairments, with
specific attention to the timing and spatial extent of those impairments.
When the problem is clearly defined, it should be relatively easy to
identify suitable indicators of impairment or restoration and thus to define
clear targets for solving the problem. The following sequence of questions
may be helpful in defining problems.

1. What are all the use impairments presently observed in the basin (or
anticipated under future growth conditions if long-term planning is
involved)? What is the geographic extent of each impairment? (Create
a long list.)

2. Why is each use considered to be impaired`? What parameters do

people use as criteria in deciding that the use is no longer viable?
What standards or objectives for the use are not currently met?

3. Which of these parameters can be measured easily? For which are

existing data available? Which can be simulated using computer
models or other methods?

4. Of the long list of use impairments developed in (1), which are the
most important to the community? (Aim for a list of two to three but
no more than live. These will become the focus of your planning
initiative.)
 5. What specific numerical targets do you wish to meet in order to
consider an impaired to be restored (e.g., total phosphorus <0.02
mg/L)'? (Be specific-identifying "bacteria" is not good enough, "fecal
coliforms" is better, but "E. c oli" is better still.)

6. During what time periods do you want these targets to apply'? All the
time'? During the summer only? During dry weather only?

7. Over what area do you want these targets to apply'? Over the full area
currently impaired or some smaller portion of that area'?

Clear and specific problem definition is also helpful in community

debate. For example, it is not uncommon to hear statements to the effect
that "industrial pollution" or "agricultural drainage" are problems in a
watershed. These terms cover a wide range of activities and a variety of
potential impacts. Typically, some of these activities do cause use
impairment-but others do not. Framing the "problem" in these terms
invites conflict without providing a practical basis for problem resolution:
it is simply not clear what needs to he fixed. On the other hand, if the
problem is described as beach closures due to elevated bacteria counts in
rural areas, few observers would argue that a problem exists or that its
source-and its solution can be found in human or animal waste
management practices.

The targets developed in Section 3.3 provide the goals toward which the
plan is directed. For each problem (impaired use), it should he possible to
state a clear goal, as well as a clear statement of the existing condition as a
benchmark of impairment. Impaired uses can rarely he restored with a few
months' activity, so this approach also creates a framework for periodic
reporting of progress toward the goal (this idea will he discussed more
fully in Section 5.5).

The final component of successful scoping is community consensus

about the relative importance of the various problems identified in the
basin. The process by which this consensus can be achieved is discussed in
more detail in Chapter 4. One approach that has been used to develop
consensus in scoping is Adaptive Environmental Assessment and
Management (AEAM). This technique is described in Section 3.4.2.

3.4.2 Examples of Scoping

Adaptive Environmental Assessment and Management (AEAM) in the

Latrobe River Basin, Australia AEAM is both a philosophical and a
methodological framework aimed at using existing knowledge of an
environmental system to develop sound management approaches for it (cf.
Holling 1978). Like many other management processes, it centers on the
development of a computer simulation model of the system and the
application of the model in assessment of alternative management
strategies.

A key feature of AEAM, consistent with the idea of community

consensus, is use of a series of structured workshops to collect the
information needed for the simulation model and to develop the model
itself. People with a range of interests attend these workshops, including
those involved in policy development, management, and technical issues,
and lay people.

Workshop participants undertake a series of structured activities to

define the following elements:

1. Model scope, including possible management actions to be simulated

2. Indicator variables of the system, to test the effectiveness of

management options

3. The required spatial scale of the model

4. The simulation time step (model time interval at which conditions

will be recalculated; e.g., daily, seasonal, annual) and overall period of
simulation (total time period to be simulated; e.g., months of April,
May, and June 1993 or 20-year period 1967-1986)
 As the process continues, subgroups are formed to examine more
detailed technical and management issues and submodels. Eventually the
various submodels are linked, and the integrated model is tested and
validated by the group. Alternative management strategies are then
evaluated, using the model in a process called "gaming" (essentially, trying
different strategies and responses to determine which is most effective).
The model can become a tool for ongoing management in the basin, and
the planning process a semicontinuous one.

Grayson et al. (1994) applied AEAM to the management of the Latrobe

catchment in Victoria, Australia. These authors note that the direct
involvement of a range of stakeholders in model development demystifies
computer simulation and ensures that outputs are understood by all. They
also note that:

The computer model is the tangible outcome of the process, but the
modeling workshops are of primary importance. They are a highly
efficient medium for the accumulation of information about the system
and require participants to focus clearly on problems and achievable
solutions from the outset. By choosing the temporal and spatial scales at
the start, it is possible to avoid the temptation to "over-mode]" those
aspects about which there is detailed information and ensure that the
complexity of sub-models is in keeping with the available data and the
questions to be asked of the final model.

Table 3.9 shows the possible management actions developed by the

Latrobe workshops and the water-quality indicator variables they would
predominantly affect. (It should be noted that in the AEAM system
management options are developed more or less at the same time as
indicator variables, through the workshop process. This approach differs
somewhat from that described in Chapter 5, on development of workable
management options, but reflects the same societal consensus as to
problems and indicators and thus achieves the same end as more structured
processes. It can be argued that the AEAM process, while less systematic
or even perhaps "scientific," has the definite advantage of developing
strong community support and tapping a wide range of information
sources in the community.)

The Latrobe AEAM process took this analysis a step further to set
priorities among the various indicator variables and to reject some as not
significant or as having insufficient data. Table 3.10 presents the results of
these decisions.

Although the Latrobe AEAM process identified a large number of

possible management actions and a wide range of "problems" (Table 3.9),
ultimately the model encompassed only total suspended solids, total
dissolved solids, total phosphorus, macroinvertebrate community index,
and index of potential fish habitat. This scope reflects the decisions of a
great many participants over a series of workshops and thus can he said to
represent community consensus. Model calibration, validation, and use in
scenario testing vary somewhat from traditional approaches, the AEAM
process employing user judgment (where the user is a group of
participants), rather than optimization of some objective function. AEAM
dictates that participants in fact "never take a hard copy of the model
output" (Grayson et al. 1994), reinforcing the view that AEAM is
essentially a consensus-building process centered on a predictive model
and its scope. As such, AEAM provides a framework within which basin
residents can meet, share information and opinions, and explore the utility
of different management approaches. In the context of this discussion,
AEAM meets all the requirements of successful scoping: a clear definition
of each problem to be solved, specific goals for restoration, and
community consensus on the importance attached to each clement of the
plan.
The Grand River Basin Water Management Study and the Don River
Water Management Strategy The Grand River and the Don River flow
through southern Ontario into Lake Erie and Lake Ontario, respectively. In
the early I980s watershed management plans were prepared for these
systems, with participation in both by many of the same agencies and
individuals (Grand River Conservation Authority 1982: Ontario Ministry
of the Environment 1987). Yet the impact of the two studies differed
greatly, in part because of differences in scoping.

The Grand River flows through southwestern Ontario, passing through

large areas of intensive agriculture and five major cities before draining
into Lake Erie. The river once supported a cold-water trout fishery, but in
recent years only less desirable warm-water species have been present. A
major impediment to the restoration of the fishery has been summertime
low dissolved oxygen levels, thought to result from excessive plant growth
encouraged by high nutrient concentrations. High phosphorus levels in
sewage treatment plant discharges were also of concern to local regulators,
who were considering limiting development in the basin until effluent
quality could he improved. A separate set of concerns centered on
persistent flooding and high flood damages in some communities along the
river.

The proponent of the Grand River Basin Water Management Study was
the Grand River Conservation Authority (GRCA), a basin management
agency that operates at the provincial level with the cooperation of local
municipalities. One of GRCA's first tasks in this study was to set up a two-
tiered advisory committee structure. The senior advisory committee
consisted of local decision makers, including local politicians and senior
government bureaucrats. A second tier of technical committees provided
specialized advice on topics ranging from hydrology to economics. These
committees were unusual at that time in that they included representation
from the general community. The advisory committees were charged with
developing a scope of work for the basin management plan, beginning
with problem definition. Over a period of months they considered a wide
range of documents from scientists, economists, and local residents. The
issues they debated were also diverse, ranging from flooding, through
degraded aesthetics (largely a result of algae blooms), to the presence of
toxic organic compounds.
 Ultimately, the advisory committees came to several important
conclusions:

1. That several of the problems were interrelated

2. That available resources would allow them to solve only three or four
of the most pressing problems.

3. That the problems most urgently in need of correction were flooding,

dissolved oxygen depletion, and elevated phosphorus concentrations
in the stream.

4. That solving these problems would likely have some indirect benefit
for other issues not directly considered in the plan

Using a purpose-built dynamic simulation model, the advisory

committees developed and tested a number of management alternatives
that would have direct benefit for flood control, would raise in-stream
oxygen levels, or would lower in-stream phosphorus. Some of these
alternatives, notably a proposal to build a major dam and inundate a large
area of farmland, were fraught with controversy. Yet through a series of
regional consultations and direct involvement of community members in
decision making, even the most controversial plans were evaluated and
screened without acrimony. In the end, the committees chose a medium-
cost alternative that included sewage works upgrades at several locations
and a variety of other channel improvements (but no dam). These
improvements were all in place by the mid-1980s, and local municipalities
continue to collaborate with GRCA as the plan and the simulation model
are continually updated to reflect changing conditions and technology.

The Don River Water Quality Management Strategy was a project of the
Ontario (Canada) Ministry of the Environment, in partnership with the
Metropolitan Toronto and Region Conservation Authority (MTRCA), a
group like GRCA charged with managing the several river basins in the
Toronto area. The problem was clearly seen to he frequent closures of
swimming beaches along the Toronto waterfront. These closures were
ordered by the Toronto Medical Officer of Health because of high
concentrations of indicator bacteria, especially fecal coliforms, and the
implicit human health risk these created. There were several sources of the
bacteria to the waterfront, primarily the number of storm and combined
sewer overflows and bypasses from the Main Toronto Sewage Treatment
Plant.

Targets for solving the problem were, however, less clear. Some people
thought that the swimming beaches ought to be open continuously
throughout the May-to-October swimming season. Others felt that it would
be enough to have beaches open within 48 hours after a rainfall event, thus
allowing time for wet-weather discharges to dissipate in the near-shore
zone. Still others felt that swimming was not an appropriate use for an
urban waterfront with a high level of industrial activity and storm-related
discharges.

The scoping process for this plan attempted to involve a range of

interested parties drawn from the six local municipalities that make up the
metropolitan Toronto area. In the early 1980s, the Ontario government,
like many bureaucracies, tended to discourage, or at least discount, citizen
involvement in public decision making. So the participants in the Don
River plan were drawn from municipal engineering and health
departments, from interested federal agencies, and from Ministry of the
Environment staff. It was expected that this group (which, as in the Grand
River study, was arranged in a two-tier committee system) was close
enough to the industrial, commercial, and residential interests in the basin
to represent them effectively.

In fact, the Don River study made several important mistakes. By

excluding nonagency representatives from the process, the government
effectively cut off access to a range of interests and solutions that might
have helped to inform the plan and clarify its scope. Although the problem
was clear, the targets for remediation were not. As a result, project staff
spent a vast amount of time and money developing and testing solutions
that might or might not be feasible. In the end, no accurate simulation
model of the basin was ever developed, and no AEAM-type qualitative
analysis was possible because of the lack of consensus on key issues. The
final plan consisted of a long and costly list of solutions that would
undeniably improve waterfront water quality. There was, however, little
consensus on who could or should pay for these, or on which could or
should be done first.

The Don River study remains to this day a paper exercise that has never
been implemented. Individual local municipalities continue to proceed. as
they always have, with their own sewer management plans. Beaches
continue to be closed throughout much of the summer. In the late 1980s,
the Don River situation was revisited through the International Joint
Commission's Remedial Action Plan (RAP) process, the Toronto
waterfront being one of the IJC's Areas of Concern under that program.
From that time until the present, the RAP process has been engaged in
redoing much of the work of the Don River study (as well as plans for
several neighboring basins). Section 4.3 describes the RAP process and
committee structure, which involves a large and active public advisory
committee and extensive public outreach activities. With the growing
awareness effected by the committee and its outreach, the RAP is
achieving a higher level of public consensus about the Don's problems and
solutions than was achieved a decade earlier and, thereby, giving
politicians public support for spending money on plan implementation.

Although the Grand and Don Rivers are separated by only a few tens of
kilometers, their basin management experience has differed widely. The
problem of scoping, and community buy-in to that scoping, lies at the heart
of these differences and explains much of their failure, or success, in
creating implementable solutions.

REFERENCES

Alpert, P. 1993. Conserving biodiversity in Cameroon. Amhio 22(1): 44-

49.

Beard, D. P. 1996. Keynote address, Conserv96 Conference: Responsible

Water Stewardship. Orlando, Florida, January 4, 1996.
Eckenfelder, W. W., Jr. 1970. Water Quality Engineering fbr Practicing
Engineers. New York: Barnes and Noble.

Goodman, A. 1984. Principles of Water Resources Planning. Englewood

Cliffs, N.J.: Prentice-Hall.

Grand River Conservation Authority. 1982. The Grand River Basin Water
Management Study. Summary Report. Cambridge, Ont.: Grand River
Conservation Authority, in conjunction with the Ontario Ministry of the
Environment.

Grayson, R. B., J. M. Dooland, and T. Blake. 1994. Application of AEAM

(Adaptive Environmental Assessment and Management) to water quality
in the Latrobe River catchment. J. Envir. Management 41: 245-258.

Heathcote, I. W. 1993. An integrated water management strategy for

Ontario: Conservation and protection for sustainable use. In
Environmental Pollution: Science, Policy and Engineering, edited by B.
Nath, L. Candela, L. Hens, and J. P. Robinson. London: European Centre
for Pollution Research, University of London.

. 1995. Conflict resolution in Ontario water resources policy. In Water

Quantity/Quality Management and Conflict Resolution, edited by A.
Dinar and E. T. Loehman. Westport, Conn.: Praeger.

Heathcote, I. W., H. R. Whiteley, and K. E. Morrison. 1996. Structuring an

Efficient Water Management System: Case studies from France, England,
and Ontario (Canada). Proceedings, American Water Works Association
Conserv96 Congress, Orlando, Florida, January 4-8, 1996. Denver:
AWWA.

Holling, C. S. 1978. Adaptive Environment Assessment and Management.

Chichester, U.K.: John Wiley & Sons.

Hutchings, J. A., and R. A. Meyers. 1994. What can be learned from the
collapse of a renewable resource? Atlantic cod, Gadus marhua, of
Newfoundland and Labrador. Canadian J. Fish. Aquat. Sci. 5I.
International Joint Commission. 1987. Great Lakes Water Quality
Agreement of 1978 (revised, as amended by Protocol signed November
18, 1987). Windsor, Ont. International Joint Commission.

Keevin, T. M. 1982. The Corps of Engineers planning process as it relates

to the assessment of the environmental impacts of low head hydro-power
development. In Energy Resources and Environment, edited by S. W.
Yuan. London: Pergamon.

Lauria, D. T., and C. H. Chiang. 1975. Models for municipal and industrial
water demand forecasting in North Carolina. Raleigh: Water Resources
Research Institute of the University of North Carolina and the North
Carolina State University.

Lund, Jay R., Vini Vannicola, David A. Moser, Samuel J. Ratick, Atul
Celly, and L. Leigh Skaggs. 1995. Risk-based decision support for
dredging the lower Mississippi River. In Integrated Water Resources
Planning for the 21st Century, edited by M. F. Domenica. Proceedings of
the 22nd Annual Conference. American Society of Civil Engineers, Water
Resources Planning and Management Division, May 7-11. New York:
American Society of Civil Engineers.

McDonald, Adrian T., and David Kay. 1988. Water Resources Issues and
Strategies. Harlow, U.K., and New York: Longman Scientific and
Technical.

Metcalf and Eddy Inc. 1991. Wastewater Engineering: Treatment,

Disposal, and Reuse. 3d ed. New York: McGraw-Hill.

MISA Advisory Committee. 1991. Water Conservation in Ontario

Municipalities: Implementing the User Pay System to Finance a Cleaner
Environment. Technical Report. Ontario: Ontario Ministry of the
Environment.

National Rivers Authority. I993a. National Rivers Authority Strategy (8-

part series encompassing water quality, water resources, flood defense,
fisheries, conservation, recreation, navigation, research and
 development). Bristol, U.K.: National Rivers Authority Corporate
Planning Branch.

National Rivers Authority. 1993b. Upper Wve Catchment Management

Plan Consultation Report. Cardiff, Wales: National Rivers Authority,
Welsh Region.

Ontario Ministry of the Environment. 1987. The Don River Wafer Quality
Management Strategy. Toronto: Ontario MOE.

Ontario Ministry of the Environment and Energy. 1994. Ontario Accessible

Standards Information System. Toronto: Ontario MOEE.

Pearse, P. H., F. Bertrand, and J. W. MacLaren. 1985. Currents of Change:

Final Report of the Inquiry on Federal Water Policy. Ottawa: Queen's
Printer.

Richardson, M. 1993. Wrestling with the preservation of the Korup rain

forest. Our Planet 5(4): 4-7.

van der Leeden, F., F. L. Troise, and D. K. Todd. 1990. The Water
Encyclopedia. 2d ed. Chelsea, Mich.: Lewis Publisher.

World Resources Institute. 1993. The 1994 Information Please

Environmental Almanac. Boston: Houghton Mifflin.

World Water 1989. Who pays what for water. World Water (Washington,
D.C.), December.
4
 The Consultation Process
4.1 THE NEED FOR PUBLIC INVOLVEMENT

A quarter of a century ago-not long in the life of regulatory institutions

public involvement was seen as unnecessary and superfluous, if not
downright invasive, by many decision-making bodies. Even a decade ago,
meaningful public involvement was rare in the development of public
policy. Yet today the public voice is heard routinely and, indeed, is
mandated by law in many jurisdictions.

What has caused this change? Seminal thinkers like Bruce Bishop
suggested as early as 1970 that water resources planning is in fact a
process of creating a program for social change (Bishop 1970). Implicit in
this notion is the idea of consensus: social change will not occur unless the
affected society agrees that change is necessary. Bishop emphasizes the
importance of focus not just on the end product of planning, but on the
planning process, "in order to produce a product which achieves a more
widely accepted solution to the wants and needs of society." Goodman
(1984) extends this idea in stating that a trend toward more open decision
making is consistent with ethical behavior in democratic societies. Bishop
believes that water development is also an instrument of social change
because of its ability to alter social and economic growth patterns.

There is no question that the trend toward openness has accelerated over
the past decade. In large part, this is due to heightened public concern
about the environment and deteriorating public trust in government
following debacles like the industrial organic contamination at Love Canal
and the pollution-related fires on the Cuyahoga River. Public opposition to
some projects has sometimes been fierce and prolonged, to the point of
forcing abandonment or, at mini mum, costly delays. Indeed, it was
Congress's concern about such delays that prompted its introduction of
action-forcing provisions in the National Environmental Policy Act of
1969, to the effect that:
 All agencies of the Federal Government shall ... identify and develop
methods and procedures ... which shall ensure that presently
unquantified environmental amenities and values may be given
appropriate consideration in decision-making, along with economic and
technical considerations (S. 102(2)(b)).

and a companion section, S. 102(2)(c):

All agencies of the Federal Government shall ... include in every

reconunenda- tion or report on proposals for legislation and other major
Federal actions significantly affecting the quality of the human
environment, a detailed statement ... on (i) the environmental impact ...
(ii) adverse environmental effects which cannot be avoided ... (iii) the
alternatives to the proposed action ... (iv) the relationship between local
short-term uses ... and ... long-term productivity, and (v) any
irreversible and irretrievable commitments of' resources.

It is unlikely that Congress foresaw the impact that these few phrases
would have on environmental management in the United States and
Canada. Together, these brief provisions now form the basis of modern
environmental assessment legislation in the United States and, indirectly,
in Canada (Gibson and Savan 1987). Furthermore, their implicit message,
that government must be able to anticipate the environmental impacts of its
actions, forced a fundamental change in the ways that environmental
information was gathered. Environmental assessment legislation generally
requires consideration not only of a proposed project but of alternatives to
that project, and thus forces a balanced and comprehensive data collection
exercise that would have been unheard of before 1969. And the collection
of such a broad database further requires consideration of a wide range of
viewpoints, which can be assayed only by extending debate outside
government, into the wider public. It can therefore be argued that with
these few sentences Congress not only changed government's preparedness
in advancing public projects, but vastly increased the transparency and
accountability of the decision-making process.

With these provisions as a backdrop, later amendments to the U.S.

Federal Water Pollution Control Act and similar legislation included clear
mention of the need for public participation in water management,
especially with respect to the setting of standards and priorities.

These changes in turn raised a need for guidance on exactly how the
public was to he involved. This need was answered by a range of
documents such as the U.S. Water Resources Council's Principles and
Guidelines (WRC 1983) and the U.S. Army Corps of Engineers' Institute
for Water Resources report Public Involvement Techniques (Creighton et
al. 1982). More important, regulatory requirements and associated
guidance documents steadily raised the pub lie's awareness of
opportunities for involvement in public policy-making. This trend, coupled
with the increasing prominence of environmental issues in the media and
on the public agenda, has continued virtually uninterrupted to the present
day.

Although opportunities for public involvement come and go, public

concern about environmental issues, perhaps especially water, remains
higher than it has ever been, while public awareness of when and how to
be heard continues to grow. Recent emphasis on government "downsizing"
and deregulation has not reversed this trend: the various interested publics
remain well informed, skeptical, and vocal.

It is perhaps unfortunate that public involvement through the 1970s and

1980s came to be viewed as adversarial: antigovernment and antiproject.
Policymakers have also expressed concern about the lengthy delays and
high costs occasioned by some public participation processes. Priscoli
(1982) notes that as adversarial positions became entrenched, participants
increasingly turned to litigation to resolve their conflicts, with the result
that extreme positions solidified, progress slowed or halted, and costs
escalated still further. In fact, Priscoli argues, this form of public
involvement subverts the goals of open planning: information sharing,
negotiation of reasonable trade-offs, and the creation of new approaches
for the development of public policy. To regulators, public involvement
may therefore appear inefficient and wasteful, and thus unnecessary.
Nothing could be further from the truth. The following section describes
how well-managed public involvement is an essential tool for determining
community consensus on community issues and, thus, for meeting
community needs.

4.2 PRINCIPLES OF CONSULTATION

4.2.1 Watershed Planning in a Pluralistic Society

In a democratic society, public decisions should reflect broad social values

and changing policy should equally reflect changing values. Implicit in this
is the simple notion of pluralism-decision making by the many, not the
few. At the watershed level, the concept is equally valid: watershed
planning, as a process to develop a program for social change, should
reflect the values of a majority of the community. In this sense, the
watershed plan should not be considered a "product" but rather a process
of achieving social change-a way of moving from the present, presumably
unacceptable, state to some future, more desirable, condition. As societal
values and priorities change, so must the plan, and the process, adapt to
accommodate them. To accomplish this end, all major viewpoints-publics-
must be heard, consensus must be built, and the views of the majority
reflected in the change that takes place. Yet for a variety of reasons relating
to time and resource demands, unwillingness to relinquish authority, or
simple ignorance of opportunities, this does not always happen.

Drawing from social and political science in his analysis of planning

processes, Bishop (1970) concluded that the ways that goals are set, and
the clarity of goals. have a direct impact on the nature of change that
occurs (Table 4.1).
 Clearly, successful achievement of change will he more difficult if the
change appears to he forced or aimless, so processes involving mutual goal
setting are to be preferred over unilateral or "bureaucratic" processes.

4.2.2 The Meaning of Involvement

Watershed planning is essentially a sequence of activities that occur over

time, each leading to the next. It is generally accepted that planning
initiatives begin with problem awareness and progress through various
information-gathering stages to a point of decision or action. Problem
awareness may come from agency representatives, from interested
professionals such as consulting engineers or planners, or from regulatory
directives, but may most frequently come from the community, whose
members are, after all, most directly affected by water resources quality
and quantity in the basin. Figure 4.1 illustrates a typical planning
sequence, showing opportunities for public involvement.

There is, arguably, a role for the public in each stage of this sequence,
yet few public involvement processes allow participation over the full
range of opportunities. Many, perhaps most, processes in fact limit public
involvement to the last two stages: stabilization of change
(implementation) and (possibly) maintenance and monitoring.

Arnstcin (1969) has described a wide range of public involvement

styles, from "tokenism" to full citizen control over decision making.
Although Arn- stein's work was less concerned with the temporal sequence
of events (and opportunities) than with the quality of opportunity at any
one stage, her hierarchy of involvement styles (Figure 4.2) illustrates the
wide range of interpretation placed on the term im'elccement.

What, then, is the ideal form of public involvement? The answer may
differ according to the perceived problem, the community and its values,
and the willingness of decision makers to delegate authority.

A variety of factors are important in designing a public participation

program. In essence, these factors relate to the concepts of democratic
decision making discussed earlier, including sharing of information,
building of trust and credibility, relationships between the public and the
decision makers, and conformance with preexisting requirements such as
laws and funding availability. Probably the most critical factors in effective
public participation are the following:

Figure 4.1 Typical sequence of public involvement in water resources

planning.
 Figure 4.2 Styles of public involvement (after Arnstein 1969).

Before the Process Begins

1. Mutual respect

2. Clearly stated expectations about:

Proposed project scope and key issues

The nature and timing of public involvement

Consultation and communication mechanisms

The level of power citizens will have in the process

Explicit proposals for selecting citizen representatives for the

planning process

3. Inclusion of all interested publics, including:

Staff of public agencies at the federal, state/provincial, regional,

and local levels of government (as appropriate for the planning
exercise)
 Elected officials at all levels of government

Private corporations and other organizations with an economic

interest in the plan

Public interest groups (that is, groups formed to represent specific

interests in the general public), including both high-profile
leaders and the more general membership of those groups

Other groups and individuals in the community, including private

citizens, legal and medical professionals, and others with a
general but not necessarily economic interest in the plan

Plan Administration

1. A single program manager and clear reporting relationships

2. Program staff who are well informed about the project and the
community, who are skilled in public involvement techniques, and
who are receptive to the ideas of community representatives

3. Specialized expertise where necessary, for instance, in conflict

resolution and facilitation

4. Adequate funding to achieve stated program goals for public

participation

Data Collection and Analysis

Joint collection and analysis of data on:

Community values, systems, and interested publics

The proposed project and any likely impacts on community

values and/or life-style

The experiences of similar projects

 The costs of proposed measures and possible funding sources for
those measures

The environmental and economic impacts of proposed measures

Communication

1. Clearly written documents and legible graphics

2. Text written in semitechnical or lay language so as to be easily

understood by all participants, and/or opportunities for technical
education during the consultation period

3. Full and unrestricted access to all data, documents, and other

materials for all participants, whether electronic or in text form

4. Well-structured advisory and consultation groups with:

A variety of skills represented (e.g., innovators, detail checkers,

encouragers, moderators, etc.)

A balance of activity between plan-related tasks and team-building

activities. (The latter, although essential for long-term group
function, should comprise no more than 40% of total activity or the
group may be unproductive.*)

5. Prompt, sensitive, and respectful review of citizen submissions and

thoughtful and timely responses from bureaucrats to public
representatives. (The goal is to create an ongoing dialogue whereby
ideas and suggestions from a variety of sources are welcomed and
thoughtfully considered, not to create an adversarial sense of "us" and
"them.")

Priscoli (1982) cites the example of a woman who attended a large

public meeting and "pour[ed] out her heart in tears over a proposed
project" in front of several hundred participants. After hearing her
submission, the chair of the meeting responded, "Thank you, ma'am. Now,
do we have any factual comments?" Priscoli goes on to say that, in fact,
the woman had just given the meeting the most important "facts" necessary
for designing successful implementation alternatives. But, he wryly adds,
"For us bureaucrats, armed with advanced engineering and scientific
degrees, this is a hard pill. To us, emotions are irrational; facts can be
separated from values."

To place this example in a different context, the woman had given the
meeting a clear and strongly expressed message about her objection to the
plan. No amount of scientific data or persuasion would be likely to change
her position on the project. Furthermore, if she felt this strongly about the
issue, it is unlikely that she was alone in her views. The watershed
planners involved would have been well advised to consider the substance
of her objections and develop alternatives that were more palatable to her
and her fellow objectors. This approach is the heart of conflict resolution,
discussed in more detail in Section 4.4.7: the respectful acknowledgment
of differing perspectives, an attempt to understand each party's true
motives, and a creative search for alternatives that accommodate all
parties' concerns.

4.3 IDENTIFYING INTERESTED PUBLICS

4.3.1 Finding Those Who Are Interested

One of the most daunting prospects in designing a public participation

scheme is to decide who should be involved. The risk of engaging too
narrow a range of participants is that some interested parties will be
excluded and will become adversaries rather than collaborators, thus
creating new obstacles in the planning process. Yet too large a group is
equally ill suited for committee work. It may also be that a large number of
possible participants will remain silent and distant from the planning
process, even though they have useful contributions to make. Creighton
(1983) notes that this silence (of potential participants) usually stems from
one of three factors:
 1. They feel adequately represented by some other group, such as a
neighborhood association or environmental public interest group.

2. They are unaware that they have a stake in the decision or somehow
view the decision as of minor importance in their lives.

3. They do not believe that they can influence the outcome of the
process.

These factors have different roots and thus different solutions. In the
first case, we can assume that such an individual is sufficiently well-
informed to make the decision not to participate; no further action may he
necessary. In the second case, the decision under discussion may or may
not have major importance to the individual; a comprehensive public
information program will ensure that people making this decision will do
so in command of all the facts. The third case, which Creighton terms
apathy, is not surprising in view of declining confidence in government.
Here again, a good public information program, focusing not only on
technical issues but on the proposed public involvement process, may allay
concerns and encourage wider participation.

Willeke (1976) and Creighton (1983) suggest three ways to identify

interested publics:

1. Sel/=idenlifirution (voluntary or to express opposition)

2. Third-porly identification (for instance, existing committees are used

to generate suggestions as to individuals and organizations who would
be appropriate participants.)

3. Staff identification; that is, identification of potential participants by

project staff based on surveys, consultation with other agencies, or
analysis of' data on community composition, associations, user
groups, geographic and demographic data, tax records, newspaper and
magazine archives, and similar sources
 All of these methods are in use to some degree. Self-identification may
be perceived as more in keeping with a spirit of openness and transparency
than identification by some (possibly biased) external group. Self-
identification may be more complex than Willeke suggests, however,
because of the difficulty in providing adequate notice to the public that
opportunities for involvement exist. This difficulty has been overcome in
various ways, each of which has advantages and disadvantages.

Open Invitation One method of soliciting participants in a public

involvement program is to issue an open invitation to the community, for
instance, to attend a public meeting. Such invitations must be widespread
to be effective and would likely include advertising in local newspapers,
television, and radio. Printed invitations can be sent to the membership of
local interest groups, Chambers of Commerce, neighborhood associations,
and other groups with potential interest in the plan. It is important to
ensure that the invitation is seen as truly open, for instance, by asking
recipients to pass along the notice to others who may be interested. There
must be no sense that only a select few have received the invitation while
others have been intentionally ignored.

Once attendees have gathered at the public meeting, the selection

process can be handled in several ways. One approach is to ask
participants to group themselves according to various constituencies-
regulators, public institutions, private corporations, nongovernment
organizations (public interest groups), and private citizens. Then ask each
group to elect a stated number of representatives. Another approach is to
hold a series of public meetings and simply allow attendance to drop off,
on the assumption that those who continue to come constitute the truly
interested public.

The advantage of the open invitation approach is that, properly

conducted, it is perceived as open, equitable, and fair. The disadvantages
are that the process of selecting individual representatives can be chaotic
and even acrimonious, and that there is no guarantee that the desired
number of representatives from each constituency will actually participate.
This method may have greatest utility in small-scale plans and close-knit
communities where the individuals and issues are well known. It becomes
more unwieldy as the potential number of participants increases and the
issues, and therefore conflicts, become more complex.

Selection byApplication Another method of soliciting representatives has

been used with success in the Remedial Action Plan (RAP) process of the
Canada-U.S. International Joint Commission. The RAP process is intended
to generate cleanup plans for each of 43 "Areas of Concern" (pollution hot
spots) in the Great Lakes system through a prescribed process of problem
identification, source identification, development of cleanup alternatives,
and selection of a preferred plan. The process, which is to be community
based, makes extensive use of public input through Public Advisory
Committees. (Technical and Scientific Advisory Committees are also part
of the process, providing detailed technical advice on issues like fisheries
biology, hydrology, and water chemistry to the agency-based RAP team.)

Representatives for the Public Advisory Committees were chosen from

a pool of applicants solicited through newspaper, television, and radio
advertising. The selection was made by the RAP team (essentially a group
of regulatory agency representatives) in each RAP area and was based on
applicants' understanding of the issues, background preparation,
availability, and other considerations. A similar approach was used with
success in the Grand River Basin Water Management Study, described in
Section 3.4.

The advantage of this approach is that it allows much better control over
the total number of participants and the number of representatives from
each interest sector. The disadvantages are minor: some administrative
effort in receiving and screening applications and the possibility that the
selection process will be viewed as biased or preferential. The latter can he
overcome if a multistake- holder committee-perhaps even a "blue ribbon"
committee of respected local citizens-conducts the screening.

Limited Invitation to Potential Opinion Leaders Any planning process

benefits from continuity of leadership and clarity of vision. Depending on
the issues under study and the community involved, it may he appropriate
to invite participation from key public and private organizations. with the
suggestion that leaders in those organizations participate in the process. It
is then up to the organization in question as to which individual is chosen
to participate, the planners having ensured the involvement of groups
whom they believe to be central in decision making.

Creighton (1983) makes the valid point that different publics may be
required throughout the planning process, some stages requiring broad
review by the widest audience possible and others having a greater need
for technical focus and continuity. As discussed earlier, this difference has
been accommodated through multitier processes. This approach is
discussed more fully in Section 4.4.

4.3.2 Finding Leaders

Ultimately, a watershed plan is intended to build public consensus about

actions that will result in an improved quality of life in the community, for
instance, through restoration of impaired water uses. As discussed in
previous sections, change can occur through mutual goal setting or can he
imposed by public agencies; in either case, it can he planned (following a
predetermined pattern or sequence) or unplanned.

Although consensus can develop gradually in the absence of discernible

leadership, there is no question that the active participation of opinion
leaders can vastly accelerate the building of a clear community agenda.
There is considerable debate in the social science literature as to what
constitutes "leadership," because people exert leadership in several ways.
This is apparent to anyone working in community activities or, indeed,
within any given organization. Those who ostensibly hold positions of
power are not always those viewed as leaders. Similarly, individuals
remote from the formal power structure may in fact be extremely
important leaders in community opinion. Some of the main categories of
leadership include the following:

Positional leaders hold elected or appointed positions at a rank that

implies influence over their constituencies. One irreverent analyst has
referred to these as "signboard" positions because the participation of
positional leaders in a process, or even a single event, is often valued more
because it implies support by the organization than because of the
individual's personal attributes or skills. Positional leaders may include
elected officials, presidents of major corporations, directors of major
public agencies, and high-profile leaders of major nongovernment
organizations.

Reputational leaders are those who are generally thought to be the key
decision makers in the community. These could be influential elected
officials in a local government, visible and respected representatives of
public agencies or private corporations (but not necessarily the most senior
or powerful members of those organizations), or any others who are
considered to have clout in the local community, for instance, through
personal wealth or family connections. Reputational leaders can include
particularly active or specialized members of nongovernment
organizations who are respected for their knowledge of key issues.

Decisional leaders have demonstrated their leadership in earlier

community decisions. While positional and reputational leaders have
potential to wield power, they may not have chosen to do so in previous
watershed planning initiatives. Decisional leaders are those who have both
the ability and-equally important-the interest to influence public opinion
on water issues, based on their involvement in similar processes in the
past.

Most processes will benefit from the involvement of leaders of all three
types. Although continuity of representation is often important, some
processes will extend over many years and, therefore, beyond the term of
office of many positional leaders and perhaps beyond the tenure of any
single reputational or decisional leader. When a change of representation
must take place, there is sometimes a formal transfer of responsibility from
one representative to another; for instance, the president of a university
may announce his or her retirement from a process, while simultaneously
introducing the incoming president as the new representative. Where
possible, this formal "handoff" is to be encouraged as a public mark of
continuity of interest and support, not just a matter of administrative
convenience.
4.3.3 Limiting the Number of Participants

Section 4.4 discusses a variety of public participation techniques and

processes, some of which can involve a large number of people. Ideally,
however, working committees should never be larger than 30 people, with
20 or fewer a much more desirable size. The reasons for this will be
obvious to anyone who has ever been involved in committee work: large
committees are difficult to administer (for instance, in arranging meetings)
and, because of the space they must occupy and the difficulty of
communicating across that space, often tend to splinter into small
discussion groups rather than to function as a single unit.

Smaller committees are also better able to build strong working

relationships and produce useful outcomes. This is particularly important
in lengthy planning processes. Deutsch (1960) and others have described
how strong teams employ a range of skills and personality types to build
and maintain the team and perform project-related tasks. It takes time for a
team to learn how to work together-how to make best use of the array of
talents available in the group and find the most appropriate style of
discussion and conflict resolution for their needs. This phenomenon may
indeed be an argument for standing committees that can provide a degree
of continuity in public decision making, rather than short-term ad hoc
committees that must he built and rebuilt each time a new issue arises.

4.3.4 The Policy Profile: Forecasting Outcomes

Coplin et at. (1983) describe a technique called policy profiling that is

useful in assessing the likely outcome of planning processes and the
probable impact of individual representatives and organizations on policy
decisions. Their technique includes the following steps:

1. Identify the issue to be decided (should be a clear and explicit

definition using terms such as "restrict," "permit," or "build," rather
than "improve" or "protect").

2. Identify the individuals, groups, and organizations that should be

included in making the decision.
3. Group together participants with the same economic interests, such as
developers. Do not group participants with similar veto power,
especially government participants. Do not group participants whose
position on the issue differs or whose power in the decision-making
process will vary significantly.

4. Structure the grouping so that the actual power distribution is

reasonably well reflected. Do not include an unreasonable proportion
of participants from one sector, thus skewing the power structure
unnaturally. If one group of participants has an immense amount of
power, divide that group into several smaller groups so that the total
power structure is accurately reflected.

5. For each actor, estimate:

Issue position: score support using numerical values such as +1,

+2, and +3 to indicate support; 0 to indicate neutrality: and -1, 2,
and -3 to indicate opposition. Larger numbers indicate more
extreme positions.

Power, expressed as a number from 0 to 3, where 0 indicates no

power or influence and 3 reflects substantial influence or veto
power.

Salience, expressed as a number from 0 to 3, where 0 indicates no

interest or concern for the issue and 3 indicates participants who
consider the issue to be of the very highest priority.

6. Calculate the weights for each participant group and the whole
system by multiplying issue position times power times salience for
each, then calculating total positive scores and total negative scores.

7. Calculate the policy profile ratio, the net weight between those
supporting and those opposing the decision under discussion. Coplin
et al. (1983) note that this ratio in a sense is a measure of the political
benefits and costs of the decision. A ratio greater than 1.00 indicates
net benefit (net community agreement) from a political and social
 point of view, while a value less than 1.00 indicates a net cost. A value
of 1.00 reflects an equal balance of benefits and costs.

An example of these authors' policy profile analysis is shown in Figure

4.3. This fairly simple analysis of participants' positions and power reveals
1.71 times as much support for the decision as opposition, and therefore
community acceptance of the proposal is predicted. The analysis also
reveals other information of interest, such as the fact that the only serious
opposition comes from the federal government and that the support of state
agencies, local environmental groups, and influential citizens is only
moderate. This gives the analyst clues that, although the decision to issue
the permit should probably be made, continued support for the measure
will depend on retention of support from these participants. The analyst
may also wish to explore the reasons for the federal government's firm
opposition to the issue, and determine whether con- Ilict resolution
(described in Section 4.4.7) may be of benefit in removing such objections.

Figure 4.3 Example of policy profile analysis (after Coplin of al. 1983).

Like many other factors in a planning process, participant interest and

position can wax and wane with time, especially if issue position is weak
(I. 0, or +I). The policy profile analysis should therefore be repeated from
time to time as the process continues. Clues as to when reassessment is
needed may include altered behavior in meetings, increased conflict
among participants who were formerly in agreement, and changes in
external forces, such as state or national politics or major community
events.

4.4 PUBLIC INVOLVEMENT TECHNIQUES AND

PROCESSES

4.4.1 Defining the Purpose of the Involvement

There are many reasons that public involvement may be desirable in a

planning process, and many times in the process when that involvement
will be useful. Yet, as discussed in Section 4.2.2, the nature of that
involvement, and thus the techniques appropriate for it, can differ
markedly from process to process. In part, this is because the planners
designing the public participation scheme have different end points in
mind. In some processes, the purpose of public involvement is simply to
give information to the public-what is often termed "public information."
No public response is sought or even desired; the agency has the facts and
has made the decision and seeks only to communicate its decision to the
public.

In other cases, the planning process may he well advanced, but a

decision point has been reached and the agency believes that input from
the public would be helpful in making that decision. The purpose of public
involvement in such an instance is for the agency to receive information, a
process that could be conducted as easily (and perhaps more thoroughly)
by nail or telephone than in a face-to-face meeting.

A third type of process intends two-way communication between the

decision makers and the public-for instance, so that each side can ask and
answer questions. An extension of this interactive process is negotiation or
conflict resolution (Section 4.4.7).

Although, generally speaking, interaction is a beneficial part of

planning, there may be a place in any planning process for simple
information giving or receiving. Whatever approach is chosen, the purpose
of the public participation may limit the techniques available. Even the
arrangement of furniture in a room may differ, depending on the goals of
the process. For example, a simple public information meeting would
likely be designed to attract as many citizens as possible and thus would
not he well suited to an informal seminar setup or even a boardroom
setting. The purpose of the event therefore dictates several important
factors:

• The number of people who should be involved

• The type of people who should be involved

• The room layout that will best accomplish the goals of the event

• The best public involvement technique(s) for the purpose of the event

The following sections discuss a range of public involvement

techniques currently in use in water management planning.

4.4.2 Techniques for Information Dissemination

As discussed previously, one form of public involvement can be termed

information giving-mechanisms by which public agencies communicate
information to the interested public. The literature abounds with
information on such techniques, probably because for many years they
were the only ways in which government formally communicated with the
public. Some of the most widely used techniques for disseminating
information to the public are given in Table 4.2.

The techniques presented in Table 4.2 are of two types: those that allow
the agency considerable control over the information the public receives,
and those that are beyond the agency's control. The former group,
including public information meetings, open houses, and various printed
materials, are useful methods of transferring technical information to the
public but can be high in cost. They are not, however, usually as effective
as the media in raising the profile of a planning exercise. For this reason,
most information dissemination programs employ a mixture of methods
and rely on the media for widespread advertising of events, key issues, and
decision points.

4.4.3 Techniques for Receiving Information

Sometimes an agency requires detailed information-feedback-from the

community, for instance, to supplement or update census data or municipal
records. Often this will require contact with as many people as possible, so
that a detailed and comprehensive database may he built. Like information
dissemination techniques, techniques for receiving information are one-
way and do not permit dialogue or negotiation. Unlike information
dissemination techniques, they tend to be private rather than public and
employ the written word rather than speech and images. Some classic
techniques for receiving information are given in Table 4.3.

4.4.4 Two-Way Communication Techniques

By far the most common public involvement techniques currently in use

provide for two-way communication. This may be informal, in the nature
of a conversation, or formal, like a citizen lawsuit against a public utility.
Two-way communication is very useful in a planning process because it
allows for the gradual modification of positions and thus may encourage
satisfactory resolution of controversial issues.
 Not all two-way techniques are alike in terms of the access they provide
to decision makers or the tools required to make them work effectively.
Some, like phone-in programs on television or radio, may simply he useful
as a sounding board for public opinion; feedback from the agency may be
limited to terse answers, and some questions may not be answered at all.
Other mechanisms, like joint field trips, may have special value beyond
their technical content, in building trust and consensus between the parties
to a decision.

Two-way techniques are of three main types: small group meetings,

large group meetings, and what might be termed detached methods,
whereby the various parties do not actually meet face to face. Table 4.4
shows a variety of two-way communication techniques suitable for use in
watershed planning.

4.4.5 Small-Group Discussion Processes

Within a small-group meeting, there are many ways to structure discussion

and achieve progress. While the most familiar of these may be so-called
parliamentary procedure, with a chair, a fixed agenda, and formal motions
that must be approved by a majority, this is by no means the only choice
available. Formal meeting structures of this type may in fact inhibit
participation by those who are shy or uncertain of appropriate procedure.
Many other methods have come into use simply because they offer more
flexibility, and are less daunting to the uninitiated, than traditional meeting
formats. Most of these are intended to open up discussion and allow
participants to be creative in identifying problems and suggesting
solutions. The challenge to the meeting organizer is, therefore, to maintain
a balance between creativity and chaos. Yet, as many observers have
pointed out, most small groups are quite effective in controlling their own
behavior, even without a formal chair. The following are some of the more
commonly used small-group discussion processes.

Brainstorming Brainstorming, a common term in everyday language,

identifies what may he the most familiar of the group processes currently
in use. Brainstorming seeks to generate a large number of creative,
innovative ideas in a short space of time. The idea is to go beyond
conventional approaches and thought patterns. To do this, the group must
decide to suspend evaluation completely, even to the point of disallowing
laughter at really farfetched proposals. The group simply suggests any
solution that comes to mind, with a facilitator recording each suggestion
on a flip chart or blackboard. The facilitator can encourage creativity by
the use of:

• Images "Here's a photograph of the earth from space. What does it

make you think of?"
 • Fantasies "In the best of all possible worlds, what solution would you
like to see here?" or "If you had all the money in the world, what
would you do about this problem'?"

• Analogies-"Think of the river as a human body. What is its heart'?

What about its hands'?"

When a long list of suggestions has been compiled, the group discusses
and rates each. A skilled facilitator can assist the group in linking ideas
from different suggestions to create new options, and can help participants
work through apparent obstacles to otherwise good suggestions.
Impractical suggestions can be discarded at this point, unless someone
cares enough about them to champion them in the discussion. Ideas that
receive no support can he quickly eliminated.

Delphi Process The Delphi process is applied, in various forms, to develop

consensus among a group of participants. It offers a quick and low-cost
way to assess group preferences using simple, intuitive techniques.

The basic approach is one of iterative questionnaires, each summarizing

and building on the results of the previous one. In one version. participants
are offered pairs of choices until all possible pairs are exhausted (pairwise
comparison). Each participant chooses one option from each pair. At the
end, all "votes" in favor of each option are totaled and the option with the
most support is adopted. A simple example of selection by pairwise
comparison is shown in Figure 4.4.

In another version, questionnaires are distributed by mail or in a

meeting setting to a group of participants, seeking their estimates,
forecasts, or preferences on key issues. Responses are then summarized
and sent to each participant for review. With all participants' responses
known, individuals are asked to estimate the probable occurrence of the
various estimates or forecasts that have been submitted, or in the case of
preferences, to rank the various possibilities. This round of answers is
again summarized and returned to the participants. Those whose answers
do not conform to the majority may be asked to supply explanations for
their positions. A final round of questionnaires may be circulated, asking
for each participant's final estimate, forecast, or preference.

Steak Fish Salad Steak Fish Salad --- Steak Steak Steak -- - Salad --- Steak
Salad

Figure 4.4 Example of pairwise comparison. In this example, the participant

is asked to indicate a preference from three menu choices: steak, fish, or
salad. The matrix summarizes all possible pairs, and the participant simply
fills in the preferred option in the open cells. The total number of "hits" is
then summed for each choice. In this case, fish is never chosen, salad is
chosen twice, and steak is chosen four times. Steak is therefore the clear
preference.

The advantages of the Delphi process are its low cost and its ability to
circumvent obstacles created by individual personalities or strong issue
positions. It has several disadvantages, among which is potential
complexity of application, especially in the second version, discussed
previously. Another disadvantage is that in applying this process
participants learn from, and may unconsciously try to accommodate to, the
positions of others, with the result that consensus may appear to have been
achieved when substantive support for the agreed position may in fact be
lacking. To be effective, Delphi processes require skilled facilitators and
careful questionnaire design, both of which may increase costs. Finally, the
group must be committed to reaching consensus and must have the power
to make commitments on behalf of their constituencies, or the process will
be simply a paper exercise.

Breakout Groups The "breakout" process, a variation on brainstorming,

uses a workshop setting in which participants are grouped into clusters of
six to eight (sometimes called breakout groups), each of which is led by a
facilitator or discussion leader. The leader poses a question (or several
questions in turn, allowing for responses after each), and participants
respond either by voicing suggestions, which are then recorded by the
leader, or by silently noting ideas on a pad in front of them. In the latter
case, each participant is then asked for a single idea (taken from his or her
notes), which is recorded on a flip chart. Several rounds of responses may
be offered before the leader moves on to the next question. The pooled
ideas are then considered one at a time by the breakout group and modified
or deleted as the group wishes. The results of the discussion are then
brought hack to the larger workshop in plenary session at the end of the
allotted time.

This process works best if the questions posed are very clear and
explicit, and if the group is strongly motivated to find solutions. Skilled
facilitation can help elicit consensus from the group, for instance, around
common elements of different possible solutions.

Breakout groups are used in many regulatory agency activities, as well

as in nonagency symposia on emerging issues or long-term planning. The
value of this process lies in the sense of teamwork created by a very small
group and the creativity that is often unleashed in such a setting. When
several breakout groups are used (the preferred format), results often differ
dramatically from group to group. Where similar positions emerge, there is
clear evidence for consensus. Where positions or solutions differ markedly,
the larger group may make use of the differences to generate new, hybrid
solutions that draw from each breakout group.

Like the Delphi process, however, breakout groups benefit from trained
facilitation, which can increase costs. When different groups develop
widely differ ent solutions, it may be difficult to find common elements
and generate useful products from the workshop, the result may instead be
confusion about which direction is best.

Values Clarification Exercise Values often underlie conflict in water

management issues. Values clarification processes attempt to make explicit
the values that underlie behavior and, thus, expectations in a planning
process. They may be most useful when opinions differ but the reasons for
these differences are not clear. Planning processes involving aboriginal
groups have found values clarification useful in resolving disputes over
land claims and resource management (e.g., fisheries) rights. The actual
process can be any of the techniques listed in this section, and may include
structured written exercises, role play, or mediated dialogue. Priscoli
(1989) gives an example of a values clarification exercise in which the
various parties are asked to mark their value positions on a figure like that
in Figure 4.5. The participants have a choice of position along a continuum
from economic development to environmental quality (along the
horizontal axis) and from low to high levels of government control (along
the vertical axis). For instance, an environmental group might take a
position close to the "environmental quality" end of the continuum and
close to the "high government control" end, reflecting its members' interest
in having strict rules to protect the environment. Business interests may
choose very different positions.

When all groups have marked their positions on the grid, value clusters
usually become apparent. These clusters can be used, as Priscoli suggests,
as a basis for understanding competing values in the watershed and for
generating creative alternatives to accommodate those values.

Figure 4.5 Example values matrix (after Priscoli 1989).

 Values clarification can be a real eye-opener in some planning
processes: participants may have thought that everyone shared common
(obvious!) values, but learn that values in fact differ widely. The technique
is therefore a useful learning tool and provides a good foundation for later
conflict resolution or negotiation. Like several of the other techniques
described in this section, values clarification usually requires trained
mediators. Also, as a relatively recent addition to the decision-making tool
kit, it may not be as well understood or as widely practiced as some other
techniques. Therefore, there is the potential for clumsy or insensitive
application, with resulting alienation of one or more parties to the process.
A more complete discussion of values in decision making is contained in
Lumsdaine and Lumsdaine (1995).

Circle Processes Circle processes are gaining favor in many difficult

negotiation situations, some of which involve aboriginal groups and strong
economic forces. The idea behind circle processes is to reduce the sense of
"us" and "them" by eliminating the role of chair or moderator. This is its
special value in strongly adversarial meetings, where any discussion leader
may be seen immediately as partisan by some members of the group.

Circle processes are intended to allow everyone who wishes to present

information or opinions, but not necessarily to resolve conflicts or arrive at
consensus. As a result, they may become an important prelude to later
processes. Circle processes are usually used with smaller groups (fewer
than 30 people) but have also been used in groups of several hundred
individuals. The larger group size is possible because the heart of the
process is a single table with five or six chairs-the rest of the room is set up
as "audience," usually with concentric rings of chairs sufficient to
accommodate the expected turnout. Aisles arranged as spokes in the
"wheel" of chairs are helpful in allowing people to move easily to and
from the central table.

The process begins with someone (for instance, a representative of the

agency who is sponsoring the meeting) calling the group to order and
announcing that the purpose of the meeting is simply for all to learn from
one another. This person should then review the rules of the process, but
have no further role in the meeting other than to call it to a close when the
allotted time has elapsed.

Typical rules for circle processes are as follows (adapted from Aggens
1983):

1. Only people seated at the central table may speak.

2. Anyone present may take an open seat at the table at any time in the
meeting.

3. Speakers may ask or answer questions, state opinions, change the

current topic of discussion, or alter the current course of discussion.

4. Speakers may interrupt other speakers or may wait for a lull in the
discussion to introduce their points.

5. Speakers may stay at the table only until they have made their points.
Once finished, a speaker is expected to leave the table so that another
person can have a chance to speak. Speakers may return to the table as
often as they like during the meeting.

6. If there are no vacant seats at the table, would-be speakers should

stand close to the table, thus signaling those at the table that a seat is
wanted.

7. Those not seated at the central table are expected to listen in silence.
They should not laugh, cheer, groan, or boo another until they are
seated at the table. Only those at the table may comment in this way.

Aggens (1983) points out that it is helpful to have one or two people
experienced with the technique to start off the discussion. Ending the
meeting is usually a matter of allowing discussion to run its course, with
the audience gradually slipping away until only the group at the table
remains.

This type of process can feel uncomfortable to those used to formal

meeting structures, but is not unusual for discussion formats in tribal
groups. It is certainly advantageous to have agency representatives present
at such a meeting, but they should be encouraged to suspend their
impatience with long-winded speakers and allow the meeting to run itself.
An implicit advantage of this technique is exactly this: that it allows the
respectful hearing-out of those who may have felt marginalized by more
formal processes.

If there is a time limit on the discussion, someone who is responsible for

meeting that limit can take a seat at the central table and remind the other
participants of the dwindling time available. When the time comes to end
the meeting, the person who opened it should approach the table and wait
for a seat. As each seat is vacated, the person should tip the chair forward,
signaling that it is no longer available for use. The audience-and the
remaining speakers-thus receive a gentle and nonthreatening reminder of
the need for closure. This process should not be rushed; each speaker
should be allowed to finish and leave the table at his or her own speed.

Circle processes, while seen as unconventional and even risky by some,

can be extremely useful in controversial processes or where there is a great
deal of pent-up emotion and thus a need for "venting." They are
inexpensive, easy to run, and usually leave participants with a good feeling
about the process and the intentions of those who sponsored the meeting.
As a result, they may be quite valuable in establishing a foundation for
later, more traditional, discussion processes.

Role Play Role play, one of the most effective techniques for teaching
conflict resolution and related topics, has real value in consultation
processes outside the classroom. In a role-play process, the participants
adopt the roles of different stakeholders in the planning situation, usually
roles that differ from their own "real life" positions. Each individual role is
usually described in a brief written handout that is seen only by the person
playing the role. In this way, the player can be given information that is not
known to others in the process. Each player may also be given props suited
to the role, and perhaps a badge stating his or her "name" and "affiliation."
The discussion leader begins the process by describing a scenario in which
the various players find themselves. The leader then sets a time limit and
describes any products or milestones expected of the group.

Role-play exercises are often most productive when they are structured
around a mock hearing or other formal process-in other words, when the
players are expected to act as they would in a real situation. This approach
provides a useful structure for group dynamics (especially a temporal
sequence of events that may be reassuring to participants) and some idea
of the type of product that is expected (for instance, a decision on a
proposed action).

Role-play exercises are usually a lot of fun (everyone likes to play

dressup) and a good icebreaker. They may not, for those same reasons, be
seen as appropriate in highly controversial situations or those in which
formal processes have been the norm. If role play is seen as beneficial in
such processes, it may be wise to use an external facilitator to introduce
and oversee the exercise. Otherwise, there is a risk that jaded agency
representatives and cynical citizens will reject the exercise as frivolous or a
waste of time.

In fact, role play is an excellent way for people to learn what it is like to
be in someone else's shoes. Citizens can learn that decision making is
perhaps more complex than they had realized, while agency
representatives can come to understand the reasons for citizens' apparently
emotional positions. Role-play exercises are, however, surprisingly time-
consuming to develop, especially if each role is well researched and
equipped. (An exercise employing 12 roles requires 12 separate
documents, one for each player, and thus 12 separate research exercises.)
Use of a professional facilitator will certainly raise costs, but will also
increase the potential of a productive outcome.

Simulation Small groups may benefit from the use of simulation games or
joint evaluation of computer simulations. Simulation games combine role
play with random forces to create a realistic sequence of events mimicking
real life. For instance, a simulation may take the form of a board game in
which various players act out roles (e.g., state government, local
government, citizens' group, industry) while various random factors (e.g.,
major storm, stock market crash, change of government) force them to take
actions affecting the community in which they live and work. Simulation
games may he proprietary-that is, available on the market-or developed by
the project team.

Many computer simulation models are also in use in watershed planning

(see Chapter 7) and can form the basis of a group process if sufficient
computer equipment is available for participants to work alone or in pairs.
In this case, participants are given some basic constraints on a system,
such as areal size and current land use, and then given free reign over
future scenarios to see the impact of alternative management approaches.
This is sometimes termed "what if?" analysis, because it allows
participants to explore the response of the system to actions within their
control: altering the characteristics of discharges, shifting the pattern of
future development, changing target water levels and water quality criteria,
and so on.

These simulations have value for several reasons. First, the very small
group size (usually fewer than live participants) allows people to get to
know each other as individuals, a good way to build trust and team feeling.
The technical content of these exercises is also important, because it
permits joint analysis of complex technical issues and the building of
mutual understanding about system response to stress and management
intervention. Simulation games are inexpensive but may not adequately
represent conditions in the real situation. Computer simulation is more
costly because it requires several computers and associated software, but
can provide more realistic insight into management issues and responses.

4.4.6 Organizational Considerations

As discussed in Section 4.2.2, public involvement can and, as many

believe, should occur throughout the planning process. Although it is
possible to obtain that involvement through a series of unrelated events,
media reports, public meetings, and similar techniques, it is often
considered advisable and efficient to establish citizen representation as a
formal part of the decision-making process.
 Typically, water management decisions are made by groups of people,
often primarily bureaucrats, through a series of meetings held regularly
over the course of a planning exercise. The beginning of the planning
process may involve more data collection and analysis (as well as
establishment of working relationships on committees), while later stages
may be more concerned with the development of consensus and the
evaluation of implementation mechanisms. One commonly used approach
for public involvement is simply to appoint public representatives on
advisory committees.

Two types of committee organization are possible, single-tier and

multitier. Simpler and shorter-term projects may be adequately served by a
single-tier committee structure, in which a single committee oversees the
entire planning exercise, including problem definition, goal setting, an
inventory of the existing system, development and testing of alternative
management strategies, and selection of a preferred approach. (The
technical work for this committee may in fact be done by staff engineers,
scientists, and economists or by external consultants. The point here is that
decision making is limited to a single committee.)

There are numerous advantages to single-tier processes, primarily

relating to administrative simplicity and the development of strong team
dynamics within a long-standing committee. In single-tier processes,
public representatives can be appointed as full members when the process
is begun and continue to serve until the planning exercise is complete.
 Figure 4.6 A typical two-tier advisory committee structure.

For more complex planning problems, larger basins, or longer planning

initiatives, however, it may he advisable to organize staff and consultative
resources into several tiers. This may be done in several ways.

The most common approach to multitier structures is to set up a senior

committee made up of positional leaders, or their delegates, and one or
more technical working committees composed of senior technical experts
from inside and outside the agency. Both committees can include
representatives from the public. A typical two-tier committee framework is
shown in Figure 4.6.

Figure 4.7 shows an alternative structure modeled on that used in the

IJC's Remedial Action Plan process. These illustrations are only intended
to show possible approaches to committee configuration. Many such
configurations are possible and will work well (indeed, in many processes
a very simple one-committee structure will serve the purposes of the
planning exercise). The advisory framework adopted by any given
planning group will depend on the goals of the planning activity, the
interests and resources of the proponent and any affiliated regulatory
agencies, and the extent to which groups outside the agencies are allowed
to participate in the process.

As discussed in Section 4.3, it is desirable to include community leaders

in the decision-making process, because these individuals are in a position
to explain and influence community opinion and thus build (or destroy)
support for change. Both committee structures illustrated in Figures 4.6
and 4.7 reflect an awareness of this point. The structure shown in Figure
4.7 emphasizes inclusion of senior bureaucrats and public officials, while
the technical committee is made up of senior technical specialists,
engineers, and scientists. In this structure, public liaison is conducted
through a designated committee, but there is no public membership on the
committees themselves.

The advisory framework shown in Figure 4.7 includes a significant

level of public involvement through a public advisory committee that is
parallel to technical and scientific advisory committees. Although this
approach gives the public a substantive voice in the process, it essentially
excludes its members from technical and scientific deliberations. In the
IJC's Remedial Action Plan sites, a typical committee structure includes a
project team (primarily, if not entirely, agency representatives) that
receives advice from a Technical Advisory Committee (sometimes also a
Scientific Advisory Committee) and a Public Advisory Committee.
Generally speaking, the Public Advisory Committee is excluded from
detailed discussions of technical matters (the province of the Technical and
Scientific Advisory Committees) and is more concerned with the review of
documents and proposals and the design of an implementation strategy.
The Technical and Scientific Advisory Committees may have nonagency
representatives, but are unlikely to include members of the general public.
In this structure the exclusion of the public, and public interest groups,
from technical decision making may limit the building of expertise in the
public and impair the building of solid public-agency relationships.

Figure 4.7 Alternative committee structure (modeled on the IJC Remedial

Action Plan process).
4.4.7 Conflict Resolution (Alternative Dispute Resolution)

Unlike negotiation, which implies distinct and entrenched bargaining

positions and the gradual yielding of concessions on each side, conflict
resolution, sometimes known as alternative dispute resolution (ADR),
seeks the common ground in a dispute. This technique has been in use for
many years but has recently gained prominence through its application in
high-profile cases such as native land claims, acrimonious labor disputes,
and timber management issues. Increasingly, conflict resolution and
mediation are being offered as formal alternatives to tribunal processes, for
instance, under the 1990 Administrative Dispute Resolution Act (U.S.) and
the 1994 Canadian Environmental Assessment Act. Conflict resolution is
based on several key assumptions:

1. That there is agreement on the vast majority (90% or more) of issues

in any dispute. Properly facilitated conflict resolution exercises will
quickly reveal this fact.

2. That each party has a true interest, or stake, in the issue. However,
this interest is not necessarily, or even usually, stated outright.

3. That each party in a dispute has a "fallback" position-the position that

will be taken if consensus on the issue is not reached. Again, this
position, which can be interpreted as the party's "threat" in the process,
is seldom stated outright.

4. That each party's true interest is valid and important in the process
and should be accommodated by whatever solution is developed by
the group.

The conflict resolution process is focused on learning each party's true

interest and likely fallback position and developing innovative solutions
that accommodate the true interests of the various parties. It is important to
realize that the positions stated at the outset are likely to change as the
process continues. This differs markedly from traditional negotiation, in
which positions are instead very likely to become more entrenched and
intractable with time. A typical conflict resolution process works like this:
I. Representatives from a range of constituencies-agencies, public
interest groups, industries, or whatever groups are involved in the
dispute are invited to attend a "retreat" over one or more days. Those
invited should have the power to make decisions on behalf of their
constituencies.

2. The meeting begins with introductory remarks from the sponsor,

including a statement to the effect that consensus is the desired
outcome of the process, that the process attempts to value and
accommodate all major viewpoints, and that everyone will have an
equal opportunity to speak.

3. Each of the participants in turn is invited to speak about the reasons

for their participation, their views on the issue, and the reasons for
those views. (Time limits may be set on this "venting" process if the
group is comfortable with that.) This step serves to clear the air and
ensure that everyone's views are known to the group.

4. The facilitator then emphasizes the importance of full understanding

in the process and invites the participants to talk to one another,
learning more about them as individuals and about their views and
values. The proviso in this step is that discussions may be held only on
a one-to-one basis. Each participant is encouraged to learn from each
other participant. The facilitator may set a time limit, probably several
hours, on this activity.

5. After this initial icebreaking and information-gathering step, the

facilitator recalls the group into a plenary session. The mood of this
session is usually much more relaxed and friendly than that of the
initial meeting, reflecting the participants' new understanding of their
colleagues and a building sense of team. The facilitator invites the
group to speak, again one at a time, on any insights they have had or
any questions that remain in their minds. The participants should,
however, be discouraged from talking about any "deals" that have
been struck (as will inevitably happen) between stakeholder groups.
(The reason for this is to avoid antagonizing those who have not yet
formed firm opinions as to possible outcomes.)
 6. The facilitator then allows the group plenty of time for free
discussion, this time allowing larger groups to converse. In this step,
what typically happens is that small clusters of people form, and as the
discussion-and consensus-begins to develop, the clusters gradually
coalesce into larger and larger groups until the whole group is
together.

7. Even at this encouraging stage, consensus may be fragile. The group

should choose a quiet and diplomatic individual as recorder. (A loud
or forceful personality may provoke argument among the group or
may be seen to be biased in representing the views of the group.) The
group then calls out the points of the consensus for the recorder to
write on a flip chart or blackboard. This list is systematically reviewed
and edited by the group until the wording is acceptable to everyone.

8. As a final step, each participant formally signs the agreed-upon

position to indicate full support.

Conflict resolution often requires more than a single day (although

surprising progress can he made in a day or less, given that the group is in
agreement on the goal of consensus) and may in fact continue over a
period of weeks or months. This time may be necessary to allow the
various parties to check back with their constituencies, to obtain necessary
funding, or simply to think. Where positions have become entrenched
through past processes, a skilled facilitator is needed to encourage the
parties to see beyond their stated positions to other possibilities still
consistent with their true interests.

Conflict resolution is the basis of the Harvard Negotiation Project's

techniques, described by Fisher and Ury (1981), and is the focus of a
growing number of publications, including the newsletters Consensus,
published by the Harvard Law School's Public Disputes Network, and
Resolve, published by the World Wildlife Fund's RESOLVE Center for
Environmental Dispute Resolution. (It should be noted, however, that the
term conflict resolution is also used in the sense of international, including
armed, conflict. The term alternative dispute resolution or simply dispute
resolution may he helpful in separating one group of sources from
another.)

REFERENCES

Aggens, Lorenz. 1983. The Samoan circle: a small group process for
discussing controversial subjects. In: Public Inrolrentent Techniques: A
Reader of Ten Years' Experience at the Institute for Water Resources,
edited by J. L. Creighton, J. Delli Priscoli, and C. Mark Dunning. IWR
Research Report 82-R1. Fort Belvoir, Va.: Institute for Water Resources,
U.S. Army Corps of Engineers.

Arnstein, Sherry. 1969. A ladder of citizen participation. .1. American

Institute of Planners 35: 216-224.

Bennis, Warren G. 1961. A typology of change process. In The Planning

(off Change, edited by W. G. Bennis, K. D. Bcnne, and R. Chin. New
York: Holt, Rhinehart, and Winston.

Bishop, Bruce. 1970. Public Participation in Planning: A Multi-Media

Course. IWR Report 70-7. Fort Belvoir, Va.: U.S. Army Corps of
Engineers Institute for Water Resources.

Coplin, William D., Donald J. McMaster, and Michael K. O'Leary. 1983.

Creating a policy profile. In Public Int'oh'ement Techniques: A Reader of
Ten Years Experience at the Institute for Water Resources, edited by J. L.
Creighton, J. Delli Priscoli, and C. Mark Dunning. IWR Research Report
82-RI. Fort Belvoir, Va.: Institute for Water Resources, U.S. Army Corps
of Engineers.

Creighton, James L. 1983. Identifying publics/staff identification

techniques. In Public Involvement Techniques: A Reader of Ten Years
Experience at the Institute for Water Resources, edited by J. L. Creighton,
J. Delli Priscoli, and C. Mark Dunning. IWR Research Report 82-RI. Fort
Belvoir, Va.: Institute for Water Resources, U.S. Army Corps of
Engineers.
Creighton, James, Jerry Delli Priscoli, and C. Mark Dunning. 1982. Public
Involvement Techniques: A Reader of Ten Years Experience tit the
Institute for Water Resources. U.S. Army Corps of Engineers Institute for
Water Resources Research Report 82-R 1. Fort Belvoir, Va.: US Army
Corps of Engineers.

Deutsch, A. 1960. The effects of cooperation upon group process. In Group

Dynamics-Research and Thearv, edited by D. Cartwright and A. Zander.
Evanston, Ill.: Row Peterson.

Fisher, Roger, and William Ury. 1981. Getting to Yes: Negotiating

Agreement Without Giving In. New York: Houghton Mifflin.

Gibson, R., and B. Savan. 1987. Environmental Assessment in Ontario.

Toronto: Canadian Institute for Environmental Law and Policy.

Goodman, Alvin S. 1984. Principles of Water Resources Planning.

Englewood Cliffs, N.J.: Prentice-Hall.

Herrmann, Ned. 1995. The Creative Brain. Lake Lure, N.C.: Brain Books,
The Ned Herrmann Group.

Lumsdaine, Edward, and Monika Lumsdaine. 1995. Creative Problem

Solving: Thinking Skills for a Changing World. New York: McGraw-Hill.

Priscoli, Jerry Delli. 1982. The enduring myths of public involvement.

Citizen Participation 3(4): 5-7.

1989. Public involvement, conflict management: Means to EQ and social

objectives. J. Water Resources Planning and Management 115(l): 31-42.

U.S. Water Resources Council. 1983. Economic (lilt/ Elivironmental

Principles and Guidelines for Water and Related Land Resources
Implementation Studies. Washington, D.C.: U.S. Water Resources
Council.

Willeke, Gene E. 1976. Identifying the public in water resources planning.

J. Water Resources Planning and Management Division (American
Society of Civil Engineers) l02(WRI) (April).
S
 Developing Workable
Management Options
Chapter 2 described the process of developing a watershed inventory:
learning how the watershed systems operate and interact, and developing a
sense of the components and processes that are most influential and highly
valued. Chapter 3 carried that analysis through to decisions about problem
definition and the "scoping" of a watershed management plan. These two
chapters therefore form the foundation of a watershed management
strategy. Along with Chapter 4, on public involvement, they provide the
basis for focused problem solving in the management area.

This chapter describes the next stage in developing a watershed

management strategy: developing workable management options-actions
that cause intervention in the existing system and result in movement
toward a desired target. Management options may include measures that
use technology or structures to change existing conditions (sometimes
called "structural" measures) and those that rely on changes in human
behavior or management practices (sometimes called "nonstructural"
measures). Generally speaking, structural measures are easier to
implement, although considerably more costly, than nonstructural
measures. Nonstructural measures, for instance, encouragement of crop
rotation or contour plowing, tend to be inexpensive but difficult to
implement, because they ask people to change entrenched behavior
patterns. Most watershed management schemes include a mixture of
structural and nonstructural measures geared to the control of different
problems and sources.

5.1 IDENTIFYING THE SOURCES

The sources of flow or pollutants to a watercourse are of four types:

1. Dry-weather point sources-discharges that occur more or less contin

uously and that enter the watercourse from a single, distinct point,
 usually a pipe or diffuser. Examples of dry-weather point sources are
sewage treatment plant effluents and treated industrial wastewaters.
The name stems from the fact that these are usually the major sources
operating during dry weather conditions.

2. Wet-weather point sources-urban discharges of stormwater runoff and

combined sewer overflows from a collection/distribution system
outflow point. Although urban stormwater runoff is diffuse in origin,
most cities now have efficient runoff collection systems and discharge
collected stormwater from pipe outlets into receiving waters. These
sources operate only during rainfall events and for a short period
following rainfall and are generally negligible during dry weather.
They are distinguished from wet-weather diffuse sources, such as
overland runoff, because of their suitability for end-of-pipe treatment
and control technologies.

3. Wet-weather diffuse (nonpoint) sources-diffuse drainage of rainwater

from urban or agricultural lands. These sources require control over a
large area, not just a single point, and may have important
implications for hank and channel stability as well as flow and quality.

4. Internal (in-stream) sources-chemical, physical, or biological sources

internal to the watercourse, such as sediment resuspension (which can
elevate suspended solids concentrations) or decaying biomass (which
can affect the levels of nutrients, solids, oxygen-demanding materials,
and other water quality constituents). These sources can be difficult to
find, quantify, and control and are therefore often omitted from
watershed planning initiatives.

Chapter 3 gave some general guidance as to the scope of the

management plan-for instance, deciding, as in the case of the Grand River
study (Section 3.4), to focus on reducing flood frequency and magnitude
and on improving instream dissolved oxygen levels. The next logical step
is to identify all sources within the basin that contribute to the problem of
interest. To avoid unnecessary effort and expenditure of resources, it is
important to focus only on sources directly related to the problem of
interest and not allow the discussion to drift into other areas of possible
interest. In the Grand River study, these might have included control of
persistent toxic substances or protection of wildlife habitat.

The scoping exercise described in Section 3.4 must provide clear

guidance, and a restricted focus, for subsequent planning activities, or
limited resources will quickly he dissipated and little progress will be
made. There must be clear agreement among the parties to the plan as to
the scope of the plan and the problems that are to be solved. If several
problems are defined, then the relative importance of each must also be
described.

If there is not clear consensus on the scope and focus of the plan-if the
problems to he solved are not clearly and explicitly defined-then the
decision making group should take time to reexamine issues and priorities
and attempt to develop a clearer focus. The techniques described in
Sections 4.4.5 and 4.4.7 may be helpful in developing this consensus.
Indeed, disagreement about project goals at this stage may suggest that
exposition of values and interests has not been complete or effective and
that there are hidden interests and agendas remaining to be discussed.

Once agreement on the focus of the plan has been achieved, the
description of sources becomes a fairly routine task. The following steps
may be used to develop a reasonably comprehensive list.

1. List the land uses in the watershed, ranking them in order of land
area.

2. For each land use, estimate the proportion of land surface that is
permeable (e.g., cropland, park land, forests, and woodlots) versus
impermeable (e.g., roads, parking lots, roofing).

3. From available sewer maps, determine which areas of the basin are
drained by storm and sanitary sewers, and the location of discharge
points for each system.

4. Identify the known dry-weather (continuous) point sources to the

system. These should include known discharges from any industrial or
 commercial operations and from any municipal sewage treatment
facilities.

5. Identify the known wet-weather (intermittent) point sources to the

system, including stormwater discharge points and combined sewer
overflows. Sewage treatment plant bypass may also occur in wet
weather and should be included in this category. Estimate the volume
and frequency of discharges from these sources.

6. Estimate the pollutant-generation potential of land uses that have a

high proportion of pervious surfaces (see, for example, Table 2.6). For
the most part, these will be diffuse sources active only, or mainly, in
wet weather.

7. Estimate the potential for internal sources or processes to generate or

remove flow or pollutants to/from the system.

The identification of these factors will often provide clear signals as to

dominant sources and even appropriate correction measures. Nevertheless,
it is useful to develop consensus about the importance of each source as
guidance for later priority setting; in other words, it is helpful to determine
which sources are likely to contribute a large portion of the total load, and
which a relatively small portion. This is best done with the advice of
agency personnel and community representatives through joint analysis of
available data.

Pollutant Loadings Versus Concentration This is perhaps an appropriate

point to introduce the concept of pollutant "load," as opposed to
concentration. Both concepts are of interest in water management, but the
terms differ in meaning and implications.

The term loading refers to the total mass of a substance, expressed

either as mass alone (e.g., kilograms or tonnes) or as mass per unit time
(e.g., tonnes per year). Concentration refers to the mass of a substance
contained in a given volume of water and is expressed as mass per unit of
water volume (e.g., milligrams per liter).
 Historically, most water quality management programs focused on
controlling pollutant concentration. There was good reason for this, in that
concentration is probably the most critical consideration in assessing
potential impact on an aquatic system. A species of fish may, for instance,
tolerate a certain concentration of ammonia in water, but die at high
concentrations. A regulatory agency may specify both an average
concentration allowed in a discharge and a maximum allowable
concentration, to avoid toxic effects on aquatic biota. Concentration
therefore gives an intuitive measure of potential impact on aquatic species.

In recent years the notion of limiting pollutant loadings has become

more important in water management. This probably reflects a growing
understanding that some pollutants, especially persistent toxic substances
like DDT and dioxin, are not readily assimilated and degraded by aquatic
systems, but accumulate over time in sediments and biological tissue. It
also reflects an increasing interest in controlling diffuse sources of
pollution through land-based activities, controls whose in-stream impact
cannot readily be predicted. For instance, conservation tillage is believed
to reduce the loss of solids and nutrients from agricultural land. But
although it is possible to estimate the total pollutant loading reduction
afforded by such a measure, it is much more difficult to predict changes in
in-stream water quality that result from activities undertaken over a large
area. By contrast, industrial effluent controls usually result in easily
measured changes in pollutant concentration at the "end of the pipe," so
both concentration and loading reductions can be estimated with accuracy.

5.2 CREATING A LONG LIST OF MANAGEMENT

OPTIONS

Once the sources have been identified and their relative importance is
known, it is possible to create a "long list"-an unedited and comprehensive
list-of all possible solutions to the problem of interest. Some of the small-
group techniques discussed in Section 4.4.5 may be helpful in this regard.
Brainstorming, for example, is frequently used to generate a list of possible
solutions. Lumsdaine and Lumsdaine (1995) make the point that creative
thinking is often limited by habits, emotions, and false assumptions. They
describe a wide range of techniques for opening up what they refer to as
the "artist's mind-set," as compared with the "detective's mind-set," which
is used in defining the problem to be solved, and the "explorer's mind-set,"
which is used to place the problem in context.

Lumsdaine and Lumsdaine's emphasis on creativity is important at this

ideageneration stage of problem solving. The challenge for the decision-
making team is to overcome preconceptions about workable options and
create, in the literal sense, a broad and imaginative range of solutions for
further investigation. Creativity is central at this stage, simply because
subsequent steps emphasize screening and reduction of options rather than
creation of new approaches. (The term long list is used to differentiate this
group of options from a "short list" of options that will be subjected to
detailed analysis and scrutiny.)

Various sources may be used to generate options for the long list. These
include experts drawn from agencies and the local community and from
more distant places that have undergone similar planning exercises. The
scholarly (peer-reviewed) and "gray" literature (e.g., government reports)
can be a rich source of inspiration and practical experience. Novel or even
apparently ridiculous proposals should not be discouraged at this stage;
indeed, it is perhaps the most appropriate time for "what if'?" and "in the
best of all possible worlds ..." speculation. This is also an important time to
clarify the values and priorities of the decision-making group and to ensure
that all values are encompassed by the evolving list of options.

Lumsdaine and Lumsdaine (1995) provide an excellent annotated

bibliography of texts on idea generation and creative problem solving. Key
among these are LeBoeuf (1980), Osborn (1963), Prince (1970), and Van
Gundy (1984). The reader seeking detailed guidance on idea generation is
referred to these and other references in Lumsdaine and Lumsdaine's
(1995) bibliography.

5.3 TYPES OF OPTIONS

The long list of potential management actions can, in theory, contain four
types of practice: the "do nothing" or status quo option; structures and
other built technologies (structural measures); vegetative approaches; and
so-called best management practices-nonstructural options. Each of these
categories is described in more detail in the following sections. Most long
lists include at least three of the four categories. Individual measures are
not described in this section and are, indeed, beyond the scope of this
work. The sources listed in Tables 5.1 through 5.4 provide additional detail
on a range of individual measures.

5.3.1 The Option of Doing Nothing

In virtually every management planning exercise, one management

strategy is to keep on doing what you are currently doing; in other words,
to maintain the status quo. This option has much to recommend it. Often,
although not always, it is cheap: no structures need be built, no education
programs funded, no additional fees or expenses to pay. It is easy for
decision makers and lay people to understand. And it provides a useful
basis of comparison with other intervention-based options. In a sense, the
"do nothing" option simply furthers the existing state of the watershed. If
selected as the desired approach in the planning exercise, it speaks to
societal consensus that change is not necessary or desired at the present
time.

In some systems, for instance, those with contaminated aquatic

sediments from historical discharges, just leaving the system alone will
result in some improvement through natural processes. In most systems, it
will be useful to model the "do nothing" option as a base case, or
foundation, against which other management scenarios can be compared.

5.3.2 Structural Measures

In most systems it is possible to build something, such as a treatment plant,

a stormwater detention pond, or a grassed waterway, that will result in
some water quality or flow improvement.
 Structural solutions are well described in the literature and, indeed, in
every basic environmental engineering text. They may include both end-
of-pipe solutions, which seek to treat or remove pollution that has already
occurred, and preventive options, which are implemented to prevent or
reduce the creation of waste within a process. Table 5.I shows some typical
structural solutions for point sources of water pollution.

Table 5.2 gives some examples of structural measures for the control of
nonpoint-source (diffuse) pollution.

5.3.3 Vegetative Practices

Sometimes included among nonstructural ("management") options,

practices that alter the vegetative cover of the land are defined by Novotny
and O1cm (1994) as a separate class of options and are treated as such
here. These include measures that change the extent, nature, and/or timing
of vegetative cover and therefore change the rate and quality of water
flowing over the land surface. Vegetative measures are almost exclusively
used in the control of non-point sources of pollution. They are particularly
important in agricultural applications because they are readily controlled
by the farm operator, are often low in cost, and can provide secondary
benefits in terms of crop production. Vegetative measures are increasingly
used in urban settings as well, particularly with the recent trend toward
"bioengineering"-the use of grading and vegetation in place of older
structural solutions such as riprap and gabion baskets in stream bank
stabilization. Table 5.3 gives some examples of common vegetative
measures.

5.3.4 Nonstructural Options (Best Management Practices)

In some systems, nonstructural measures-sometimes called "best

management practices" to differentiate them from best management
technologies may be as or more effective than structural measures. They
are, however, often harder to implement because they require people to
change the way they behave. A classic example of a low-cost, highly
effective, but nevertheless often unsuccessful management measure is a
"stoop and scoop" ordinance or bylaw for the control of pet excrement in
urban systems. Despite the obvious efficacy of this measure, it has been
difficult to implement in many cities because of public resistance to
change and because municipalities simply cannot enforce such ordinances
effectively. Successful implementation has occurred where the ordinance
has been combined with a comprehensive public education program.
 In agricultural systems, best management practices may include
measures such as conservation tillage and contour plowing. Here again, the
water manager may meet with resistance from the farm operator, even
though the measure is low in cost and likely to be effective. Family
traditions, for example, may dictate that a particular tract of land is plowed
in a given way. More commonly, farmers express an understandable
reluctance to risk lower productivity by changing tried-and-true farming
methods. Sometimes financial incentives such as grants, loans, or tax relief
can be helpful in encouraging a shift toward more protective practices.
Table 5.4 gives some examples of nonstructural, or "management,"
practices commonly used in agricultural and urban systems.
5.4 DEVELOPING MUTUALLY EXCLUSIVE
MANAGEMENT ALTERNATIVES

In most cases the management goals of a watershed plan can be

accomplished in many different ways. In one situation a "high-tech"
solution, employing a range of built technologies, may work best; in
another, simply changing operating practices will be the preferred
approach. But how are these strategies developed in the first place`?
Although management strategies are, in practice, developed in many ways,
often ad hoc, careful and systematic formulation of alternatives can vastly
simplify subsequent analysis and improve decision clarity.

The principles of developing alternatives are straightforward:

1. All possible management options must be listed separately (these can

be considered separate and independent "projects" in that each would
have a unique budget, schedule, and associated considerations).

(Sources: Novotny and Olem 1994; Switzer-Howse 1982; Waterhouse 1982)

2. All possible combinations of options are listed (for the purposes of

this discussion, we can consider that these will he the feasible
 combinations of options-the "short list" to be developed through
screening procedures discussed in Chapter 6).

3. All options (and comhinations of options) are evaluated on the basis

of a common planning period, discount rate, and basis of comparison
(evaluation criteria).

This framework ensures that all options are enumerated and that all
management strategies, whether comprising one or several individual
actions, are mutually exclusive. This is important, so that acceptance of
one management strategy clearly precludes acceptance of another.

An example may illustrate this important point. A hypothetical river

basin currently experiences heavy loadings of sediment and nutrients,
primarily from agricultural activities. Nutrient enrichment has, in turn, led
to excessive aquatic plant growth (cutrophication), with associated
nighttime oxygen depletion and fish kills. The following (simplified)
individual management actions are possible:

1. Do nothing.

2. Construct riparian buffer strips to capture sediment and attached

nutrients before they enter the stream.
(Sources: Novotny and Olem 1994; Switzer-Howse 1982; Waterhouse 1982)

3. Construct livestock exclusion fencing to prevent cattle from entering

the stream and trampling stream banks.

4. Encourage conservation tillage in the basin.

These individual options may then be combined in various ways to

create several new management strategies:

5. Construct riparian buffer strips and livestock exclusion fencing.

6. Construct riparian buffer strips and encourage conservation tillage.

7. Construct livestock exclusion fencing and encourage conservation

tillage.

8. Construct riparian buffer strips and livestock exclusion fencing and

encourage conservation tillage.
 This list of options is exhaustive (of all possible combinations of
options), and the alternatives listed are mutually exclusive-that is,
acceptance of one automatically precludes acceptance of another. The fact
of this complete list thus allows systematic and objective comparison of
each possible management strategy to determine which is the preferred
alternative. Table 5.5 illustrates the process graphically.

This example does not include the case of alternatives that are not
independent-that are in some fashion interdependent. An example of a
dependent alternative is provided by a sewage treatment plant upgrading
scheme. One alternative is, of course, to do nothing to decide not to
upgrade. Another option is to upgrade from primary treatment to secondary
treatment. A third option is to upgrade from primary treatment to tertiary
(advanced) treatment. Although the upgrade-to-secondary option is clearly
independent of the other two, the upgrade-to-tertiary option implies
installation of secondary treatment as well: the upgrade-to-tertiary option
is therefore dependent on the upgradeto-secondary option.

Dependencies between management alternatives are not common but do

arise in the development of remedial strategies. They must he treated
carefully, so as to avoid double counting or, as is more likely to be the
case, omission of linked options in evaluation.

Further discussion of this topic is available in most introductory

engineering economics texts (which are generally concerned with the
economic analysis and comparison of engineering projects) and in major
works such as Bussey (1978).

5.5 EVALUATION CONSTRAINTS AND CRITERIA

The problem of determining which of a long list of management options is

"best" occupies much of Chapters 6, 7, and I I in this book. Fundamental to
that screening is a need for evaluation constraints-limits to the solution-and
criteria-measures of the effectiveness or suitability of a possible
management action. In some decision processes, these constraints and
criteria are not made explicit, with the result that participants disagree
about the acceptability of an option without a clear understanding of the
reasons for their dissatisfaction. Explicit discussion of design constraints
and evaluation criteria encourages better citizen understanding and more
focused decision making and can strengthen (or at least clarify) agency
support.

There is an extensive literature available on multiobjective planning

techniques. Some introductory engineering texts (e.g., DeGarmo et al.
1997) offer a simplified overview of this literature, and there are also
review papers such as that by Cohon and Marks (1975) that summarize
and evaluate possible approaches for use in situations where there may be
multiple decision makers and disparate planning goals.

5.5.1 Development and Application of Planning Constraints

Constraints are the practical limitations to a solution, the limits beyond
which the manager may not go. Constraints are typically expressed as
"musts," rather than "shoulds," and are usually stated in terms of a specific
numerical or spatial boundary. The most obvious example of a constraint is
financial: there are only so many dollars available to achieve the goals of a
project. Other con straints may include time (the project must be
completed within a specified time period), space (must fit a specific site),
or performance (must meet a minimum standard-for instance, of effluent
quality).

In many countries, including the United States, Canada, and the United
Kingdom, federal and state/provincial regulations limit the use of certain
kinds of land or the development of certain resources. Typical among these
are protections on wetlands and estuaries, wilderness preserves, scientific
reserves of various kinds, and similar lands. As discussed in Chapter 2,
valued historical or cultural resources in a watershed may be protected in
ways that preclude certain water management options. Rare or endangered
species may similarly be protected, often under law, from the impacts of
near or encroaching development. These kinds of regulatory limits
therefore place clear constraints on a plan.

Application of constraints is the first stage of option screening. It is the

step in which infeasible options are excluded from further analysis and the
long list of possibilities is reduced to a short list of feasible options that
will be subjected to detailed scrutiny.

A simple example illustrates the application of constraints. Using the

hypothetical river situation presented in Section 5.4, we can further assume
that solutions must meet the following constraints:

• Annualized capital and operating costs less than $1,000

• Fully implementable within four months of plan approval

Table 5.6 presents some hypothetical data for the several mutually
exclusive management alternatives under consideration.
 Table 5.6 illustrates that if costs must be kept under $ 1,000, and any
solution must be fully implemented within four months of plan approval,
only three alternatives are really feasible: do nothing, construct buffer
strips, or construct livestock exclusion fencing. The rest are either too
costly or take too long to implement. This analysis thus allows us to reject
the remaining five alternatives and proceed with detailed analysis of the
three feasible options. (This simple example also illustrates the need for
constraints to be chosen carefully. Is it really essential that costs be kept
under $1,000'? Or that all implementation he completed within four
months'? If so, these are appropriate constraints. If not, the analyst would
be well advised to reconsider the limits to the solution and determine
which are truly necessary for the planning situation.)

5.5.2 Choosing Appropriate Evaluation Criteria

Having eliminated options that are clearly not feasible, the analyst can
proceed to evaluate the worth of the remaining options using a set of
evaluation criteria. The term criteria seems intuitive: we know, for
instance, what criteria we might use to choose a new car. But careful
thought reveals that effective evaluation criteria must-even in the case of a
new car-have two important characteristics. First, they must be measurable
by some agreed-upon method. In this sense, "capital cost" is readily seen
as a good criterion; we can easily imagine the components of capital cost
and how they might he measured. By contrast, "social impact" is a more
nebulous criterion: do we mean employment rate'? average income? or
some other measure such as "quality of life"? "Environmental impact" is
just as problematic-it can be taken to mean anything from habitat
disruption to the presence of persistent toxic chemicals. A criterion is most
useful when it expresses a single, quantifiable attribute.

The second important characteristic of an effective criterion is that it

does, in fact, separate alternatives. Again, "capital cost" is effective in this
regard, because no two alternatives will have exactly the same capital
costs. But a criterion such as "phosphorus removal capability" is less
useful if all options under consideration have similar removal capability.
 This second characteristic is often harder to achieve. In some cases, it
will not immediately be apparent until the decision process is well
advanced that the chosen criteria are not effective in separating available
options. Even late in the game, however, it is possible (and indeed
advisable) to refine, replace, or simply drop problematic criteria, as long as
the change has the support of the full decision-making group.
 Lumsdaine and Lumsdaine (1995) suggest that criteria should attempt to
answer the following concerns:

• Motivation: Why would people want to accept the option`?

• People: How will people be affected by the option?

• Cost: What will the costs be to you and others?

• Support: What support is available for implementation?

• Values: What social values are involved? What will be the benefits to
people?

• Time: Will the option take a long time to implement?

• Effects: What will be the consequences of the option?

Table 5.7 gives some examples of typical evaluation criteria.

If these criteria seem familiar, it is because they, and others like them,
are drawn directly from the targets set in Chapter 3. This is as it should be:
if the scoping process described in Chapter 3 is intended to reflect a
growing societal consensus about the current state of the watershed, and
about some desired future state, then the targets for improvement should be
used as direct measures of progress toward the societal goal.

it is important to understand that there is no single evaluation criterion

that is universally appropriate (although it must be admitted that cost is
almost ubiq uitous in evaluation). The correct mix of criteria should reflect
not only the problem under consideration, but also the attitudes and values
of the community undertaking the evaluation.

Table 5.8 shows an extension of the simple river example used in

Sections 5.4 and 5.5.1, illustrating the analysis of alternatives using a
common set of evaluation criteria.

Table 5.8 illustrates a common problem in multicriteria analysis: that a

single option rarely is "best" across all evaluation criteria. (It is, however,
clear from this analysis that Alternative 1, the "do nothing" option, is less
effective than the other two and probably should be excluded from further
consideration.) Sections 5.5.3 and 5.5.4 describe several ways to deal with
the problem of multicriteria decision making.

5.5.3 Weighting Evaluation Criteria

In many decision processes, a single criterion will emerge as most

important. Sometimes, although not as often as might be expected, that
criterion is cost. Equally often, however, considerations such as control of
a particular pollutant, or reduction of flood frequency or severity, eclipse
others in reaching a final decision.

Assigning weights to each evaluation criterion serves at least two

purposes. First, it allows the analyst to make full use of available
information about community interests and priorities. And second, it has
the result of "spreading" the field-accentuating differences between
alternatives so as to make a final choice more clear-cut.

Weights can and should be chosen by consensus, not arbitrarily. It is

helpful if all weights in a set of criteria sum to unity, not just because the
idea of percentage weight is intuitive, but also because it avoids excessive
overweighting of high-priority criteria and thus skewing of results. Using
the data presented in Table 5.8, we can see that three criteria are under
consideration: cost, suspended sediment reduction efficiency, and
phosphorus reduction efficiency. Although we may believe cost to be more
important than the other two criteria, it would not be reasonable to assign
that criterion 10 times the weight of the others. A more realistic approach
might be to assign a weight of 40% to cost, and 30% to each of the other
criteria. This would yield a modified determination of criteria like that
shown in Table 5.9. But it introduces a new concern: although we can
readily imagine how to multiply each score by the weight of a particular
criterion, how do we compile a composite score that mixes costs with
pollutant removal efficiency? Section 5.5.4 discusses this problem in more
detail.

5.5.4 Multiattribute Decision Making

Inevitably, watershed management requires decision making that

encompasses a variety of considerations measured in different ways. The
following brief discussion of multi attribute decision making provides an
introduction to the more detailed techniques presented in Chapters 6, 7,
and 11.

The fundamental problem of multiattribute decision making is that of

comparing apples with oranges. Some researchers have tried to address
this difficulty by reducing all considerations, including not only costs but
environmental and social criteria, to dollar values. Although this approach,
termed "compensatory analysis," is certainly viable-the literature is full of
examples-it is not the only possible approach, nor is it necessarily the best.

The clear advantage of compensatory models is that they are intuitive

(everyone understands the meaning of a unit of money) and allow trade-
offs among alternatives. For instance, if alternative A costs more but
creates more dollar benefits (say, in terms of crop production) than
alternative B, it is easy to see that net benefits-benefits minus costs-are
higher for A, making it the preferred approach.

Trade-offs of' this kind are not possible when we are comparing costs
and, say, unemployment rate, or the incidence of swimmer's itch. When the
attributes of interest cannot, or should not, he reduced to a single measure
such as cost, other methods of comparison must be found.

How does one make a decision when "value" cannot be reduced to

dollar terms? Several "full-dimensioned" or "noncompensatory" models
suggest possible approaches.
Dominance Where one alternative is clearly better than all others for all
attributes, that alternative is said to dominate the process and is obviously
to he preferred. Regrettably, this rarely occurs. More often an alternative
performs well on some criteria (for example, cost) but less well on others
(for example, phosphorus removal capability). The decision maker is thus
faced with an unhappy choice between cost and treatment effectiveness.
Some of the following methods are helpful in this more common situation.

Feasible Ranges In this method, an acceptable range of performance is

established for each attribute (for example, cost between $1 and $2
million; phosphorus removal capability between 65(/c and 100%).
Alternatives whose performance falls outside the acceptable range on any
individual attribute are rejected from further consideration. Alternatives
that remain after the screening process therefore meet or exceed the
specified standard of performmance on every decision criterion.

There are several problems with this method. One is that it can he
difficult to decide what constitutes an acceptable versus unacceptable
range for any single attribute. Perhaps its greatest failing, however, is that
it yields satisfactory, but not optimal, management strategies.

Lexicography Lexicogr-aphyy is a term sometimes used for a stepwise

process of option screening. It requires that attributes he weighted. The
most important attribute is then used to choose the "best" option. If there
are several options tied for performance on this attribute, the second and
third most important attributes are used to break the tic(s).

This approach places heavy emphasis on the primary attribute, even to

the exclusion of other considerations, and so must he used with care to
ensure that all possible information is used in the analysis.

Standardization (Nondimensional Scaling) It may he possible to reduce

different kinds of measurements to a standardized scale, say from 0 to 10,
or from 10 to +10. (Simple ranking is one approach to this problem, but it
sacrifices information that is possible to retain with ordinal scaling.) Then
the standardized variables can he manipulated as if they were indeed a
common measure. If different attributes are weighted differently, each
nondimensional value can he multiplied by the appropriate weight and the
weighted sum of scores on all attributes calculated. (The weighted-sum
approach is, of course, also applicable to compensatory models, where all
attributes are measured in the same way.)

Standardization of the data presented in Table 5.9 allows the completion

of alternative screening and selection of a preferred alternative, as shown
in Table 5.10.
 This analysis demonstrates that, within the specified constraints and
using the given evaluation criteria and their weights, the long list of
alternatives presented in Table 5.5 can be reduced to a single, preferred
alternative: 3, construction of livestock exclusion fencing. This analysis
also reveals several other interesting aspects. Among these is the fact that
there is not a lot of difference between the alternatives in terms of their
weighted scores; in other words, although Alternative 3 proved to be
preferable, it did not "win" by a large margin. This result is a useful clue to
the analyst that some test of robustness of the solution may be desirable.
For instance, if cost is weighted more heavily, say at 50%, with only 25%
each for the other two criteria, the scores would be:

Under these conditions, Alternative 3 remains the preferred approach, but

Alternative 1 ("do nothing") appears preferred over Alternative 2.

On the other hand, if we reversed the weights of the criteria, weighting

cost at only 20% and the other two at 40% each, we find that Alternative 3
still has the lowest score and remains the most attractive:

These results suggest (not surprisingly, given the simplicity of the

example) that the analysis would be strengthened by the addition of other
evaluation cri teria to "spread" the final results and make selection of a
preferred alternative more clear-cut. Or the results of the analysis can he
used "as is" to instruct decision makers: if cost is clearly the most
important criterion, then Alternative 2 is the preferred approach. If
pollutant reduction efficiency is more important than cost, then the
decision makers should select Alternative 3. Alternative I is attractive only
if cost is the sole consideration and pollutant removal efficiency is
excluded from the analysis.

This brief discussion is intended to introduce the concepts of

constraints, criteria, weighting, and multiattribute decision making.
Specific analytical techniques for the evaluation of management
alternatives and the problem of' selecting a "best" management plan for the
watershed are described in more detail in Chapters 6, 7, and I I of this
hook.

REFERENCES

Bussey. L. E. 1978. The Economic Analvsis o/ Industrial Projects.

Englewood Cliffs. N.J.: Prentice-Hall.

Cohon, J. L., and D. H. Marks. 1975. A review and evaluation of

multiobjective planning techniques. Water Resources Research 11(2):
208-220.

Davis. Mackenzie L. and David A.. Cornwell. 1991. Introduction to

Environmental Engineering. 2d ed. New York: McGraw-Hill.

DeGarmo, E. Paul, William G. Sullivan, James A. Bontadelli. and Elin

Wicks. 1997. Engineering Economy. 10th ed. Upper Saddle River, N.J.:
Prentice-Hall.

LeBoeuf, Michael. 1980. linag>incering: Hon, to Profit fronr Your Creative

Powers. New York: Berkley Books.

Lumsdaine, Edward. and Monika Lumsdaine. 1995. Creative Problem

Solving: Thinking Skills for a Changing World. New York: McGraw-Hill.

Metcalf and Eddy Inc. 1991. Wa,ttc water Engineering: Treatment,

Disposal, and Reuse. 3d ed. New York: McGraw-Hill.
Novotny, V., and H. 01cm. 1994. Water Qualihv: Prevention. Identification,
and Management of Diffuse Pollution. New York: Van Nostrand
Reinhold.

Ontario Ministry of the Environment. 1987. Technical Guidelines for

Preparing a Pollution Control Plan. Report from Urban Drainage Policy
Implementation Committee. Technical Sub-Committee No. 2. Toronto,
Ont.: Ontario Ministry of the Environment.

Osborn, Alex F. 1963. Applied lnwgination The Principles amid Problems

of Creative Problem-Solving. 3d rev. ed. New York: Scribner's.

Prince, George M. 1970. Practice of'Creativiiv. New York: Macmillan.

Schwab Glenn 0., Delmar D. Fangmeier, William J. Elliot, and Richard K.

Frevert. 1993. Soil and Water Conservation Engineering. 4th ed. New
York: John Wiley & Sons.

Switzer-Howse, Karen. 1982. Agricultural Management Practices for

Improved Water Quality in the Canadian Great Lakes Basin. LRRI
Contribution #82-10. Ottawa: Agriculture Canada Research Branch.

Van Gundy, Arthur B. 1984. Managing Group Creativity: A

ModularApproach to Problem Solving. New York: American
Management Associations.

Waterhouse, James. 1982. Water Engineering for Agriculture. London:

Batsford Academic and Educational Ltd.
 Simple Assessment
Methods
Chapter 5 described the development of alternative approaches for water
management problems. As discussed in that chapter, any given option will
usually be considered very promising in some respects, but disappointing
in others. Weighing diverse criteria in the evaluation of management
alternatives is therefore one of the most challenging aspects of developing
a water management plan. The amount of effort and resources put toward
this task will depend on the goals of the process, the availability of
resources and information, and the expertise of the analysts.

Several approaches are possible. A cursory, or preliminary, screening

can inexpensively provide rapid insight into the probable effectiveness of
different management strategies. It may not, however, give the kind of
detailed, quantitative results necessary to justify fiscal commitment or
definitively sway public, or political, opinion. Detailed assessment
techniques, on the other hand, are timeconsuming and costly and may not
be feasible within the window of opportunity afforded by a particular
funding program or political administration.

Any screening program should have the following goals:

1. To determine which of the available alternatives is feasible-that is,

meets the constraints imposed by cost, space, time, or technology

2. To determine which of the remaining alternatives performs best in

terms of specified evaluation criteria (see Section 5.5).

In fulfilling these objectives the analyst will, by default, accomplish two

other goals:

1. To ensure that complete and accurate data are available for all
alternatives under consideration.
 2. To rank individual alternatives (and component projects) in terms of
their overall performance.

In many, perhaps most, cases watershed managers will have a limited

budget and time frame within which to develop management
recommendations. Resource limitations will therefore restrict the amount
of' new data that can be collected in the planning process, and the
development of new analytical tools such as computer simulation models
for evaluating management alternatives. In other cases a preliminary or
"feasibility" level analysis may be desired in advance of more detailed
work, and perhaps to justify the resource allocation that would be
necessary for more detailed data collection and analysis.

In these circumstances the analyst can employ it range of simple

analytical techniques, with little or no collection of new data, and produce
a general picture of problems, mechanisms, and solutions. The
effectiveness of this analysis will depend very much on the quality of
available data and the extent to which watershed processes are already
understood. At the very least, a preliminary analysis should reveal:

• Significant data gaps

• The need for additional technical analysis

• The need for clarification of study goals and objectives

• Recommendations for immediate pollution abatement actions

Figure 6.1 illustrates the main components of a preliminary watershed

analysis, some of which have already been discussed in Chapters 2 to 5. It
will be seen from Figure 6. I that a preliminary (simple) analysis includes
the same steps of inventory, scoping, development and testing of
alternatives, and outputs as does a detailed analysis. It differs from a
detailed assessment primarily in the level of detail used, the quality (and
often age) of data, and thus in the precision and accuracy of outputs. The
following discussion illustrates this point for each component of the
analysis.
 The following discussion illustrates this point for each component of the
process.

6.1 THE WATERSHED INVENTORY

The introduction to Chapter 2 makes the point that water movement in a

system is affected by many physical, chemical, and biological features and
processes. Preparing an inventory of a watershed is therefore directed at
developing an understanding of these features and processes. At the level
of a simple assessment, most information can be gleaned from existing
data, from the literature, or from simple observations.
 Figure 6.1 Components of a preliminary watershed analysis.

6.1.1 Using Existing Data

A simple inventory of watershed resources can be compiled with the
following tools, most of which are usually readily available in major
towns:

1. A foundation map of the watershed, showing

Topography

Location of major features with potential to influence water

management, such as dams, locks, protected areas, areas of
cultural or spiritual value, habitats of rare or endangered species

Overview of land use (e.g., residential, industrial, open space,

forested, agricultural, etc.)

Location of major point source discharges (e.g., municipal

sewage treatment plant(s), industrial effluent discharges)

Spatial relationships between potential sources and receiving

waters

Political boundaries

2. Existing data on specific physical, chemical, and biological

characteristics of the system (for instance, from state agencies)

3. Existing data on specific physical, chemical, and biological

characteristics of pollutant sources (from public or industrial sources)

4. Media reports and public commentary about the key water use
impairments and use conflicts currently present in the basin

Table 6.1 lists the types of data that are usually necessary for a basic
watershed inventory.

One simple but very useful inventory technique is to use a topographic

map of the watershed as a foundation and overlay that map with
transparencies showing other features such as important recreational areas,
key sources, and so on. (Geographic Information Systems are of great
value in this undertaking and may already be in use in some jurisdictions.
They are not, however, necessary to achieve the required overlays.) The
overlays should reveal areas that are under particular stress, such as a
major point source located near fragile wildlife habitat, or residential
development encroaching on sensitive wetlands.

A composite map can be compiled for the present condition and, using
land use and population growth projections, for various planning horizons
in the future.
6.1.2 Filling Data Gaps

Data gaps are unavoidable in any watershed analysis, but they are perhaps
less critical in a simple assessment than in more detailed analysis. Because
a simple analysis itself is less demanding, alternate data sources can be
used to good advantage. Table 6.2 suggests some other ways to collect
basic watershed information if the data suggested in Table 6.1 are
unavailable.

6.2 SCOPING

6.2.1 Defining the Problem

For the purposes of a simple assessment, the most urgent problems in the
watershed can be determined in several simple ways:

1. Through a listing of existing (or perceived) use impairments and their

causes

2. By comparing existing receiving water or sediment quality with

published national or state/provincial targets

3. Through key informant interviews and media reports

Processes like Adaptive Environmental Assessment and Management

(AEAM) (Section 3.4.2) can be useful in clarifying community consensus
on watershed problems without detailed quantitative analysis. AEAM
draws from community knowledge of the basin and its resources and can
be a very powerful tool in problem definition. The various techniques
described in Chapter 4 may also be useful, but some may require resources
beyond those available for a simple assessment.

6.2.2 Identifying the Sources

Identification of key problems often simplifies the identification of

sources, especially if such a problem is limited to a particular area within
the basin. Reference to a composite map should quickly reveal the
presence of possible point sources or of land uses that may generate diffuse
pollution during wet weather.

In some cases there will be several sources of a single pollutant. It

makes sense to determine which of these is the most important contributor
to the total load and which would repay the greatest benefit per dollar
spent on remediation. The two are not necessarily the same: it may be very
costly to reduce pollutant loads from a large contributor such as a sewage
treatment plant. On the other hand, inexpensive measures like livestock-
exclusion fencing may achieve greater pollutant reductions for every dollar
spent. Depending on the situation, it may therefore be better to install
many kilometers of fencing rather than upgrade a sewage treatment plant
to advanced treatment levels. The problem of evaluating the relative
importance of sources is discussed in the next section.
6.2.3 Evaluating the Relative Importance of Sources

Pollutant (or flow) source analysis at a coarse level is designed to set

priorities on sources for further evaluation or immediate action. Source
analysis can be of two types: total mass and critical levels (which, in the
case of pollutants, would be critical concentrations). Section 5.1 discusses
the difference between mass loading and concentration. Either or both
types of analysis may be appropriate in any given situation; the decision
will depend on the nature of the problem(s) under investigation, applicable
laws, and the interests of the investigating agencies.

The screening techniques discussed in Section 6.3 have value in terms

of estimating loads and concentrations under status quo conditions. Tables
6.1 and 6.2 gave a number of suggested sources of data on dry-weather
(continuous) point sources, wet-weather (intermittent) point sources, and
nonpoint (diffuse) sources. In general, for a simple evaluation, the
literature provides a rich supply of information on all categories of source
and on potential control approaches. Data drawn from the literature are
rarely directly applicable to the area under study, however. Some judgment
is required in applying literature values to the problem of interest. Local
experts from government agencies, consulting firms, or universities can be
helpful in making these decisions.

Ideally, the analyst should attempt to develop source analyses like that
shown in Figure 6.2 for every pollutant or material of concern. (The
graphical analysis shown in Figure 6.2 could, of course, be replaced with
tables containing the same or similar data.)

6.3 DEVELOPING AND SCREENING MANAGEMENT

ALTERNATIVES

6.3.1 Developing Alternatives

Chapter 5 has described the general process of developing alternatives. At

the simple assessment level, the analyst should strive to:
 • Identify the sources that contribute most to the problem under study

• For each nontrivial source (those contributing, say, more than 5% of

the total load), identify at least one management option in each of the
following categories:

"Do nothing"-retaining status quo

Structural measures-built technologies or structures

Vegetative approaches

"Best management practices" (nonstructural measures)

Figure 6.2 Pollutant source analysis for a hypothetical planning area:

estimated total phosphorus loads to receiving water.

The techniques described in Chapters 4 and 5 can he used to generate

these ideas. This suite of four options provides a minimum set for coarse
screening. The following is an example of such a set of measures for the
control of solids in urban stormwater runoff:

• "Do nothing"-Retain status quo.

• Structural measure-Construct stormwater detention pond at the

intersection of Oak Street and Willow Road.

• Vegetative approach-Construct artificial wetland for stormwater

treatment at intersection of Oak Street and Willow Road.

• `Best management practice"-Institute semiannual cleaning for all

existing catch basins throughout the city.

This brief list is clearly a bare minimum; the tables of options given in
Section 5.3 provide many more, as would the extensive literature on each
source type. The point is to develop a set of possible solutions for each
source of each pollutant or material of concern.

The next step is then to apply any planning constraints (Section 5.5) and
eliminate any options that are clearly inappropriate or impractical for the
planning situation. The remaining "short list" of options can then be
subjected to screening, using a range of simple screening procedures, as
discussed in the following section.

6.3.2 Simple Procedures for Estimating Pollutant Loadings from Pervious Surfaces

Section 2.3.3 described several methods for estimating stormwater runoff.

Such estimates are an important first step in estimating the pollutant
loadings associated with runoff. As discussed in Section 2.3.3, runoff rates,
volumes, and quality are closely related to the perviousness of the soil
surface. In urban areas most rainwater falls on paved or roofed surfaces
and is diverted through sewer systems and to treatment or discharge points.
In rural areas more of the land surface is pervious, so more of the rainwater
falling over a rural area infiltrates into soils, is taken up by growing
vegetation, or is returned to subsurface flows. Runoff prediction models
take these differences into account.
 The quality of runoff is also different in urban and rural areas. Table 2.6
illustrated some differences in pollutant loadings from different land uses.
Table 6.3 presents certain values for pollutant concentrations associated
with different runoff sources (note that these values can be used as values
for the variable "C" values in Schueler's "Simple Method," shown in Table
6.5).

The much higher pollutant concentrations associated with older

residential areas are attributable to poor "housekeeping," including
accumulation of refuse and debris, decaying building fabric, high traffic
volumes, and poor maintenance of open spaces and lawns.

It is certainly possible to generate detailed estimates of pollutant

loadings using quantitative methods. These methods are, however, usually
far too complex and time-consuming to be useful in a simple assessment.
A number of authors have suggested simple techniques for estimating
pollutant loadings, especially for wet-weather, intermittent sources. (These
are usually the most difficult to estimate, continuous dry-weather sources
being far more predictable in terms of flow and quality.)

Pollutant loads from areas with predominantly pervious surfaces (most

rural areas and urban open space) are easily estimated with the Universal
Soil Loss Equation (USLE) (Wischmeier and Smith 1978):

A = RK(LS)CP (6.1)

R = rainfall and runoff erosivity index for location of interest

K = soil erodibility factor

L = slope length factor

S = slope steepness factor

C = cover management factor

 P = conservation practice factor

Potential values for the USLE variables vary greatly. Graphs, tables, and
nomographs showing recommended values for different conditions are
widely available in the literature. Novotny and Olem (1994) provide an
excellent dis cussion of the USLE and its modifications, including a full
range of values for the USLE variables.
 The USLE cannot be used to predict loads of solids directly. To do so,
the analyst must calculate potential soil loss using the USLE and then
multiply that value by the delivery ratio, DR. Table 6.4 provides an
overview of DR values for a range of land use types.

Table 6.4 clearly shows that as land becomes impervious and natural
drainage (and infiltration) patterns are overridden by sewer systems, a
higher proportion of eroded sediment is delivered to the receiving water.
This may seem counterintuitive: after all, surely a higher degree of
imperviousness and controlled drainage would reduce erosive forces. But
upon understanding that sediment delivery ratio refers to the movement of
sediment that has already been eroded, it becomes easy to see that paved
and roofed surfaces are smooth and would not trap and retain sediment as
well as pervious rural lands. In a city, therefore, sediment eroded or
washed off (for instance, from construction sites) is rapidly transported to
and through storm sewer systems and into a lake or stream. Unless
treatment facilities have been installed to trap sediment, most (perhaps all,
as Novotny and Olem suggest) of the eroded sediment will in fact reach
the receiving water.

6.3.3 Simple Procedures for Estimating Pollutant Loadings from Impervious Surfaces

The preceding discussion illustrates the point that urbanization, with its
paving, roofing, and sewerage, changes and overrides the natural
hydrologic processes of a watershed. Typically, in systems with a high
proportion of impervious surface, pollutants accumulate on surfaces during
dry weather and are washed off those surfaces into stormwater runoff
during wet weather. Stormwater is then conveyed through a system of
pipes and traps and ultimately is discharged into a lake or stream. It is
therefore possible to estimate the unit area pollutant loads from urban areas
by combining estimates of pollutant accumulation on the land surface
("buildup") with estimates of runoff potential ("wash-off"). Many authors
have suggested simple methods for the estimation of pollutant loads from
urban areas. Most of these incorporate buildup/wash-off relationships.
 Most streets are constructed with a slightly higher profile in the center
of the roadway, sloping down to curbs or medians on either side. The
purpose of this construction is, of course, to encourage drainage of
stormwater away from traffic and toward storm drains. A variety of studies
have shown that as a result of this construction, most solids tend to
accumulate within a 1-meterdeep zone along the curb. This phenomenon is
sometimes termed "curb storage," with solids accumulating in this way
expressed as grams of pollutant per unit length of curb. The accumulation
of solids on a road surface has been demonstrated to be nonlinear in most
cases, with solids volume rising quickly to approach a pseudo-equilibrium
level, where accumulation apparently levels off, probably as a result of
displacement or movement of solids to neighboring areas. The "steady
state" curb storage the "buildup" of the buildup/wash-off relationship-
ranges from about 20 to 100 gym of curb (Novotny and Olem 1994).

Wash-off occurs when rainwater falls on an impervious surface,

dislodging particles and conveying them into the stormflows. The quantity
of pollutant in stormflows can therefore be estimated as a fraction of curb
storage:

W = CD ) Sr (6.2)
 S = stored curb load of the pollutant during the preceding dry
period (gym curb length)

CD curb density (km/ha)

A = subwatershed area (ha)

r = wash-off factor (related to runoff energy)

Sartor et a]. (1974) represented solids wash-off using a simple first-

order relationship:

dP = k„rP dt (6.3)

which integrates to:

P, = PO~(l - exp(-k„rt) (6.4)

k„ = urban washoff coefficient (constant, derived from street

surface characteristics)

P = amount of solids remaining on the surface

PO = initial mass of solids in storage

P, = mass of solids removed by rain with intensity r and duration t

This equation was further modified by staff of the Hydrologic

Engineering Center (1975) in their development of the storm runoff model
STORM, by adding an availability factor, A, to the equation by Sartor et
al., on the assumption that not all particles are available for removal:

P, = APO (I - exp(-k„rt) (6.5)

where A, the availability factor, has a value between 0 and 1.0 and is
determined as:

A = 0.057 + 0.04(ri. ] ) (6.6)

where r is rainfall intensity.

These equations can be used to estimate the total mass of solids that is
removed by rainfall, but not the mass of pollutants removed. The most
common approach to estimating pollutant loading is to multiply the mass
of solids by some factor expressing the pollutant concentration of those
solids. Novotny and Olem (1994) term these multipliers "potency factors."
For example, if 57 grams of solids per meter of curb are estimated to have
washed off during a rainstorm, and if the analyst believes that total
phosphorus loads are roughly one-third of total solids loads, then the mass
of total phosphorus (TP) removed by rain (and presumably available to
receiving waters) would be:

This method, although widely used in watershed planning, cannot be

used to predict pollutant loadings with any accuracy, simply because the
relationship between solids loading (or concentration) and the loading (or
concentration) of any given pollutant is never constant nor even linear.
Nevertheless, it offers a "quick and dirty" way of estimating pollutant
loads from intermittent sources and is likely to be useful in comparing the
relative impacts of different management alternatives. Zison (1980)
provides a discussion of the potency factors for different pollutants in
urban and rural areas.

Several authors have developed simple screening methods, based

largely on the Universal Soil Loss Equation (for rural, pervious areas) and
huildup/wash- off relationships (for urban areas). Most of these take the
form:
 Table 6.5 presents some of these methods, which may be most helpful for
the estimation of flows and pollutant loadings from urban stormwater and
combined sewer overflows and agricultural drainage.

The following example demonstrates the application of Schueler's

Simple Method in the evaluation of development alternatives for a
suburban site.

Example: Prediction of Pollutant Loading from Development Alternatives:

1. Assume a site 23 ha in area, currently forested, almost fully pervious

(assume 2%% imperviousness).

2. Assume that there is interest in developing this site, either as a

townhouse community (45% impervious surface) or as a low-density
housing development (30% impervious surface).

3. Estimate the current and projected loadings of total phosphorus and

nitrogen under each of these planning scenarios using Schueler's
Simple Method.

Step l: Estimate parameters for pre- and postdevelopment conditions

(this example uses the values drawn from Table 6.3:
 Step 2: Estimate annual pollutant loads from stormwater runoff. (Note
that Schueler's Simple Method equation works equally well whether using
acres and inches, or hectares and centimeters. In either case, the product
will he expressed in pounds and must therefore he multiplied by 2.2 to
obtain an estimate in kilograms.)
(Refer to Table 6.5 for explanation of equation)

For total nitrogen:

For total phosphorus:

Although crude, this analysis shows that planners can expect significant
increases in nutrient export from the site with either townhouse or low-
density residential development. This information is useful in several
ways. It can guide the analyst as to which alternative will generate the
lowest nutrient loads (or create the smallest change from existing
conditions). It can point to the need for mitigative measures to reduce
nutrient export under development conditions-in other words, to reduce
runoff from the site by encouraging retention on-site, by improving the
quality of runoff by on-site treatment, or by similar measures. And it may
reveal differences so dramatic (as in this simple example) that additional
data collection (for instance, on runoff quality and volume) may be
warranted before development goes ahead.

Simple screening methods of the type shown in Table 6.5 and in this
example are not, however, useful for making accurate estimates of existing
or future conditions. So the analyst could not, for example, draw the
conclusion that exactly 18 kg/year of phosphorus will be exported from
this site if low-density residential development is constructed. Rather,
these estimates should be considered "order of magnitude" values:
predevelopment phosphorus loads are probably in the range of 0 to 10
kg/year, while townhouse development would likely increase those loads
by a factor of 10.

Figure 6.3 Simple graphical analysis of pollutant sources and in stream

impacts for a hypothetical planning area.
 6.3.4 Simple Procedures for Estimating Receiving Water Impacts

The simplest method of determining cause-and-effect relationships in a

stream is to plot the average concentration of one or more key parameters
against potential sources. Figure 6.3 illustrates this method for a
hypothetical planning area. Even this type of simple analysis often reveals
important relationships between pollutant sources in in-stream impacts. It
will also yield clues as to the relative importance of sources for the
parameter of' interest.

Figure 6.3 provides a clear and visual demonstration that industrial

discharge to the stream is much less important than agricultural runoff or
the municipal sewage treatment plant, even in dry weather. In wet weather,
the four most important sources are agricultural runoff, combined sewer
overflows, the municipal sewage treatment plant effluent, and stormwater
discharges. Of these, agricultural runoff and the municipal sewage
treatment plant may he somewhat larger contributors than the two wet-
weather sources, especially since they operate more or less continuously.

Simple quantitative methods like those described earlier for estimation

of pollutants can also be used to predict receiving water impacts, such as
exceedances of water quality criteria. Again, several approaches are
available, mostly using simple equations for dilution with or without in-
stream degradation. The following discussion, drawn from MOE (1987)
summarizes these methods for conservative substances like chloride or
heavy metals, which do not break down or react in the water column;
reactive substances such as bacterial densities or BOD, whose
concentration may change depending on physical, chemical, or biological
conditions; and "coupled" substances such as ammonia, whose
concentration depends directly on the concentration of another parameter
(in the case of ammonia, pH).

(i) Conservative Substances (e.g., Chloride) Discharged From a Point Source

C= Co+ Q (6.7)
 COQ = concentration at distance x = 0

W = point source loading rate (mass/time)

Q = flow (volume/time)

(ii) Conservative Substances Discharged from Nonpoint Sources

C Co+w (6.8)

CO = concentration at distance x = 0

w = non-point source loading rate (mass/time)

x = distance

Q = flow (volume/time)

(iii) Reactive substances (e.g., BOD) discharged from a point source

L = Loe K, %/(/ + (W/Q)e K,.k/U (6.9)

where: L = concentration of reactive substance

L0 = concentration at x = 0

x = distance

K,. = rate of pollutant removal

U = velocity

W = point source waste loading rate

Q = flow

(iv) Reactive Substances Discharged from Nonpoint Sources

 L = L,)e K,"X/u + w (I - e K,X/u) A K, (6.10)

LO - concentration at x - 0

x = distance

K. = rate of pollutant removal (e.g. BOD removal)

U = velocity

w = non-point source waste loading rate

A = cross-sectional area of stream

(v) Coupled Substance Discharged From a Point Source

D Doe K"X/u + L►~ K,I le K,-X/u e K"X/U K„ - K. W K,I K' -X; I/ K„X i,
+ Q K K le e l (6.11)

Doi - coupled substance concentration at X = 0

X - distance

U = velocity

Laa = concentration of reactive substance at X 0

K,1 = BOD oxidation coefficient

K,. = rate of pollutant removal (e.g., BOD removal)

K, - reaeration coel'ticicnt

W - point source waste loading rate

 Q = flow

(vi) Coupled Substance Discharged from Non-Point Sources

Doi - coupled substance concentration at X - 0

K, - dissolved oxygen reaeration rate (i.e., rate of removal of

deficit)

X - distance

U - velocity

L0 concentration of reactive substance at X 0

K,I -- BOD oxidation coefficient

K. - rate of pollutant removal (e.g.. BOD removal)

K„ reaeration coefficient

w = non-point source waste loading rate

A = cross-sectional area of stream

Although some of these equations are lengthy, they do not require

advanced mathematical techniques nor detailed information about
discharges or receiving water.
 It is important to note that the various reaction rates in these equations
are important in influencing predictions. Basic engineering texts such as
Metcalf and Eddy (1991) provide some information about reaction rates,
and more detailed information is widely available in the scientific
literature. Examination of the literature, however, usually reveals a
substantial range in reported values. The analyst is therefore advised to
calculate expected in-stream impacts using both low and (in a separate
calculation) high rates, thus arriving at a likely range of outcomes rather
than a single uncertain value.

(vii) Prediction of Exceedances of In-Stream Water Quality Criteria

Schueler (1987) gives an example of how his Simple Method can be
adapted for use in predicting exceedances of in-stream water quality
criteria. The following is adapted from that example:

Example: Prediction of Water Quality Impacts of Development Alternatives

1. Assume that a developer wishes to build a partially sewered

residential development 30 ha in area within a larger tract of 130 ha of
forested land.

2. Assume that the percentage of imperviousness of the developed area

is 27%, while that of the forested area is 2%.

3. Assume that the receiving water into which storm runoff will drain
has a hardness of 100 mg/L CaCO3.

4. Assume that the concentrations of trace metals in runoff from the

forest site are negligible.

5. Determine whether the projected concentrations of lead, zinc, and

copper in this development situation would exceed established water
quality criteria.

Step /: Calculate the expected volume of runoff from the developed and
forested portions of the watershed for a 2.5 cm storm using the first term of
Schueler's Simple Method equation (previous example), as follows:
(Note that this calculation does not contain the term P1. As discussed in
Table 6.5, P1 is not used for a single storm. Assume that R„ has been
calculated to be 0.07 for the forested area and 0.29 for the development
site.) Schueler's equation was designed for use with acres and yields
volume in acre-feet. If hectares are used for area, the product must be
multiplied by 488.18 to obtain cubic meters:

Runoff volume from the two land uses for a 2.5-cm storm (i.e., 2.5 cm of
rainfall) can then he calculated as:

The total runoff to be expected from the mixed-use watershed would

therefore be 362 + 261 = 623 cubic meters.

Step 2: Calculate the dilution ratio of developed site runoff to watershed

runoff:

Step 3: Determine the trace metal concentration in runoff that is

typically exceeded in 5°/r of storms (about three storms a year): see Table
6.6:

The "5%" column shows that, based on NURP data (US EPA 1983), a
lead concentration of 0.109 mg/L was exceeded in the runoff from 5% of
storms. The equivalent concentrations for copper and zinc were 0.055 and
0.216 mg/L, respectively.

Step 4: Calculate expected in-stream concentration of trace metals.

 The runoff concentrations shown in Table 6.6 are rapidly diluted by
cleaner flows. In Step 2, we calculated the dilution ratio for this site to be
58% (in other words, the flow from the development site constitutes
58%Yc of the total watershed runoff). We can therefore assume that
contaminated runoff from the development site will he diluted in this
proportion when it reaches the receiving stream:

or, for lead:

for zinc:

and for copper:

These values (again, rough estimates) can be compared against available

criteria for ambient water quality to determine whether in-stream
conditions will approach or exceed recommended levels following
development.

Although techniques like Schueler's Simple Method were developed for

single-site analyses, they can also be applied over larger areas. The larger
variability-in rainfall, in runoff, in pollutant concentration-over a larger
area means that estimates prepared this way may be even less accurate
than those prepared for a single site. They may, nevertheless, be useful for
comparing broad categories of management alternatives-for example:
increase residential development in the southern part of the watershed by
15% over the next 10 years; or encourage light industrial development in
that area, or retain as a mixture of open space and rural residential
development.

(viii) The Special Case of Lakes The foregoing techniques illustrate the
range of simple methods available for estimation of receiving water
impacts in streams. Generally speaking, estimation of impacts in lakes is
more complex, especially when a lake is large, deep, or has complicated
circulation patterns.

Simple models such as those developed by Vollenweider (1976) and

Dillon and Rigler (1974) use estimates of gross pollutant loads and lake
residence time to determine the probability of eutrophication. Chapra
(1979) developed a modified version of this approach for use in the
embayment areas of large lakes. Thomann and Mueller (1987) provide a
more detailed discussion of this approach and formulate Vollenweider's
model as follows:

dp V (it -W - v,A,p - Qp (6.13)

p = total phosphorus in the lake, expressed in mass per unit

volume (e.g., µg/L)

Q outflow (volume per unit time)

 A, = lake surface area

W = allochthonous source of phosphorus expressed in mass per

unit time

(e.g., g/sec)

v., = settling rate of phosphorus (distance per unit time, e.g.,

cm/sec)

At steady state, dp/dt= 0 and division by the surface area of the lake
yields:

W, p Qc + U, (6.14)

W' = allocthonous phosphorus leading

y,. = hydraulic overflow rate Q/A, = hp

V. = settling rate of phosphorus

Q = outflow

p = Q/V == 11-T,, = flushing rate

A,, = lake surface area

H water depth

T,,, = detention time in the lake

V = volume of the lake

Novotny and 01cm (1994) point out that the fundamental weakness in
this approach is a lack of knowledge about the settling rate of phosphorus.
Vollenweider (1976) used the following to approximate u,:
 u.,=HV/ P (6.15)

which in turn yields:

W, CI Hp I +)

Most of the information needed for these calculations is readily

available from pollutant loading estimates and simple water balance
calculations. This type of approach is most useful for long-term planning
purposes (it has sometimes been applied in the determination of allowable
development density along a lake shoreline). It is not intended for detailed,
short-term, or site-specific predictions.

Figure 6.4 illustrates the practical application of the Vollenweider

model.
 Figure 6.4 Example of critical phosphorus loading plot for lakes (after
Vollenweider 1976, MOE 1987).

Some Cautionary Notes It will be seen from this discussion that available
simple screening models are easy to use and require little in the way of
input data. Their simplicity means that they have ignored or minimized the
importance of factors such as basetlow runoff, which can be significant
over larger areas, and any pollutant loadings attached to that runoff. In
some systems, the baseflow can account for more than half the total runoff,
so use of these simple screening methods could significantly underestimate
the pollutant load that would be associated with specific management
actions.

A second caution relates to the generalized data that are commonly used
with screening models. The NURP data (US EPA 1983), for example,
represent the combined results of studies on more than 300 runoff events.
They do not, however, include every condition that might he encountered
in watershed planning, nor do they apply to every land use configuration or
runoff situation. These methods are best used to compare management
alternatives with fundamentally different influences on the land surface
and water flow over it (such as in the preceding examples). It would not be
appropriate to compare two very similar alternatives-for instance, a 250-
townhouse development versus a 275-unit development-using these
methods. The techniques simply are not sensitive enough, nor accurate
enough, for accurate prediction of the differences between closely related
alternatives.

Third, most of these models have been developed for a particular

geographic location or set of data (for instance, Schueler's method was
based on urban runoff data from recently stabilized suburban watersheds).
They may be less accurate when applied in different situations, such as
areas undergoing construction, industrial areas, nonfarming rural
development, or agriculture.

In summary, simple screening methods are intended to he only that:

simple techniques to assist the analyst with limited time and budget in
assessing management alternatives. More complex basins, and/or more
complex problems, will require detailed watershed and receiving water
simulation, with the extensive input data required by those tools. These
methods are described in more detail in Chapter 7. The reader should,
however, be cautioned that even the most complex simulation models are
sometimes based on the very simple screening models discussed earlier,
not on precise mechanistic representations of hydrologic or chemical
phenomena. The reader is advised to examine carefully the algorithms on
which such models are based, so as to understand the assumptions and
mechanisms used to make model predictions.
6.3.5 Simple Screening Procedures

Armed with an overview of the watershed system and some basic

information about pollution sources and their relative magnitude, the
analyst can proceed to screen management alternatives.

The first step in screening is application of constraints. Graphical

methods are helpful here. Table 5.6 illustrates a tabular method, but it is
also possible to represent the information given in Table 5.6 in graphical
form. The advantage of such an approach is that it can allow the
presentation of more complex information than is possible in a table.
Figure 6.5 shows how this might he done for the example given in Table
5.6. This example illustrates only two variables, cost and time. With a
larger number of variables, graphical methods allow visual patterns to
emerge across a single management option, or a single constraint.

Other graphical approaches are possible. Figures 6.6 to 6.8 show

methods that are helpful in displaying information with a higher degree of
uncertainty-in other words, variables in which performance may be
feasible under some conditions, clearly not feasible under others, but
marginally feasible in some intermediate zone. Graphical methods allow
this gray area to be included in the analysis.

6.4 OUTPUTS OF A SIMPLE ASSESSMENT PROCESS

The outputs of a simple assessment process may be of several types.

Egregious problems may have been identified in the data collection and
plan screening process, leading to recommendations for immediate
abatement action. Frequently, the level of data quality and analytical
capability possible in a simple assessment is found to be disappointing or
in some way inadequate for the purposes of the planning initiative. In such
a case, an outcome might be recommendations for the collection of
additional data and/or a commitment to more detailed, and probably more
costly, analysis.
Figure 6.5 Example of graphical methods for application of planning
constraints.
Figure 6.6 Example of gradient bars used in the application of planning
constraints (adapted from Schueler 1987).
 Sometimes the output of a simple analysis is simply the realization that
the true sources of use impairments are not what they were originally
imagined to be, or that the scope of the initiative needs to be altered. New
or previously underestimated use impairments may emerge during the
course of the assessment, or the scale of a particular problem or source
may turn out to be far greater than originally assumed.
 In a recent watershed plan in Ottawa, Ontario, for example, bathing
beach closures were thought to be caused by discharges of bacteria from
storm and combined sewers, even during dry weather. The planners
believed that the continuous nature of the loads (which had been measured
in field studies) pointed to illegal cross-connections between storm and
sanitary sewers. These cross-connections would have allowed untreated
sewage to flow from residential sanitary lines into storm sewers, and then
into the Rideau River, without benefit of treatment or disinfection. During
the Rideau River Stormwater Study in the early 1980s, it was discovered
that nesting waterfowl and pigeons-present in large numbers along the
river, particularly on and under bridges were contributing significant fecal
loads to near--shore waters. This finding led to more detailed data
collection and revised terms of reference and, ultimately, to a different
suite of management actions than might have been contemplated at the
outset of the study.

Finally, most planning initiatives extend beyond the present day into
some imagined future condition. So another possible outcome of the
assessment is that present and future problems and sources differ. This
information is useful for long-term planning purposes and can be updated
as more information about watershed problems and sources becomes
available with time. Table 6.7 summarizes some possible outcomes of a
simple assessment process.

6.5 AN EXAMPLE OF THE APPLICATION OF SIMPLE

ASSESSMENT PROCEDURES

Tables 6.8 and 6.9 summarize the findings of a hypothetical simple

assessment process for a watershed that is currently experiencing impaired
swimming opportunities as a result of bacterial contamination and algae
blooms at local beaches. It is thought that the algae blooms are directly
related to elevated phosphorus concentrations in the river. Table 6.8
presents the findings for the present case; Table 6.9 presents predictions for
a time 25 years into the future.
 Tables 6.8 and 6.9 do not recap basic information about system
characteristics nor state where data on these elements were obtained. It can
be assumed that a mixture of state and federal agency sources, private
sources (for instance, industrial records or septic system installation
records), key informant interviews, and simple observation were used to
generate the information given in the tables.

For this example, it can further be assumed that the planning horizon is
25 years, over which period 30% growth in population (with associated
demand increases) is expected. The goal of the process is to restore the
river to "swimmability" throughout the period of May to September.

Tables 6.8 and 6.9 obviously do not incorporate all elements of the
watershed inventory, study scoping, the development and screening of
alternatives, and the generation of output. They do, however, illustrate the
following key features of a simple assessment:

• Focus on a small number of impaired uses

• Use of a small number of indicators with which to evaluate

improvements in impaired uses

• Comprehensive inventory of sources

• Identification of key sources

• Step-wise elimination of infeasible options using systematic

evaluation techniques

• Use of present and future scenarios to capture likely trends over time

• Specific outputs, including recommendations for immediate action, for

deferred action, and for additional data collection and analysis
 The simple analysis illustrated in Tables 6.8 and 6.9 does not yield
detailed predictions about pollutant loadings from specific sources, nor
does it attempt to make accurate predictions of receiving water quality.
Desktop methods are used to generate rough, even order-of-magnitude,
estimates that allow the analyst to determine which sources are most
urgently in need of attention and which can safely be ignored. The analysis
of options is similarly simple, avoiding detailed prediction of expected
performance levels or loading reductions. A long list of options is
identified, feasible options are chosen from that list (probably on the basis
of cost and rough estimates of pollutant removal), and, in many cases, one
or more preferred options are identified for further action.

This level of analysis can be completed with available data, using

desktop methods, in a very short period of time, perhaps two to three
weeks. Yet its benefits are immediately obvious. By clarifying the roles of
the different sources of the use impairment, it helps to focus public and
agency attention where remedial actions will do the most good. In some
cases, this will mean a request for resources to undertake more detailed
assessments of specific system components (see Chapter 7) or of the
system overall. In either case, this simple and inexpensive analysis will
provide a useful foundation for dialogue and future action.

REFERENCES

Baltimore Regional Planning Council (BRPC). 1986. Technical Surnnuuy:

Jones Falls Urban Stormwater Runoff Project. Baltimore, Md.: BRPC.

Chapra, S. C. 1979. Applying phosphorus loading models to embayments.

Limnologv and Oceanography 24(l): 168.

Dillon, Peter J., and F. H. Rigler. 1974. A test of a simple nutrient budget
model predicting the phosphorus concentration in lake water. J. Fish Res.
Bd. Can. 31: 1771.

Haith, D. A., and L. L. Shoemaker. 1987. Generalized watershed loading

functions for stream flow nutrients. Water Resources Bull. 23(3): 371-
 478.

Hydrologic Engineering Center. 1975. Urban Stornnrater Runoff STORM.

Davis, Calif.: U.S. Army Corps of Engineers.

Metcalf and Eddy Inc. 1991. Wastewater Engineering: Treatmnent,

Disposal, Reuse. 3d ed. New York: McGraw-Hill.

Metropolitan Washington Council of Governments (MWCOG). 1983.

Urban Runoff in the Washington Metropolitan Area-Final Report.
Washington, D.C.: Area Urban Runoff Project.

Mills, W. B., et al. 1982. Water Quality Assessment: A Screening

Procedure for- Toxic and Conventional Pollutants. US EPA Report 600/6-
82-004a and h. Athens, Ga.: U.S. Environmental Protection Agency.

Mills, W. B., et al. 1985. Water Quality Assessment: A Screening

Procedure for Toxic and Conventional Pollutants. rev. ed. US EPA Report
600/6-85-002a and h. Athens, Ga.: U.S. Environmental Protection
Agency.

Novotny, V., and H. Olen. 1994. Water Quality': Prevention, Identification,

and Management o/'Diffuse Pollution. New York: Van Nostrand
Reinhold.

Occoquan Watershed Monitoring Lab (OWML). 1983. Final Contract

Report: Washington Area NURP Project. Prepared for Metropolitan
Washington Council of Governments, Manassas, Virginia.

Ontario Ministry of the Environment (MOE). 1987. Technical Guidelines

for Preparing a Pollution Control Plan. Report of the Urban Drainage
Policy Implementation Committee, Technical Sub-committee No. 2.
Toronto: Ontario Ministry of the Environment.

Sartor, J. D., G. B. Boyd, and F. J. Agardy. 1974. Water pollution aspects of

street surface contamination..1. WPCF 46: 458-465.
Schueler, Thomas R. 1987. Controlling Urban Runoff: A Practical Manual
fbr Planning and Designing BMPs. Washington, D.C.: Washington
Metropolitan Water Resources Planning Board.

Shelley, P. E., and D. R. Gaboury. 1986. Estimation of pollution from

highway runoff-initial results. In Urban Runoff Quality, edited by B.
Urhonas and L. A. Rocsner. New York: American Society of Civil
Engineering.

Sullivan, W. D. 1978. Evaluation of the magnitude and Significance of

Pollution Loadings fi-omr Urban Stormuater Runoff in Ontario. COA
Report #81. Toronto: Ontario Ministry of the Environment.

Thomann, R. V., and J. A. Mueller, 1987. Principles of Surfawe Water

QualihY Modeling and Control. New York: Harper & Row.

U.S. Environmental Protection Agency (US EPA). 1976. Area-Wide

Assessment Procedures Manual, Vols. 1, 2, and 3. US EPA Report No.
600/9-76-014. Washington. D.C.: US EPA.

U.S. Environmental Protection Agency (US EPA). 1983. Results of the

Nationwide Urban Runoff' Program, Vol. 1. Final Report. Washington,
D.C.: US EPA, Water Planning Division.

Vollenweider, R. A. 1976. Input-output models with special reference to

the phosphorus loading concept in limnology. Schweiz. Z. Hvclrol. 37:
53.

Wischmcier, W. H., and D. D. Smith. 1978. Predicting Rainfall Erosion

Leases-A Guide to Conservation Planning. USDA Handbook 537.
Washington, D.C.: Government Printing Office.

Zison, S. W. 1980. Sediment-Pollutant Relationships in Runoff From

Selected Agricultural, Suburban, and Urban Watershed. A Statistical
Correlation Stud v. US EPA Report 600/3-80-022, Tetra-Tech, Inc.,
Lafayette, Calif. Athens, Ga.: Environmental Research Laboratory.
 Detailed Assessment
Methods
Every step described in Chapter 6 for a simple watershed assessment has
its parallel in a detailed watershed assessment. The difference between the
two, as discussed previously, lies in the level of detail used, the quality
(and often age) of data, and thus in the precision and accuracy of outputs.
In a detailed assessment, a commitment has usually been made to collect
additional data, at additional cost and over an extended period of time.
Detailed assessments often, perhaps usually, make use of computer
simulation models, which in themselves take time to develop, test, and
apply. While the duration of a simple assessment may be weeks or, at the
most, months, a detailed assessment process usually takes months or even
years to complete.

Detailed assessments may be required when one or more of the

following conditions is present:

• Pollutant sources are highly variable in quantity and quality and thus
demand a more detailed understanding of their behavior in space
and/or time.

• The dynamics of the receiving water are complex, for instance, in a

large river or lake, requiring assessment of impacts in two or three
dimensions as well as over time.

• Available data are inadequate to characterize sources or predict the

impacts of management measures on the receiving environment (as in
the case shown in Tables 6.8 and 6.9 with suspected in-stream
resuspension of bacteria and phosphorus from river sediments).

• There is a need to optimize the behavior of a complex system such as

a storm or sanitary sewer system, or a sewage treatment plant, where
several unit processes or steps may be involved between influent and
effluent.
 Whereas a simple assessment relies on rules-of-thumb and simplified
equations, it detailed assessment may (but does not always) use data and
algorithms tailored to the site of interest, as well as two- or three-
dimensional analysis of physical and chemical properties. The "but" arises
because, as discussed in Section 6. I. in some of the computer simulation
models described later in this chapter, the predictive equations upon which
the models are based are in fact quite simplistic. The analyst may not,
however, realize this, because those simple equations may be solved for
many locations in a basin, and for every hour or half-hour of model time.
In other words, the "complexity" of the model cones not from its intrinsic
structure, but from the computational demands of hundreds or thousands of
recalculations.

It is absolutely essential that the user understand the strengths and

weaknesses of any models to he used in watershed assessment. High cost
and high "complexity" does not guarantee high accuracy. As SWMM guru
Bill James has often commented, "All models are wrong. Some are useful.
The trick is to know why, and how."

7.1 THE DETAILED WATERSHED INVENTORY

As described in Chapter 2 and Section 6.1.1, the purpose of a watershed

inventory is to develop an understanding of the many physical, chemical,
and biological features and processes that affect water movement in a
system. Data on features alone, such as water flows or quality, are of use
only in describing the current condition. They cannot be used to predict
changes in future land use or population situations or under the influence
of potential management actions. To this end, it is also necessary to
understand watershed processes and to use that understanding to inform
predictions of watershed conditions in different situations.

A watershed inventory serves two purposes:

1. To determine the nature and extent of use impairments in the basin,

and
 2. To identify the causes of existing use impairments.

As such, the inventory provides the foundation for scoping the

watershed study-that is, for deciding which problems are most urgent and
which sources contribute most to those problems.

7.1.1 Using Existing Information

Most detailed assessments begin in the same way as simple assessments,

with compilation of existing data (see Tables 6.1 and 6.2). As discussed in
Section 6.1, these data are seldom sufficiently detailed or site-specific for
the purposes of most planning initiatives. In a simple assessment, the
analyst could nevertheless proceed, using simplifying assumptions or
modifying data from neighboring systems. In a detailed assessment, there
is likely an expectation that the analysis will yield:

• Identification and detailed characterization of specific sources

• Quantitative evidence in regard to the performance of different

management alternatives

• Elucidation of processes, and thus cause-and-effect relationships

within the basin (e.g., proof that algal blooms are primarily linked to
high phosphorus levels)

• Detailed and quantitative projections about the impact of specific

remedial measures on in-stream hydrology, water quality, and
biological systems

The fundamental premise of a detailed assessment is that it will provide

definitive answers to key planning questions, and although it may be
repeated periodically, there is no implication that there will be "follow-up"
data collection or analysis. The detailed assessment aims to collect and
analyze all the data necessary to answer the questions that are posed at
present. Thus, it is usually necessary to proceed beyond existing data
sources to the collection of new data.
7.1.2 Filling Data Gaps

For a complex computer simulation model with lengthy input data

requirements, existing data are seldom adequate for a detailed analysis. A
typical dynamic simulation model requires input data (for instance
meteorological conditions, stormwater quality and quantity, and effluent
flows) on a one- or two-hour time step; that is, input data files that contain
real or realistic values for variables such as hourly (or two-hourly) solar
radiation, rainfall, and storm flows. In addition, a typical dynamic model
requires definition of rate constants and limiting factors for the wide range
of processes that are to be reflected in the simulation. The need to collect
such a huge amount of data is often a daunting prospect. Indeed, it is
primarily in attempting to fill gaps in the watershed inventory that the
water manager becomes aware of the incremental costs and time
associated with a detailed assessment. (Data collection can, in fact, be so
costly and timeconsuming that it limits the computer simulation methods
that can be used. This problem is discussed further in later sections of this
chapter.)

Whereas a simple assessment uses mostly off-the-shelf information,

however incomplete or dated, a detailed assessment usually involves
extensive field work to complete and/or update databases on some or all of
the following:

• Use impairments and water conflicts (for instance, using user surveys,
aerial mapping, satellite imagery, or in-depth analysis of agency
records)

• Hydrology (often requiring the collection of new or more detailed

streamflow data, including stream cross sections, analysis of critical
low and high flow levels, lake bathymetry, aquifer flows, and similar
data)

• Water quality (which may incorporate both routine sampling, perhaps

weekly or monthly, to determine "average" conditions at one or more
depths and at multiple stations, and intensive sampling during wet
weather to determine the shape and height of the hydrograph and
 associated pol- lutographs; detailed data collection may also be
targeted at extending the number of parameters for which data are
available-for instance, to add a range of industrial organic substances
in addition to "conventional" pollutants like solids and phosphorus)

• The rates of physical, chemical, and biological processes in the system

and the factors that influence those rates

• Biological systems (mapping of aquatic community structure,

including algae, rooted plants, invertebrates, fish, and amphibians;
terrestrial communities, including woody and herbaceous plants,
mammals, birds, reptiles, and insects; habitat types; and the presence
of rare or endangered systems)

• Population and land use (using agency records, census information,

aerial mapping, satellite imagery, and "ground truthing" to confirm
interpretations of remote data)

• Facilities and infrastructure (including detailed characterization of the

behavior of sources in space and time, such as video surveillance of
sewer systems to determine the condition of the physical fabric,
intensive sampling of effluent quality and quantity, collection of data
on bypass frequency and volume, and detailed projections of the size
and nature of facilities proposed for the future)

7.1.3 Sampling Design Considerations

Which data should be collected, and where, and how frequently, will
depend on the questions being asked. Accurately mapping current
conditions in a watershed is essentially a problem of sampling design:
choosing an appropriate set of samples in space and time that will reflect
actual conditions with reasonable accuracy, but without unnecessary cost
or effort. As stated in Section 2.7.3, clear formulation of the hypothesis to
be tested provides valuable guidance in sampling design. A clear statement
of hypothesis can allow the application of statistical tools in the
development of sampling strategies and the analysis of analytical results.
 Very often, the hypothesis is something like "Current average river
concentrations for parameter X exceed desired target values." To test this
hypothesis, the analyst must first determine the "average" concentration of
parameter X and then apply a statistical test (for instance, Student's test) to
determine whether, at a given confidence level, the hypothesis should be
supported or rejected.

The average determined from field measurements is simply an estimate;

the "true" average value (based on a complete knowledge of the system)
could be higher or lower than this average. To quantify this "confidence
interval," the analyst must have some information about the variability of
the system. A good understanding of the range of variability, and the
factors that affect it, is fundamental to the development of cause-and-effect
relationships and, thus, to the testing of alternative management
approaches. It can also help the analyst determine likely high and low
conditions. In many cases the collection of additional data for a detailed
screening process therefore focuses on characterization of variability (as
opposed to average conditions) in flow or quality.

It is beyond the scope of this book to describe detailed sample collection

procedures for specific situations. The interested reader is referred to
Ostler and Holley (1997) for additional guidance. The following discussion
is intended merely to outline some major considerations in designing data
collection programs for time-variable systems.

The central problem of any sampling program, whether or not related to

water, is to obtain a reasonably accurate picture of a complex system from
a small set of observations. The fewer the observations, the greater the
chance that we have guessed incorrectly as to the true condition of the
system. For example, if there are a thousand cars in a parking lot and we
wish to determine the relative proportions of different makes of car, we
could select a small number of cars at random and use this sample to draw
conclusions about the whole. It can readily be imagined that if we choose a
hundred cars at random, we could make a pretty good guess about the
relative proportions of makes in the parking lot. But if we choose only
three cars (after all, three is a small number and would not require much
effort or time for sampling), we would very likely draw incorrect
conclusions about the system overall.

The same is generally true of water and wastewater sampling: the larger
the number of samples, the more accurate the estimate. (There is, however,
some risk in drawing conclusions from very large samples in that higher
variability-an artifact of the sampling process-may be apparent in the
sample than in the true condition.) A single sample is never enough to
characterize a time-variable stream. At minimum, a sampling program
should include at least three samples, representing low, mid-range or
typical, and high conditions. A better approach is to use available
information (however limited) about the variability of the system to inform
the development of a sampling plan. Figure 7.1 illustrates some problems
in field sampling of time-variable systems. It shows the differing
conclusions that can be drawn from different sampling frequencies-
differences of more than 100% from the lowest to the highest estimate.

More heterogeneous systems generally require more detailed and

therefore more costly-characterization. A good characterization provides
information on:
Figure 7.1 Common problems with characterization of variable systems
(hypothetical parameter measured in parts per million).

• Probable maximum and minimum values

• Level of variation at different scales (e.g., for time: minutes, hours,

days, weeks, months, seasons, years, multiyear cycles)

• Potential correlations with other variables (e.g., peak pollutant

loadings in spring could be associated with snowmelt and runoff:
 multiyear flow cycles could be related to larger meteorological
phenomena)

• Insight into cause-and-effect relationships

Some industrial effluents (for instance, those from petroleum refineries)

are relatively constant through time. Sampling of these sources may
require only a few days' effort to map normal operating ranges. In other
industries (for instance, organic chemical manufacturing) effluent quality
is highly variable because different products may be made every day of the
week. Characterization of a highly variable effluent may require months of
sampling to capture the full range of products, and thus effluent quality,
from this source. The same may be true even of municipal sewage
treatment plant effluents, depending on the nature of the sewer system and
its users. Systems with primarily residential users will have reasonably
constant flow and quality (unless there are substantial inputs from
stormwater via combined sewers or from inflow/infiltration). On the other
hand, systems that receive industrial wastewater flows (very common in
larger cities) may have highly variable quality, with occasional high pulses
of solids, BOD, metals, or toxic organic substances, depending on the
behavior of the contributing industries. Treatment plant bypass is similarly
unpredictable and difficult to characterize without detailed sampling.
Figure 7.2 illustrates the variability of effluent from a mining operation.

Ostler and Holley (1997) provide the following method for estimating
the number of samples required to properly characterize the quality of a
wastewater discharge or continuous waste generation process where
several preliminary analysis results have already been obtained. The same
approach can be used for characterizing a lake or river zone that is thought
to be reasonably homogeneous through space but variable over time. (Note
that in a body of water this zone does not necessarily extend to the bottom
of the water column, but merely to a level where conditions cease to be
homogeneous-for instance, of different temperature and, therefore,
density.)

N = (t2s2)/(RT - X) 2 (7.1)
 Figure 7.2 Example of variability in industrial effluent: suspended solids
concentrations in the final discharge from the Rio Algom (Stanleigh) mine,
Ontario (adapted from MOEE 1992).

t = t value taken from Students' t table (consult statistics text)

s2 = variance = (standard deviation)'- (see Eq. 7.2)

RT = regulatory threshold (target or standard)

Y - mean

The variance, s2, is the square of the standard deviation. The standard
deviation, s, may be calculated as:

_x; - value of individual observation

 K = mean of the sample = (Ex, )fin

This equation can he simplified for computational purposes as:

or

It will be seen that equation 7.1 is very sensitive to changes in variance.

In other words, if the system is known to be highly variable, more samples
will he required for adequate characterization than if it is relatively
constant through time. This point is illustrated in Table 7.1.

Example:

Given the following live preliminary sample results from a wastewater

stream, how many samples would be required to determine whether the
stream meets a regulatory threshold of 6.0 mg/L?

Using these values, the standard deviation can be calculated as:

Therefore the variance of this data set is simply the square of s or 1.55.

For 4 degrees of freedom (= nI = 4), and a 90% confidence level (i.e., a

5% chance that the characterization is too low, and a 5% chance that the
characterization is too high = 10% chance of having made the wrong
characterization), Student's t value (obtained from statistical tables) is
2.132.

We can therefore calculate the required number of samples as

or

so about five samples are required to determine with 90% confidence

whether the stream in question meets the regulatory standard of 6.0 mg/L.
Table 7.1 illustrates how this number of samples would change at higher
and lower variances (note that because the variance is a square, differences
in standard deviation need not be large to create large changes in variance).

Note that the method proposed by Ostler and Holley requires that
sample values be normally distributed ("bell curve" distribution). If, as is
frequently the case, values are log-normally distributed (in other words,
require a logarithmic transformation to obtain a normal distribution) or are
distributed in some other way, it may be difficult to apply statistical tests
and sample size estimation methods like this.
7.1.4 Sampling Methods

Although also beyond the scope of this book, the problem of which
sampling method to use is important to note. Basically, samples can be of
two types: discrete or composite. In an effort to reduce the costs associated
with laboratory analysis, water managers sometimes mix aliquots
(subsamples) of different discrete samples to create a "composite" sample.
In this way, they believe, an average value can be obtained over the
sampling period. Although it is true that composite samples blend, or
average, the results of several individual samples, the blended sample does
not reveal any of the variability that would have been present in the
discrete measurements. Depending on the goals of the study, this may or
may not present a problem. If only "average" water quality is required,
composite samples may be a cost-effective sampling method. But if it is
important to know the true range of results the maximum and minimum
values (for instance, to evaluate the risk of toxicity to aquatic biota) then a
composite sample will be inadequate for the purpose. The analyst must be
cautious in trading off sampling and analytical cost against the information
to be gained from the sampling effort.

Flow-weighted sampling is sometimes conducted during periods of

rapid change in flow or quality, for instance, during wet weather events. If
significant variation is anticipated, the frequency of sampling should he
increased until the system is stable again. Metcalf and Eddy (1991)
recommend intervals as short as 10 to 15 minutes for variable wastewater
flows; MOE (1987) recommends intervals of 0.5 to 2 hours for in-stream
 sampling during wet weather. If flow is monitored simultaneously, the
flow and concentration data can be combined to yield loading data for each
time interval. Simultaneous monitoring of sources and receiving water
during periods of high variability will provide very useful insights into the
relationships between the two. Marsalek (1976) provides a helpful
discussion of required equipment and sampling techniques appropriate for
sampling intermittent sources.

7.1.5 Understanding Watershed Processes

Understanding the rates of watershed processes, and the factors that

influence them, is more difficult than simply mapping existing conditions.
Chemical and biological processes are largely controlled by the
thermodynamics of fundamental reactions; physical processes are strongly
influenced by the physics of water, air, and solids.

Analysts attempting to understand the dominant processes affecting

water flow and quality within a watershed should aim to answer the
following questions for each parameter under study:

Sources

1. What are the primary mechanisms by which materials are transported

from sources to the receiving water (e.g., advection, diffusion,
volatilization, atmospheric deposition, etc.)?

2. What are the primary mechanisms (if any) by which source materials
are transformed between the source and the receiver (e.g.,
sedimentation, resuspension, adsorption, die-off, etc.)?

3. Which of these mechanisms has a significant effect on the material?

Which has a trivial effect (and thus can safely be neglected)?

4. What is the time scale at which important mechanisms operate (e.g.,

seconds, minutes, hours, days, weeks, months, seasons, years,
multiyear cycles)?
 5. What is the spatial scale at which important mechanisms operate
(e.g., microns, millimeters, centimeters, meters, kilometers, etc.)?

6. What equations can be used to represent these processes for the

purpose of predicting system response within altered management
scenarios'?

7. What factors affect the rate at which these processes occur'?

8. What data are available to substantiate the operation of processes that

are thought to be important, and the rate at which these processes
occur (e.g., rainfall, runoff, and infiltration rates, temporal variation in
industrial effluent flow and quality, erosion rates, etc.)?

9. What data remain to be collected? What sampling frequency and

spatial distribution would be appropriate for such a data collection
program?

Receiving Waters

1. What are the primary mechanisms by which materials are transported

within the receiving water (e.g., advection, diffusion, volatilization,
atmospheric deposition, etc.)?

2. What are the primary mechanisms (if any) by which source materials
are transformed within the receiver (e.g., sedimentation, resuspension,
adsorption, die-off, etc.)?

3. Which of these mechanisms has a significant effect on the material?

Which has a trivial effect (and thus can safely be neglected)?

4. What is the time scale at which important mechanisms operate (e.g.,

seconds, minutes, hours, days, weeks, months, seasons, years,
multiyear cycles)?

5. What is the spatial scale at which important mechanisms operate

(e.g., microns, millimeters, centimeters, meters, kilometers, etc.)`?
 6. What equations can be used to represent these processes for the
purpose of predicting system response within altered management
scenarios?

7. What factors affect the rate at which these processes occur?

8. What data are available to substantiate the operation of processes that

are thought to be important, and the rate at which these processes
occur (e.g., stream hydrographs, patterns of species diversity, annual
cycling of algal populations, fish spawning patterns and timing, etc.)

9. What data remain to be collected? What sampling frequency and

spatial distribution would be appropriate for such a data collection
program?

In most cases, a knowledge of watershed processes is used to generate a

one-time or step-wise accounting of mass, number, or quality in the
system. (If the accounting is repeated many times-for instance, once per
model hour, the result is a more or less continuous approximation of
watershed conditions.)

In a detailed assessment, development of the watershed inventory

therefore cannot be separated completely from the requirements of the
methodology that will be used to evaluate the performance of different
management alternatives. In other words, once a commitment has been
made to conduct a detailed assessment, data collection and development of
a methodology for testing alternatives often proceed more or less
simultaneously. Indeed, a computer simulation model may be selected
simply because it requires only data that are readily available or easily
collected.

A good example of this problem can be found in the prediction of runoff

and soil erosion from agricultural fields. Ideally, the watershed manager
may want to simulate conditions on an hourly basis, 365 days of the year.
Such a simulation would demand very detailed knowledge of existing
conditions and dominant processes-on an hourly basis. Yet data like these
are scarce and would be prohibitively expensive to collect. So in many
cases the manager will instead opt for a model that simulates "average
annual" conditions (thus losing information about system maxima and
minima), or perhaps one that simulates "event" conditions-that is, those
associated with a single rainfall event (thus losing information about other
events or soil saturation conditions resulting from previous storms). In
either case, the analyst has made a decision to trade off the information to
be obtained from detailed modeling against the cost of collecting the data
necessary to drive a detailed model.

The physical processes most affecting water flow-runoff, evaporation,

and infiltration-have been discussed in Chapter 2. The following are the
most important processes governing the transformation of water quality in
aquatic systems.

The general equation for mass balance in any given volume of water or
sediment, or in any area of land (for plants and animals) is:

with each term expressed in units of mass per unit time. (If numbers of
plants or animals are used, the units are number per unit time.) For the
purposes of the following discussion, water quality constituents are used to
illustrate the general theory. Metcalf and Eddy (1991) describe the
processes affecting the rates of change of water quality constituents as
discussed in the following paragraphs.

Transport processes are of two kinds, (1) advection, or transport of

material by the flow of water into or out of a control volume, and (2)
diffusion, or transport of the constituent by turbulence in the water.

For advection into or out of an infinitesimally small control volume, the

following equations apply:

Rate of mass increase in the control volume

 = 8C dx dv dz. (7.3)

Rate of mass entering the control volume

= CUdv dz (7.4)

Rate of mass leaving the control volume

I C+ 'C dx) (U+ SU dx) dv dz (7.5)

U = water velocity in x direction (distance per unit time)

dx, dy, dl= dimensions of the control volume in x, y, z directions

t = time

With the addition of flow components in the y and z directions, this set
of equations can be simplified and substituted to yield:

where V and W are the velocity components in the y and z directions.

Using the equation of continuity from fluid mechanics, this equation can
be further reduced to the effect of advection on concentration changes with
time, as follows:

8C 8x - V SC - W 8C (7.7)

Diffusion is caused by small velocity variations acting in conjunction

with concentration gradients. If advection is equivalent to stirring in a cup
of coffee, diffusion is the gradual mixing that occurs without the stirring
force. The corresponding equations for diffusive transport are as follows:
 Rate of mass increase in the control volume

8C = St dr dv dz. (7.8)

Rate of mass entering the control volume

EA -- A dz 8C bx (7.9)

Rate of mass leaving the control volume

~E, SC + S ( E SC ) dx~ dv & S_x S.a (7.10)

where: E, = diffusion coefficient (or "diffusivity")

With the addition of flow components in the vv and z directions, this set
of equations can be simplified and substituted to yield:

Trans%ornmtioli processes depend on the underlying chemical reactions

and thus are constituent-dependent. Usually, transformation occurs
independently of transport, and thus the effects of transformation and
transport are additive.

Some of the most important transformation processes in water quality

are:

1. The oxidation of 13Ol) (i.e., consumption of available oxygen by

oxygendemanding compounds), a first-order process in which the
oxidation rate is proportional to the amount of BOD present. The oxidation
of carbonaceous BOD can be expressed as:

r( _ Kc,L(* (7.12)

and that of nitrogenous BOD as:

 I"N -KNL,v (7.13)

L(, LN = carbonaceous and nitrogenous (respectively) BOD

concentrations (mass per unit volume)

K('. KN = rate constants for carbonaceous and nitrogenous

(respectively) BOD oxidation (expressed as units of time 1)

2. Surface reaeration occurs when the dissolved oxygen concentration of

water is below saturation and oxygen from the atmosphere enters the water
surface. This flux is proportional to the oxygen deficit and can be
calculated as:

rR=kR A (C,-C)= kR H (CS- C)=K2(Cs-C) (7.14)

kR = reaeration flux rate (distance per unit time)

A = free surface area of the control volume

V = control volume

Cs = saturation concentration of dissolved oxygen

C = dissolved oxygen concentration

H = depth of control volume

K2 = surface reaeration rate (1/time)

Although kR is more closely linked to the physics of oxygen transfer, K,

is used more frequently. O'Connor and Dobbins (1958) and Metcalf and
Eddy (1991) provide formulas for K2. Methods for field measurement of
reaeration coefficients are described in the literature; they usually employ
release/recapture of volatile gases, which can he assayed in sample
volumes of water.

3. Sediment oxygen demand is an important consideration in systems

with high loadings of organic matter that may settle out and decay-often
the case downstream of sewage treatment plant effluent discharges, for
example. For a control volume above bottom sediments, the sediment
oxygen demand can be calculated by:

ks rs = H (7.15)

ks - sediment oxygen demand uptake rate (mass per unit area per
unit time)

H = depth of control volume

Like reaeration rate, sediment oxygen demand can be measured directly

in the field using a closed chamber (sometimes termed a "respirometer") to
separate a control volume of water and its associated bottom sediments
from overlying water. Continuous measurement of dissolved oxygen
concentration within the chamber yields an estimate of oxygen depletion
over time. Metcalf and Eddy (1991) suggest values for r,5 ranging from 2-
10 gym' near municipal wastewater outfalls, to 0.05-0.1 g/m2 for mineral
soils.

4. Settling q/'suspended solids can be represented by the simple

relationship:

w rI)=HS (7.16)

S = suspended solids concentration (mass per unit volume)

 w = average settling velocity of the solids (distance per unit time)

H = depth of control volume

Settling velocity is difficult to measure but may be in the order of

0.0001 cm/sec for turbid natural waters or mixed primary/secondary
wastewater effluent.

5. Bacterial die-off is highly variable in natural waters and is affected by

a wide range of factors including temperature, salinity, incident light,
available nutrients, and the growth characteristics of the species under
consideration. The general equation for bacterial die-off is first-order:

r11 = -KI,CR (7.17)

K1 = die-off constant (I/time)

CB = bacterial density (number per unit volume)

The die-off constant, KI;, is determined empirically and is used to

capture species-specific response to the various aforementioned factors.
These rates are highly variable, even for the same species in the same
water body. Metcalf and Eddy (1991) report ranges of KB from 0.12 to 26
per day for freshwater systems, with much more rapid die-off in seawater.

6. Adsorption to solids was discussed in Chapter 2 as an important

mechanism by which water quality is altered as water passes through
porous media such as soils. Here again, removal rates are highly variable,
depending on the characteristics of the solid and the chemical constituent
affected. A more complete discussion of adsorption kinetics can he found
in the literature on groundwater contamination.

7. Volatilization is an important process for certain constituents-for

instance, methyl mercury in open waters (such as sometimes occurs in
impoundments behind water control structures), and for volatile organic
compounds (such as benzene) in general. The relationship here is similar to
that in surface reaeration (discussed earlier), except that material is
transferred out of the water surface rather than into it, and the atmospheric
saturation concentration of most compounds (the equivalent value to Cs in
the equation for surface reaeration) is usually close to zero because the
partial pressure of the material is usually close to zero. Here again, the
relationship follows first-order kinetics, with the challenge being the
determination of the volatilization constant, Kv:

rv - KVC (7.18)

Kv = volatilization constant (Mime)

C = concentration of substance in water (mass per unit volume)

If estimates of the surface reaeration rate rate, K2, are available, Kv can
be approximated by relationship to that value:

Kv Kz D„ Dc, (7.19)

D„ = diffusion coefficient for oxygen in water (area/time)

D,. = diffusion coefficient for substance of interest in water

(area/time)

8. Photosynthesis and respiration are critical processes in many systems

where the growth of rooted aquatic plants and/or algae is a prominent
force. These processes have a fundamental effect on dissolved oxygen
concentrations and are themselves affected by concentrations of nutrients,
especially phosphorus and nitrogen, but also trace elements such as iron.
They are also affected by biological processes such as crowding, tissue
aging and death, and variable growth rates. Specialized models are
available for the prediction of plant growth with associated dissolved
oxygen effects (see, for example, Lopez-Ivich 1996), although this area of
endeavor is still in its infancy. The majority of available models generate
crude estimates of biomass accumulation in response to changing nutrient,
light, and temperature conditions. A few (such as that discussed by Lopez-
Ivich) provide species-specific representations of the stoichiometry of
photosynthesis and respiration, species-specific growth, deterioration,
death, and washout rates, and the presence of additional considerations
such as "luxury" uptake of excess nutrients for storage in plant tissue. A
full discussion of these processes, and how they may be accounted for in
the testing of watershed management alternatives, is beyond the scope of
this book.

This discussion has introduced some of the most important processes

that must be understood and accounted for in compiling a watershed
inventory. Most of these processes are in fact highly complex and variable
through time and space. The foregoing discussion suggested a fairly
consistent approach to representing transformation processes primarily as
simple first-order removal mechanisms. The same approach can be used
for other processes not discussed here, such as photolysis, with the same
concerns, such as determination of removal rates. This approach
unquestionably simplifies what can in fact be an extremely complex
physico-chemical system, but it is widespread in the literature and forms
the foundation of most water quality simulation models. It is based on the
conservation of mass equation, it second-order partial differential equation
that is difficult to solve in three dimensions but that can be simplified in
various ways.

The general equation, incorporating both transport and transformation

processes, is:
 i = transformation process index

j = input identification index

Ez = diffusion coefficient in the x direction

E,. = diffusion coefficient in the v direction

E= diffusion coefficient in the z. direction

U = water velocity in x direction

V = water velocity in v direction

W = water velocity in direction

r; = rate of transformation (production or consumption)

In summary, the challenge of compiling a watershed inventory for a

detailed assessment is to collect enough information to answer the
fundamental management questions under study, without wasting scarce
human or fiscal resources. Usually, this means:

1. An evaluation of the current condition of the watershed ecosystem

(components )

2. Identification of the physical, chemical, and biological mechanisms

that influence water quality and flow in the watershed

3. An assessment of the quality and quantity of available data on

watershed components (both sources and receivers) and mechanisms

4. Preliminary selection of a methodology for evaluating the

performance of alternative management strategies
 5. Comparison of the data requirements of the methodology with
available data

6. Development of a field program to collect new data or supplement

missing or outdated data

7. Evaluation of the resource requirements of the proposed field

program

8. If adequate resources are available, execution of the program

followed by scenario testing (Section 7.3.7) as appropriate

9. If adequate resources are not available, reevaluation of the proposed

scenario testing methodology (or selection of a different methodology)
with a view to simplifying the representation of the watershed system
in space or time, so as to reduce required input data

10. Development and evaluation of a revised field program

1 I . Continued iteration of the methodology-data requirements-data

requiredresource availability analysis until an acceptable balance of
resource expenditure and management information has been reached

7.2 SCOPING

7.2.1 Defining the Problem

Section 6.2 outlined the basic methods by which the most urgent problems
in a watershed can be determined:

1. Through a listing of existing (or perceived) use impairments and their

causes

2. By comparing existing receiving water or sediment quality with

published national or state/provincial targets

3. Through key informant interviews and media reports

 Scoping for it detailed assessment can begin in the same way, with the
proviso that detailed analysis allows the luxury of site-specific
comparisons, collection of extensive new data, and longer time for
surveys, interviews, and data reduction. Table 7.2 gives some examples of
how scoping techniques may differ fora detailed assessment as compared
with a simple assessment.

The use of current, detailed, and site-specific data allows the detailed
assessment to develop very precise statements about the timing and extent
of use impairments and other potential factors in scoping. This precision is
necessary to support the detailed questions posed to, and answers required
of, the planning process.

The difficulty with detailed analysis is that the analyst may be faced
with an overwhelming amount of data, so that the answers to even simple
questions are not always easy to find. The three questions central to
scoping are:

1. Which uses are currently impaired?

2. Of these, which are the two or three that are most important to the
community?
 3. For each of these two or three use impairments, what are the best
indicator parameters of a healthy condition? What specific (even
numerical) targets should be met, to consider the impaired use as
having been restored (e.g., total phosphorus <0.02 mg/L)?

It is easy to lose sight of these three "beacons" as more and more data
are collected. This problem does not exist in a simple analysis. For
example, a simple analysis may involve a few key informant interviews
and a review of historical documents and media reports, concluding that
most uses are unimpaired but that there is a problem with "swimmability"
in the summer. In a detailed analysis, existing data may be supplemented
with intensive sampling at specific bathing beaches, demonstrating that
sometimes desirable conditions are met and sometimes they are not. What
should the analyst conclude from this?
 Figure 7.3 illustrates how the relative contribution of a single
"important" source can change, depending on the season, weather, or even
time of day.

Figure 7.3 illustrates the need to develop reliable and specific indicators
of progress toward desired watershed conditions-the evaluation criteria
discussed in Section 5.5.2. The higher costs and staff requirements of a
detailed assessment therefore demand care in defining problems, in
identifying planning constraints, and in selecting appropriate indicators of
progress.
 Given the vast range of problems and sources that may be found in any
given planning area, it is difficult to generalize about scoping
methodologies. The following points may be helpful in understanding the
different challenges involved in scoping a detailed study as compared with
a simple study.
 Scoping is usually considered to comprise two separate tasks:

1. Boundary setting, and

2. Focusing.

Boundary setting refers to the need to place limits on the time period,
spatial extent, and populations that will he evaluated in the planning
initiative. Focusing means choosing the two or three most important issues
or problems, which will form the focus of most of the sampling and
analytical effort.

In a simple analysis, boundary setting is usually crude and general, and

the justification for the choice of boundaries may he simple consensus. For
instance, agency staff undertaking a simple analysis may simply decide on
a 20-year time frame, because that is the expected life of the local sewage
works, and to use the watershed boundaries as a geographic boundary (for
analytical simplicity), excluding groundwater considerations (because data
are too scarce and analytical techniques too difficult), and focusing only on
human populations (again because of' concerns about potential resource
expenditure if many species were assessed).

By contrast, a detailed analysis usually involves careful and detailed

scoping, as described in the following paragraphs.

Boundary Setting-Time Time boundaries (the planning horizon) may be

determined through a process of infrastructure evaluation (expected
facility life), surveys of public opinion (determined by in-person or mailed
surveys, public meetings, workshops, etc.), and an iterative series of
boundary-setting proposals that are successively reviewed and revised by
stakeholders. Planning may be conducted over several time periods, for
instance an immediate future ( I to 2 years post-project), a short-term
future (3 to 5 years post-project), and a long-term future (10 to 50 years
post-project or even longer). The differing requirements of these planning
levels demand different types of information, both at the inventory stage
and in the development and testing of alternatives.
 Many major projects will involve construction activities and the use of
heavy machinery. This construction phase will have its own special
impacts and concerns (for example, dust, noise, and vibration) that are
separate and different from those that will occur during system operation
or closeout. It is important to understand how long such a construction
phase might last and the nature of the impacts that may occur in this stage.

A different set of concerns is likely to arise during the operational phase,

the time following completion of construction when the system or device
is operat ing normally. The operational phase is usually much longer than
the construction phase-perhaps 25 years, as compared with 3 or 4 years for
construction.

Finally, the analyst must understand the possible implications of

"closeout" or decommissioning of the project. When the operational life of
a structure or facility is finished, some action has to be taken to close it
down and seal it off from the environment. In some cases, this will mean
dismantling it; in others, it may mean capping an underground structure.
Whatever the necessary action, the analyst will need to understand what is
involved with closeout, how long it will take, and whether ongoing care is
required. If perpetual care is needed, as it would be in the case of a landfill
or abandoned mine, the post-closeout phase, with its associated impacts,
could continue for tens or hundreds of years.

In certain cases different time boundaries may be set for different

environmental compartments or media; this is seldom required in a simple
analysis. For example, planners may decide to look at surface water quality
and flows over a 20-year time horizon, but groundwater quality and flows
(especially where there has been groundwater contamination) over 100
years or longer. In such a situation, predictive simulations over the longer
term would focus only on the compartment of interest, in this case
groundwater.

Choosing one or more planning horizons takes work and thought but is
seldom a highly controversial issue. The exception may be where
contamination by persistent, toxic, or radioactive substances currently
exists or is predicted for the future. In such instances people who hold
more protective, resourceconservative values may prefer a very long
planning horizon, anticipating the potential for cumulative or
intergenerational effects. Others may hold conflicting values, perhaps
emphasizing careful expenditure of funds or anticipating the development
of better treatment technologies in the distant future. Such a conflict may
arise, for example, in regard to deep-water disposal of tailings from mining
and milling operations. Deep-water disposal offers a way to "bury" tailings
where they will not cause acid mine drainage-in the deepest waters of a
lake or reservoir. Yet if these materials are especially toxic, as in the case
of uranium mine tailings, downstream water users may prefer an analysis
that extends over hundreds or thousands of years to be sure that predicted
releases from deep waters are low or nil.

Boundary Setting-Space The task of setting spatial boundaries on an

analysis can be both more difficult and more controversial than setting
time boundaries. For one thing, impacts of current or proposed
management actions may have both local effects (such as disturbance of
fish habitat) and very widespread effects (such as changes in local and
regional traffic patterns). It may not be clear whether it is preferable to
adopt a very local, intensive scope or a much larger, and perhaps more
general, scope. As with time boundaries, it is possible to conduct an
analysis at several scales. A very local scope (a radius of less than a
kilometer around the proposed site of an activity) could be used for
situations where there is a single major source or where intensive activity
is planned for the future. A larger scope (for instance, the watershed
boundaries) could be used to evaluate alternative management scenarios
for the whole basin. And if major changes are anticipated within the basin,
it may be necessary to adopt a scope that extends beyond the watershed
boundaries; this may be especially true for groundwater flow and surface
traffic patterns. Different levels of detail, and even different watershed
components, can be examined at each scale of analysis.

It should perhaps be added here that in urban centers, sewer systems in

fact override natural drainage patterns and may thus supersede natural
watershed boundaries in scoping. In other words, the analyst may need to
use the "sew- ershed" boundaries rather than those of the natural
watershed.

Boundary Setting-Populations A third important element in boundary

setting relates to the choice of which species of plants and animals will be
included in the analysis. If the analyst is seeking to determine current and
future impacts of alternative management actions, the question arises,
"Impacts on whom?" It seems to be implicit in most planning initiatives
that the species of most, perhaps sole, concern is the human species. But
this is by no means always the case. Fish species in particular may form an
important part of the analysis, as may macroinvertebrate assemblages (see,
for example, the case of watershed planning for the Latrobe River basin,
discussed in Section 3.4.2; Grayson et al. 1994).

There are potentially thousands of plant and animal species in an area to

be assessed, including not only mammals and birds but insect, amphibian,
and reptile species that may be of special interest or value. Domestic
animals such as livestock and house pets may also be contributors to, and
affected by, water problems in the basin.

The task of setting boundaries on the populations to be assessed may be

the most complex and the most controversial of any boundary-setting
exercise. Even if there is consensus that the human population is the only
one of interest, there may he conflict in regard to which humans are
affected, and how. Furthermore, within what appears to be a single user
group (for instance, recreational boaters) there may be conflict between
Stillwater boaters, such as sailors and canoeists, and users of motor craft. If
resources permit, a detailed analysis should include all significant users. If
these populations-or, in the case of nonhuman animals and plants,
advocates for those populations-are allowed to speak for themselves, the
scoping exercise becomes more straightforward. Problems, and conflict,
arise when one group, such as agency staff, attempt to make these kinds of
scoping decisions without input from the community. The techniques
described in Chapter 4 are especially helpful in developing scoping
guidelines that are acceptable to all water users.
Focusing Implicit in Chapters 3 and 4, and in the foregoing discussion in
this chapter, is the notion that community needs and values should drive
the scoping exercise. That is, there is no single correct outcome toward
which planners should strive. Rather, an appropriate scope for the detailed
study will emerge only after discussion, preferably ongoing, with the
community affected by the plan. This premise reflects Bishop's (1970)
sense of watershed planning as an mechanism for social change-for
moving from the current, unacceptable condition to some condition that is
desired by the community.

Focusing is the logical extension of that concept. Focusing recognizes

the problem of too much data to analyze meaningfully and demands that
the analyst examine the issues that are truly key to the community. These
are usually strongly felt issues and, thus, factors that may, if treated
cavalierly, adversely affect the implementation success of a management
plan. Focusing asks the community to decide which problems it wants
solved first and then directs the planners to expend time and resources on
those problems, with the result that secondary issues are treated in much
less depth or may even be ignored.

In a detailed assessment, with the luxury of time and money to spend on

data collection, careful focusing should emerge as it matter of course: the
key issues in the basin will be mentioned again and again by those
involved in the process. Key informant interviews (discussed in the
following paragraphs) provide an excellent mechanism for developing a
clear focus for the planning initiative. It must be emphasized, however,
that focusing is not a scientific process. That is, no amount of data analysis
will reveal with certainty the key foci of public concern. Data analysis will
certainly help to reveal areas where beneficial uses are, or soon may be,
impaired, but it will not show which issues are considered by the public to
be most serious or most urgent. Focusing is therefore not an analytical
exercise, but rather a consensus-building process. The community must
decide which problems to fix first and then agree to live with that decision.

The key informant interview has many applications in watershed

management-for instance, in the quantification of intangible benefits (see
Chapter 8). Interviews with positional, reputational, and decisional leaders
(Section 4.3.2) are almost essential in a detailed analysis. Table 4.1
illustrated some possible outcomes of a watershed planning initiative,
depending on the nature of goal setting, which includes focusing. The table
illustrates Bishop's (1970) important point that the ways in which goals are
set, and the clarity of goals, have a direct impact on the nature of change
that occurs. Inclusion of community leaders in the focusing process is
central to collaborative goal setting.

Key informant interviews should include people from all sides of an

issue (in this case, people who are likely to support the project and those
who are opposed to it, as was the case in policy profiling, discussed in
Section 4.3.4). Appropriate key informants may include government
representatives, public interest group staff, local residents' association
leaders, and professional staff from local industries or consulting
operations.

In the interview, the analyst should be open and receptive, stating that
he or she is looking for advice on scoping and would welcome new ideas
and viewpoints. Then the key informant can be invited to discuss the
important issues in the basin, boundary-setting considerations, sources of
data, and suggestions for additional contacts. The larger the number of key
informants interviewed, the more comprehensive the analyst's
understanding of issues will be and the stronger the rationale for boundary
setting and focusing.

Conflicts may certainly arise in the process of scoping. They can he

dealt with in various ways, including:

1. Technocrulic decision. The agency undertaking the study simply

makes a decision and proceeds with the analysis. Those not satisfied
with the decision are likely to withdraw their support from the
planning process and may become vocal opponents.

2. Conflict resolutions. (Alternative Dispute Resolution). Conflict

resolution (Section 4.4.7) can be used to find equitable solutions to
disagreements on boundary setting, scoping, or the impacts or benefits
 of proposed management actions. The technique has broad application
and has been used in the resolution of aboriginal land claims, forest
management issues, social and economic impacts of dams and
impoundments, and similar controversies. However, it will be
successful only if participants can be encouraged to abandon
entrenched positions and think creatively about alternative solutions.
This may not he possible if there are longstanding disputes and a
history of failed negotiations in the basin. In such instances, new
ground may he found using circle techniques.

3. Circle processes. As described in Section 4.4.5, circle processes allow

the exchange of information, including values and priorities, without a
formal meeting structure or chair. The self-regulating nature of circle
processes makes them less daunting to groups that consider
themselves marginalized by the process and may thus reveal attitudes
and solutions that would not emerge within a more rigid and
hierarchical format.

4. Other processes. Section 4.4.5 describes a variety of other small-

group processes that may be helpful in resolving conflicts. Adaptive
Environmental Assessment and Management (Section 3.4.2), for
instance, uses a series of workshops to develop boundaries, foci, and
methodologies for the assessment process.

Also part of scoping is the identification of planning constraints and the

development of evaluation criteria. These elements were discussed in
Sections 5.5 and 6.3 and are not treated at length in this section. The
central considerations relating to planning constraints and evaluation
criteria in a detailed analysis are as follows:

Constraints must be:

• Exhaustive, inclusive of all factors that may place practical limits on

the planning exercise.

• Specific and quantitative; if possible, expressed as numerical limits so

as to permit a "pass-fail" evaluation.
 • Endorsed by the decision-making group and by the wider community.

Evaluation criteria must be:

• Able to capture the vision of the community in terms of a desired

future condition.

• Useful in separating alternatives; in other words, a criterion on which

all options score equally is of little use as a screening tool.

• Explicit and specific: measurable in some objective way, or at least

permitting qualitative ranking, such that participants in the decision-
making process can agree on the performance of each option.

7.2.2 Identifying the Sources

A detailed watershed inventory will quickly reveal a range of pollutant

sources and should suggest their relative contributions to the problem of
interest. The problem of characterizing variable sources and receivers was
discussed in Section 7.1.3. As in a simple assessment, the goal of source
identification and prioritization is to determine where remedial efforts will
have the greatest impact. A source analysis such as that shown in Figure
6.3 can form the basis of such a decision.

Figure 7.3 illustrates the problem of identifying sources when the

"problem" appears to vary from time to time and place to place. Table 7.3
illustrates how clear scoping can assist in the identification of sources.

As observed in the simple analysis, and as shown in Table 7.3, there are
frequently several sources of a single problem. Careful scoping, including
attention to the true interests of the plan, can help to narrow the field of
investigation. For instance, the flooding case presented in Table 7.3 has
one primary source, rapid urbanization in the basin, but may have many
solutions. The analyst must be very clear about the goals of the study: in
this case, whether to reduce flood damages rather than the depth of
flooding. Reducing the depth of flooding would demand land management
measures upstream, because existing flood control structures are
considered adequate. Given the extent of flooding observed, changes in
land management practices would have to be extensive, and thus costly, to
be effective. If, on the other hand, the goal of the study is to reduce the
impact of flooding-the dollar costs, the injuries and death, the social
disruption-then other "sources" of the problem (in this case, building on
the floodplain) could be examined along with changes in land management
practices.

7.2.3 Evaluating the Relative Importance of Sources

Source evaluation has seven major steps:

1. Explicit identification of the issue to he resolved, usually expressed

as a use impairment.
 2. Explicit identification of the target(s) to be met (i.e., the point at
which the problem will be considered "solved" or the use restored).

3. Explicit identification of measures of progress toward desired targets,

sometimes termed "indicator variables."

4. An inventory of all sources within the basin that affect the levels of
these indicator variables.

5. Description of the behavior of these sources in space and time.

6. Ranking of these sources in terms of their overall contribution to the

problem (usually expressed as total mass, for pollutants, or volume,
for flows, contributed or as a percentage of the total mass or volume).
If a source is highly variable through time, accurate calculation of
total contribution may require subcalculations for individual events
(e.g., rainfall events, industrial spills, etc.), even on an hour-by-hour
basis using flow-proportional weighting.

7. Elimination of trivial sources (usually those contributing less than 5%

of the total load) from further consideration.

These steps parallel those followed in a simple analysis, varying only in

that the extent of detail involved is much greater. In particular, Step 6,
ranking of sources, is a complex task if done thoroughly, taking into
account the timeand space-variable nature of sources and receivers.

Much will hinge on the outcome of a detailed analysis: expenditures of

time and money, impacts on people and structures, long-term changes in
life-style and taxation. It therefore behooves the analyst to undertake a
detailed analysis of each suspected source and its behavior in space and
time. Often, this will demand both careful sampling and the use of an
effective computer simulation model to extrapolate from available
observations to a wider range of operating conditions. The computer model
is first calibrated (tuned) to a set of observations, then validated (without
further tuning) by the analyst's trying to predict a different set of observed
conditions. The predictive error of that validation run (the difference
between observed and predicted values) gives some indication of how well
the model will perform in simulating future management scenarios.

Figure 7.4 illustrates the results of a hypothetical calibration and

validation sequence for a continuous hydrologic model. Calibration was
performed on data obtained from one storm event (Figure 7.4a), and the
"tuned" model was then applied without further changes in input variables
to simulate a second storm event (Figure 7.4b).

For the purposes of understanding the relative importance of sources, a

good simulation model can provide the analyst with a way of exploring
how sources contribute to a problem under different meteorological, flow,
or operating conditions. Although available data alone give some insight
into the behavior of each source, a predictive model allows the analyst to
pull together data on several sources simultaneously and investigate their
relative influence on in-stream conditions.
 Figure 7.4 Example calibration and validation sequence for a continuous
hydrologic model (after James 1992).

In a sense, a simulation model provides the basis for developing an

understanding of current cause-and-effect relationships in the basin. It can
reveal how the relative contributions of various sources change during a
major rainfall event, for example, or can allow the analyst to examine
source behavior during high-risk-low-frequency events such as hurricanes
or major industrial upsets.

Chapter 3 discussed the need for systematic comparison using a

predetermined set of planning constraints and criteria. These criteria, in
particular, become critical in the evaluation of sources in the basin, and in
subsequent testing of management options (Section 7.3). It is therefore
recommended that planning constraints and evaluation criteria be reviewed
carefully and endorsed by the decision-making body before further
analysis of remedial actions begins.

7.3 DEVELOPING AND SCREENING MANAGEMENT

ALTERNATIVES

7.3.1 Developing Alternatives

As in a simple analysis, the development of management alternatives in a

detailed analysis should flow naturally from an inventory of sources. As
discussed in Section 6.3.1, the steps to be followed are:

• Identify the sources that contribute most to the problem under study.

• For each nontrivial source (those contributing, say, more than 5% of

the total load), identify at least one management option (i1' possible)
in each of the following categories:

"Do nothing"-retaining status quo

Structural measures-built technologies or structures

 Vegetative approaches

"Best management practices" (nonstructural measures)

Ideas for management options can be obtained from the literature, from
local experts such as agency staff', consultants, or university researchers,
or can be formulated de novo by stakeholders or decision makers. The
various techniques described in Chapters 4 and 5 are helpful in structuring
idea-generation processes for maximum productivity.

A more challenging task is evaluating the performance of different

management options. The outcome of the detailed analysis is likely to he
real dollars spent on real, tangible improvements, not just completion of a
paper exercise. Each step of option screening must therefore be carefully
justified, including a full description of anticipated predictive accuracy.
Statements of relative effectiveness ("better than," "not as good as," etc.)
will not provide the justification necessary to support decisions about
resource allocation and expenditure, which must often be approved by
elected officials who are accountable to the larger population. For these
reasons, computer models are now routinely used in detailed watershed
studies and are required for the preparation of certain discharge permit
applications under the U.S. Clean Water Act.

Although the time and effort needed to build and calibrate a good
simulation model may be daunting, the rewards can be tremendous. Some
of these benefits include:

• Insight into basin processes and cause-and-effect relationships

between sources and receiving water impacts

• Compilation of an extensive database that may reveal other issues or

trends not previously recognized

• A powerful, flexible tool for the comparison of management

alternatives and for extending the understanding of the system to
periods for which data are unavailable.
 Section 7.3.2 describes some approaches to simulating existing
conditions and predicting the impacts of management changes on a current
or future system.

7.3.2 Types of Simulation Models

The mid -to late 1970s and early 1980s saw the development of dozens,
perhaps hundreds, of sophisticated computer simulation models for use in
water management. These models were costly to develop and costly to run,
requiring the use of minutes or hours of processing time on a mainframe
computer. Through the later 1980s some of these models were converted
for use on desktop microcomputers, although most remain Fortran-based
and, to a large extent, inconsistent with newer software technology. Many
models were developed for a given task and/or geographic location and
may have limited application in other systems. The huge range of models
now available may be confusing to the inexperienced user, perhaps
especially because there is seldom a clear "right" choice. The following
discussion may be helpful in assisting water managers in selecting a model
appropriate for their needs. Figure 7.5 illustrates how different models may
be used at different stages of the planning process.
Figure 7.5 Use of computer simulation models in the development of
watershed management plans.

Models can be of various types:

1. Simple statistical relationships (such as the frequencies of violation,

as shown in Table 6.6, or predictions based on the expected frequency
 distribution of flows or pollutant loads).

2. Complex statistical or empirical models incorporating nonlinear

processes or time-variable rates.

3. Unit area loads (e.g., expected sediment loss per hectare from a given
land use).

4. Simple relationships incorporating interaction between basic

processes such as slope, runoff, and roughness (e.g., the Universal Soil
Loss Equation).

5. Simple steady-state deterministic models (which are likely to be

based on the same relationships as in (3) but which incorporate more
detailed hydrologic functions, allowing prediction at multiple
locations in the system under study.

6. Dynamic deterministic models incorporating step-wise changes in

input and generating step-wise changes in output. The time step
employed in a dynamic model may vary from a few minutes to days
or weeks, but is typically on the order of one to two hours. These
models usually incorporate more detailed and comprehensive
representations of meteorological, hydrologic, chemical, and
biological phenomena rather than lump complex processes into a
handful of algorithms.

Figure 7.6 illustrates the components of a typical dynamic, deterministic

hydrologic model. It is important to note that models like this simulate
flows, runoff, and infiltration processes in a detailed fashion. By contrast,
simulation of water quality is often very crude and weak. For instance,
sediment buildup and wash-off may be estimated by the simple
relationships described in Section 6.3.3. The load of a given pollutant (e.g.,
total phosphorus or total zinc) may then be estimated simply as a
proportion of total sediment load. This approach (discussed in regard to the
use of potency factors in buildup and wash-off simulation) was also
described in Section 6.3.3. This is a good example of the point made in
Chapter 6 that even the most computationally complex models may reduce
complex natural processes to very simplistic representations. Failure to
understand the weaknesses, not just the strengths, of deterministic models
may lead the analyst to make naive decisions or to misrepresent findings to
the decision-making body and the wider public.

Figure 7.6 Components of a typical continuous, deterministic hydrologic

model.

Table 7.4 summarizes the characteristics of major water quality

simulation models.

There are many other simulation models available, and those outlined in
Table 7.4 are continually revised and improved. Novotny and O1em (1994)
describe some other types, including stochastic and neural network
models. The interested reader is encouraged to contact model authors or
responsible agencies for current information on the models in Table 7.4.

Models are not always used singly. It is not uncommon to find several
models, perhaps of different types, linked together in the modeling
process. (Indeed, several of the models described in Table 7.4 incorporate
submodels or model linkages.) An example of a linked model system is
shown in Figure 7.7. Clearly, the challenges involved in building,
calibrating, and validating a model increase substantially when more than
one model is in use. Nevertheless, use of linked models allows the analyst
to tailor simulation tools to the particular task at hand. In the work of
Lopez-Ivich (1996), for example, the US EPA channel model WASP5,
which contains good two-dimensional deterministic hydrologic and water
quality routines, was linked to the plant growth simulation model ECOL,
which contains more sophisticated algorithms representing plant growth
and nutrient utilization dynamics for three different aquatic plant species.
In this way, Lopez-Ivich was able to reduce computational complexity
while making optimum use of each model's strengths.

7.3.3 Selection of an Appropriate Modeling Methodology

Given the enormous range of modeling approaches possible, the analyst is

faced with a difficult decision: which model will provide the desired level
of information within the desired time frame and budget? Table 7.4
illustrates that the choices to be made include the following:

1. Dimensionality: To simulate the system in one, two, or three

dimensions?

2. Time period: To simulate the system for a single rainfall event (in that
case, which event?), for several individual but not necessarily
contiguous events, or continuously'?
 3. State: To simulate the system at "steady state" (a single set of
conditions) or "unsteady state" (flow and other conditions varying as
they naturally would)?

The analyst must answer a series of other questions as well:

Figure 7.7 Example of a linked model system: flow chart for the Grand
River Simulation Model (GRSM) (after Willson 1981).

1. What parameters will be used to make decisions; i.e., what

parameters must be modeled?

2. What kinds of information will be useful in making decisions (e.g.,

load reductions from land sources? in-stream water quality? average
contributions from a given source, etc.)?

3. Over what time period (days, hours, seasons, years) would

predictions be helpful? In other words, is it enough to know that
average annual phosphorus will decline by 0.2 ppm in Scenario 4? Or
is it necessary to know the maximum and minimum phosphorus
concentrations that are likely to occur during a particular rainstorm`?

4. What computer models or other predictive tools are available for the
parameters that must be evaluated? Do available models predict the
vari- able(s) of interest in a realistic fashion? (i.e., are the algorithms
upon which the model is based sufficiently realistic to represent the
system, and its response to management intervention, accurately?)

5. What are the input data requirements for each model that could be
used?

6. What data have already been collected for the watershed system of
interest?

7. What hardware/software does each model require? Is the required

system available? Is sufficient memory, hard disk space, etc.
available? Is an experienced user available to operate the model or to
advise a novice user?

8. Of all the available models for the time period and parameters of
interest, which will be easiest to use and/or require input data closest
to that which is available and/or be most realistic or accurate in its
predictions?
 James (1992) has the following suggestions to make regarding model
selection:

1. Identify all the hydraulically and hydrologically significant elements

and processes in the basin so that the model(s) selected can be shown
to include all relevant processes.

2. Develop study objectives that relate objective functions to design

options.

3. Identify important performance criteria so that minimum modeling

effort can be used to select the best option.

4. Analyze systematic and random error in field data to reduce model

effort.

5. Review available models.

6. Assess study resources, including time, money, and staff.

7. Compile and review existing data. Review existing technical

expertise. Choose the minimum-cost model to deliver the required
accuracy for the desired objective functions.

8. Use a simplified data set to verify each identified process.

9. Produce objective functions for each hydrologic element, including

design options. Draw a system schematic.

10. Estimate parameters. Discretize homogeneous subspaces.

I1. Conduct a sensitivity analysis: produce a family of plots illustrating

model response to changes in input variables. Use extreme high and
low values for input variables, and assess impact on predicted values.

12. Calibrate the model using both simplified and full data sets. Choose
optimum values for the most sensitive parameters. Simulate the
system over the maximum possible number of rainfall events.
 13. Validate the model against an independent data set.

14. Interpret the output: discuss the magnitude and direction of probable
errors. Plot ranges, showing output and expected confidence hands.

15. Document the results. Identify the model and version used, archive
input, output, and source coding.

Figure 7.8 illustrates this process graphically.

There is an old saying in computing: "Garbage in, garbage out." A

model is only useful as a planning tool if the user understands how, and
how well, it represents the natural world. To gain this understanding, the
analyst must go through four steps:

1. Model verification-Determining whether the algorithms used by the

model are appropriate for the system. or should be replaced with
others more suitable.

2. Model calibration-Tuning the model so that it simulates a set of

observed conditions with a high degree of accuracy.

3. Model validation-See how well the tuned model predicts a different

data set, without changing any input variables or rate kinetics.

4. Sensitivity analysis To determine which input variables most

influence model output and must therefore be chosen with great care.

The first of these steps, verification, is usually straightforward, if

somewhat subjective. The rest will be discussed in the following sections.

7.3.4 Model Calibration

Calibration is usually the most challenging step in model testing,

particularly with a complex continuous model. Bicknell et al. (1993)
define calibration as "a test of a model with known input and output
information that is used to adjust or estimate factors for which data are not
available." The more complex the model, the more numerous the
adjustments that must be made, and the more complex the input data set.
Both input data and variable parameters (such as diffusion or removal
rates) can be in error. The model may therefore contain dozens or even
hundreds of error sources that singly or in combination can make the
model system fail to replicate the real system. Figure 7.9 illustrates some
of these sources of error.

The challenge in calibration is to determine the probable sources of

error and to eliminate them systematically, leaving (it is to be hoped) a
model with only very minor deviation from observed processes. This task
requires a combination of system understanding, common sense, and
random luck.
Figure 7.8 Sequence of tasks in model selection and application (adapted
from James 1992).
 Figure 7.9 Sources of error in model building and application.

Models may be calibrated manually by changing one parameter at a

time and trying to understand how it affects the model overall or by using
an automated (computer-based) system. Automatic calibration is much
faster than manual calibration. (Al-Abed (1997) estimates that manual
calibration requires five times more time than computer-assisted
calibration.) Most automatic calibration programs operate on the basis of
optimizing some objective function using maximum likelihood theory or a
least-squares approach for parameter estimation. Although more efficient
than manual calibration, automatic calibration may produce parameter
values that are "optimal" but that make little sense in terms of the
mechanics of the system. By contrast, manual calibration is tedious, but
provides good insight into the behavior and credibility of individual
parameter values.

A typical calibration sequence works like this:

1. The model code is compiled, and a model run is executed using either
a simplified data set or a full data set of observed input conditions.

2. Observed data are compared with (usually plotted against) model

predictions. (This step may be done manually or with computer
 assistance.)

3. The differences between observed and predicted data are evaluated,

and a decision is made as to whether the calibration is complete or
should continue. (This is seldom a clear-cut decision and may require
considerable subjective judgment.)

4. This run-compare-evaluate sequence is repeated until the differences

between observed and predicted data are within acceptable ranges.

5. The model is considered complete and ready for validation. No

further adjustments are made to model parameters.

Figure 7.10 Results of a partial calibration of the ECOL model.

Example of a Partial Calibration Figure 7. 10 illustrates the results of a

partial calibration of ECOL, a model for the simulation of three species of
aquatic plants and algae in temperate river systems. It shows a situation
common in manual calibration: certain oddities in the simulated results
suggest areas for further parameter adjustment or model revision. In the
model illustrated in Figure 7.10, diurnal variations in dissolved oxygen are
generally captured by the model for the first six weeks or so of model time.
Starting at about the third week in July of the year being simulated, the
model fails to reproduce observed oxygen peaks. This gives the modeler
several pieces of information. First, the fact that the model is reproducing a
diurnal oxygen cycle with an appropriate timing indicates that the basic
mechanisms of photosynthesis are adequately modeled. Second, the fact
that the general relationship between simulated and observed data changes
over a short period of time suggests that some process or source
originating at that time has been omitted from the model. Third, the
magnitude and timing of the shortfall gives the modeler clues as to what
might be missing.

In this case observed oxygen levels continue, on a diurnal cycle,

throughout the problem period, so the modeler can conclude that plant
growth (and thus photosynthesis) is responsible for the observed peaks.
The large difference between observed and simulated oxygen maxima
suggests that some major plant population (or increase in the
photosynthetic rate of a smaller number of plants) is causing high oxygen
maxima to continue throughout July. The model simulates the die-off of
one species, the attached alga Cladophora, in response to rising
temperatures, while two rooted aquatic plant species, Potamogeton and
Mvriophyllum, begin to take over habitat formerly occupied by
Cladophora. In fact, simulated hiomass curves clearly show this transition
taking place, consistent with field observations. So three possibilities
remain:

1. There was more biomass in the stream than was apparent from field
observations.

2. The growth kinetics of the rooted aquatic plants are not correctly
represented; in other words, the model is not growing the plants fast
enough to generate the observed levels of oxygen.

3. There is some other source of photosynthesis in the stream that has

been omitted from the model.
 Subsequent calibration demonstrated that changes to the growth kinetics
of the two rooted aquatic species produced only modest changes in
predicted oxygen levels. Field data were reviewed and judged to be
accurate and robust. There remained the possibility that some other source
of photosynthesis had been overlooked. Consultation with field naturalists
revealed that large numbers of insect larvae emerge from aquatic systems
in late June in this system. In the natural system, these larvae feed off
diatoms and other algae. When they leave the system, this "grazing
pressure" is greatly reduced and algal growth can increase rapidly creating
the high dissolved oxygen levels observed in the stream.

Calibration offers the analyst the opportunity to learn how well the
model system reproduces the real system, and to explore its strengths and
weaknesses. These insights are important in understanding model output,
especially during scenario testing (Chapter I I ).

7.3.5 Model Validation

A calibrated model has been carefully tuned to a particular set of observed

conditions. In Figure 7. 10 these were the summer months of 1993 in the
lower Grand River, Ontario. The user of the model has, however, no clear
idea of how the model would perform if asked to simulate, say, the winter
of 1957 in the same system or the summer of 1993 in a different system.
Similarly, the user cannot know how accurately the calibrated model
would reflect conditions that may occur in the future. (It is perhaps wise to
add here that model output looks official-it looks like fact-and the
incautious user can he tempted to communicate it as fact when it is, in
reality, only guesswork. Model predictions should always be shown with
some estimate of error or uncertainty, to give the user information about
how to interpret them. The AEAM process described in Section 3.4.2 has
as a cardinal rule that participants must never make a hard copy of the
output, so that they can never treat the output as fact.)

The modeler can learn more about how a model would perform under
other circumstances by testing it, without further adjustment, on a different
data set. This process is called validation. Validation results give the user
an understand ing of the probable predictive error in the model, and thus
its probable reliability in simulating future conditions with and without
management intervention. The data set chosen for validation purposes
should be similar in most respects to that used for calibration. The most
common approach in validating continuous simulation models such as
WASP or SWMM is to use data from a different year.

7.3.6 Sensitivity Analysis

The final step in model building is sensitivity analysis. This step allows the
analyst to determine which parameters are sensitive, and which relatively
insensitive, to changes. Sensitivity in itself can provide valuable insight
into the model's adequacy in representing the natural system. For instance,
the ECOL model whose results were shown in Figure 7.10 was very
sensitive to changes in species-specific temperature optima (the
temperature at which growth is best). The modeler must judge whether this
is a reasonable finding (which, in this case, it was) or not. If the response
of the model to a parameter seems odd or unrealistic, recalibration may he
required.

Sensitivity can be judged qualitatively ("very sensitive," "fairly

sensitive," "insensitive," etc.) or as a percentage change in output, given a
percentage change in input. Sometimes these results are plotted in
graphical form, as illustrated in Figure 7.11.
 Figure 7.11 Example plot of sensitivity analysis.

In sensitivity analysis plots like Figure 7.11, parameters that have

relatively little influence on model output (parameters for which the model
could be said to be insensitive) should plot as more or less horizontal lines:
very large changes in the parameter create very small changes in the model
objective function (for instance, predicted mean annual flow). Parameters
that have a major influence on model output (those to which the model is
sensitive), by contrast, will appear as steeply sloping lines (large changes
in model output as a result of small to large changes in input values).

Sensitivity analysis is a useful adjunct in understanding model results

and in anticipating predictive error in scenario testing. It can also point to
areas where the model is deficient mechanistically or where lumped
parameters should be separated into individual values.

7.3.7 Scenario Testing

Once an appropriate model has been selected, calibrated, and validated, it
is available for evaluating alternative management approaches in the basin.
Scenario testing is usually the foundation for plan development and
resource expenditure. Generally speaking, individual management options
are tested first, and their performance evaluated against the decision
criteria. Those that do not perform at an adequate level are rejected. When
all options have been tested in this way, individual options are combined
into management strategies, which may be "thematic" in nature. That is,
one strategy may, for example, encompass urban management options,
while another groups agricultural management options. Mixed strategies
may also be developed. Typically, the analyst will evaluate high-cost, high-
benefit strategies, mid-cost, "typical" management strategies, and least-
cost, lower-benefit strategies, so that decision makers have a range of costs
and performance to evaluate in arriving at a best management approach.

The topic of scenario testing and selection of a preferred management

strategy is discussed in greater detail in Chapter 11.

REFERENCES

Al-Abed, Nassim. 1997. Modelling Water Quality and Quantity on a

Watershed Scale Using Automatic Calibration and GIS. Ph.D. thesis,
University of Guelph.

Bicknell, B. R., J. C. Imhoff, J. L. Kittle, Jr., A. S. Donigian, Jr., and R. C.

Johanson. 1993. Hydrological Simulation Program Fortran: Users manual
for- Release 10. Environmental Research Laboratory, Report
#EPA/600/R-93/174. Athens, Ga.: US EPA.

Bishop, Bruce. 1970. Public Participation in Planning: A Multi-Media

Course. IWR Report 70-7. Fort Belvoir, Va.: U.S. Army Engineers
Institute for Water Resources.

Donigian, A. S., and H. H. Davis. 1978. Users Manual for Agricultural

Runoff Management (ARM) Model. Athens, Ga.: US EPA.
Grayson, R. B., J. M. Dooland, and T. Blake. 1994. Application of AEAM
(Adaptive Environmental Assessment and Management) to water quality
in the Latrobe River catachment. J. Emir. Management 41:245-258

James, William. 1992. Workbook for Stormwater Management Modeling.

Guelph, Ont.: CHI.

Lopez-Ivich, Karina. 1996. Feasibility of modeling phosphorus dynamics

in stormwater wetland. Master's thesis, University of Guelph.

Marsalek, Jiri. 1976. Instrumentation for Field Studies of Urban Runoff

COA Report No. 42. Toronto: Ontario Ministry of the Environment.

Metcalf and Eddy Inc. 1991. Wastewater Engineering: Treatment,

Disposal, Reuse. 3d ed. New York: McGraw-Hill.

Novotny, V., and H. Olem. 1994. Water Quality: Prevention, Identification,

and Management of Diffuse Pollution. New York: Van Nostrand
Reinhold.

O'Connor, D. J., and W. E. Dobbins. 1958. Mechanisms of reaeration in

natural streams. Transactions of the American Society of Civil Engineers
123:641-666.

Ontario Ministry of the Environment. 1987. Technical Guidelines fur

Preparing a Pollution Control Plan. Report of the Urban Drainage Policy
Implementation Committee, Technical Sub-committee No. 2. Toronto:
Ontario Ministry of the Environment.

Ontario Ministry of the Environment and Energy (MOEE). 1992. Status

Report: The Metal Mining Sector Effluent Monitoring Data for the Period
February 1, 1990 to January 31, 199/. Toronto: MISA Program, MOEE.

Ostler, Neal K., and Patrick K. Holley, eds. 1997. Sampling and Analysis.
Vol. 4, Environmental Technology Series. Upper Saddle River, N.J.:
Prentice-Hall.
Pitt, Robert. 1998. Unique features of the source loading and management
model (SLAMM). In Advances in Modeling the Management
of'Storinrrater Impacts, Vol. 6. Edited by William James. Guelph,
Ontario: Computational Hydraulics Inc.

Pitt, R., and J. Voorhees. 1995. Source Loading and Management Model
(SLAMM). National Conference on Urban Runoff Management:
Enhancing Urban Watershed Management at the Local, County, and State
Levels. U.S. Environmental Protection Agency Report No. EPA/625/R-
95/003. Cincinnati, Ohio: U.S. Environmental Protection Agency, Center
for Environmental Research Information.

Willson, Keith. 1981. Grand River Simulation Model. Grand River Basin
Water Management Study Technical Report Series, Report #30.
Cambridge, Ontario: Grand River Conservation Authority.
8
 Costing and Financing
Cost is a central consideration in the development of most watershed
management strategies. Water managers are usually operating under some
sort of fiscal constraints, so there is only a certain amount of money to be
spent on watershed projects. Difficult issues also arise in the costing of
individual measures, in the quantification of benefits, and in the selection
of discounting rates and analytical methods for economic considerations.
Sometimes it is important to include uncertainty-probability-in cost
analyses, adding another complication to the study.

This chapter reviews fundamental principles of cost estimation,

comparison of alternatives on the basis of costs, application of uncertainty
in economic analyses, and similar issues relevant to watershed planning
and management.

8.1 SCOPE AND MEASURES

8.1.1 Concepts of Cost

What should be included in an economic analysis of watershed projects?

Obvious components are capital costs-the labor and materials costs
required to build or implement water management measures-and operating
and maintenance costs for those structures or measures. "Costs" might also
be taken to include items that are sometimes called "disbenetits," or
impacts-for instance, impacts on the environment. These costs are much
more difficult to quantify than the costs of lumber and concrete, because
they often include intangible impacts such as reduced beauty-impacts that
are hard to express in dollar values. Section 7.5 discusses some approaches
for quantifying intangibles.

Several types of cost may he of interest:

Fixed Costs costs that are not affected by changes in production

or activity, such as taxes, insurance, salaries for management
 and administrative staff, license fees, and interest on borrowed
capital.

Variable Costs-costs that vary with the quantity of output or other

measures of activity, such as the costs of labor and material to
produce a product or service. Variable costs often constitute the
primary difference between two alternatives, because their
fixed costs will be similar.

Incremental Costs-the additional costs or revenues that will result

from increasing the output of a system by one or more units.
Incremental costs can be estimated only if the analyst can
identify a "normal operating range" beyond which incremental
costs might be expected to apply. An example of incremental
costs may be those associated with hiring summer students to
conduct a special one-time-only field investigation; such labor
would not be considered part of routine operations.

Costs can also be categorized by their frequency of recurrence. This

consideration is important in developing cash flows for long-term projects,
where money will flow into, and perhaps from, projects over many years.
These categories include:

Recurring Costs-costs that are repetitive, occurring in a

foreseeable cycle. Recurring costs include fixed costs such as
rent and insurance.

Nonrecurring Costs-costs that occur only once or infrequently, at

intervals that are difficult to predict over the long term.
Nonrecurring costs include those associated with the purchase
of major equipment or real estate to develop a new facility
(rather than to operate an existing facility).

Costs may also be categorized as direct or indirect. This distinction

becomes important in calculating the true "cost" of an alternative,
including not only the obvious, direct costs, but also less obvious costs that
are nevertheless attributable to the project. Direct and indirect costs may
be defined as follows:

Direct Costs-costs that can be measured and that are obviously

attributable to a particular product or work activity (e.g..
lumber required to build a piece of furniture and the labor
required to build that item).

Indirect Costs-costs that are shared and thus difficult to attribute-

for instance, the cost of tools that are shared among several
workers, general supplies such as lubricants, equipment
maintenance costs, heating and lighting of a shared facility, and
so on. Indirect costs are usually allocated as a proportion of
direct labor hours, direct labor dollars, or direct material
dollars, but other methods of allocation may he used.
Sometimes indirect costs are referred to as "overhead" costs.

A variety of other cost concepts are relevant to watershed planning and

management. Some of these relate to accounting issues, such as keeping
track of the "value" of an asset as it depreciates over time. Among these
are:

Cash Cost-a cash cost involves a cash payment. Cash costs are
those that are used in calculations of future expenditures.
Usually, only cash costs are used in making economic decisions
about water management. A cash cost can be differentiated
from the following types of costs.

Book Cost the estimated cost of an asset, including any

depreciation.

Sunk Cost --a cost that is behind you, that no longer figures in
current planning or analysis. Examples of sunk costs are the
costs of defunct structures or machinery. Sunk cost can also be
thought of as the difference between the book cost (estimated
value) of an asset and the price that a buyer will pay for it.
 Opportunity Cost the cost of losing an opportunity to use the
money elsewhere. An example of an opportunity cost can be
found in the development of a landfill site: the opportunity
costs are those of foregoing the chance to develop the land for
residential or industrial purposes (e.g., from the municipality's
perspective, taxes that would have accrued to the municipality,
etc.).

The concept of "life cycle costs" refers to the fact that costs are incurred
over the entire life of a project, from need definition through preliminary
design, detailed design, construction, operation, production of goods and
services, transportation and delivery, operation of facilities, retirement of
structures, and postretirement activities such as long-term monitoring. The
true cost of a structure therefore legitimately includes not only the costs of
building and maintaining it, but also the costs of designing it, operating all
aspects, and retiring it from service. The greatest potential for cost errors
lies in the early stages of the life cycle, for instance, at the conceptual
design phase. Undetected errors at this stage, carried through the life of the
facility, can have enormous cost implications. By comparison, errors in
costing the retirement phase will likely have a much smaller impact on
total project cost. The clear lesson here is to pay particular attention to cost
analysis for the early stages in a project's life cycle, and to have cost
estimates for these (and other) stages carefully reviewed by objective
analysts to assure their completeness and accuracy.

A third concept of interest is the break-even point. Although more usual

in manufacturing contexts, the break-even point also has application in
watershed management. It is a simple and familiar concept: that the
revenues from a project should equal or exceed its costs, or the project is
not viable:

Profit or loss = total revenue total costs

This can be thought of in terms of demand for some product or service

for instance, water supply. At a given demand level, revenues will be
sufficient to recoup costs. At higher demand levels, the operator of the
utility will realize a profit. At lower demand, costs will exceed revenues
and a net loss will result. If the unit costs and revenues for the product (in
this case, water) can be determined, it is a simple task to solve the equation
for the break-even demand point.

8.1.2 The Time Value of Money

At the heart of economic analysis is the notion that the value of money
changes with time. Two factors operate on the value of money over time:
interest, paid by a bank or other custodian of an investment, and inflation,
the phenomenon by which prices for goods and services are not constant
over time but fluctuate in response to a variety of social and economic
conditions.

Interest has been part of human society for more than 4,000 years and is
recorded in documents of ancient societies such as Babylon. In a sense,
interest is a fee paid by the borrower or user of money (or some other
commodity) for the privilege of that use. Legitimate interest rates in
history have ranged between a few percent and 25%; higher interest rates
were considered socially unacceptable "usury." The Bible (Exodus 22:25)
prohibits usury. Over the past two decades, interest rates in Canada and the
United States have ranged as high as 18% and as low as 2%, depending on
the time and the type of investment. Often, higher interest rates are paid for
riskier investments.

Simple interest is interest paid as a percentage of the original loan. For

example, a loan of $1,000 at 10% interest over five years would require the
following interest payments:

Year 1 10% of $1,000 - $100

 Total interest paid would therefore be $500. Simple interest is almost
never used in modern business or government.

Compound interest is the more commonly used form. In compound

interest, the "fee" is paid on the original loan amount (the principal) plus
any accumulated interest charges. Using the preceding example, a schedule
of compound interest payments would be:

Year 1 10% of $1,000 = $100

(At the end of Year 1, principal plus interest = $1,000 + $100 = $1,100)

(At the end of Year 2, principal plus interest = $1,100 + $110 = $1,210)

(At the end of Year 3, principal plus interest = $1,210 + $121 = $1,331)

(At the end of Year 4, principal plus interest = $1,331 + $133.10 -

$1,464.10)

(At the end of Year 5, principal plus interest $1,464.10 + $146.41

= $1610.51)

So in the case of compound interest, the total interest charges paid on the
loan would be $610.51, as compared with $500 using simple interest. This
concept can be expressed by an exponential relationship, as follows:

F= P(1 +1)" (8.1)

 P Present value, the principal of the loan, the amount borrowed
(orloaned)

i = the interest rate expressed in percentage

n = number of compounding intervals (if compounded once a

year, this would be number of years)

This simple relationship is the foundation of all time-value-of-money

calculations. It can be used to calculate the amount payable on a loan after
n years at i% interest, or the amount earned by an investment over n years
at I % interest (in other words, from the perspective of the borrower or the
lender).

Inverting the relationship yields the formula to obtain the present value
of a sum, given its future value:

P- F - 1 (I +i1~' (8.2)

This is a helpful formula if you know the amount that will be needed at
some point in the future and wish to find out how much you must invest
today to ensure that the necessary funds are in hand at the time they are
needed.

Other useful formulas can be derived from these basic equations. The
four most commonly described relate present and future values to annual
payments. Their application is obvious, because most financing requires
periodic payments into or out of some fund, not a simple lump sum
payment. These four equations are:

F'-A[ (8.3)

This formula allows the user to find the future value of a series of
uniform payments (for instance, monthly payments into a bank account:
how much money will have accumulated after five years?).
 The inverse is also possible to calculate (that is, the individual payment
amount required to accrue a total of F dollars at the end of the period of
investment). The amount inside the brackets is sometimes termed the
"sinking fund" factor:

A = F a (I+i)"- 1 (8.4)

Finally, we can calculate the present value (the value in today's dollars)
of a series of uniform payments. This calculation may be of interest if the
investor wishes to withdraw a certain amount in each of several years and
wants to know how much money must be invested now to support those
withdrawals:

P - A i(1 +i)" (8.5)

Again, the inverse may be calculated to determine the value of a series of

uniform payments required to repay a loan of present value P:

A=P (I +i)" 1 (8.6)

These concepts of present value, future value, and payment amount

(sometimes termed annual value, where there is a single annual payment or
revenue to be considered) are ubiquitous throughout economic analysis.
Indeed, the fundamental task of economic comparison of alternatives is to
reduce all options to a common basis of comparison, which can be present
value, future value, or payment amount (annual value).

Goodman (1984) presents an economic analysis of two engineering

options for an electric power project. The two options under consideration
are a 1,500,000-kilowatt coal-fired thermal power plant and a 400,000-
kilowatt hydroelectric power project. The scenario presented in this
example is typical of (although probably simpler than) most public sector
projects. One option, the hydropower project, will not require fuel or other
similar variable costs. By contrast, although the thermal plant will incur
fuel costs, it is larger and will thus satisfy a larger demand. Either project
may be financed by private investors or from government sources; private
investment costs are higher because higher interest rates must be paid in
the private sector. There are other differences between the two projects as
well. The hydropower project is estimated to have a 50-year life span,
while the thermal plant is expected to last only 30 years. For planning
purposes, it is assumed that both plants will operate at an "average plant
factor" (ratio of average output to rated capacity) of 55(11o. This
assumption may prove to be false over the long term-for instance, if
operation of the hydropower plant can be made more efficient, or if the
efficiency of the thermal plant declines as it ages and equipment
deteriorates.

Table 8.1 summarizes the economic aspects of the decision to he made

in this example, assuming that funding comes from private sources. Table
8.2 provides the same analysis, assuming government sponsorship of the
project.

Which of these options is "best"? The answer lies in:

• Choosing a common time period (decision horizon) over which to

compare the two options

• Choosing an appropriate discount rate (applicable interest rate, or

blended rate reflecting both interest and inflation over the planning
period)

• Choosing a common basis of comparison, for example, total cost

expressed in present-day dollars, or annual cost over the planning
horizon; various measures are possible

The following paragraphs describe the analyst's approach to each of

these steps for the example under discussion.

Choosing a Planning Horizon The choice of a time period over which to

evaluate economic aspects may be determined by land use planning needs
(e.g., regional planning forecasts extend 50 years into the future), by the
expected useful life of a major structure (such as a dam), or by other
social, economic, or political considerations. There is seldom a single
correct approach to choosing a planning horizon, although planning in the
public sector seldom extends over less than 5 or more than 100 years.
Typical water resources planning horizons fall in the 20- to 50-year range,
because this is the horizon over which population growth forecasts and
basin hydrology can be predicted with some degree of accuracy. Forecasts
over very short time periods are rare, because major public planning
exercises usually involve the expenditure of large sums of money and
public agencies must consider financing logistics, environmental impacts,
and benefits over a much longer term. The analyst is well advised to seek
the opinion of the community and interested agencies in the choice of a
planning period and to ensure that all interests and concerns will be
captured by the time period chosen. (For example, cumulative
environmental impacts such as reservoir siltation would be underestimated
if the planning period were only 5 years; 20 or 50 years would be a better
choice.)
 In the present example, the analyst has elected to evaluate costs on an
annual basis, thus eliminating the need to choose a common planning
period. This decision is discussed further in a following section, "Choosing
a Common Basis of Comparison."

Choosing an Appropriate Discount Rate As with the choice of a suitable

planning period, there is no clear answer to the problem of choosing an
appropriate discount rate the interest rate to be used in Eqs. 8.1 through
8.6. This rate can be a blended rate, reflecting both the positive action of
interest accumulation and the negative action of inflation. Typical discount
rates for major public projects probably range between 6% and 12%,
depending on prevailing economic conditions and the expected source of
funding (private sector funding usually entails higher interest rates). The
choice of an appropriate discount rate may be the most controversial
element in the economic analysis of major public projects. (Private
projects, such as those undertaken by major corporations, generally
employ discount rates and minimum attractive rates of return set by
company policy and endorsed by shareholders.) As in choosing a planning
period, the analyst should seek the approval of community and agency
representatives before proceeding to adopt a particular discount rate.
Sensitivity analysis, in which several analyses are conducted using high
and low interest rates, can give insight into the impact of prevailing
interest rates on economic viability.

In the present example, the analyst has chosen a blended (interest +

inflation) rate of 10.50%, termed the "cost of money," for projects under
private sponsorship. For government sponsorship, lower interest rates
would apply, but applicable rates may differ depending on the type of
project. The analyst has chosen a rate of 7.5% for the hydroelectric project,
and 7% for the thermal plant. Appropriate rates for government-sponsored
public projects can be determined through consultation with interested
agencies.
Choosing a Common Basis of Comparison The economic performance of
an option can be measured in various ways. The most obvious of these is
its gross or net cost expressed in present-day dollars; this is termed its
"present worth" or "present value," so comparing alternatives on the basis
of present value is called comparison using the present-worth or present-
value method.

Similarly, alternatives can be compared on the basis of their future

value. This may be a good approach if revenues are expected over the life
of the project and the analyst wishes to determine which option has the
larger value at the end of the project. This type of comparison is termed the
future-worth or.future-value method.

Any series of cash flows can be reduced to a present or future value or

converted into a uniform series of annual payments the series that would
yield a value equivalent to that represented by the lump sum. Essentially,
this method, termed the annual-worth or annual-value method, reduces a
series of nonuniform expenditures and receipts to an equivalent uniform
series of cash flows.

Options can also be compared on the basis of their internal rate

of'return, a measure of their rate of growth over the planning period, or on
their benefit-cost ratio, a measure that compares the total value of
"benetits," including revenues, to total project costs.

Each method of comparison has advantages and disadvantages. The

presentworth and future-worth methods, for example, are easily
understood and easily calculated. However, they do not provide insight
into cash flow exigencies that may arise over the project life, and they may
require the analyst to work with a very long planning period to
accommodate the various life spans of different project structures. Annual
worth is also intuitive and obviates the need for selection of a common
planning period (the year is the time unit on which all options will be
compared), but gives no sense of the financial impact of the project at the
outset, when financing issues may be crucial to success. Internal rate of
return is critical in most private sector analyses, where economic
performance may make or break a project, but it is relatively difficult to
calculate and less easily understood than present or annual worth. Internal
rate of return is used less often in public projects, which are usually
intended to provide some necessary public service and for which revenues
may be a small or nonexistent component of the project cash flows.
Benefit-cost analysis is required by some environmental and planning
legislation, but it continues to be controversial because of the difficulty of
quantifying intangible costs and benefits and the problem of deciding what
to include in, and exclude from, the analysis.

In the present example, the analyst decided to examine annual costs,

including both the annualized portion of capital (construction) costs and
the routine operation and maintenance costs that can be expected each
year. This eliminates the need to choose a common planning period,
because both projects can be evaluated on an "annualized basis" using the
following calculations.

Total project costs include both capital costs and operating costs, and
thus are incurred at irregular intervals over the entire life of the project.
Each individual expense can be converted to present-day dollars using Eq.
8.2, and these "present values" then summed over the project lifetime to
create a total cost expressed in present-day dollars. For example, expected
operating expenses in year 12 can be converted to a present value, and
expected expenses at close-out or decommissioning can also be converted
to a present value. But the two calculations would involve different values
for n in Eq. 8.2, to reflect the different times at which they occur. The
calculation for year-I2 expenses would set n == 12, and that for close-out
costs would set n - 30 or 50 (the expected project life span), depending on
which option was under consideration. Revenues can be calculated in the
same way. If both revenues and costs are included in the calculation, costs
should be shown as negative values and revenues as positive values.

The present values of individual costs (and revenues) are then summed
to create a net project cost expressed in terms of present-day dollars. This
present value or individual cash flow values can then be converted to an
annual equivalent over the life of the facility, using Eq. 8.6.
 If the analyst had instead decided to use total project costs as a basis of
comparison, a decision would have to be made about an appropriate time
period for comparison. Because the life spans of the two projects differ, we
cannot simply use either 30 or 50 years. Several other approaches are
available. The easiest is probably to choose the lowest common multiple of
years, 150 years in this case, and assume that each project would be
replaced during that time as the useful life of the structure ended. In this
case, the thermal plant would have to be replaced five times (30-year life x
5 replacements = 150 years) and the hydropower project three times (50-
year life x 3 replacements - 150 years). The analyst would then examine
expected cash flows over the 150-year period, including the capital costs
involved in replacing these major structures several times.

The lowest-common-multiple approach is often used when the expected

life span of an option is less than the desired planning period. This
frequently occurs in major public projects, where government agencies
may wish to plan over 50 or 100 years, but project structures such as
sewage treatment plants may have expected useful lives of only 20 or 25
years.

The example illustrated in Tables 8.1 and 8.2 shows how apparently
disparate management options can be reduced to a common basis of
comparison. In summary, Table 8. I shows that, under private sponsorship,
the thermal plant is overall a cheaper option despite its higher fuel and
maintenance costs, on a per-kilowatthour basis. Under government
sponsorship (Table 8.2), the difference between the two options becomes
much smaller, so the decision may be driven by factors other than
economics (for instance, social impacts or aesthetics).

8.1.3 Equivalency and the Problem of Inconsistent Ranking

In theory, a series of payments and receipts can be expressed as a present

worth, a future worth, or an annual worth, with each form of expression
equivalent to the others. This notion of equivalence is fundamental to
economic analysis. Table 8.3 illustrates this idea for four different
repayment schemes.
 In each of the schemes illustrated in Table 8.3, $8,000 in principal will
be repaid. In each, 10% interest will apply. And in each, the payments will
be made over a four-year period. The differences between the total
payments arises because of the tuning of payments and, thus, the interest
payable on the outstanding balance. In theory, therefore, the analyst should
be able to propose any of these four scenarios as equivalent to the others.
The same is true in economic analysis of major public projects: the analyst
should be able to represent the economic value of the project in any of
several different, equivalent, forms.

In general, this notion will be found to be true. If rate-of-return methods

are used, however, alternatives may sometimes appear to rank differently
using present-, future-, or annual-worth methods than with the rate-of-
return method.

The internal rate of return for an alternative is the percentage growth of

that alternative over the period of study; in this sense, it is rather like an
interest rate and is expressed as a percentage of the original investment. To
use a simple example, an individual may wish to determine whether it was
better to invest $1,000 in a bank deposit at 51/c interest or to invest the
same amount of money in a hot dog stand: which is the better investment`?
Analysis of the projected cash flows from the hot dog stand can be used to
estimate the "internal rate of return" (IRR) for that investment. If the IRR
is higher than 5%, the individual is better to invest in the hot dog stand
than to put the money in the bank. If the IRR for the hot dog stand is less
than 5%, the money is better in the bank.

Internal rate of return analysis is seldom used in major public projects.

But where it is, it can lead to serious errors in ranking alternatives if the
correct analytical procedure is not used. This is because IRR is a growth
rate, not an absolute dollar value. Calculation of IRR assumes that any
positive cash flows will be reinvested at the calculated rate of return, not at
the prevailing interest rate, as would be the case in present-, future-, or
annual-worth methods. Thus, an alternative might be ranked as "best"
using the IRR method, but second or third using one of the other methods.
This is termed the "inconsistent ranking problem."
 To get around this problem of inconsistent ranking, engineering
economic texts recommend incremental analysis of alternatives. Simply
put, this means examining alternatives one at a time, taking the lowest-cost
alternative first and moving progressively toward the highest-cost
alternative. This allows the analyst to see the performance of the base
(starting) alternative, and then of each incremental cost that would be
incurred by moving to a higher-cost option. The recommended steps are as
follows:

1. Rank the options in order of increasing first (capital) costs.

2. Take the lowest-cost alternative as the base case, or current "best"

alternative. Evaluate the present, future, or annual worth of this
option, or its IRR.

3. Determine whether the performance of this base option on the chosen

measure is adequate, or whether the money would be better left in the
bank (but note that for major public projects, the "do nothing" option
may not be viable, whether or not the economic performance of the
base option is adequate; see Section 8.2). The option will be viable
economically if its present (or future or annual) worth is positive or if
its IRR exceeds some predetermined target, such as the prevailing
bank interest rate.
 4. Identify the next most expensive option. Calculate the performance of
the incremental cash flows for this option using the same measure
applied in (2) for the base case. That is, subtract the costs of the base
case from the costs of the next most expensive option, and conduct the
analysis only on these incremental costs (and revenues).

5. Determine whether the expenditure of the additional funds to achieve

the second option is warranted. In other words, determine whether the
present (or future or annual) worth of the incremental cash flows is
positive or the IRR of the incremental cash flows exceeds some
predetermined target, such as the prevailing bank interest rate.

6. If the expenditure of the incremental funds is deemed to be viable,

the second alternative becomes the "current best" alternative. If the
expenditure of incremental funds is not deemed to be viable, the
original, base option remains the "current best" and the second
alternative is discarded.

7. Repeat steps 4 through 6 until all alternatives have been examined.

8. Select as the preferred alternative the last one for which the
incremental cash flow analysis was favorable.

This analytical sequence may appear cumbersome, but it provides a

useful framework for the systematic comparison of alternatives and guards
against the common error of inconsistent ranking. Incremental analysis
must, for this reason, be used with rate-of-return methods; it is, however,
optional with equivalent-worth methods.

8.2 COSTING MAJOR PUBLIC WORKS

8.2.1 Special Needs in the Economic Analysis of Major Public Works

As mentioned in Section 8.1, the economic analysis of major public works

differs from that of private projects in several major aspects:
 1. The capital investment required for major public projects is often
very large, as compared with private projects.

2. The ratio of fixed to variable costs is often very high for major public
projects.

3. Public utilities are required to provide service within established rate

schedules.

4. Major public works are subject to close public scrutiny and must be
highly accountable to the public and to regulatory agencies. They must
therefore keep abreast of, and incorporate, any technology that would
reduce costs.

5. Major public projects generally aim for full-cost recovery (which can
include setting aside funds for maintenance and replacement) but do
not generally seek excess revenues (profit).

6. Public utilities usually have stable demand, and therefore income, as

compared with private industries, which may experience fluctuating
price, demand, or both.

7. Major public works are usually financed with 50% to 60% or more
debt (borrowed) capital. A proportion this high would be unusual in
the private sector.

8. Major public assets can involve longer write-off periods than

privately owned structures or equipment.

The combination of long planning period, high public accountability,

and focus on cost recovery but not profit means that economic analyses of
major public projects often must include a number of elements that would
not be required in a similar analysis of a private project. Among these
elements are the following:

1. There is emphasis on the interests of the customer, as compared with

the interests of the utility.
 2. There is usually a need to "do something," so the "do nothing"
alternative is generally not viable and may even be omitted from the
analysis.

3. Administrative and supervision costs may be excluded from

economic assessments if these costs will be incurred in any case.

4. The costs of money, depreciation, income taxes, and property taxes

are usually expressed as a percentage of total capital invested.

Goodman (1984), De Garmo et al. (1997), and other engineering

economic texts contain more detailed discussion of the basic principles of
economic analysis and the application of those principles to major public
projects.

8.2.2 Utility Rate Setting

Economic analysis of major public projects can include the need to

estimate revenues, for instance, from public utilities. This is by no means a
simple task, even for the utility involved. Rate setting is the subject of
many major works. The following discussion is intended only as an
overview of the problem.

The main challenge in rate setting is to include all the fixed, variable,
direct, and indirect costs that are legitimately attributable to the provision
of the service being sold. So, for example, development of a drinking
water supply entails the costs of drilling and maintenance of wells, or
systems to extract water from surface water sources; treatment and
disinfection of the water; labor and materials; construction and
maintenance of buildings in which to house workers, equipment, and
supplies; construction and maintenance of a water distribution system; and
similar costs. All of these costs must be factored into the price charged the
customer, or the utility will end up subsidizing costs to some degree. It is
not at all uncommon for prevailing water rates to fall short of the actual
costs of service. The MISA Advisory Committee (1991) reported that, on
average, Canadian water prices were about 65% of the real cost of
providing water supply and sewerage. The difference was made up by
government grants and subsidies that are invisible to the consumer but are
often buried in the residential tax bill. The MISA Advisory Committee
endorsed the notion of "full-cost" pricing that would eliminate grants and
subsidies and make the full cost of service transparent to the customer.
While this would undoubtedly increase water costs, the consumer could
also be expected to realize a tax saving.

A typical rate-setting exercise uses the following calculation (annualized

values implied) (De Garmo et al. 1997):

Rate base = total capital costs of the plants in service

-- accumulated depreciation + materials and supplies

+ fossil fuel inventory + working capital allowance

- deferred income taxes - deferred investment tax credit

+ construction work in progress

Current thinking suggests that rates should also include provision for
sinking funds to cover future maintenance and replacement costs. In
Ontario, which is typical of the systems the MISA Advisory Committee
was describing, no such provision had been made over many years, with
the result that replacement costs for water supply and sewer systems are
now approaching $50 billion. It may be far better to build these costs into
rates when they are established.

Table 8.4 summarizes the components of current water and sewer

charges in the city of Guelph, Ontario. Guelph meters residential water
use.

Annual water and sewer charges in Guelph average about $200 a year
per household. This is less than half the equivalent charges in the city of
Kitchener, located less than 25 kilometers away. Kitchener currently faces
severe water supply shortages and has undertaken an aggressive demand-
management program to encourage water conservation. Among these
demand-management measures is higher pricing for water and sewer
services.

8.3 BENEFIT-COST ANALYSIS

8.3.1 Overview

Benefit-cost analysis is a relatively simple method of evaluating the

economic performance of a project. It was popularized by the U.S. Flood
Control Act of 1936 and has since been widely used (and misused) in the
assessment of largescale public and private projects. It has particular value
in the public sector, however, where profits are seldom sought, but where
maximum benefits are desired for each dollar spent.

The calculation of benefit-cost ratios is a relatively simple matter. More

controversial is the problem of what to include in, and exclude from, the
analysis. This problem is dealt with later in this section.
 The conventional benefit-cost ratio is calculated as the ratio of
equivalent worth of benefits to equivalent worth of costs. Any of the three
equivalent worth measures (present worth, future worth, or annual worth)
may be used, but present worth and annual worth are by far the most
common. The basic formula for conventional benefit-cost ratio is:

B benefits C costs (8.7)

For example, if present worth its the desired measure, Eq. 8.7 would look
like:

B P Whcneets C P Wants (8.8)

As discussed in Section 8.1, the term "costs" can include both capital
(i.e., construction, investment) costs and operating and maintenance costs;
thus:

B P Whenefits C P WCC + P WoM (8.9)

where: CC = capital costs

OM = operation and maintenance costs

According to the principle of equivalence, Eq. 8.9 can equally well be

expressed in terms of annual worth:

B A WheOCIits C A WC(- + A WoM (8.10)

An alternative formula, termed the "modified benefit-cost ratio," is

sometimes used and should generally yield the same results as the
conventional formula. For a present-worth analysis, this formula can be
written as:

B PWhccfit, - PWOM C PWcc (8.11)

In this case the denominator of the equation contains only the present
worth of project capital costs. The numerator contains the present worth of
benefits, less the costs of operation and maintenance.

The most usual, but not the only, approach in benefit-cost analysis is to
maximize the B/C ratio. Depending on the type of project and the goals
and objectives of decision makers, it may sometimes be important to
maximize or minimize components of the benefit-cost analysis. Some
common targets are to:

1. Maximize total benefits-choose the alternative that has the highest

present worth of total benefits, for instance, to emphasize the positive
impact a project would have on a community.

2. Maximize net benefits-choose the alternative that has the highest net
benefits (net benefits = total benefits - total costs), to draw attention to
the net (incremental) benefit to the community.

3. Minimize total costs-choose the lowest-cost alternative that still

demonstrates a B/C ratio > 1.0, perhaps because total available funds
are limited.

4. Minimize capital (investment) costs-choose the alternative that has

the lowest initial costs, but which still demonstrates a B/C ratio > 1.0,
perhaps because a fixed sum is available for project construction.

5. Minimize operating and maintenance costs-choose the alternative that

has the lowest operation and maintenance costs, but which still demon
strates a B/C ratio > 1.0, perhaps to minimize the annual financial load
of a project.

6. Maximize rate of return--for some projects like major irrigation

projects, which have significant expected revenues, cost recovery or
even cost surpluses may he possible. In these cases, decision makers
may opt for the alternative that shows the highest rate of return.

Table 8.5 illustrates how the outcome of a decision-making process

might differ depending on the goal of the process.
 Table 8.5 illustrates the following differences under different decision
crite- ri a:

1. To maximize benefit-cost ratio choose Alternative I.

2. To maximize total benefits-choose Alternative 2.

3. To maximize net benefits-choose Alternative 1.

4. To minimize total costs--choose Alternative 4.

5. To minimize capital (investment) costs-choose Alternative 3 or 4

(choice would likely depend on other factors).

6. To minimize operating and maintenance costs choose Alternative I.

As with other types of decision criteria, the analyst must make explicit
the goals of the decision-making process and the targets that are to be met
for a management alternative to be considered "acceptable." Table 8.5
reveals that any one of the four alternatives could be considered
"acceptable" or "unacceptable," depending on the criterion used to judge
performance.

An Example of Benefit-Cost Analysis The following example of a benefit-

cost analysis is adapted from DeGarmo et al. (1997). It employs a
conventional B/C ratio method for the analysis of a flood-control and
power project on the White River in Missouri and Arkansas. The two
management options under consideration were to:
 1. Build a reservoir only.

2. Build a reservoir and associated channel improvements for flood

protection.

Table 8.6 presents an overview of pre-project flood-related losses in the

White River basin. Table 8.7 presents an analysis of costs and benefits for
the reservoir-only option and the reservoir-plus-channel option.

This example illustrates that both options are viable (B/C ratio > 1.00)
but the multipurpose project has significantly higher benefits, and thus a
significantly higher B/C ratio, than the single-purpose darn. Significant
benefits accrue in the Mississippi River basin, as compared with the target
White River basin; flood losses (which would be reduced by either
management option) are not shown in Table 8.7. Finally, the "benefits"
incorporated in this analysis are only the obvious (and measurable) flood
loss reductions (prevented flood losses) and the value of the power from
the power generation facility. They do not include a wide range of other
considerations such as increased earning power among the people
employed in construction, spin-off benefits in industries affected by that
buying (e.g., house construction, automobile manufacturing, etc.). This
problem of what to include in, and exclude from, a benefit-cost analysis is
discussed in Section 8.3.2.

8.3.2 The Problem of Scoping a Benefit-Cost Analysis

In theory, analysis of benefit-cost ratios is straightforward: a project is

considered economically viable if benefits outweigh costs, or B/C >_ 1.0.
In practice, however, this analysis is often more complicated by several
factors. As mentioned earlier, there is the problem of deciding what to
include as "benefits" and as "costs." Benefits in particular can be
problematic, because of the difficulty in attaching dollar values to
intangible qualities such as aesthetics or quietude. This problem is
addressed further in Section 8.5.

The problem of scoping is also a difficult one, even for the most
experienced analyst. It is not uncommon for an agency to request a second
benefit-cost anal ysis if the first seems inadequately scoped. For example,
a conservative benefitcost analysis of an irrigation scheme might include
only measurable benefits, such as increased crop revenues, and only
capital and construction costs. An alternative analysis might comprise a
range of benefits, including quantification of improved "quality of life" for
the people who no longer have to haul water to the field, perhaps improved
project aesthetics, and secondary or indirect benefits such as the sale of
electricity for pumping. The alternative analysis might also include
quantification of a range of environmental "costs" or benefits, such as
altered water quality or flow regimes. The scoping of benefit-cost analysis
is not a simple problem and should be discussed with the community and
interested regulatory agencies. In one recent analysis of a major public-
sector project, the first, more conservative, benefit-cost analysis showed
that only $800,000 worth of benefits would accrue from more than $3
million in expenditures: clearly, the project was not viable. When this
analysis was repeated by a different firm, using a wider net of secondary
and indirect benefits, the benefit-cost ratio was found to be 3 : 1, and the
project clearly justifiable.
 The U.S. Water Resources Council (1983) provides a series of
guidelines for the evaluation of the benefits of major public water projects.
The types of benefits it considers, and proposed measures to evaluate these
benefits, are summarized in Table 8.8.

8.3.3 Classification of Benefits and Costs

Other complications in benefit-cost analysis include the problem of

whether to count certain types of benefit explicitly as benefits or as
decreased costs. The salvage value of equipment is one such problem. It
may be treated as a positive benefit, in the numerator of the equation, or as
a reduced cost, in the denominator of the equation (normally, the latter is
the case). De Garmo et al. (1997) have pointed out,

An arbitrary decision as to the classification of a benefit or a cost has no

effect on the acceptability of a project ... regardless of the classification
of a cash flow item as an additional benefit or a reduced cost, the
present worth of the project will be the same and the acceptability of the
project based on the B/C ratio will be unaffected.

8.3.4 Impact of Discount Rate on Decision Outcome

As discussed in this and earlier chapters, the choice of a discount rate can
be a very controversial issue in water management planning. The reasons
for this relate to the timing of costs and benefits over the course of the
project and can be easily illustrated by an example.

Assume that a major public project will extend over five years. Capital
costs are, of course, highest at project initiation, but benefits (for example,
flood control benefits, as shown in Table 8.7) will accrue over the life of
the project. Operation and maintenance costs will also occur over the
project life. The following simple example (developed by Reid
Kreutzwiser, Department of Geography, University of Guelph) illustrates
the effect of discount rate on the present worth of costs and benefits, and
thus on the benefit-cost ratio. Values are in thousands of dollars.
 The following table shows the impact of discount rate on the benefit-
cost ratio and thus on the "viability" of the project:

To obtain a discounted value for costs or benefits, the value for any
given year is taken (e.g., costs of $10,000 in Year 2) and converted from
that "future" value to a present worth using Eq. 8.2. Discounted benefits
are then totaled, as are discounted costs, and the ratio of the two
calculated.

The reason that the benefit-cost ratio (and thus the project viability)
changes depending on discount rate is that the benefits and costs occur at
different times throughout the project. Most of the costs occur at the
beginning of the project, so the total discounted value of costs is not much
different from the total of undiscounted costs ($254,800 discounted at 12%
vs. $275,000 undiscounted). By contrast, the benefits accrue only after the
project has been in place for some time-and a dollar five years in the future
is worth much less than a dollar today. The difference between total
discounted benefits ($249,300 at 12%) and undiscounted benefits
($385,000) is therefore much larger. As the discount rate increases, the
difference between the two also increases.

This discussion serves to illustrate that, while useful in many

applications, and indeed required by law in some, benefit-cost analysis is
far from being an objective analytical tool. It requires careful application
by the water resources analyst-and cautious interpretation by the user-to be
a reasonable representation of project pros and cons.

8.4 ALLOCATION OF COSTS AMONG MULTIPURPOSE

PROJECTS
The flood-control/power generation project described in Table 8.7
illustrated that the multipurpose project offered more benefits than either a
dam alone or (not shown in the table) channel improvements alone. Table
8.7 shows the total costs and benefits accruing from the dam-plus-channel
improvement project, but does not attempt to allocate costs and benefits to
the two component projects. The most common method for making such
allocations is called the "separable cost-remaining benefits" (SCRB)
method.

The SCRB method allocates costs on the basis of "purposes"-that is,

intended impacts of the project (e.g., irrigation, power generation, flood
control, etc.). The separable costs of a purpose are those that are clearly
linked to that purpose, and then can be calculated as the difference
between the total costs of a multipurpose project and the costs of a project
in which the purpose is excluded. In the example shown in Table 8.6, the
separable costs for the channel improvement project are the costs of the
dual-purpose project minus the costs of the dam. The purpose is also
assigned a portion of the total joint costs of the multipurpose project; this
allocation is made on the basis of remaining benefits (described in a
subsequent paragraph). The total joint costs are the difference between the
total costs of the Multipurpose project and the sum of all the separable
costs for all purposes.

The separable and joint costs allocated to any purpose cannot exceed the
"adjusted benefits" or "justifiable costs" of that purpose, which are either
the benefits attributable to the purpose by the most favorable method
available or the benefits evaluated as the costs of the most economic
alternative (usually a single-purpose project), whichever is least.

Finally, the `remaining benefits" for each purpose are the adjusted
benefits (justifiable costs) minus the separable costs. The remaining
benefits are used to allocate a portion of the joint costs to each purpose, in
proportion to the amount of each purpose's remaining benefits.

Methods of cost allocation are not perfect and, like benefit-cost analysis,
can be misapplied. Loughlin (1977) and the U.S. Water Resource Council's
Principles and Guidelines (1983) provide an overview of cost-allocation
methods, including illustrative examples, and explain some of the
shortcomings of the SCRB and other methods.

8.5 QUANTIFYING INTANGIBLES

Implicit in the notion of quantifying benefits is the problem of how to

attach dollar values to intangible considerations such as quality of life,
aesthetics, silence, and so on. While it can he argued that no such
quantification is legitimate (who can put a price on peace of mind, for
example?), the fact is that such analyses are often part of water
management planning and are routinely incorporated into benefit-cost
analyses. An understanding of how they are developed is helpful in
identifying weaknesses in the analysis and in developing new estimates.

Generally speaking, quantification of intangibles aims to determine the

socalled consumer surplus-that is, the value to the consumer over the price
that would normally be paid. There are several methods of estimating the
consumer surplus. All are fraught with error, but each can be useful if
carefully applied. Using a combination of methods may help to offset the
weaknesses of an individual method.

8.5.1 Contingency Valuation: Consumer Surveys of Willingness-to-Pay

As discussed in Chapter 4, consumer surveys are useful tools in

determining public opinion. Survey forms can include mailed, telephone,
or in-person questionnaires, but because of the need for one-to-one
interaction they tend to be costly to prepare and distribute. There is usually
a need to analyze large numbers of questionnaires to obtain representative
results; this also adds to the cost.

A well-designed survey can produce valuable insights into willingness-

topay, for example. Willingness-to-pay is a measure of how highly a
resource or asset is valued in the community. A survey may, for instance,
ask individuals how much they would be willing to pay to retain an
existing resource or to obtain it if it is currently absent. Survey design is a
fine art, however, and should be undertaken by an expert. Improperly
worded questions may result in answers that are difficult to interpret or that
do not address the issue of interest. In terms of willingness-to-pay, for
example, there is a risk that the person being interviewed will interpret the
questions as an attempt to determine the amount that should be charged for
access to a resource or asset. An example can be found in quantifying the
value of bird-watching opportunities. A survey of public willingness-to-
pay, if not correctly worded or conducted, could lead the interviewee to
believe that in the future an admission fee will be charged for access to
bird habitat. Properly conducted surveys of willingness-to-pay have been
used in decision making relating to increased water treatment levels in
major municipal centers, because they give an idea of how much extra the
public would be willing to pay (for instance, in property taxes) to obtain a
higher quality of drinking water.

Contingency valuation techniques can also include iterative bidding in a

personal-interview context. The respondent is asked to respond "yes" or
"no" to values suggested by the interviewer to obtain a specified increment
in the desired resource. The highest price the respondent is willing to pay
is the respondent's bid for the incremental improvement in the commodity.

8.5.2 The Travel-Cost Method

Some situations can benefit from application of the travel-cost method,

which, simply put, evaluates how much an individual on average or at
maximum will spend to visit a particular resource or asset, above the actual
costs of gaining admission. A good example here can be found in tourism
at national parks. The travel-cost method may identify the person who has
traveled farthest to reach the park-for instance, a visitor from Europe-and
evaluate the travel costs (airfare, car rental, gasoline, etc.) and time costs
(value of the individual's time-for instance, on an hourly or daily basis)
necessary for that person to reach the park. It may be found that the travel
and time costs incurred by this traveler approach $2,000, while admission
to the park is only $10. The consumer surplus would then be evaluated at
$1,990. Tobias and Mendelsohn (1991) used the travel-cost method to
assign a value to ecotourisni in Costa Rica. They used multiple regression
techniques to express visitation rate as a function of distance traveled,
population density, and illiteracy rate, plus some random error. Their
results suggest that travel by Costa Ricans to ecological reserves in that
country has a consumer surplus value of about $116,200 US/year, or about
$35 per visit, based on a typical travel cost of $0.15 per kilometer. These
values translate to a present worth for the ecotourism industry of between
$2.4 and $2.9 million US for domestic visitors only. Adding international
visitors brings the total consumer surplus value to over $10 million US, or
about $1,250 per hectare of ecological preserve. This is roughly 10 times
the price that is currently paid for the acquisition of new land for
ecological preserves in Costa Rica.

8.5.3 Hedonic Price Approach

The hedonic price approach can be used in cases where it can be proven
that people have historically paid more to have a certain attribute or access
available to them. The method requires a comparison of historical real
estate prices of matched properties with and without the attribute. For
example, the consumer surplus associated with ravine aesthetics may be
estimated by comparing the recent sale prices of homes that backed on a
ravine with those of similar homes that did not back on a ravine. The
difference between the sale prices can be taken as the consumer surplus
value of that resource.

8.5.4 Alternative Cost Approach

Where it is possible to identify another way to supply the resource or asset

to be quantified, the cost of that alternative can be used to estimate the
value of the consumer surplus. For instance, if existing fish habitat is to be
destroyed by some management action that will result in excessive
siltation, new habitat could be created elsewhere to replace that which is
destroyed. The cost of developing the new habitat, including construction,
vegetation, and maintenance, could be used to estimate the value of the
habitat being destroyed. Alternatively, the value of a clean drinking water
supply could be estimated as the cost of creating and maintaining an
alternative water supply-for instance, through the use of trucked or bottled
water.
8.5.5 Qualitative Techniques

A range of qualitative techniques can be used in estimating the value of

intangibles. Typically, these techniques involve ranking or scoring, but not
direct measurement. In this regard, consider the example of visits to a
national park. Visitors may be asked to rate the reasons for their visits on a
scale from I to 5. Categories evaluated may include considerations such as
solitude, exercise opportunity, time with family, nature enjoyment, and so
on. Taken over a large number of visitors, responses to a survey like this
can reveal that most peo ple visit the park to enjoy nature, indicating that
park management may wish to extend visitor services in this area. If, by
contrast, most people want "time with family," group camping or campfire
activities may be promoted. While not-strictly speaking-quantification of
intangibles, these methods are nevertheless a helpful adjunct to more
quantitative methods and provide useful insight into public opinions and
values.

8.6 INCORPORATING RISK AND UNCERTAINTY IN

ECONOMIC ANALYSES

It is frequently the case that some components of the analysis will be

subject to chance events, particularly when planning encompasses future
conditions that cannot be known with certainty. The further into the future
an event will happen, the less precise our estimates about it will be. These
estimates may include considerations such as prevailing interest rates,
utility rates, capital investment requirements, market share, inflation rates,
and taxation rates. All of these factors will certainly vary with time, and all
can in turn affect estimates of cash flow. Some are correlated with others;
some are autocorrelated (that is, values at a given point in time are
correlated with values that occurred some time earlier).

Two situations may occur. In the first, the possible outcomes are known,
and their respective probabilities are also known. This condition is called
decision making under risk. In the second situation, possible outcomes
may be known but their probabilities of occurrence are not known. The
analyst must therefore decide whether or not to incorporate uncertainty
into the decision making (decision making under uncertainty) or whether
to assume that some condition will occur with 100% probability.

Attitudes to risk and uncertainty vary with the population and with the
sponsor agency. Individuals, for instance, may willingly accept the risk of
heart and lung disease by smoking cigarettes, but adamantly oppose
exposure to much lower health risks from nuclear power generation.
Formal risk analysis, which incorporates estimates of probability with
estimates of impact, are sometimes used in the evaluation of potential
human health impacts caused by chemical pollutants, of high-impact low-
frequency disasters such as earthquakes or nuclear accident, or to develop
priorities for environmental remediation.

Various methods are employed to deal with decision making under risk
and under uncertainty. These methods are reviewed in the following
sections.

8.6.1 Decision Making Under Risk

Decision making under risk tends to be easier and more robust than
decision making under uncertainty. The basic approach to incorporating
risk is called "risk aggregation." It can be illustrated with a simple
example.

Assume that a project will result in a series of expenditures and

revenues over time, but that the precise value of those cash flows cannot
be known with certainty. The analyst probably has some insight into the
relative likelihood of various levels of cash flow, however, and can
translate that intuitive understanding into a rough probability distribution.
Figure 8.1 illustrates this process.

Figure 8.1 simply records the analyst's "gut feel" for the project over
time. For instance, the analyst may feel that a very pessimistic prediction
might he $2 million in net revenues, while a very optimistic prediction
would he $2,350,000. The most likely outcome is probably somewhere
around $2,200,000. (Although this analysis may seem arbitrary, it should
be borne in mind that an analyst with considerable experience in the
watershed under study, and with good know]edge of water resources
management techniques, is usually able to guess these outcomes with fair
to good accuracy.)

We can then assign probabilities to these three conditions-for instance,

211 each to the low and high conditions and 35% to the "most likely" case.
We can add two or three more high and low guesses to round out the range
of probabilities.

This rough probability distribution can then be converted into a

cumulative probability distribution (Figure 8.2), which in turn allows the
analyst to interpolate probabilities for intermediate values of net revenue.
From an intuitive feeling for the project, the analyst has created an
approximate probability distribution that can be used in subsequent risk
analyses.

Table 8.9 illustrates the values that can be estimated from this kind of
analysis.

Recall that each of the entries in Table 8.9 is simply a guess by an

experienced analyst. These values can, however, be used to estimate an
overall present worth for the cash flow series, based on the premise that
the expected present worth of a series of cash flows is equal to the sum of
the present worths of the expected values of the individual cash flows.
Recalling Eq. 8.2,
Figure 8.1 Hypothetical probability distribution based on analyst's intuition.

we can convert anticipated cash flows over time to a present worth.

Similarly,

E(P) = EE(F) { I 1 (1 (8.12)

where E is the expected value of present worth (P) or future worth (F).
Figure 8.2 Hypothetical cumulative probability distribution based on
analyst's intuition.
 If two projects are being compared, the analyst can estimate pessimistic,
optimistic, and expected (most likely) cash flow values for each. The
outcome is often as shown in Table 8.10.

Given these results, the analyst can make an informed decision as to

whether to accept the project with the higher potential worth (but much
higher risk) or the one with lower worth but more certainty.

Analysis of this type makes uncertainty explicit in the analysis and thus
allows the analyst and other participants in the planning exercise to make
maximum use of available information. Uncertainty in several variables
can be considered simultaneously, rather than independently, creating an
analysis that is more realistic. Although it is computationally convenient to
develop a solution under assumed certainty (e.g., interest rates will he
12%) rather than the more realistic risk (e.g., interest rates will lie between
8%/c and 14%, with the following probabilities for each interest rate ...), in
real life probabilities are seldom known. An understanding of the probable
"spread" of results can be just as important, or more important, than
estimating the "expected" value. This point is made very clearly in the
simple example in Table 8.10. Possibly most important, an explicit
analysis of risk forces the analyst to confront unknowns one at a time, and
to separate what is known from what is unknown or uncertain.

Several different methods can be used to assess problems under risk.

These are briefly discussed in the following paragraphs.

Simulation (Monte Carlo) Approach The Monte Carlo simulation

technique is often used in a complex situation with many variables. It can
also be helpful in validating the results of analytical solutions, although it
can introduce a random element that may not be present in those solutions.
 The general approach is to develop a range of estimates for revenues
and expenses and attach probabilities to each estimate. Random numbers
are then assigned to the estimates, with the number of random values
proportional to the probability assigned to the estimate. A computer
program is then written that draws at random from the pool, assigning
values to the calculation based on the random numbers drawn. Many
calculations are performed in this way, usually hundreds or even
thousands, and the analyst develops it probability distribution for present
worth using the results. Table 8.11 illustrates how a simple Monte Carlo
simulation might he set up.

Many decision techniques employ analytical solutions based on the

following elements:

"States'-Designated as Sk States are the possible future conditions that may

occur, numbered from I to k. Examples of states are different levels of
drinking water demand (e.g., S, = 250 L/person/day; S, = 300
L/person/day; S3 = 350 L/person/day, etc.) or number of bulldozers
required for a particular construction job (e.g., S, = 3 bulldozers; S, = 4
bulldozers; S3 = 5 bulldozers, etc.). States are components of an analysis
over which the analyst has no control, and which are subject to chance
occurrence.

Feasible Alternatives-Designated as Aj Feasible alternatives are the various

management decisions one could make (e.g., supply enough drinking
water for 300 L/person/day; buy four bulldozers, etc.). Each is the factual
situation that would exist under a given management decision.
Outcomes-Designated by Bak Outcomes are the combinations of "state"
and "alternative." For example, suppose that random chance has led to a
situation where 250 L/person/day is required for drinking water supply, but
the management decision that was made resulted in a supply sufficient for
350 L/person/day. Or suppose it turns out that you needed five bulldozers
("state"), but in fact you bought only three (the management alternative
selected).

Value of Outcomes-Designated by B;k The value of each outcome is the

measure of its cost or benefit. For example, if you decided to supply
enough drinking water for 350 L/person/day but needed only 250
L/person/day, you have overdesigned your utility and paid more than you
needed to, this outcome can be assigned a value, which may be either a net
cost or a reduced revenue over what would have occurred had you
designed the utility to exactly the right capacity. Similarly, if you bought
three bulldozers but needed five, you would have to lease the additional
bulldozers at extra cost. This outcome can also be assigned a value, which
again would represent a net cost or reduced revenue over what would have
occurred if you had bought exactly the right number of bulldozers.

Probability of Each State-Designated by pk The various states, which are

subject to chance occurrence, can be assigned probabilities based on the
analyst's understanding of the problem and other sources (for streamflow,
for example, long-term hydrologic records are a good source of
information about the probability of different flow levels).
 We can set up a hypothetical decision problem in matrix form, as shown
in Table 8.12.

Table 8.13 gives an example of the types of cost estimates that may be
developed in a decision under risk.

We will examine this simple example in illustrating several different

analytical methods for decision making under risk.

Assumed Certainty The analyst can simplify the analysis (while still
having gained something from the explicit statement of probabilities) by
assuming that one state is 100% certain and the others have a probability
of zero. In this case, the analysis becomes the same as if only a single state
were considered. This approach may he used in cases where corporate
policy dictates a certain outcome (for example, corporate policy may be to
use a discount rate of 12.5%, regardless of actual prevailing interest rates).
With this approach, Table 8.13 is altered, as shown in Table 8.14.

Expectation-Variance Principle If the analyst wishes to retain the separate

states and their respective probabilities in the analysis, a weighted sum of
outcome values can be calculated, as follows:

E(A1) = EV(B;ti)l)k (8.13)

or, effectively:

Using this approach and the values in Table 8.12, we obtain the
following expected annual worth in millions for Alternative 1:

The equivalent value for Alternative 2 is:

and for Alternative 3 it is:

If the cash flows for these alternatives are positive (that is, revenues),
the best choice is Alternative 2, with an expected annual worth of $2.22
million. If the cash flows are negative (expenses, costs), the best choice is
Alternative 3, with an expected net cost of $2.035 million. If two
alternatives were tied, the analyst might opt for the one with the lowest
variance.

Most Probable Future Principle In a situation where one state has a much
higher probability of occurrence than others, that state can be assumed to
be certain, thus reducing the case to one of assumed certainty as discussed
earlier and illustrated in Table 8.14. This approach should be applied only
where probabilities are greatly different, with one probability very high. In
a case such as that shown in Table 8.13, this technique would not likely be
justified unless other factors, such as corporate policy, prompted an
assumption of' certainty.

Dominance Where one alternative is always preferred regardless of the

state that prevails, that alternative can be said to dominate the analysis. In
Table 8. 13, if the goal of the decision were to minimize costs (minimize
the value of the outcome), Alternative 3 would be dominant because it has
the lowest value in every state.

Aspiration-Level Principle If there is an ideal outcome level (for instance,

total costs of $2.1 million or less, or profits greater than $2.25 million), the
analyst can evaluate each alternative in terms of its probability of meeting
that threshold. This is done by summing the probabilities (not the outcome
values) of outcomes that meet the criterion. Using the example in Table
8.13, with a desire to ensure that total costs are less than $2.1 million, we
find that:

• Alternative I has a probability of 30% that costs will he less than $2.1
million (because only State 1, interest rate - 8%, meets this criterion,
and that state has a probability of 0.30),

• Alternative 2 has a probability of 0 of meeting the criterion (all

outcome values exceed the threshold), and

• Alternative 3 has a probability of 100% of meeting the criterion (all

three states have outcome values <_$2.1 million).
 Under this criterion, Alternative 3 would be the best choice. If the
values in Table 8.13 referred to revenues, and we wanted to maximize the
likelihood of those revenues exceeding $2.25 million, we would find that:

• Alternatives I and 2 each have a probability of 20% of meeting this

threshold (State 3 only, with a probability of 20%), and

• Alternative 3 has zero probability of meeting this threshold (all values

less than $2.25 million).

So either Alternative I or 2 would be a reasonable choice; again, the

selection may he made on the basis of lowest variability or some other
factor.

8.6.2 Decision Making Under Uncertainty

Decision making under uncertainty is more difficult than decision making

under risk, and the results are less satisfactory. Goodman (1984) suggests
several steps before attempting to make decisions when several factors are
unknown:

• Collecting more detailed data to reduce measurement error and

improve understanding of the system

• Using more refined analytical techniques

• Increasing safety margins in design

• Selecting measures with better-known performance characteristics

• Reducing the irreversible or irretrievable commitments of resources

• Performing a sensitivity analysis of the estimated costs and benefits of

alternative plans

Nevertheless, situations may arise in which the analyst has no

information about the relative likelihood of various possible outcomes, but
would like to incorporate in the analysis some consideration of uncertainty.
The following methods are useful in decision making under uncertainty.

Equal Probability The simplest approach to detailing with uncertainty is to

assume that all possible future states have equal probability. If we applied
this notion to the values in Table 8.13, we find that the expected annual
worth of Alternative I is:

The equivalent value for Alternative 2 is:

and for Alternative 3 it is:

So if we are minimizing costs, the best choice is Alternative 3, with an

expected annual cost of $2.013 million. If we are maximizing revenues,
the best choice is Alternative 2, with an expected annual worth of $2.211
million. If two alternatives were tied, the analyst might opt for the one with
the lowest variance.

Maximin and Minimax Principles if some information about the range of

outcomes is available, but the analyst cannot specify probabilities for each,
it is possible to "hedge one's bets" by choosing the option that performs
best under worst-case conditions, or that which performs worst under best-
case conditions. Taking the values in Table 8.13 as the basis, if the goal is
to maximize profits, the worst-case condition is probably that with the
lowest interest rate, State 1. Examining only the column for State 1, we
find that the alternative that performs best under those conditions is
Alternative 2, at $2.2 million; the other two alternatives can be expected to
generate only $2 million. So the best choice would he Alternative 2. The
analyst has maximized the minimum revenues, hence the term maximin, as
applied to analysis of this type.
 If the goal of the project is to minimize costs, the worst case would
likely be State 3, with the highest interest rate. Examining only that
column, we find that the alternative that performs best (has the lowest
costs) is Alternative 3, so that would be the preferred option. The analyst
has minimized the maximum cost, hence the term ininimar is applied to
this type of analysis.

Maximin and minimax analyses are pessimistic in that they attempt to

find the option that would perform best under the worst conditions. It is
also possible, but less common, to examine the most optimistic case,
finding the option that earns the most revenues under best-case conditions
(maximax) or that which costs least under best-case conditions (minimin).

Hurwicz Principle While the maximin and maximax methods allow the
analyst to examine performance under the most pessimistic or most
optimistic conditions, the Hurwicz principle allows consideration of
intermediate levels of optimists or pessimism. The method employs an
"index of optimism," designated by a, the value of which is chosen by the
analyst based on an understanding of the problem and its context.

To apply the Flurwicz principle, the most optimistic performance of an

alternative is multiplied by cx and the most pessimistic by (I - (x). The two
products are added for each alternative, and the largest (for revenues) or
smallest (for costs) sum is chosen as the best alternative.

Using the examples in Table 8.13, and assuming that we are reasonably
optimistic about the outcome of the project, we could assign an index of
optimism of a = 0.70. If we are seeking to maximize revenues, then the
performance of each alternative can be calculated as:
 So the best choice is probably Alternative 1, although it is sufficiently
close to Alternative 2 that the option should probably also he examined as
viable.

8.6.3 Sequential Decisions and Decision Trees

Concepts of decision making under risk can he applied to situations in

which sequential decisions must he made-for example, in a project that
will be built in several stages. Decision trees represent this time sequence
graphically and allow the analyst to take advantage of information as it
becomes available over the course of the project. A common application of
decision trees is found in decision making where new and uncertain
technology is contemplated.

Assume that the problem under scrutiny is to decide whether to install

traditional pump technology or a new type of pump. The relevant data for
the two technologies appears in the following table:

Traditional technology may be completely reliable, but the

manufacturers of the new technology warn that it has only 65% reliability.
If the new type of pump fails, it will have to be replaced with traditional
technology, incurring "penalty" costs of, say, $3,000 resulting from added
delays. The money for the new technology will be refunded, but the
installer will have to pay the full price of the traditional technology as well
as the penalty, for a total of $18,000. Depending on the cost of the new
technology, it may nevertheless be worth taking the risk of failure. The
situation can be analyzed as follows.

The decision to be made can be represented in a decision tree, as shown

in Figure 8.3. There are three "branches" on the right side of the tree. In
evaluating this decision, the correct approach is to work from right to left,
reducing the branches to a single either-or choice. In this case, we can
reduce the "works" and "fails" branches to a single value using the
expectation-variance principle (essentially, calculating a weighted sum of
the expected costs):

We can then compare the cost of Alternative 1, installing traditional

technology ($15,000), with the cost of Alternative 2, installing new
technology and taking the chance of its failing ($14,100; see previous
calculation), and find that trying the new technology is probably justified.

We can extend this analysis to make better use of information in the

future by examining the impact of adding a field test of the equipment.
This approach is illustrated in Figure 8.4.

Figure 8.4 illustrates that installation of the traditional pump technology

without testing remains the most expensive option ($15,000) and can be
discarded. This leaves two alternatives to consider: Alternative 2, proceed
to install the new technology immediately ($14,100), and Alternative 3,
conduct a test and then decide which technology to install. The cost of
conducting a preliminary test can be calculated as the weighted sum of the
two options, plus the actual cost of the test:
 Figure 8.3 Decision tree for pump technology decision.

where T is the cost of the test.

Recall that the expected cost of Alternative 2 is $14,100. Figure 8.4

demonstrates that the company can spend up to $ 1,050 ( $14,100 -
$13,050) on the test and Alternative 3 will still he as low in cost as
Alternative 2. Therefore, if the testing can he done for $1,050 or less, the
company should choose Alternative 3 and conduct a preliminary test
before deciding which technology to install.

Decision trees capture a sequence of events, as well as the chance

events that can influence outcomes over time, in a way that traditional
economic analysis does not. They are particularly useful in evaluating
high-impact, low-risk (low frequency) events such as extreme storm
conditions, dam failure, and other "disasters". They allow the decision
maker to keep track of many possible outcomes and incorporate the
possibility of adding new information as time passes.

Figure 8.4 Amended decision tree for pump technology decision.

8.7 CAPITAL FINANCING (SOURCING)

The decision about where the money will come from is a difficult one,
particularly in private-sector projects. In major public projects, financing is
usually a combination of grants and loans from government and some
independent debt financing.

There are two main sources of capital: debt capital (i.e., borrowed
money) and equity capital (i.e., money already available in treasury). As
anyone who has ever bought a house or a car knows, there are major
differences in these sources.

Financing with debt capital binds the project proponent to paying

interest to the lenders of the capital, and to repaying the capital at some
fixed point in time. There may be conditions placed on the loan that restrict
how or when the money can he used. Under some circumstances, the
interest payable on a capital loan is an allowable deduction from income
taxes. The most common type of debt financing in major public projects is
through the sale of bonds.

Financing with equity incurs no additional interest costs. As a result,

there are no "costs" of equity financing that can be used to offset income
and reduce income tax payable. Equity capital can include funds raised
through the sale of common or preferred stock, from retained earnings, or
from depreciation funds. The following discussion is drawn largely from
De Garmo et al. (1997).

8.7.1 Bond Capital

Probably the most common source of debt capital for major public works
is bond capital. Bonds are also very common in private corporations. They
differ from "stock" or "shares" in that a bond represents only indebtedness
on the part of the borrower and does not confer any ownership in the
company on the lender.

Bonds are long-term loans made by an investor (lender) to the borrower.

The bond document specifies the interest rate to be paid, the term of the
bond, and other repayment conditions. Bonds are usually issued in
denominations of thousands or five thousands. The denomination of a
bond is known as its "face value" or "par value." and the interest rate
quoted on the bond is the "bond rate." The interest payable in any payment
period is simply the face value multiplied by the bond rate for the period of
interest. Bonds are said to be "retired" when their term has expired and the
face value has been repaid.

In public project financing, bonds are often issued for very long terms,
perhaps 10 or 20 years. Over this period, interest rates may fluctuate
widely. During the bond period, the bond rate may therefore become more
or less attractive relative to prevailing interest rates. Investors may freely
buy and sell bonds (subject to bond conditions) during the bond period, but
the actual price paid for a bond will rise or fall depending on prevailing
economic conditions and the term remaining on the bond.

8.7.2 Equity Capital

Common Stock Private corporations have the capability of issuing

common stock shares in the company-as a way of raising equity capital.
The price of common stock is set, essentially, by what the market will bear.
It can be affected by prevailing economic conditions, the history of the
company, the nature and potential of the product or service the company
offers, and similar factors.

Preferred Stock Preferred stock can also be issued by a private corporation.

Preferred stockholders are guaranteed a dividend before one is paid to
common stockholders. If the company is dissolved, obligations to
preferred stockholders must be met before those to common stockholders.
Preferred stockholders may also have special privileges-for instance, the
right to elect senior officers of the corporation which are not available to
common stockholders. The dividends payable to preferred stockholders are
usually specified as a fixed percentage of the stock par value. Preferred
stock can therefore be a more secure investment than common stock and
may be less susceptible to market fluctuations.

Retained Earnings Profits made by a company may either be paid as

dividends to stockholders or kept for reinvestment in the company. If kept,
these profits are referred to as "retained earnings" and are available as a
source of equity capital for future projects. It is seldom the case that all of
a com pany's profits are kept as retained earnings, simply because
stockholders usually demand a share of the profits.

Depreciation Funds Depreciation funds are those set aside from profits as a
source of capital to replace aging equipment. Until such time as they are
needed for that purpose, they are available for investment and thus
constitute a source of equity capital for the company. Care must be taken
to ensure that equipment-replacement demands are protected, however, so
that future capacity is not limited by present capital demands.

Leasing Traditionally, leasing has been thought to offer an income tax

advantage over debt financing, because the full amount of lease payments
is usually tax deductible, while only the depreciation of purchased assets
can be claimed as an income tax deduction. Under recent income tax
amendments, however, the advantage of leasing and debt financing has
diminished. Some leases incorporate maintenance costs and may therefore
offer savings, so that saved funds can be reinvested elsewhere. The
greatest advantage of leasing may, however, be that it allows a company to
keep abreast of fast-changing technology without tying up capital funds in
obsolete equipment.

8.7.3 Allocation of Capital

In a company or government with many projects under way

simultaneously, capital must be allocated among projects in a systematic
fashion. The following is a typical sequence of capital allocation (De
Garmo et al. 1997):

1. Preliminary project planning and screening

2. Determination of the cost of enterprise capital

3. Review of capital expenditure policies and evaluation procedures

4. Preparation of an annual capital expenditure budget

5. Project implementation and review

 6. Communication of results

Various techniques are used to decide which projects should, and should
not, be allocated capital. Among these are the methods described in
Section 8.1 (i.e., present-worth method, future-worth method, annual-
worth method, internal rate of return, benefit-cost analysis, etc.). If several
projects are under consideration simultaneously, optimization techniques,
including linear programming methods, may be helpful in determining the
best combination of investments. Linear programming was used, for
instance, in a recent study of point and nonpoint source control measures
in the Bay of Quinte, Ontario, to determine which combination of
agricultural and sewage treatment plant controls would result in the most
cost-effective reduction in total phosphorus loads to the Bay of Quinte
(D.W. Draper and Associates Ltd. et al. 1997).

REFERENCES

DeGarmo, E. Paul, William G. Sullivan, James A. Bontadelli, and Elin A.

Wicks. 1997. Engineering Economy. 10th ed. Upper Saddle River, N.J.:
Prentice-Hall.

D. W. Draper and Associates Ltd., M. Fortin, Bos Engineering and

Environmental Services, and 1. W. Heathcote. 1997. Phosphorus Trading
Program Evaluation and Design: Final Report. Prepared for the Ontario
Ministry of Environment and Energy, Environment Canada, the Lower
Trent Region Conservation Authority, the TrentSevern Waterway, and the
Bay of Quinte RAP Implementation Advisory Committee, Kingston,
Ontario.

Goodman, Alvin S. 1984. Principles o/ Water Resources Planning.

Englewood Cliffs, N.J.: Prentice-Hall.

Loughlin, James C. 1977. The efficiency and equity of cost allocation

methods for multipurpose water projects. Water Resources Res. 13(l).

MISA Advisory Committee. 1991. Water Conservation in Ontario

Municipalities: Implementing the User Pay System to Finance a Cleaner
 Environment. Technical Report. Toronto: Ontario Ministry of the
Environment.

Tobias, Daniel, and R. Mendelsohn. 1991. Valuing ecotourism in a tropical

rain-forest reserve. Anibio 20(2): 91-93.

U.S. Water Resources Council. 1983. Economic and Enriromunental

Principles and Guidelines far Water and Relates! Land Resources
Implementation Studies. March 10. Washington, D.C.: U.S. Water
Resources Council.
9
 Legal, Institutional, and
Administrative Concerns
Humans are social animals and live together in social groups. In every
society, no matter how advanced or how primitive, conflicts arise and must
be resolved if the society is to continue to exist peacefully and sustainably.
The framework for this resolution usually consists of a system of written
or unwritten rules that state the boundaries of acceptable behavior within
the society and specify sanctions, the penalties that will be imposed on
those who violate the rules that prevail in their society.

These rules are not cast in stone. As internal and external forces shape a
society, laws may come to be perceived as unjust or immoral. Gradually,
members of the society may begin to disregard a law or to actively protest
it. As opposition becomes widespread, initiatives are begun to revise or
replace the law. In democratic societies, for instance, the legal system itself
often contains provisions to amend or abolish outdated rules. A society's
rules-its legal system-therefore reflects a sort of social consensus as to
what that society currently judges to be morally correct and just.

Current legal and administrative frameworks in English-speaking

nations reflect such a consensus-and an ongoing debate about how much
government protection of the environment is "enough." Our laws also
reflect the development of a thousand years of rule making in a variety of
societies, as well as the social and economic pressures that shaped those
rules. The following discussion provides an overview of that development
from prehistoric times to the present. The author believes this history to be
important in understanding the development of current views on water
management, water allocation, and water pollution control. Readers who
are interested primarily in current legislative and administrative systems
are directed to Section 9.2 and following sections.

9.1 THE EVOLUTION OF MODERN ENVIRONMENTAL

PROTECTION LEGISLATION
9.1.1 Behavioral Codes in Primitive Societies

Behavioral codes have probably existed from the earliest beginnings of

human society, and they are found everywhere. Typically, these systems
are based either on compensation-the precept of "an eye for an eye"-or on
punishment.

Under compensation-based systems, only adequate compensation will

satisfy the injured party and put an end to a dispute. These systems take
many forms, depending on the values of the society that developed them.
They are generally established by custom and handed down orally from
one generation to the next. Bohannan (1967) offers the following examples
of compensation-based behavioral codes, sanctions, and retribution among
various peoples of the world. The Yukon Indians of Northern California,
for instance, believed that every invasion of privilege or property must be
exactly compensated, with compensation taking the form of wealth,
including dentalium shells, woodpecker scalps, obsidian blades, and
deerskins. In this system, an individual accused of killing a man of social
standing would be required to pay 15 strings of dentalium, with perhaps a
red obsidian and a woodpecker scalp headband, and would have to hand
over a daughter to the family of the slain man. (Killing a common man
would carry a penalty of only 10 strings of dentalium.)

In punishment-based systems, sanctions imposed against the offender

are generally weighted in proportion to the misdeed. In modern aboriginal
societies in Australia, societal codes of behavior are unwritten but appear
to be based on the principle of equivalent retaliation and compensation. In
general, a man pays less to members of his own clan than he pays to
outsiders as compensation for injuries, because as a clan member he is
entitled to share in all compensation. Some groups use older members of
the tribe as the agents of public opinion in enforcing sanctions against
offenders, requiring the parties to meet, and directing the proceedings.

Compensation and punishment rituals may be complex, involving the

facing of opposing parties, the singing of songs by the accused and his
family, and the exchange of ritualized feints and blows. Bohannan notes
that the purpose of these complex rituals seems to be as much to bring an
end to the dispute as to bring compensation to the injured party.

These elements of conflict resolution, compensation, and punishment

are present in modern, written legal systems as well. Indeed, they are the
founda tion of "common law," which in turn underlies much of the
statutory law-and common practice-in English-speaking countries around
the world. Even the use of elders as arbiters in a dispute has its parallels in
the Supreme Court system and in modern rabbinical practice.

9.1.2 The Emergence of Written Law Codes

The history of law is discussed in most basic law and administration texts,
such as Fitzgerald (1977), Kernaghan and Siegel (1991), Stephenson
(1975), and Jennings and Zuber (1991). The following discussion draws
from a variety of sources.

Written law codes appeared first in the Middle Eastern kingdom of

Babylonia in the twenty-second century B.C. The Hammurabic code,
assembled in about 1700 B.C., established a long list of rules governing
specific offenses, such as adultery, theft, and faulty workmanship by a
house builder. In a sense, these codes were formalized forms of earlier
unwritten customs. Their written form, however, made them more
straightforward to enforce and less flexible to apply.

By about 500 B.C. written codes had emerged in the civilizations of

India and China, and the complex social and religious laws of the Israelites
had been assembled into a written code. This so-called Mosaic code
reflected the teachings of Moses, the great Israelite leader of the 1200s
B.C., and, like the Indian and Chinese codes, stressed moral principles and
the obligations of a citizen under the law. The Mosaic code, which includes
the Ten Commandments traditionally considered to have been given to
Moses by God, was later incorporated into the Hebrew religious writings
and the first books of the Christian Bible. Rabbinical interpretation of this
code continues to form the basis of rabbinical courts and parallels the
English system of common law and the power of the courts to make law.
 The first Greek law code was drafted by the Athenian politician Draco
in 621 B.C. Draco's code was, again, a list of specific offenses and the
(usually harsh) penalties attached to them. The Draconian code was
replaced with a less severe code in the 590s B.C. At the same time, the
Athenian assembly was made more representative and given increased
lawmaking powers. With time, these elected assemblies of Athenian
citizens gained power in decision making and are considered by many
historians to represent the founding of democratic government.

The Greek system of law was, however, hampered in its development

by an overabundance of lay administrators and by large and cumbersome
decisionmaking bodies. Roman law was able to evolve further through the
efforts of individuals (jurisconsults) who specialized in interpreting the
law. Roman law was founded on a list of customary rules called the Laws
of the Twelve Tables, written about 450 B.C. This simple system was
frequently amended and enlarged over the next four hundred years, to the
point where an average citizen was not capable either of understanding or
of applying it. Specialists in law, like the jurisconsults and the rabbinical
schools (and like the common-law judges discussed later), used traditions
of wisdom, reason, and interpretation of general principles to develop
concrete guidance for problem solving.

In the sixth century A.D., the Roman Emperor Justinian introduced his
Corpus Juris Cirilis (body of civil law), a code that covered the whole field
of law. The Justinian code persists to this day as the hasis of much canon
(church) law in the West and, ultimately, as the foundation of the law in
most civil-law countries.

9.1.3 The Emergence of Feudalism: Ties to the Land

In the late 400s A.D., the West Roman Empire, which had its capital in
Rome, fell to invading Germanic tribes, who brought with them strong
clan loyalty and primitive law codes consisting mostly of fines for specific
offenses. By the 800s, most of the legal and cultural institutions developed
by the Romans had been replaced by a system of allegiance to individual
lords rather than to a central government. The legal and economic systems
that developed through the Middle Ages developed from this landlord-serf
relationship and centered on a system of large "manors" or estates owned
by the wealthy and worked by peasants. Under this system, peasants were
bound to the land they worked-essentially a part of the property and
inseparable from it if the property were to he sold. As the feudal/manorial
system grew, cities and towns declined and a network of large rural estates
burgeoned. Remnants of this system persist throughout Europe and Great
Britain in large ancestral land holdings.

Feudal law remained the fundamental legal system in Western Europe

until about 1300. As trade grew and human activities became more
complex, however, peasants began to leave their manors to seek work in
urban areas and cities once again expanded. With this changing social
structure came a revival of an economic system based on payment with
money for goods and services and the need for a legal system sufficient (as
the feudal system was not) to cope with complex property and commercial
disputes. In Europe, scholars at the University of Bologna began to train
law students from across Europe in the principles of Justinian's Corpus
Juris Cirilis. Interest in the code soon spread to other European universities
and, with time, Roman law began to replace feudal law throughout
mainland Europe.

In England the legal system took a different direction. England already

had a strong legal tradition based on the rights of judges to make decisions
suited to local customs and conditions. By the early 1 100s, this system
had grown chaotic and there thus began an effort on the part of the
monarchy to establish it system of royal courts to apply similar rulings in
similar cases. This court system soon established a body of "common law,"
law that would apply equally anywhere in England. As English common
law developed over the years, it established many precedents that limited
the powers of government and the monarchy and protected the rights of the
people.

9.1.4 The New World and Free Enterprise

Europe in the eighteenth century therefore exhibited three major legal
traditions: common law, Roman law, and feudalism. In England, the
common-law system was, and continues to be, well established and
powerful. Laws need not be written to have force: what is necessary is that
a court upholds long-standing traditions of property rights and personal
justice. Roman law has formed the basis of most Western European law
codes, including those of France and Spain, from the fourteenth century.
Even feudalism, a relic of the Middle Ages, persists in some forms. For
example, in early nineteenth-century Norway, servants constituted the
largest single category of the farm population (Starr and Collier 1989). In
1835 there were 124,600 servants, 103,000 freehold farmers, and 100,000
cotters. The relations between the farmer and the cotter, and the conditions
of the servant, were based on contracts. Two decrees, in 1750 and 1752,
stipulated that if the cotter himself had cleared the plot allocated to him, he
was entitled to a lifelong contract and the right of his widow to remain on
the land. The cotter's rights and duties were not specific in these statutes,
and the contracts differed greatly between various districts and regions,
depending on local custom.

The French and American Revolutions, and the opening of the New
World to settlement, proved to be a turning point in the history of property
law. In Europe many factors, including feudalism and a scarcity of arable
land, had contributed to the evolution of strict controls on land transfer.
This framework was ill suited to an environment where rapid development
was to be encouraged. Early North American law codes therefore tended to
abandon restrictions on land transfers and instead focus on the rights and
obligations involved in buying and selling land. Contract law was similarly
important in the New World, where free enterprise was highly valued. In
the growing free enterprise system, business people wanted to have the
freedom to regulate their dealings largely by contract without the
interference of government.

9.1.5 Settlement and Law in the New World

Early European settlers in North America faced daunting challenges in

clearing forests and crossing natural obstacles such as great rivers and deep
ravines. As a result, early development proceeded slowly, hampered by
reliance on hand tools and animal-drawn equipment. By about 1800, for
instance, Upper Canada's (now Ontario) rural population numbered about
15,000, scattered in tiny settlements through the dense forest and often
remote from developing urban centers (Careless, 1984). By contrast, the
town of York in that province had fewer than 700 residents.

Governments and private interests saw great potential in the vast natural
resources of these new territories and sought to exploit them as quickly as
possible. By 1850, the United States government owned 80% of the land in
the country, most of it appropriated from Native American tribes. By 1900
more than half of this land had been sold or given away to private interests.
Ostler et al. (1996) note that by artificially lowering prices, these land
transfers encouraged wastefulness in resource exploitation, the rate of
which increased tremendously following the corning of the railway in the
middle of the century. From the 1850s rail lines spread rapidly to major
centers in the United States and Canada, linking water traffic (including
that from the Erie Canal, opened in 1825) with upland routes. For urban
centers, the railway brought improved communications and accelerated
economic growth. It also opened access to rural areas, allowing the rapid
establishment of farm communities with cities as their economic core.

The railway had two important impacts on the rate and pattern of
settlement. First, railway access greatly increased the rate at which trees
could be brought out of logging areas and transported to urban centers for
shipment throughout the United States and elsewhere. It thus vastly
accelerated the rate at which watersheds were stripped of trees and erosion
could accelerate. And second, the establishment of the railway network in
turn encouraged the development of industrial centers like Chicago and
Toronto, which were close to rail and water routes, for the efficient
distribution of goods and services.

The new railway cities had busy port and manufacturing areas and
increasingly dense urban development. Their harbors were filled with all
manner of shipping vessels and the assorted wastes and debris from fast-
growing populations. The intense pace of industrial activity took
precedence over maintenance and sanitation, and in many cities decaying
structures and debris were allowed to accumulate in harbor waters,
sometimes obstructing ship passage so that ships had to be towed to their
berths.

Municipal water supply and sewage services lagged well behind

industrial development. Most homes did not have piped water supplies
until the beginning of the twentieth century, but continued to draw water
directly from streams or from private wells. As the population increased,
so did the number of backyard privies and leaking cesspools, and thus the
contamination of wells and water supplies. Untreated sewage flowed from
homes to lakes and rivers in a haphazard system of pipes originally
designed to drain roadways. Urban growth also meant rapid population
growth and an increase in the typhus and cholera that would plague North
American cities until the last quarter of the nineteenth century.

There were several direct consequences of this industrial growth and

wasteful resource extraction. Ostler et al. (1996) describe the warnings
published by contemporary authors like Ralph Waldo Emerson, Horace
Greeley, and Henry David Thoreau. The state of urban lakes, rivers, and
harbors grew so atrocious that it could no longer be ignored, and citizens
became more vocal in questioning the quality of their drinking water. In
1882 , an editorial in the Toronto (Canada) Globe called the city's water
supply "drinkable sewage."

By the mid-1880s public health officials had become more aware of the
causes of infection and better able to combat them. Dedicated public health
agencies began to emerge, and with thetas regular inspections of
unsanitary conditions in homes and businesses, particularly leaking privies
and cesspools, refuse dumps and unscrupulous food suppliers.

These two concerns, about excessive resource wastage and the

protection of public health as a responsibility of government, were driving
forces in the development of modern environmental and water
management legislation. They also prompted the formation of
nongovernment organizations, like the Sierra Club in 1892 and the
Audubon Society in 1905, which remain important in watershed
management today.

Legislation such as the U.S. River and Harbors Act (later the Refuse
Act) of 1899 and the Ontario Public Health Act of 1884 arose from
concerns about the maintenance of public water supplies, navigation, and
sewage disposal. From these statutes grew much of modern environmental
legislation. These statutes were not originally intended as environmental
conservation instruments, but were rather concerned with public or private
nuisance (a common law concept). Nevertheless, they reflected a growing
realization on the part of the public that resources are finite and that the
government (and the law) has, after all, a legitimate role in resource
conservation and public health.

9.1.6 The Beginnings of the U.S. Legal System

When the American colonists declared their independence from England in

1776, they based their claims partly on ancient Greek and Roman ideas of
natural law. These ideas, developed in detail by various French
philosophers of the 1700s, promoted the idea that the natural law gives all
people equal rights. The U.S. Declaration of Independence echoed this
idea in the phrase "... all men are created equal [and] are endowed by their
Creator with certain unalienable Rights."

English common law was, however, at the heart of the American claims
for independence. Many American leaders of the time were lawyers who
had been trained in the common law. Common-law principles (e.g., the
rights of the people) also influenced the development of the Declaration of
Independence, the U.S. Constitution, and the Bill of Rights.

Although American courts in theory had the same power to make laws
that English courts had, a series of U.S. Supreme Court decisions in the
early 1800s strengthened this power. The Court's decision in 1803 in the
case of Marhury v. Madison was especially important. In this decision, the
Court declared a federal law unconstitutional for the first time. The
principle of judicial review was thus firmly established, enabling U.S.
courts to overturn laws they judged unconstitutional.
 Despite the fact that English common law had provided the foundation
for the Declaration of Independence and the Constitution, parts of the
common law, particularly those related to property transfers, were
impractical for the new, rapidly expanding nation of the United States.
Land was scarce in England, and so common law narrowly restricted the
transfer of land from one owner to another. But much of the land in the
United States was unsettled, and the nation was constantly expanding its
frontiers. To ensure the nation's growth, people had to he free to buy and
sell land. American property law therefore began to stress the rights and
obligations involved in land transfers, and the English laws that restricted
such transfers were discarded.

By the early 1800s, Americans had begun to develop a flourishing

economy based almost entirely on free enterprise. The rapid growth of the
U.S. economy in the 1800s brought an enormous increase in contract law,
which is used in it free enterprise system to regulate commerce.

9.1.7 The Beginnings of the Canadian Legal System

Canada's legal history dates from the legal system established by the first
French settlers in the 1600s. The French set up a civil-law system in the
areas they colonized, including what is now the province of Quebec. They
based their system on one of the major local law codes in France, a code
known as the Custom of Paris.

In 1763, Great Britain gained control of France's Canadian possessions

and introduced a common-law system. But French Canadians objected to
giving up their legal traditions. In 1774, the British parliament passed the
Quebec Act, which allowed French Canadians to follow their traditional
system in private-law matters. Common law, however, remained the basis
of all other law in Canada. In 1866, Quebec adopted a private-law code
based on the Code Napoleon.

In 1867, the British North America Act, passed by the British

Parliament, created the Dominion of Canada and gave that country limited
self-government and a constitutional framework. Under the British North
America Act (now called the Constitution Act 1982), jurisdiction (law-
making power) was split between the federal and provincial governments,
based on a common-law framework at the federal level and on prevailing
legal systems in each province. Each province except Quebec elected to
base its legal framework on common law; Quebec retained its civil-law
system in matters of private law.

9.1.8 The Development of Modern Legal Frameworks

In the nineteenth century the emphasis on free enterprise and the rights of
the individual encouraged a spirit of entrepreneurialism in Canada and the
United States and indirectly set the stage for the development of public
health and resource management legislation. Three interrelated factors
contributed to this development: the rate of uncontrolled resource
exploitation, the advent of the railway, and the rapid growth of urban
centers.

The main civil- and common-law systems in the United States and
Canada remained largely unchanged throughout the nineteenth century. By
1900, U.S. private law dealt mainly with the protection of property rights
and businesses, with freedom of contract a central doctrine. Complete
freedom of contract had served the needs of America's rapidly expanding
economy during the 1800s, but by 1900 many businesses in the United
States were using this freedom to increase their profits at the expense of
their employees, stockholders, and customers. For example, factory
owners claimed that efforts to protect the rights of workers interfered with
the owners' rights to contract freely with their employees. Employees often
had to accept unfavorable contracts or lose their jobs.

During the 1800s, most Americans rejected the notion that the law
should be able to interfere in private business matters, but as concerns
about social welfare increased through the end of the nineteenth century
and the beginning of the twentieth, public attitudes toward the law began
to change. Today the prevailing societal view is that the private interests of
some members of society should not deprive other members of their rights.
Legislation and court decisions during the 1900s have reflected this belief,
especially by stressing the social aspects of contract law. For example,
Congress and the state legislatures have passed many laws to help ensure
the fairness of employment contracts. Some of these laws regulate working
conditions and workers' wages and hours. Other laws guarantee the right of
workers to organize and to strike.

Legislation and court decisions also changed many features of property

and tort law during the last century. The social obligations of property
owners have been enforced by zoning laws and by laws prohibiting
environmental pollution. During the 1800s tort law held that a person
could collect for an injury only if another person could be proved at fault.
But the development of private and public insurance programs during the
early years of this century helped to establish that a person should be paid
for accidental injuries regardless of who was at fault. This "no fault"
principle has made it unnecessary to sue for damages in certain cases.

Today, the United States, Canada, and other former British colonies
continue to base their national legal systems on the common law, although
to a large extent statutes have been written in an attempt to clarify and, in
some cases, alter interpretations of (unwritten) common law. These
systems are in contrast with those in civil-law countries, whose legal
systems are based exclusively on the Justinian Code or the Napoleonic
Code-systems of written statutes. In civillaw countries, and in the U.S.
state of Louisiana and the Canadian province of Quebec, which were
colonized by France and whose legal systems are patterned after the
French civil-law system, every legal decision must be based on an existing
statute, and not on precedent. These countries organize their written laws
(statutes) into codes, which provide the final answer in any question of
law. In common-law countries the judges in effect make the laws, and old
commonlaw traditions of personal rights are embedded in constitutional
law.

There are several important points to be drawn from this discussion:

• Canada and the United States (with the exceptions of Louisiana and
Quebec) share a tradition of common law and value highly the rights
of the individual.
 • In both countries laws are made in the courts through the decisions of
judges, based on legal precedents.

• Some aspects of common law have been extracted and written down,
either to clarify them or to alter permanently their interpretation; these
written laws, called statutes, include many environmental and water
laws.

• Notwithstanding the existence of statutory law, common-law rights

and responsibilities continue to bind people in the United States and
Canada.

9.2 COMMON-LAW CAUSES OF ACTION

As discussed earlier, common-law rights and responsibilities continue to

bind people even when statutory law also appears to exist in the same
domain." For example, a citizen has the right to charge a polluter under
statutory law such as the U.S. Clean Water Act or the Canadian
Environmental Protection Act, but that individual may also have the right
to bring a private action against the offender in nuisance, under common
law.

There are several main differences between common law and statute
law. One the fact that common law is unwritten has been described. A
second major difference lies in the relief that is available under each type
of law. Under common law, the age-old principle of compensation an "eye
for an eye"-can apply (cf. Section 9.1.1). Under common law,
compensation is often termed "damages." Alternatively, a plaintiff can seek
an injunction to stop a defendant from continuing an offensive action. This
injunction may either be temporary or permanent, depending on the
circumstances of the case. It is a long-held tradition in common law that an
injunction will never be granted if damages can give an adequate remedy
in the case. Under statute law, fines are paid to the government, not to
individuals (but the cost of prosecution is also borne by government.
Statutes can, like common law, require the cessation of an offensive
activity.
9.2.1 Common-Law Causes of Action in Water and Environmental Management

Under common law, a "cause of action" is a right to sue an individual or

group-that is, to bring a complaint (suit) before a judge and ask for
compensation or an injunction to stop the offense. More than one cause of
action can be cited in a single complaint.

There are numerous common-law causes of action. In terms of water

and environmental management, the most common of these are nuisance,
riparian rights, strict liability, trespass, and negligence, which are briefly
discussed in the following paragraphs. The interested reader is referred to
Estrin and Swaigen (1993) for a more detailed discussion of common-law
rights and remedies; much of the following discussion is drawn from that
source.

Nuisance Nuisance is a very old concept under common law. It embodies

the societal precept that one person should not interfere with another's
peaceful enjoyment of property. Suits in nuisance are useful remedies for
air and water pollution, noise, vibration, smells, soil contamination,
flooding, and similar intrusions into peaceful enjoyment of property
(Estrin and Swaigen 1993). Nuisance may be either "private"
(unreasonable interference with another person's use or enjoyment of
owned or occupied land) or "public" (unreasonable interference with the
use of public lands and water; affecting more than one or two people).

Traditionally, nuisance is measured by the degree of impact on

"peaceful enjoyment." Liability is deemed to exist if the activities of the
defendant have caused real harm to another person's health or property.
The plaintiff must be able to prove that the defendant could have foreseen
the consequences of his or her action but did not take steps to prevent the
nuisance from occurring. Examples of actions in nuisance have included
suits against farmers who spread manure over saturated land, thus allowing
contaminated runoff to flow into receiving waters, and against companies
who sprayed nearby land with pesticides, allowing the spray to drift over
the plaintiff's land. Nuisance can, therefore, be the result of either direct or
indirect action; the test of nuisance is that harm must be proved to have
occurred. Property may be either owned or occupied (for instance, as a
tenant) by the plaintiff; in other words, the plaintiff need not be the owner
of the land.

Public nuisance actions are often brought by public agencies (in

Ontario, for instance, by the attorney general) rather than by an individual.
They may therefore be subject to political influences and, indeed, may not
come forward at all because of partisan concerns. Generally speaking,
public nuisance actions must show that a public nuisance is also a private
nuisance and, therefore, that at least one person's property or health has
been significantly affected by the offense.

Traditional defenses against nuisance include the following:

• Statutory authority. The law required that the defendant take the action
that caused the effect; the effect was therefore unavoidable.

• Prescription. The activity had been conducted for a long time in the
same place without complaint and should be allowed to continue.

• Acquiescence by plaintiff. The plaintiff knew of the situation and

failed to make a complaint within a reasonable period of time.

Defenses that are not usually successful in nuisance actions include:

• That the land was used for a reasonable purpose

That the "best available technology" was used to avoid the effect, but
the effect occurred despite the use of this technology

• That the activity existed before the plaintiff arrived on the scene

Under this common-law precept, therefore, individuals have a

responsibility to their neighbors (society at large) to conduct themselves in
such a way that they do not interfere with others' "peaceful enjoyment" of
water or property. In court, the test of nuisance is whether harm actually
occurred, and whether the defendant tried hard enough to avoid that harm.
This latter concept, sometimes called "due diligence," runs throughout
common law and underlies most public and private obligations under
statutory law as well.

Riparian Rights Like nuisance, riparian rights is a very old concept in

common law. In essence, riparian rights are the rights of a property owner
(or tenant) to use the water (stream, river, or lake) flowing through or past
that person's land. Under riparian rights, the landowner (or tenant) has the
right to enjoy the continued flow of the water in its natural quantity and
quality, undiminished and unpolluted. Typically, riparian rights refer to
visible flows of water such as those in a defined channel. They are not,
therefore, usually taken to encompass groundwater flows. In some areas of
the United States and elsewhere, the doctrine of riparian rights has been
supplanted by prior appropriation as a way of allocating water rights. This
distinction is discussed in more detail in Section 9.4.4.

Strict Liability Strict liability simply says that the owner of dangerous
goods or materials is responsible for seeing that those goods or materials
do not cause harm to neighboring people or properties. Strict liability as a
cause of action implies that some abnormal use is being made of the land,
otherwise dangerous materials would not he held there. The classic
judgment in a strict liability action is usually taken to be the eighteenth-
century decision in Rvlands vs. Fletcher in England. In this case an
individual dammed a large quantity of water (the "dangerous substance")
on his land. The darn broke, allowing the water to escape and flood a
neighbor's mine. The owner of the flooded mine therefore had a legitimate
action in strict liability against his neighbor.

Because of the implication of "dangerous materials" and unusual use of

land, many pollution episodes can be taken as instances of strict liability.
Examples of successful actions in strict liability include those involving
the release of noxious gases, mercury, radioactivity, and lead into the
environment.

Trespass Trespass is a familiar concept, and its common usage is consistent

with its meaning under common law: individuals should not enter others'
property, nor should they deposit materials there, without the express
permission of the property owner or tenant. Legal precedents in trespass
actions have demon strated several important elements in this cause of
action. First, there must be physical intrusion by the defendant onto or over
the land owned or occupied by the plaintiff. (There is some question as to
how far above the land this doctrine applies; for instance, are jets flying at
35,000 feet over the land surface "trespassing"?) The physical intrusion
can take the form of people entering the land or of objects or materials
deposited on the land. To have an effective action in trespass, there must
also be intent on the part of the defendant to enter the property; in other
words, the intrusion must have been intentional and negligent. Finally,
unlike an action in nuisance in which indirect or consequential harm is a
legitimate cause of action, in trespass there must be direct harm resulting
from the defendant's intrusion onto the property. Intrusion alone is not a
sufficient cause of action. To have a successful defense, the defendant must
prove that the entry was unintentional; usually it is argued that the property
was not marked "private" in such a way that the defendant could be
expected to realize he or she was entering private lands. This is why many
landowners post ,,no trespassing" signs at regular intervals on fenced
property boundaries. And the landowner must demonstrate that real and
direct harm resulted from this intrusion. So, for instance, entry onto private
lands for the purposes of birdwatching might not constitute trespass unless
the bird-watcher damaged plants or structures, or left behind litter, during
the intrusion.

Negligence Negligence has been defined as "conduct that falls below the
standard regarded as normal or reasonable in a given community" (Estrin
and Swaigen 1993). Clearly, this standard will vary from community to
community and from business to business. For this reason, the court will
usually judge the standard of conduct expected of a business operator
against usual conduct in other businesses of the same type. For example, if
a judge was trying to determine whether a pulp and paper mill had been
negligent in allowing the release of partially treated effluent into a
receiving water, that judge would compare the precautions taken by the
mill against typical operating procedures in place at other pulp and paper
mills. In cases where materials that are particularly hazardous are in use, or
the receiving environment is particularly sensitive, higher standards may
reasonably he expected to apply. Once again, this is the idea of due
diligence: did the individual, or company, try hard enough to ensure that
the pollution did not occur'?

There are several difficulties with assessing due diligence in real cases.
For one thing, business operations differ significantly, and no two can be
expected to be exactly alike. Second, pollution control systems can be
highly complex and tailor-made for the particular site or receiving water.
Precautions that would be reasonable, and indeed necessary, at one
location may therefore be unreasonable or even unnecessary at another.
Third, given the complexity of natural systems (for instance, the difficulty
of predicting meteorological conditions with associated rainfall and
flooding), no individual can be expected to anticipate, and guard against,
everything that might possibly go wrong, so it is not always clear as to
what is due diligence and what is negligence.

These common-law principles seem familiar and comfortable to us.

They are the rules we learn in kindergarten: you do your very best not to
hurt other people, not to damage other people's property, and not to annoy
people unnecessarily. If you break these rules, society will require you to
compensate the offended party or to halt your activity altogether.

These same principles underlie all of the environmental statutes

discussed in Section 9.3. In large part, these are requirements that are
implicit in common law. Writing them down as statute law allows the
regulatory agency to impose very strict and specific interpretations on
common law. It also allows governments to impose sanctions on offenders
that benefit the public purse, punish the offender, or both, rather than
sanctions that benefit only the plaintiff or cause the offense to cease.

9.3 THE MAKING OF LAWS

9.3.1 The Legislative Process

A law is a binding rule, whether unwritten but widely endorsed (as in the
case of common law) or written and formally endorsed through a public
affirmation process (as in the case of statute law). Statutes are, in effect,
modifications or clarifications of common law that are written down,
approved by elected governments, and enforced by public agencies. In
both the United States and Canada, proposed statutes are scrutinized by all
political parties and undergo a multistep approval process. As a result, they
cannot easily be modified without a second multistep, multiparty approval
process.

Generally speaking, there are two types of statutes. Criminal laws

prohibit the kinds of actions that our society deems fundamentally wrong
and unacceptable-acts like theft, murder, and rape. Public welfare laws
address more "everyday" matters, regulating acceptable conduct (for
instance, on the part of engineers), setting out the powers of major public
bodies (for instance, universities), granting decision-making power, and so
on.

New statutes or regulations of any kind (including amendments to

existing laws) are made through the legislative process. The multistep
process in each country is intended to provide checks and balances on
lawmakinb in other words, to ensure that no single voice is able to create
new laws that will bind the wider society.

The legislative process differs between Canada and the United States, so
the two systems must be described separately. In each country, each of the
three branches of government-legislative, executive, and judicial-has a
role, although this role differs somewhat in the two nations. The legislative
branch in the United States is Congress; in Canada it is the House of
Commons and Senate. State and provincial legislatures are also considered
part of the legislative (lawmaking) branch of government, as are municipal
councils. The executive branch of government is the group of elected
representatives and their staffs, including the president of the United
States, the prime minister of Canada, and their senior advisors. The
executive branch also includes the mayors, executive committees, or
boards of control of municipalities. The executive branch of government is
usually made up of people who are chosen from among a group of elected
representatives, but may also include individuals who are appointed rather
than elected. The judicial branch of government consists of the courts and
judges. As shown under common law, the judicial branch is central in
lawmaking in that it protects the rights of the individual citizen, is (in
theory) independent and impartial, and interprets the law and legal
precedents in judgments about new cases.

The following paragraphs describe the first of these branches, the

legislative branch, in the United States and Canada.

United States The legislative process in the United States involves

Congress (comprising the Senate and the House of Representatives) and
the executive branch (primarily the president). Of these, Congress has the
most central role. It was granted power under the Constitution to "make all
laws which shall be necessary and proper" to carry out the various powers
delegated to it under the Constitution. These powers include the "power of
the purse" (power over government taxation and spending) and the power
to coin money, regulate trade, and wage war.

Congress consists of two bodies, the Senate (100 members) and the
House of Representatives (435 members). Every state has at least one seat
in the House. Both major political parties (Republicans and Democrats) are
represented among this membership. All members of these bodies are
elected. In addition to making laws, Congress approves the president's
choices for senior government appointments and Supreme Court justices
and approves (or rejects) treaties with other countries.

Congress uses a system of committees to draft, review, and revise

proposed legislation ("bills") so that the actual time spent in congressional
debate is used efficiently. Bills may be either public, affecting the public
good, or private, affecting a single individual. If a bill is favorably received
by a congressional committee, it is referred to the Rules Committee, which
decides which bills will be debated on the floor and what conditions may
apply in that debate. The Rules Committee has the power to block floor
debate on a proposed bill and thus, indirectly, can block a bill from passing
into law. If a bill is allowed into floor debate, it may be approved,
amended, or rejected, all by majority vote or unanimous consent (the
preferred mechanism). It can also be sent back to committee for further
discussion and revision.
The main steps in the process are as follows:

1. A law is proposed, by any member of Congress, by anyone on a

member's staff, by executive officers of government, by the president,
by a political party at a national convention, or by any other interested
party. 2. The proponent of the new law must seek a sponsor, a member
of Congress who will formally introduce the bill. Several members of
Congress may cosponsor a single bill.

3. The new bill is assigned a number with a prefix indicating its origin:
S for the Senate and H.R. for the House. Bills are also often
nicknamed after their sponsors or topics.

4. The bill is formally introduced in Congress.

5. The bill passes to an appropriate congressional committee for debate.

There are currently 16 standing committees of the Senate and 22
standing committees of the House of Representatives. Special ad hoc
committees may also be appointed for special tasks. Standing
committees cover areas such as agriculture, finance, foreign relations,
labor and human resources, and so on.

6. The congressional committee either "reports" the bill (that is, passes
it on to the next step) or "tables" it (that is, shelves it, intending no
further action). The committee may hold public hearings to determine
what action to take on a bill. The task of the committee is to screen out
bills that are unlikely to have broad public support and will therefore
fail in congressional debate. Thus, most bills that reach the Senate or
the House are likely to succeed and pass into law.

7. If the bill is reported, it passes to the Rules Committee for decision

about calendaring (scheduling for formal debate on the floor). The
Senate has only one calendar for public and private bills. The House
has five calendars: for bills that raise or spend money, for other public
bills, for private bills, for noncontroversial bills, and for motions to
remove a bill from committee.
 8. The bill is passed to the House or Senate (as appropriate for its place
of origin) and debated. It may be passed, amended, rejected, or passed
back to committee for further discussion.

9. After the bill has passed one house of Congress, it is referred to the
other house. Approval must be obtained from both houses for the bill
to become law. The second house often approves a bill without
change.

10. After approval has been obtained from the second house, the bill
passes to the president for approval. The president has 10 days to sign
or veto the bill. If the president fails to sign or veto the bill within this
period (which excludes Sundays) and Congress is in session, the bill is
passed into law. The president may choose this course if the executive
does not wish to be a party to a particular law but sees that there is
widespread public support for it. If, on the other hand, Congress has
adjourned, the bill does not become law; this is sometimes called a
"pocket veto." Presidential vetoes are rare, reflecting strong feeling on
the part of the president that the bill is somehow unconstitutional, too
expensive, or will be too difficult to enforce. Vetoes can be overridden
in the legislature.

This lengthy debate ensures that all members of Congress have the
opportunity to scrutinize proposed legislation, request changes, and satisfy
themselves that the bill will serve the needs of their constituencies. In this
way, the process ensures that the public, through their elected
representatives, has a say in determining which bills become law and
which are rejected. State governments tend to parallel the federal system
but are not necessarily identical in terms of powers and processes.

Although government departments are created by Congress, once

created they become part of the executive branch. As such, their interests
and objectives may sometimes differ from those of elected representatives.
Administrative arrangements are discussed in more detail in Section 9.5.

The strength and centrality in lawmaking granted to Congress under the

Constitution ensures that lawmaking power in the United States will
clearly reside at the federal level, although powers can be, and have been,
formally delegated to the states by Congress. As a result, most U.S.
legislation of importance to environmental and water management
interests is federal legislation. Although additional state laws may also
apply in a given situation, federal law provides a clear and unequivocal
foundation for all water management activities in the country.

In Canada, this is far from the case, because of the division of powers
under the British North America Act of 1867, now the Constitution Act
1982. These differences are discussed in the following section.

Canada The Canadian legislative system, like the U.S. system, is

bicameral-that is, consisting of two houses. It, like the parliaments of other
former British colonies like Australia, India, and South Africa, is modeled
on the parliamentary system of Great Britain. In Canada, Parliament has
two houses, the Senate and the House of Commons. Senators are appointed
by the government in power and generally have less power than their
counterparts in the House of Commons. Members of Parliament are the
elected representatives that serve in the House. As in the U.S. Congress,
Members of Parliament (often termed MPs) represent a constituency of the
public. Every Canadian province has a number of seats in the House of
Commons, and all major political parties (Conservative, Liberal, New
Democratic, Reform, Bloc Quebecois) are represented there.

As in the U.S. process, bills may be proposed by any member of the

public, by elected officials, by bureaucrats, or by any other person, but
they must be brought forward to the House by a sponsor. A bill must pass
through three "readings" in the House of Commons and must be approved
in each reading. As part of this process, a law is usually "gazetted"-that is,
distributed for wider scrutiny among members of the legal profession. If
approved, the bill passes to the Senate for endorsement and then to the
governor general for "royal assent." Royal assent in fact means approval
by the queen; the governor general, as the queen's representative in
Canada, is delegated the power to approve laws on behalf of the monarch.
The main steps in the Canadian legislative process are as follows:
1. A law is proposed, by any MP, by anyone on a member's staff, by
executive officers of government, by a member of the government
bureaucracy, by a member of the public, by a political party at a
national convention, or by any other interested party.

2. The wording of the bill is then drafted and reviewed by the political
party who will endorse the bill. Private members' bills are also
possible and need only he drafted and approved by the sponsoring
member.

3. Most often, hills are introduced by the political party currently in

power. When this occurs, the bill undergoes extensive scrutiny in
Cabinet, the committee of senior ministers of the party in power.

4. Cabinet may refer the bill to committee for further discussion. The
interests of the various members of Cabinet may vary widely. For
instance, the federal minister of industry may support a bill that would
encourage aggressive development in Northern Canada, whereas a
minister of the environment may be opposed to such a bill. Normally,
these differences are resolved in Cabinet so that the party in power can
present a united position in the House. (Failure to secure this support
can result in split voting and failure of the bill in the first or
subsequent readings.)

5. When approved by Cabinet, the bill passes to the legislature for

approval. The bill must receive three readings in the legislature. The
first reading is essentially a mechanism for introducing the hill to the
elected representatives.

6. After the second reading, the MPs may debate the contents of the bill,
propose amendments, or reject all or part of it. As in the U.S. system,
there is provision in this process for public hearings in the event that it
bill is particularly far-reaching or controversial.

7. The third reading presents the bill, with any revisions that were
incorporated after the second reading, to the House for a vote. No
 further discussion occurs after the third reading, and it is rare for
changes to occur at this stage.

8. The Senate reviews and usually ratifies the decision of the legislature
regarding the proposed bill.

9. The bill is passed to the governor general in the federal process, or (in
the case of provincial legislation) to the provincial lieutenant
governor, to receive approval from that individual as the queen's
representative ill the process (royal assent).

10. Having received royal assent, the bill becomes law. It may not,
however, he binding until it is "proclaimed in force." Proclamation in
force normally occurs upon granting of royal assent. but may also he
delayed if Cabinet believes that the bill may be onerous or costly to
implement. Some hills have taken up to six years after being passed in
the legislature to be proclaimed in force.

Like the U.S. process, the Canadian legislative process is designed with
a system of checks and balances. All parties have an opportunity to review
proposed legislation and request changes in it. Thus, the resulting statute,
in theory, has the support of a majority of citizens, as represented by their
elected representatives in the House of Commons.

9.3.2 Subordinate Legislation

Statute law is intended to be carefully thought out, robust, and difficult to

change. As a result, it usually contains wording that is clear and simple and
that has broad public support. Such a law may not include the details of,
for instance, allowable effluent quality from an individual discharger, or
the ways in which certain types of waste must be managed. These details
are often appended to the law in regulations or other so-called subordinate
legislation. In Canada, for example, the Canadian Environmental
Protection Act simply states that the governor general "in council" has the
power to make laws regarding, for instance, the control of pollution from
industrial sources. The statute itself does not specify which industries,
which types of pollution, or which types of controls may be appropriate;
these details (where available at all) are contained in regulations appended
to the Act. Regulations may be extremely detailed, to the point where
individual manufacturing facilities and even individual effluent discharge
pipes may be listed, with allowable discharge levels for a long and detailed
list of contaminants.

Subordinate legislation, which includes documents such as discharge

permits as well as regulations and "Codes of Practice," is not, unlike
statute law, drafted with the cooperation of all political parties.
Subordinate legislation is prepared by bureaucrats, often without external
scrutiny. It may receive careful attention by members of the party in
power, but may never be seen by those outside government. For these
reasons, while statute law is difficult to change, subordinate legislation-
permits, regulations, and so on-is relatively easy to change and perhaps
more subject to political forces than statute law. In many cases,
subordinate legislation reflects only the views of the party currently in
power, rather than a wider societal consensus.

9.3.3 The Division of Powers

As described earlier, the United States Constitution vested considerable

lawmaking power in the U.S. Congress, and it is clear even two hundred
years later that the body remains the central and dominant legislative force.
While Congress can, and does, delegate power to the states, there is still a
strong federal oversight role. If a given state is not sufficiently assiduous in
enforcing its own and federal laws, federal agencies often step in and see
that the enforcement is carried out.

The situation is considerably different in Canada, where the division of

powers is much less clear. The Canadian Constitution was drafted in 1867,
at a time when people were more concerned with resource exploitation and
resource ownership than they were about environmental protection or
human health. As a result, the Constitution is silent on many issues that are
important today, such as water quality and water diversions. On the other
hand, resource ownership and oversight of business practices are clearly
spelled out. In terms of water management, the result is that federal and
provincial jurisdiction (the power to make laws) over the environment and
water resources sometimes, even often, overlap.

In general, the Constitution grants jurisdiction over matters of national

significance, and which cross provincial boundaries, to the federal level of
government. These areas include the power to:

• Regulate international/interprovincial trade and commerce

• Regulate navigation and shipping

• Regulate seacoast and inland fisheries

• Impose taxes and spend money raised by taxation as it chooses, as

long as it does not interfere with the rights and responsibilities of the
provinces

• Make criminal laws (interpreted also as protection of public health)

• Regulate works that are interprovincial or international in nature:

shipping, railways, telegraphs, interprovincial pipelines

• Make laws for the "peace, order and good government of Canada"
(generally interpreted to mean laws of national significance for
instance, control of nuclear power)

The provinces were granted jurisdiction over matters of business and

property and were allowed to retain ownership of their natural resources.
Provincial powers include:

• Control of natural resources.

• Management and sale of public lands belonging to the province and

the timber and wood on those lands.

• Establishment and control of municipal institutions (in other words,

the power to set up municipalities and delegate responsibility to
them).
 • Power over property and civil rights (power to make contracts and
carry on businesses).

• Power over all matters of a local or private nature in the province.

Over the past quarter-century, water management responsibility has

been split between the federal and provincial levels of government. The
provinces (with their municipalities) traditionally have had primary
responsibility for water supply and pollution control through constitutional
rights over their own natural resources. In recent years, however, the
federal government has taken an increasingly active role in regulation
writing and enforcement, but this role may again be altered under proposed
"harmonization" and "regulatory efficiency" initiatives, begun in
December 1994 and intended to make federal and provincial laws mutually
compatible.

9.4 EXISTING LEGAL FRAMEWORKS FOR WATER AND

ENVIRONMENTAL MANAGEMENT

Most authors date the rise of the modern environmental movement, and
public awareness about environmental degradation, to the publication of
Rachel Carson's book Silent Spring in 1962. Over the next decade, several
fundamental pieces of environmental legislation were enacted in Canada
and the United States, including the U.S. National Environmental
Protection Act (1970) and the Canadian (federal) Fisheries Act (1970). The
1960s also saw rapid growth in the field of ecology, as well as widespread
data collection efforts to map ecological systems in lakes, rivers, and
oceans. Older agencies with narrow mandates (for instance, sanitary
engineering) were transformed into environmental protection agencies. In
1970, the US Environmental Protection Agency was established by
President Richard Nixon. In Canada, federal and provincial departments of
the environment were set up.

A detailed discussion of current environmental legislation is beyond the

scope of this book, and in any case would be quickly outdated as statutes
and subordinate legislation are continually changing. The following is
intended merely as an overview of major legislation pertinent to water
management at the time of writing.

9.4.1 Major United States Federal Statutes

A number of United States federal, state, and local agencies have some
level of jurisdiction over water and water-related activities. At the federal
level, for example, the U.S. Environmental Protection Agency's Office of
Water, Office of Pollution Prevention and Toxic Substances, Office of
Prevention, Pesticides, and Toxic Substances, and the Pollution Prevention
Council all have jurisdiction regarding water use and water pollution. The
National Oceanic and Atmospheric Administration within the U.S.
Department of Commerce has jurisdiction over certain water issues. The
Natural Resources Conservation Service, and various agricultural agencies
(including the Agricultural Extension Service and the Forest Service, both
of the U.S. Department of Agriculture) have varying degrees of authority
over agricultural activities and, thus, over nonpoint-source pollution.

At the state level, various state departments, including those interested

in natural resources, environmental quality, and agriculture, are involved in
the management of water resources. At the local level, planning
commissions and zoning boards can affect water bodies and point and
nonpoint sources of pollution through ordinances and building regulations.
County governments are also involved in regulation-for example, in
floodplain and shore land management and rural zoning-and often in
implementation of animal waste laws. Many agencies therefore share the
responsibility for water management. As is the case in Canada, these
various agencies have over the years developed ways of separating their
mandates to minimize overlap and unnecessary resource expenditure. The
current major U.S. water-related legislation is discussed in following
paragraphs.

The Clean Water Act The Federal Water Pollution Control Act (commonly
referred to as the Clean Water Act, or CWA) was first enacted in 1972 in
an attempt to consolidate several water pollution control statutes, including
the River and Harbors Act of 1899. The Clean Water Act, which was
reauthorized in 1987, aims to improve water quality for the "protection and
propagation of fish, shellfish, and wildlife" and for "recreation in and on
the water." Originally, the Act intended to eliminate the discharge of all
water pollutants by 1985, but in fact contained provisions only for the
control of point-source pollution. The significant contributions of nonpoint
sources and groundwater flows were largely overlooked when the Act was
drafted.

Since 1977, the Clean Water Act has contained provisions for civil
litigation against people who violate pretreatment standards for wastewater
discharges. The following discussion about this and subsequent Acts is
largely drawn from EcoLogic International (1995).

The Act contains seven major provisions for the control of water
pollution:

1. A requirement that point-source dischargers of' wastewaters hold

valid discharge permits (the National Pollution Discharge Elimination
System, or NPDES, permit system).

2. Requirements for the control and prevention of accidental or

intentional spills of oils and toxic substances.

3. Regulations pertaining to dredged and fill materials and dredging and

filling activities, including those on wetland areas.

4. Grant and subsidy programs for sewage treatment plants ("publicly

owned treatment works," or POTWs).

5. A requirement that states establish a planning process for point and

nonpoint sources of pollution.

6. A provision that states must share the responsibility for regulating

land use practices in order to pursue national water quality objectives
outlined in the CWA. (States are, however, allowed discretion in
choosing a process for carrying out this requirement, since much of
the planning is intended to be regionally oriented.).
 7. A framework for the states concerning the structure of nonpoint-
source management programs (although no requirement that states
adopt such a program: a management program is required only if a
state wants to he eligible for Section 319(b) funds).

Normal agricultural activities, such as plowing, seeding, cultivation,

minor drainage, harvesting, farm pond construction, and irrigation ditch
maintenance, are exempt from the permitting process required under the
Clean Water Act. These broad protections for agriculture also exist in
Canada.

The Great Lakes Critical Programs Act (CPA) amends Section 118 of
the Clean Water Act and requires the US EPA to develop water guidance
for the Great Lakes system. This guidance is to consist of not only
numerical limits on pollutants in the Great Lakes aimed at protecting
human health, aquatic life, and wildlife, but also direction to the Great
Lakes states on the minimum water quality standards, antidegradation
policies, and implementation procedures necessary to achieve the
aforementioned numerical limits.

The result of the CPA's mandate is the Great Lakes Water Quality
Guidance (also termed the Great Lakes Initiative), originally published in
the Federal Register on April 16, 1993. This guidance focuses primarily on
point-source loadings to the Great Lakes. For this reason, the US EPA and
the Great Lakes states have agreed to establish a multimedia approach to
identify and address gaps in effectively preventing, controlling, or
eliminating toxic loadings to the Great Lakes Basin from non-NPDES
sources. This effort, called the Great Lakes Toxics Reduction Effort,
addresses five pathways omitted from the Great Lakes Water Quality
Guidance, including contaminated sediments; atmospheric deposition;
storage, handling, and transport (spills); stormwater, combined sewer
overflows, and urban runoff; and waste sites. At the time of writing, the
Great Lakes Initiative does not address agricultural nonpoint-source
pollution.

Coastal Zone Act Reauthorization Amendments of 1990 Section 6217 of

the Coastal Zone Act Reauthorization Amendments of 1990 (CZARA)
requires coastal states with approved coastal zone management programs
to address nonpoint-source pollution sources that impact or threaten
coastal waters. The legislation specifies that states must submit Coastal
Nonpoint Pollution Control Programs for approval to both the US EPA and
the National Oceanic and Atmospheric Administration (NOAA). Failure to
submit an approvable program will result in a state's losing a portion of its
federal funding under Section 306 of the CZMA and Section 3 19 of the
CWA.

Section 6217(G) of the legislation requires that the US EPA, in

consultation with NOAA and other federal agencies, publish guidance
specifying management measures to restore and protect coastal waters
from specific categories of nonpoint-source pollution, including
agriculture. In January 1993, the US EPA published its Guidance
Specifying Management Measures for Sources of Nonpoint Pollution ill
Coastal Waters.

The Safe Drinking Water Act The Safe Drinking Water Act (SDWA) was
first passed in 1974 and most recently amended in 1986. It establishes
mandatory and nationwide drinking water standards, which are
administered by state and local authorities. The 1986 amendments
significantly increased EPA's requirements for disinfection, filtration,
contaminant monitoring, and ground water protection. (The Lead
Contamination Act of 1988, which is not described here, further amends
the SDWA to provide additional safeguards against contamination of
drinking water with lead.) The Act also limits deep-well disposal of liquid
hazardous wastes where those wastes may impact on drinking water
supplies and establishes a Wellhead Protection Program for public water
supplies drawn from groundwater.

Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA) The 1947

Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA), as amended
in 1972, authorizes the US EPA to control pesticides that may threaten
groundwater and surface waters. FIFRA mandates registration and
approval by the US EPA prior to use of any pesticide, sets labeling and
applicator licensing requirements, and makes improper use of pesticides a
misdemeanor. The US EPA uses a step-by-step approach to determine the
appropriate regulatory approach for specific pesticides: (1) the Agency first
determines the pesticide's potential for leaching into ground and surface
waters; (2) if such a potential exists, the US EPA considers whether
establishing national labeling restrictions, enforceable under FIFRA,
would adequately address these leaching concerns; (3) if restrictions would
not adequately address these concerns, the US EPA must determine
whether providing states with the opportunity to develop Pesticide
Management Plans for the chemical will address the risk from pesticide
contamination; (4) in the event that state Pesticide Management Plans
would not address the problem, the US EPA is authorized to suspend or
cancel the registration of these pesticides.

Necessary components of state plans include state philosophy and goals,

state roles and responsibilities, legal authority, resources, assessment and
planning, monitoring, prevention, response, enforcement, public awareness
and participation, information dissemination, and records and reporting.

Application of FIFRA has resulted in the suspension or cancellation of

pesticides containing some of the most toxic chemicals, such as DDT,
dieldrin, aldrin, chlordane, heptachlor, mercury, kepone, chlorobenzilate,
endrin, DBCP, and 2,4,5-T silvex. To ban a substance, the US EPA must
prove that a particular chemical causes "unreasonable adverse effects on
the environment." This process requires extensive study on a compound-
by-compound basis and thus can result in extensive delays in the decision-
making process.

Resource Conservation and Recovery Act (RCRA) The Resource

Conservation and Recovery Act of 1976, as amended in 1984 and 1986, is
the primary piece of federal legislation addressing solid waste disposal.
Hazardous waste disposal is mandated under this Act, although some
categories of solid waste that are not considered hazardous generally fall
under state, rather than federal, solid waste management plans. State
management plans required under RCRA must (I) identify general
strategies the state has chosen to protect its citizens against adverse effects
of solid waste disposal; (2) describe how the state will provide adequate
sanitary landfill capacity, and (3) show that there are adequate institutional
arrangements for implementing these strategies. Although some states,
such as California, require localities to prepare solid waste management
plans that address agricultural production and processing wastes, including
manure, prunings, and crop residues, other states, including most of the
Great Lakes states, exclude agricultural wastes from their solid waste act
requirements.

The Comprehensive Environmental Response, Compensation, and

Liability Act (CERCLA or "Superfund') CERCLA, which was drafted in
1980 and amended in 1982, was intended to respond to the environmental
disasters of the 1960s and 1970s, like those caused by leaking hazardous
waste disposal sites at Love Canal, Times Beach, and Stringfellow, and to
perceived weaknesses in RCRA in respect to problems of this kind.
Particular problems had been identified with abandoned sites where the
costs of cleanup might be enormous. CERCLA granted the US EPA the
power to carry out cleanup in situations where public health and welfare
was urgently in need of protection. Public trust funds are used in these
situations, and where possible the costs of cleanup are recovered from
those responsible for the wastes. CERCLA does not provide for
compensation to individuals, either for property damage or for direct or
indirect health effects.

The Marine Protection, Research, and Sanctuaries Act (Ocean Dumping

Act) The Ocean Dumping Act, enacted in 1972, provides guidelines for the
dumping of wastes and other materials into the oceans. It also imposes
limits on who can do that dumping. Under the Act, certain sites are
designated as approved locations for ocean dumping, and other sites as
sanctuaries, to be patrolled by the Coast Guard. The Act prohibits the
disposal of radiological, chemical, or biological warfare agents in the
oceans, restricts the ways in which U.S. vessels can discharge wastes into
the oceans, and prohibits unauthorized transportation from U.S. territory
with the purpose of ocean dumping. This Act also provides oversight of the
National Estuary Program, which uses existing legislation as a framework
for the protection of 17 estuarine sites in the United States.
The National Environmental Policy Act The National Environmental
Policy Act (NEPA) of 1969 has the intention to create a unified
government approach to environmental protection. It requires
Environmental Impact Statements (EIS) by federal agencies before they
undertake actions with potential to impact the environment, including
direct, indirect, and cumulative impacts. The NEPA also establishes a
Council on Environmental Quality, with specific regulations covering EIS,
and provides for the preparation of annual environmental quality reports to
the president and Congress on the status of the U.S. environment and the
development of policies for the improvement of environmental quality.

The Pollution Prevention Act The Pollution Prevention Act of 1990 reflects
a growing societal concern that, where possible, pollution should be
prevented before it is created, not just managed after the fact The Act
establishes national targets for recycling and establishes an Office of
Pollution Prevention as an information clearing house. It establishes an
emphasis, under the law, on the pollution prevention hierarchy: prevention
(first), recycling, treatment, and disposal (last resort).

The Toxic Substances Control Act (TSCA, often pronounced "Tosca") The
Toxic Substances Control Act, passed in 1976 and last amended in 1986,
establishes several programs relating to the introduction of new and
potentially hazardous substances into commerce. It lays out requirements
for premanufacture testing and notification and empowers EPA to require
manufacturers of certain chemicals to undertake extensive testing of those
chemicals if there is a high risk of exposure or if existing data are
inadequate. Through TSCA provisions, EPA has the power to limit the
manufacture and release of chemicals until it is satisfied with the results of
industrial testing. EPA also has the power, under TSCA, to ban chemicals
of known risks and has done so in the case of PCBs, chlorofluorocarbons,
and asbestos. TSCA does not have authority over products covered by
FIFRA or over tobacco, nuclear materials, food or drugs covered by the
Food, Drug and Cosmetics Act, or certain other materials.

The Soil Conservation Act The Soil Conservation Act of 1935 exists as
one of the major pieces of' federal legislation governing soil conservation.
The close relationship between soil erosion and transport of various
pollutants, including nutrients, heavy metals, and pesticides, therefore
makes this an important mechanism for the control of nonpoint-source
pollution. The intent of this Act is to foster. develop, and implement
practices that will positively impact conservation of soils. The Act
originally focused on land treatment activities, but subsequent amendments
have authorized small watershed management pro- granms, rural
development activities, and so on. The Act is now the central basis for
provision of technical assistance to landowners through cost-share
programs and education for soil and water conservation purposes. Under
this legislation, farmers with highly erodible land are required to plan and
apply soil conservation systems in order to remain eligible for farm
program benefits.

The Clean Air Act Atmospheric sources comprise a large portion of the
total loadings of some pollutants, including dioxins and furans, mercury,
and lead. Section 112(m) of the 1990 amendments to the Clean Air Act
requires the US EPA to identify and assess the extent of atmospheric
deposition of hazardous air pollutants to the Great Lakes and other
designated "Great Waters'' such as Lake Champlain and the Chesapeake
Bay. The Great Waters program is intended to examine sources and
deposition rates of air pollutants, the relative contribution of atmospheric
pollution to total pollutant loadings, and potential adverse health effects
resulting from the volatilization and reemission of chemicals from
agricultural lands and subsequent deposition to these waters.

Food Security Act of 1985 (also referred to as the Farm Bill) The 1985
Food Security Act (FSA) contained provisions which, though not
specifically water quality-oriented, have the potential to reduce agricultural
nonpointsource pollutant loadings. The FSA was enacted to establish
greater consistency between farm production and soil and water
conservation programs. Essentially, the Act provides for more forceful
obligations for soil and water conservation in the form of cross-compliance
for those farmers dependent on government commodity supports and other
subsidy or support programs. A few pilot projects were conducted under
this legislation, but no nationwide program was ever developed. Wetlands
reserve conservation easements are not currently funded under the Act.
The Act was amended in 1990 and again in 1995 as the 1995 Farm Bill .

Great Lakes Water Quality Agreement Although not a statute, the U.S.-
Canadian Great Lakes Water Quality Agreement (GLWQA), as amended
in 1987, is a treaty-based obligation with some weight under the law. The
Agreement, which arises from the Boundary Waters Treaty of 1909
between the United States and Canada, sets out guidelines for reducing
point and nonpoint sources of pollution into the "boundary waters" shared
by the two nations, including binational channels such as the St. Croix
(Maine and New Brunswick) and the entire Great Lakes Basin. The
GLWQA specifies water quality targets for surface waters, sets out a
research agenda, and hinds the United States and Canada to a range of
control measures for point- and nonpoint-source pollution, including
sewage treatment measures, reduction in phosphorus loadings, control of
toxic substances, and similar programs. Annex 13 requires the two
countries to identify land-based activities contributing to water quality
problems described in Remedial Action Plans (RAPs) or Lakewide
Management Plans (LaMPs) and to develop and implement watershed
management plans, consistent with the RAPs and LaMPs, to reduce
pollutant inputs.

Clearly, U.S. federal legislation contains many provisions for the control
of point and nonpoint sources of water pollution. Generally speaking, these
provisions are consistent in tone and intent, facilitating cross-compliance
and program consistency. At the federal level, the Clean Water Act
provides broad authority to maintain and preserve the integrity of the
nation's waters. Collaboration between the various federal, state, and local
agencies is clearly vital to the implementation of effective water
management programs. For example, the planning and implementation of
controls for nonpoint-source pollution requires collaboration between US
EPA, USDA (U.S. Department of Agriculture), and NOAA at the federal
level, including the provision of joint technical and financial assistance
mechanisms. In addition, these agencies must engage in regular
communication regarding agricultural nonpoint-source pollution control so
as to avoid the promulgation of overlapping or contradictory regulations.
Most important, the federal government must provide the necessary
financial support to the states to ensure that they are able to implement any
nonpoint-source management programs that may potentially be required
under federal law. The same interagency, interjurisdictional collaboration
is necessary in most other water management situations.

9.4.2 Major Canadian Federal Statutes

In Canada, as discussed in Section 9.3, regulation of water management

involves all levels of government: federal, provincial, and, to a lesser
extent, municipal. As a result, many departments and agencies have
authority over similar water-related activities. At the federal level,
Environment Canada, Fisheries and Oceans Canada, Natural Resources
Canada, Agriculture Canada, and Health and Welfare Canada all have
varying degrees of control over water and waterbased activities.

de Loe (1991) identifies five agencies at the Canadian federal level, and
at least six at the provincial level, with clear mandates in the management
of Ontario water resources. He describes these institutional arrangements
as developing in an ad hoc fashion and lacking integration. While he notes
that this pattern is typical of water management in many jurisdictions, it is
clear that the lack of an overall guiding strategy, delivered by a single
agency, complicates the development of integrated water management
policy in that province.

In addition to federal agencies, most provinces have agencies with

authority over water management. As in the United States, these include
ministries of environment (or environmental protection), natural resources,
agriculture, and energy. At the municipal level, planning procedures,
zoning, and bylaws can also have an effect on water resources
management activities. With at least three levels of government (four, if
regional municipalities are included), and multiple departments having
control over the same activities, there is significant potential for the
legislation governing these activities to overlap and become fragmented.
Numerous authors have attested to this fragmentation (e.g., Heathcote
1993, de Loe 1991, Estrin and Swaigen 1993).
 The federal level of Canadian legislation has jurisdiction over
nonpointsource pollution from agricultural sources mainly through the
departments of' Environment Canada, Health and Welfare Canada,
Agriculture Canada, and Fisheries and Oceans Canada. The Canada
Fisheries Act and the Environmental Protection Act are the most
significant federal statutes in regard to water pollution control at this level.
Although neither Act directly mentions nonpoint sources or groundwater
pollution, the general prohibitions outlined in the statutes are thought to
encompass these issues.

The Canadian Environmental Protection Act The Canadian Environmental

Protection Act (CEPA) of 1988, administered through both Environment
Canada and Health and Welfare Canada, has the potential to impact point-
and nonpoint-source pollution in that "the purpose of this Act is to provide
for the protection and conservation of Canada's natural environment."
CEPA provides for the protection of the environment through the
application of preventative and remedial measures and the establishment of
levels of environmental quality that function to protect the environment
from the release of toxic substances. The Act purports to adopt an
ecosystem approach to regulations, in that it attempts to control the release
of toxic contaminants throughout the environment rather than dealing
separately with air, land, and water issues.

CEPA creates a framework within which nationwide environmental

standards controlling toxic substances in Canada can be developed.
Section 8 of the Act outlines the authority of the minister to formulate
nonregulatory instruments, such as release guidelines and codes of
practice, that will provide environmental guidance to industries and
regulators. The Act also provides a framework for the federal environment
minister to suggest standards and practices to fellow cabinet members to
govern pollution from activities within other federal departments.

The Fisheries Act The Fisheries Act, administered by Fisheries and Oceans
Canada in conjunction with provincial agencies, is intended to protect fish
and other marine animals and their respective habitats in all internal and
coastal waters of Canada. The Act was first passed in 1868 for the
protection of fisheries resources. Its modern form was first enacted in 1970
and amended in 1991 to increase the penalties for violations under the Act.

Section 35 prohibits the "carrying on of any work or undertaking that

results in the harmful alteration, disruption or destruction of fish habitat."
Section 36(3) of the Act further states, "No person shall deposit or permit
the deposit of a deleterious substance of any type in water frequented by
fish, or in any place under any conditions where the deleterious substance
may enter any such water." A deleterious substance, as outlined in
Subsection 34(1), is defined as "any substance that if added to any water
would degrade or alter or form part of a process of degradation or
alteration of the quality of water so that it is rendered or is likely to be
rendered deleterious to fish or fish habitat or to the use by man of fish that
frequent that water." The term deposit has also been defined by Subsection
34(I) to mean any discharging, spraying, releasing, spilling, leaking,
seeping, pouring, emitting, throwing, dumping, or placing. It should also
be noted that under Subsection 40(5) a "deposit" as defined in Subsection
34(1) takes place regardless of whether or not the act or omission resulting
in the deposit is intentional. Under the Act there is also a duty to report the
deposit of a deleterious substance. Subsection 38(5) requires prevention
(where possible) and mitigation of adverse effects. The Act further
empowers the federal Minister of Fisheries and Oceans to request plans,
procedures, schedules, analyses, samples, and other information relating to
any work or undertaking that results in or is likely to result in the
alteration, disruption, or destruction of a fish habitat or the deposition of a
deleterious substance in waters frequented by fish. With this information,
an evaluation of the potential impact can be made and the measures, if any,
that would prevent the deposit or mitigate the effects thereof can be
examined.

Under the Fisheries Act, the governor-general in council is empowered

to make regulations for the proper management and control of the seacoast
and inland fisheries, the conservation and protection of fish, the
obstruction and pollution of any waters frequented by fish, and the
conservation and protection of spawning grounds.
 The Fisheries Act remains one of Canada's most powerful
environmental statutes. Its power derives, in part, from its simple wording
and from the fact that it is usually easy to demonstrate that fish were, or
could be, present in a given water body and that the materials discharged to
that water body are in some way deleterious to those fish. There is no need
to prove that fish, in fact, died or that contaminants were discharged in
toxic amounts. As a result, the Fisheries Act remains a cornerstone of
Canadian water management.

The International Rivers Improvements Act The International Rivers

Improvement Act regulates the construction, operation, and maintenance
of dams and diversions in international rivers. Essentially, the Act requires
that such "improvements" proceed only with the formal approval of the
federal Minister of Transport. Like the Navigable Waters Protection Act, it
has become important not so much because of its specific provisions, but
because it can trigger the federal environmental assessment process and
thus interfere with provincial dam-building or diversion activities.

Navigable Water Protection Act The objective of the Navigable Water

Protection Act, administered by Transport Canada, is to prevent work that
affects navigable waters from being undertaken unless first approved by
the Minister of Transport. This Act's relevance to water management
relates to a general prohibition, to the effect that no person shall deposit
any material or rubbish into navigable water that is liable to interfere with
navigation.

Fertilizers Act The Fertilizers Act is administered by Agriculture Canada

and is intended to prevent the sale or import of any fertilizer or supplement
that has not been registered, packaged, or labeled as prescribed in the Act.
Section 9 of the Act prevents the sale of any fertilizer or supplement
product that is harmful to plant growth. Section 9 states, "No person shall
sell any fertilizer or supplement that contains destructive ingredients or
properties harmful to plant growth when used according to directions
accompanying the fertilizer or supplement or appearing on the label of the
package in which the fertilizer or supplement is contained." Although this
Act prohibits the use of fertilizers that are harmful to plant growth, it does
not contain any prohibitions against the use of fertilizers that are harmful
to the environment in general.

The Canadian Environmental Assessment Act The Canadian

Environmental Assessment Act (CEAA) was passed in 1995, in response
to nationwide calls for reform in environmental assessment procedures. It
replaced less formal policies that had previously existed and that had been
administered by Environment Canada. The CEAA established the
Canadian Environmental Assessment Agency and specified four types of
assessment: screening, comprehensive study, review by a mediator, or
review by a panel. All projects receive an appropriate level of assessment
based on their complexity. The federal government undertakes annually
about 6,600 environmental assessments of all types. More than 99% are
screenings. Since 1973 the federal government has referred more than 50
environmental assessments to a panel for public review.

The Canada Water Act The Canada Water Act of 1970 provided for
research and planning of water management programs by the federal
government. Part III of the Act was incorporated into the Canadian
Environmental Protection Act in 1988; the remaining portions of the Act
have been used only to set up federal-provincial or interprovincial water
management agreements.

Canada Wildlife Act The Canada Wildlife Act, administered by

Environment Canada, imposes controls on point- and nonpoint-source
pollution in areas designated for the protection of wildlife. In such
protected areas, this Act outlines prohibitions on activities deemed harmful
to wildlife. These activities may include the use of vehicles, the dumping
of wastes, and the disturbance of soil or vegetation. Numerous scientific
projects, such as environmental impact studies and research into the effects
of pollutants on wildlife, are performed by the Canadian Wildlife Service
under the Canadian Wildlife Act.

The Pest Control Products Act The Pest Control Products Act,
administered by Agriculture Canada, is the principle statute governing
pesticides in Canada. The intent of the Act is to prohibit any person from
manufacturing, storing, displaying, distributing, or using any pest control
product under unsafe conditions. Under the Act, a pest control product is
defined in Section 2 as "any product, device, organism, substance, or thing
that is manufactured, represented, sold, or used as a means of directly or
indirectly controlling, preventing, destroying, mitigating, attracting, or
repelling, any pest and includes any compound or substance that enhances
or modifies or is intended to enhance or modify the physical or chemical
characteristics of a control product to which it is added and any active
ingredient used for the manufacture of a control product."

The Pest Control Products Act regulates the import, sale, and use of pest
control products in Canada through the use of a registration system. All
pest control products in Canada must be registered. Section 4 of the Act
prohibits any person from importing or selling any pest control product
unless it has been registered, packaged, and labeled according to
prescribed conditions.

The Minister of Agriculture has the responsibility of registering

pesticides when he or she is of the opinion that the control product has
merit and will not pose an "unacceptable risk of harm" to public health,
plants, or the environment. During the registration process the onus is on
the pesticide manufacturer to demonstrate product safety and merit through
numerous scientific tests on the control product's effectiveness, toxicity,
environmental hazard, residue persistence, bioaccumulation,
carcinogenicity, and other related matters. The scientific data are then
reviewed by a board consisting of representatives from Health and Welfare
Canada. Environment Canada, Fisheries and Oceans Canada, and Forestry
Canada. Although information and data on the control product are
reviewed by a number of departments, Agriculture Canada has the final
decision concerning registration. Public involvement in the registration
process has so far been prevented. The Act has been criticized for being
dated, in comparison with other Canadian statutes, and indeed has not been
significantly amended since 1969. Numerous problems and concerns have
been identified in regard to the Act and its control product registration
process. Such problems include the lack of public involvement in the
registration process, the lack of public access to information on health and
safety data concerning control products, and the use of questionable testing
techniques by chemical manufacturers in the registration process. These
concerns are currently under review by a government panel charged with
recommending revisions to the Act.

In summary, Canadian federal statutes governing water-related activities

provide a comprehensive, if fragmented, framework for water
management. In general, they underemphasize or entirely overlook
contributions from nonpoint sources of pollution, as well as to and from
groundwater. The emphasis in most statutes is clearly on management of
point sources to surface waters. Indeed, a major weakness in this
framework is the specific exemptions and other forms of relief that have
been granted to agricultural operations. Almost without exception,
exemptions for agriculture (e.g., exemption from permitting requirements,
exemption from normal waste disposal requirements) are granted in all
primary federal and provincial environmental statutes. These exemptions
keep agricultural operations free from the environmental regulations and
constraints imposed on other sectors of Canadian industry.

The split responsibility for water management in Canada demands close

and continued interaction between the various federal, provincial, regional,
and local agencies charged with water management. In a binational
channel like the Niagara River, for example, effective management
demands, at minimum, liaison with U.S. federal and state agencies.
Environment Canada, the Ontario provincial Ministry of Environment and
Energy, the regional municipality of Niagara (an "upper tier" or regional
municipal body), local municipalities (individual cities like Fort Eric,
Welland, and Niagara Falls), and a variety of other agricultural, natural
resources, tourism, and industrial agencies at several levels of government.
In general, this collaboration works well, and multiagency planning
initiatives are the rule rather than the exception in Canada. This is a
situation that may be unfamiliar to the novice, however, and one that
requires careful management in the establishment of planning structures
and processes.

9.4.3 State and Provincial Legislation

The first point that must be noted in regard to state and provincial
legislation is the vast variability from one jurisdiction to another. In part,
this variability arises from the differing social and economic histories of
the various regions and the social priorities that have arisen as a result of
those forces. For instance, in water-poor areas, considerable emphasis may
be placed on allocation of water rights, while in water-rich areas the law
may be silent on this issue.

Most states and provinces have a general "environmental protection"

statute that is designed to protect air, water, and terrestrial ecosystems from
pollution or mismanagement. Most also have environmental assessment
legislation (see Chapter 10), requiring environmental assessment of major
public projects. In some cases environmental assessment requirements may
be built into general environmental statutes. Some jurisdictions have
specific legislation dealing with the management of public utilities (water
supply, sewerage, electric power generation, and so on). Finally,
occupational health and safety legislation is found in most regions and,
although not expressly directed at environmental concerns, often contains
provisions for the control of pollutants in the indoor environment and in
the bodies of workers.

It is beyond the scope of this hook to inventory the huge number of state
and provincial statutes, regulations, and policies that may hear on water
resources management. It can, however, be stated that optimal water
management almost always requires action, and therefore legislation, at the
local or, at least, regional level. Often, however, these local ordinances (for
instance, land use planning restrictions) are poorly enforced because local
government and soil conservation district officials are hesitant to restrict
landowners' property rights (Holloway and Guy 1990) and may even be
subject to legal challenge by landowners.

As was observed at the federal level, state and provincial protections for
groundwater, and control of nonpoint sources of pollution, sometimes lack
the force of those for surface water protection and point sources of
pollution. In general, however, state and provincial water pollution control
laws, wetlands protection laws, soil and water conservation laws, and
zoning laws provide the statutory authority necessary to create and
implement sound water management programs.

9.4.4 Allocation of Surface and Ground Water Rights

Riparian Rights and Reasonable Use in Surface Water Allocation Systems

of water allocation have, in large part, developed in response to local water
disputes. There are two fundamental systems of allocating water rights.
The oldest of these is riparian rights, discussed in Section 9.2. Under
riparian rights, a landowner is entitled to use the water that flows through
his or her property, unimpeded in quantity and unimpaired in quality. The
term riparian refers to the banks of a lake or stream (usually the latter).

English settlers brought the riparian rights doctrine with them and
applied it in early settlements in eastern North America. Over time,
however, the restrictions of riparian rights -a landowner could not
adversely affect the quality or quantity of water flowing through his land-
became too burdensome for the entrepreneurial American spirit. In the
eastern United States, riparian rights were gradually replaced with the
doctrine of "reasonable use," which permitted landowners to modify water
quality and flows as long as the use was deemed to be reasonable for the
purpose to which the land was put.

Although "reasonable use" in theory permits withdrawals consistent

with land use, some jurisdictions have chosen to modify the doctrine by
imposing a permitting system on water extractions. In this system, the
landowner must secure it permit from the appropriate state agency before
withdrawals can he made. In some jurisdictions, permits need only he
obtained if extractions are expected to exceed some threshold level.
Permits may restrict the amount of water that can he extracted, the place
where extractions can he made, the time period over which water is
extracted, or any combination of these factors. Permits may specify the use
to which the water is to be put and may prohibit changes in that use
without prior approval from the permitting agency. Finally. permits may
specify water use priorities in the event of water shortages; in other words,
they may require that under certain conditions the permit holder may or
may not extract water, depending on the proposed water use.

Today, riparian rights (reasonable use) water allocation is usually found

in areas that are relatively rich in water for instance, in eastern Canada and
in much of the eastern United States. In these areas, when water shortages
occur, landowners are expected to reduce their water use equally, not in
order of seniority. There is usually no restriction on the ways that water
can be used in a riparian rights system.

Appropriation Rights in Surface Water Allocation In water-poor areas of

the United States, particularly the West and Southwest, state economies
were built on water-intensive activities like mining and agriculture
(Viessmann and Welty 1985). In some of these areas, water sources were
located far from the place where water was needed, requiring diversion
and pumping over land that might be held by another party-and affecting
that landowner's riparian rights. The settlers felt that their interests were
not well served by riparian rights in these cases and, instead, developed a
doctrine called "appropriation rights." Under this doctrine, appropriators
are assigned rights based on their priority of application for those rights.
The phrase "first in time, first in right" is sometimes used to describe
appropriation rights.

Under an appropriation rights system, a would-he water user must apply

to the appropriate state agency, stating the beneficial use to which the
water will be put. If the application is successful, it is assigned a date that
determines its priority. Applications are usually successful if they do not
conflict with existing water uses. Conflicts arise in times of water shortage,
when there simply may not be enough water to go around. At these times,
the most recent appropriator may not be allowed any water at all.
Appropriators must use their water regularly or they will risk losing their
water rights through nonuse. Appropriation rights may be sold or
transferred as long as the purported use of the water does not change. If it
does, the appropriator may be required to apply for a new permit.

There are several difficulties with appropriation rights. First, they place
heavy emphasis on "beneficial uses," which historically have included uses
with economic value, such as mining, agriculture, and municipal
extractions. They have not traditionally included maintenance of minimum
flows for the purpose of preservation of aquatic organisms or habitat
(although some western states are now incorporating such definitions).
And finally, the "use it or lose it" emphasis in appropriation rights creates
an incentive to use water rather than to conserve it, even though
appropriation-rights areas are often those that would benefit most from
water conservation efforts.

Allocation of Groundwater Rights Groundwater is less visible, less easily

mapped, and more variably described than surface water. As a result,
various jurisdictions define it differently and allocate its use using different
principles. In general, these principles differ somewhat from those used for
the allocation of surface water rights.

Under common law, groundwater is not encompassed by riparian rights,

and traditionally landowners were able to withdraw unlimited quantities of
water for any purpose without fear of retribution. This cavalier attitude
may have arisen from a time when technology allowed only small
extractions and there were few users sharing any given aquifer. In recent
centuries, however, population densities have increased and technology
has improved to the point that very large water extractions can be made
easily, and many users may share a groundwater source.

Three groundwater-allocation doctrines are commonly used in the

United States. The first is reasonable use, which, as in surface water
extractions, implies sensible, conservative use, with all users having equal
rights. The second doctrine is correlative rights, whereby rights are
allocated in proportion to the extent of ownership of overlying land. Under
this doctrine, the more land you own, the more groundwater you may
extract. Finally, prior appropriation is used for groundwater allocation as it
is for surface water allocation. The user must apply for a groundwater
extraction permit from the appropriate agency and, in being granted
approval, is also given seniority in water rights.

Certain areas such as Native American reservations, national parks,

national forests, and other federal lands have rights to the water necessary
to carry out their functions. Native American water rights are separate and
distinct from those allocated under the various aforementioned doctrines.
Aboriginal water rights are normally granted first-in-time status for the
water body serving the needs of the reservation.

Table 9.1 lists the current surface and groundwater allocation systems in
the 48 contiguous United States. In Canada, the pattern is much the same,
with eastern provinces employing a reasonable use approach, with or
without permitting, and western provinces relying on appropriation rights.

Source: Viessman and Welty 1985.

9.5 ADMINISTRATIVE AND INSTITUTIONAL SYSTEMS

Administrative law consists chiefly of (I) the legal powers that are granted
to administrative agencies by the legislature and (2) the rules that the
agencies make to carry out their powers. Administrative law also includes
court rulings in cases between the agencies and private citizens. Basically,
under their respective Constitutions, nations like the United States and
Canada have granted powers to various levels of government. Through
those powers, they have further allocated responsibility ("jurisdiction")
among government departments and agencies.

Although administrative arrangements may appear to be trivial forces in

comparison with environmental legislation, they can in fact be very
important in promoting or blocking the flow of information and in
addressing national, regional, and local concerns.

The United States and Canada are countries that cover large geographic
areas. It is not, therefore, surprising that their administrative systems
reflect a decentralization of responsibility. Both countries have "head
offices" located in their respective capitals of Washington, D.C., and
Ottawa, Ontario. Each also has a wide network of satellite offices located
throughout the country, and those regional offices work closely with local
citizens and industries on issues of regional importance.

The following discussion illustrates current administrative structures

within US EPA and Environment Canada as representative of other
structures at the federal and state/provincial levels in those countries.

US EPA The United States Environmental Protection Agency (EPA) is an

independent agency, or department, of the executive branch of government
(Ostler et al. 1996). It is headed by one administrator, a deputy
administrator, and nine assistant administrators, all of whom are appointed
by the president with confirmation by Congress. EPA has 10 regional
offices, each directed by a regional administrator.

EPA's mandate is to administer and enforce a range of major

environmental statutes and, in certain circumstances, to administer and
enforce state and local environmental laws. The laws administered by EPA
include the Clean Air Act, the Clean Water Act, the Safe Drinking Water
Act, the Ocean Dumping Act, the Pollution Prevention Act, TSCA, FIFRA.
CERCLA, and RCRA. Where existing state and local enforcement
programs are deemed to he ineffective, EPA has the authority to step in and
enforce them in support of EPA's mandate. Enforcement usually consists
of notification of violation, followed by informal negotiations if the
violation is not corrected, followed by administrative hearings and, finally.
by legal action in U.S. district court. Penalties for violations of statutes
enforced by EPA can include heavy fines, criminal conviction, and even
imprisonment. EPA can revoke permits and licenses for programs under its
mandate and can impose cleanup orders to correct environmental damage.

EPA also administers national effluent-discharge permit programs for

industries, publicly owned treatment works, storinwater runoff, and a
variety of other programs such as those related to wetland protection and
spill prevention.
 Information about EPA and its activities can be found in EPA's World
Wide Web page: www.epa.gov.

Environment Canada Environment Canada, like US EPA, is an

independent agency, or department, of the executive branch of the federal
government. It is administered by a minister, who is a senior (elected)
member of Parliament appointed to the position by the prime minister. This
appointment does not receive formal ratification like that provided by the
U.S. Congress for senior administrators in that government. The minister is
supported by a staff which he or she personally appoints, and which is
expected to resign when the minister steps down. Members of the
minister's staff are not elected officials but may be present or former
bureaucrats or other individuals. Reporting to the minister is the deputy
minister, the most senior bureaucrat in the agency. The deputy minister is
not an elected official and usually remains in the position even after the
minister has stepped down. Under the deputy minister are a number of
assistant deputy ministers, usually with responsibility for specific program
areas such as Finance and Management, Human Resources, and Corporate
Policy. Also reporting to the deputy minister, and not necessarily through
an assistant deputy minister, are several directors general. The number of
these changes with departmental reorganization, but there are typically
four to seven in total. Each director general is responsible for a program
area or geographic area; for example, there is a regional director geneal for
each regional office of Environment Canada.

Environment Canada administers a number of environmental agencies

with specific mandates, including Action 21 (Canada's response to Agenda
21), the Atmospheric Environment Service, the Canadian Environmental
Assessment Agency, the Canadian Ice Service, the Canadian
Meteorological Centre, the Climate and Water Information Branch, the
Ecological Monitoring and Assessment Network, the Environmental
Effects Monitoring Program, and the National Water Research Institute.

The federal statutes administered by Environment Canada include the

Canada Water Act, the Canada Wildlife Act, the Canadian Environmental
Assessment Act, the Canadian Environmental Protection Act, the
International River Improvements Act, and the Interprovincial Trade Act.
Environment Canada also assists in the administration of the Arctic Waters
Pollution Prevention Act, the Export and Import Permits Act, the Fisheries
Act, the Pest Control Products Act, and the Transportation of Dangerous
Goods Act.

Traditionally, Environment Canada has deferred to the provinces in

matters of environmental control and enforcement, preferring to take an
arm's-length, advisory role instead. Over the past 10 years, however, the
agency has increased its regulatory presence through the Canadian
Environmental Protection Act and the Canadian Environmental
Assessment Act, both of which are duplicated in most provinces'
environmental legislation. Accompanying this change has been an increase
in enforcement activities, which tend to follow the same sequence as those
in US EPA: notification of violation, informal negotiations, formal
hearings, and, finally, legal action.

Information about Environment Canada and its activities can be found

in Environment Canada's World Wide Web page, the "Green Lane":
www.doe.ca/envhome.

9.6 TRANSBOUNDARY WATER ISSUES AND FREE

TRADE

Relationships between the United States and Canada, and the United States
and Mexico, under the North American Free Trade Agreement (NAFTA)
are still evolving. Various authors (e.g., Eaton 1996, Heathcote 1996, and
Holm 1988) have expressed concern about the potential for impacts on
state and provincial laws, on water diversions, on water as a salable good,
and similar issues under NAFTA. Water-poor areas see the trade agreement
as a mechanism for obtaining water for food production and other essential
services. Water-rich areas are worried that the agreement will make water
diversions easier and drain resources away from those who currently own
them.
 Although Canadian federal analysts have argued that water is not
covered by NAFTA or the environmental side agreement, private analysts
(e.g., Linton and Holm 1993) conclude that the agreements cover "all
natural water other than sea water" and not merely bottled water, as has
sometimes been suggested. While certain items are specifically exempted,
water is not among them: "ice, snow and potable water not containing
sugar or sweetener" are considered tariff items in the agreement.

Authors such as Makuch and Sinclair (1993) note that the agreement
does not hind the states or provinces and thus cannot be used to enforce
state or provincial environmental laws. Furthermore, the definition of
environment in the agreement excludes laws regarding natural resource
management, so no enforcement of such laws is possible under NAFTA.
Some analysts therefore fear that the agreement will encourage accelerated
extraction of natural resources, including water.

Canadians' greatest fear in this regard is that NAFTA will facilitate

major water diversions such as the GRAND Canal scheme, a northern
diversion/ hydroelectric project whose ultimate goal is to generate
electricity for sale in the United States. Accelerated extraction of
groundwater for bottled water sale is another major concern. Conservative
Canadian analysts, including those at the Fraser Institute in British
Columbia, justify water exports to the United States on the grounds that
the water will be used to irrigate vegetables grown for export hack to
Canada. Other authors are less optimistic that Canada will reap benefits
proportional to the costs of such diversions.

At present there is consensus that major water diversions are an unlikely

prospect, if only because of their high cost and complex environmental
impacts. Yet it appears clear that if water were to be diverted from a river
and reserved for the use of Canadians only, NAFTA would allow the
United States to launch a trade challenge to obtain a proportional share of
the resource.

Transhoundary issues, including trade agreements, are emerging as

especially important in the implementation of multi jurisdictional
watershed management plans. Chapter 12 discusses these problems in
more detail, presenting case studies from the Danube River, the Rhine
River, and other nutltijurisdictional basins.

REFERENCES

Auhcrt, Vilhclm. 1989. Continuity and Developmmw in Lan, and Socie (v.
London: Norwegian University Press.

Bohannan, Paul. 1967. Latin and Wel%are. Garden City, N.Y.: The Natural
History Press.

Cairns, J. 1992. Restoration of Aquatic Ecos_cstems. Washington, D.C.:

National Research Council (U.S.), National Academy Press.

Careless, J. M. S. 1984. "Toronto to 19/8: An Illustrated History. Toronto:

James Lorimer and Co., Publishers, and National Museum of Man.

de Lob, R. 1991. The institutional pattern for water quality management in

Ontario. Can. Water Resources Journal 16(1):23-43.

Eaton, David .J., ed. 1996. The hnpact of Trade Agreements on State and
Provincial Laws. Proceedings of a conference sponsored by the U.S.-
Mexican Policy Studies Program, University of Texas at Austin,
November 10, 1995.

Ecologic International Inc. 1995. Agricultural nonpoint source pollution

regulatory framework: United States federal and state. In An Agricultural
Profile of the Great Lakes Basin: Characteristics and Trends in
Production, Land-Use and Enuironnien- tal Impacts, edited by the Great
Lakes Commission. Ann Arbor, Mich.: Great Lakes Commission.

Estrin, David, and John Swaigen. 1993. Environment on Trial: A Guide to

Ontario Environmental Law and Policy. 3d ed. Toronto: Emond
Montgomery Publications Ltd.

Fitzgerald, Patrick. 1977. This Law of Ours. Toronto: Prentice-Hall.

Government of Canada. 1991. The Slate of Canada's Environment.
Toronto: Queen's Printer.

Hayman. D., and C. Merkley. 1986. Upper Thames River Rural Beaches
Strategy Program-19X6. Toronto: Ontario Ministry of the Environment.

Heathcote, Isobel W. 1993. An integrated water management strategy for

Ontario: Conservation and protection for sustainable use. In
Environmental Pollution: Science, Policy and Engineering, edited by B.
Nath, L. Candela, L. Hens, and J. P. Robinson. London: European Centre
for Pollution Research, University of London.

Heathcote, Isobel W. 1996. Canadian Water Resources Management. In

Environment and Canadian Society, edited by T. Fleming. Toronto:
Nelson Canada.

Holm, Wendy. 1988. Water and Free Trade: The Mulroney Government's
Agenda for Canada's Most Precious Resource. Toronto: J. Lorimer.

Holloway, James E., and Donald C. Guy. 1990. Emerging regulatory

emphasis on coordinating land use, soil management, and environmental
policies to promote farmland preservation, soil conservation, and water
quality. Zoning and Planning Law Report 13:49.

Jennings, Winfield Holmes, and Thomas G. Zuher. 1991. Canadian Law.

5th ed. Toronto: McGraw-Hill Ryerson.

Kernaghan, Kenneth, and David Siegel. 1991. Public Administration in

Canada: A Text. Scarborough, Ont.: Nelson Canada.

Linton, J., and W. Holm. 1993. NAFTA and Water Exports. Canadian
Environmental Law Association special report. Toronto: CELA.

Makuch, Z., and S. Sinclair. 1993. The NAFTA Environmental Side

Agreement: Implications for Canadian Environmental Management.
Canadian Environmental Law Association special report. Toronto: CELA.
Ostler, Neal K. and John T. Nielsen (eds.). 1996. Environmental
Regulations Overview. Volume 2, Prentice Hall's Environmental
Technology Series. Englewood Cliffs, N.J.: Prentice Hall.

Saxe, Dianne. 1994. Ontario Environmental Protection Act, Annotated.

Aurora, Ont.: Canada Law Book Inc.

Starr, June, and Jane F. Collier. 1989. History and Power in the Study of
Law. Ithaca and London: Cornell University Press.

Stephenson, Ian S. 1975. The Law Relating to Agriculture. Westmead,

England: Saxon House.

Swaigen, John. 1990. The "Right-to-Farm Movement" and environmental

protection. Canadian Environmental Law Reports (New Series) 4(2):121-
128.

Viessman, Warren, Jr., and Claire Welty. 1985. Water Management:

Technology and Institutions. New York: Harper & Row.
10
 Environmental and
Social Impact Assessment
Environmental assessment (EA) has been called one of the more
successful policy innovations of the twentieth century (Sadler 1996).
Begun as a practical response to a persistent administrative problem, EA is
now applied in more than 100 countries around the world to help decision
makers make informed judgments about projects that have potential to
impact the environment.

10.1 THE HISTORY OF ENVIRONMENTAL ASSESSMENT

POLICY

The roots of EA lie in congressional concerns in the late 1960s. At that

time many government agencies were frustrated by the waste of time and
money that resulted when a major public project that was well underway
encountered public opposition. Often, the concerns expressed by the public
revealed that project planners had not considered important indirect or
cumulative effects of the project. These deficiencies were difficult to
rectify once the project was under construction. Congress wanted to
establish a legislative requirement that federal agencies would carry out a
complete and balanced assessment of impacts before construction was
allowed to proceed. Congress was also concerned about the social and
economic well-being of individuals and communities in addition to, and as
linked with, biophysical impact. Of particular concern were cumulative,
secondary, and "off-site" environmental effects (Gibson and Savan 1987).
Although reactive legislation (and common-law remedies) were available
to punish those who adversely affected the environment, Congress sought
to prevent those impacts by careful data collection and informed decision
making.

To this end, Congress drafted two action-forcing clauses, which, at the

time, must have seemed modest if not downright unimpressive, but which
have since had far-reaching consequences around the world. In 1969,
Congress passed the National Environmental Policy Act, which includes
an "action-forcing" provision (Section 102(2)(b)) to the effect that:

All agencies of' the Federal Government shall ... identify and develop
methods and procedures ... which will ensure that presently
unquantified and environmental amenities and values may be given
appropriate consideration in decision making, along with economic and
technical considerations.

Section 102(2)(c) of the Act contains the basis of most modern

environmental assessment legislation:

All agencies of the Federal Government shall ... include in every

recommendation or report on proposals for legislation and other major
Federal actions significantly affecting the quality of the human
environment, a detailed statement ... on (i) the environmental impact ...
(ii) adverse environmental effects which cannot he avoided ... (iii) the
alternatives to the proposed action ... (iv) the relationship between local
short-term uses ... and ... long-term productivity, and (v) any
irreversible and irretrievable commitments of resources.

Congress intended these provisions to ensure that proposed projects

were evaluated on the basis of clearly stated purposes, adequate and
balanced consideration of alternatives, and careful evaluation of potential
impacts on the social, economic, cultural, and biophysical environment. In
the past projects had not been exposed to this rigorous evaluation, with the
result that problems had often cropped up during or after construction,
after large quantities of time and money had already been expended.
Congress also hoped that by forcing comprehensive and early data
collection, more cumulative, secondary, and off-site environmental effects
could be foreseen before costly projects were actually built.

Many countries have now adopted EA as a central component of their

environmental policy framework. Although the form and process of EA
vary from country to country, the intent of the legislation is virtually the
same all over the world. So fundamental was the National Environmental
Policy Act in proposing this approach, in fact, that the U.S. Council on
Environmental Quality (1993) has called it the "Magna Carta" of the field.

In both Canada and the United States the environmental assessment

process is open to public scrutiny at all stages. This means that the scope
of the assessment, and the way that each element is analyzed, must be well
justified and competently performed. If the analyst is careless about what
is included, or how impacts are assessed, the government has the right to
halt progress on the project until an adequate assessment has been
compiled. As a result, the EA process sometimes functions as a mechanism
for building social consensus about major public activities-what Baber
(1988) has called "procedural democracy."

In essence, EA forces the proponent of a major project to evaluate the

proposed undertaking in light of:

• A clearly stated purpose

• Adequate consideration of alternatives

• Comprehensive evaluation of the potential impacts on the social,

economic, cultural, and biophysical environment of the undertaking
and of alternatives to the undertaking

• Preventive or mitigative measures that would be required to reduce or

eliminate anticipated impacts

In addition to the many countries that have adopted EA as part of their

policy framework, international lending agencies like the World Bank and
international aid organizations employ EA in the evaluation of projects in
client countries. In 1985, the European Union (EU) established minimum
provisions for environmental impact assessment (EIA) in a directive to
member states (these provisions were not binding until 1988). Canada,
New Zealand, and Australia were among the first countries to follow the
lead of the United States in implementing EA. Originally, Canada opted
for an administrative process (Australia chose a legislated approach),
which was replaced in 1992 with the Canadian Environmental Assessment
Act. Now all 10 provinces and both territories have EA processes that bind
provincial (territorial) activities. Indeed, Ontario's definition of
environment under the Ontario Environmental Assessment Act remains the
most powerful in that province's environmental legislation and may be the
most comprehensive in Canada. In Ontario's Environmental Assessment
Act, enviromnent is defined as:

• Physical features (air, land, water) and

• Buildings, structures, machines, or other devices or things made by

man;

• Any solid, liquid, gas, odor, heat, sound, vibration, or radiation

resulting directly or indirectly from the activities of man;

• Biological subjects (plant and animal life, including man) and

• Human and ecological systems (the social, economic, and cultural

conditions that influence the life of man or a community); and

• Any part or combination of these elements

10.2 OVERVIEW OF THE EA PROCESS

There are many forms of EA. Vanclay and Bronstein (1995) report at least
a dozen categories and subcategories, including environmental impact
assessment, social impact assessment, technology assessment, policy
assessment, economic assessment, health impact assessment,
environmental auditing, and sustainability assessment. Sadler (1996)
defines four key terms in EA:

L Environmental assessment (EA) is a systematic process of evaluating

and documenting information on the potentials, capacities, and
functions of natural systems and resources in order to facilitate
sustainable development planning and decision making in general, and
to anticipate and manage the adverse effects and consequences of
proposed undertakings in particular.
 2. Environmental impact assessment (EIA) is a process of identifying,
predicting, evaluating, and mitigating the biophysical, social, and
other relevant effects of proposed projects and physical activities prior
to major decisions and commitments being made.

3. Social impact assessment (SIA) is a process of estimating the social

consequences that are likely to follow from specific policy and
government proposals, particularly in the context of national EA
requirements.

4. Strategic environmental assessment (SEA) is a process of prior

examination and appraisal of policies, plans, and programs and other
higher-level or preproject initiatives.

Environmental assessment often encompasses three stages: preliminary

assessment, detailed assessment, and follow-up. Preliminary assessment is
used to determine whether a project is covered by EA legislation or policy,
whether an EIS is required, the necessary nature and extent of the EA
process, and scoping. Scoping, in terms of EA, generally includes the same
tasks as described in Chapter 3: determining the boundaries of the plan,
and identifying the issues, processes, and components that are most
important in the analysis. Detailed assessment includes analysis of the
impacts and mitigation necessary for the undertaking and a range of
alternatives to the undertaking, usually including the "do nothing" option.
Follow-up includes monitoring and audit functions to determine the actual
impacts of the project and to ensure that necessary mitigation measures are
in place. A typical EA process is illustrated in Figure 10.1.

Although this process appears to be straightforward, it can be very

difficult for the practitioner to decide what to include in the assessment,
what to exclude, what to emphasize, and so on. Figure 10.2 illustrates a
hierarchy of principles that can assist in the successful application of EA.

The difficulty in applying a theoretical hierarchy such as that shown in

Figure 10.2 lies in determining whose values should be incorporated in the
EA process. Ideally, of course, there should be social consensus on values.
Who, after all, could argue with "integrity," which Sadler (1996) defines as
conforming "with accepted standards and principles of good practice"? Yet
in environmental assessment processes such as the Canadian federal EA
process in regard to the vast proposed Voisey's Bay nickel mine in coastal
Labrador, aboriginal interests and values may not at all align with Western
values. "Accepted standards" of behavior in the affected Innu and Inuit
societies differ markedly from those in typical English-speaking groups.

Consider the following simple example of this problem. A white,

English speaking hydrologist may make an inventory of water resources
and conclude that there are primary and secondary streams, small lakes,
wetland systems, estuaries, and oceans. By contrast, an Innu elder may
identify 20 different categories of "small lakes": grassy lakes, shallow
lakes, muskeg lakes, swampy lakes, chains-of-lakes, small deep lakes with
no apparent source, small lakes, very small lakes, and so on (Clement
1995). "Good practice" to the elder may include a category-by-category
analysis of impacts on each type of lake. The appearance of the water, and
the plants and animals it supports, would be an important part of this
analysis. (The Innu are traditional hunter-gatherers who rely heavily on the
food resources of the back country.) To the white hydrologist, good
practice may consist simply of assessing probable changes in lake levels or
streamflows, with a statistical assessment of expected high and low values.
To the Innu, deep lakes with no apparent source are particularly special,
because they are thought to have spiritual or even magical properties.
Damage to these lakes (for instance, through tailings disposal practices)
might be viewed by the Innu as far more important than, say, damage to a
muskeg lake. This distinction would be lost on the average English-
speaking hydrologist."
Figure 10.1 Typical environmental assessment process.
Figure 10.2 Hierarchy of environmental assessment principles (adapted from
Sadler 1996).

Values also become critical when conflicts arise and a choice must be
made. For instance, a mining engineer may select technology A over
technology B because A is lower in cost but still does a reasonable job of
pollution control. But if technology A is much noisier than technology B
(and would thus affect caribou migration, for example), or if technology A
will affect those deep, sourceless lakes more than technology B, the Innu
elder might opt for B instead. Who is "right"`?

Many EA texts implicitly adopt Western values without even

acknowledging the fact that consensus on values rarely exists. There is a
large literature on "traditional ecological knowledge" (TEK), some of
which addresses this issue. The interested reader is referred to Inglis
(1993) and Freeman and Carbyn (1988). Clement (1990, 1995) provides
detailed reviews of Innu ethnobotany and eth- nozoology.

Finally, business interests sometimes object to the time and cost

involved in full-scale EA. Yet over the long term, EA can provide
important guidance as to what businesses should expect to encounter-and
to pay for. The World Business Council on Sustainable Development
(1995) has said:

Running a business encompasses a broad range of activities, not least of

which is the management of potential risks associated with failure to
handle environmental matters adequately. Industry does this by
managing its environmental affairs with due care and by being
increasingly eco-efficient, taking account of scientific, technical, and
economic factors, and the requirements of environmental legislation as
a starting point.

At the heart of sound environmental management is the assessment of

effects, real or potential, on the environment as a consequence of
business activities and the planning and implementation of measures to
avoid or mitigate that damage. Environmental assessments can assist
companies in their quest for continuous improvement by identifying
ways of maximizing profits through reducing waste and liabilities,
raising productivity and demonstrating a company's sense of duty
towards its customers and neighbors.
 Section 10.3 describes some of the most common methodologies for
EA. Sadler (1996) provides a discussion of the broader context of EA and
an evaluation of the effectiveness of current practice.

10.3 CHALLENGES IN ENVIRONMENTAL ASSESSMENT

10.3.1 Scoping

The problem of scoping was discussed at length in Chapter 3, and many of

the concepts presented there also apply to EA. Scoping sets the boundaries
for EA, it asks the important questions, and it identities the information
needed for successful decision making. Sadler's (1996) review shows that
all jurisdictions have difficulty with scoping, especially in prioritizing
issues and identifying the impacts of real concern. In part, this may be
because underlying values and expectations differ among the participants
in an EA process. If these fundamental values and expectations are not
clarified and the debate on scoping drags on, a number of problems are
likely to occur:

EIS (Environmental Impact Statement) documents become large and

poorly focused in an attempt to satisfy the range of interests.

• Important issues are identified late, forcing major revisions to the EIS
and thus significantly increasing expenditures of time and cost.

• Impact analysis becomes a "laundry list" or catalog rather than a

focused assessment.

Environmental assessment is now required for most major public

projects. Guidelines developed by the U.S. Council on Environmental
Quality (CEQ) (1981) provide advice on all aspects of EA. In terms of
scoping, the CEQ guidelines advise:

1. Start scoping when you have enough information to present an initial

list of alternatives and environmental issues.
 2. Prepare an information package so that participants can make an
informed contribution to the process.

3. Design the scoping process.

4. Give the public adequate notice of opportunities for participation in

the scoping process.

5. Hold a public meeting to hear public opinions.

6. Review comments received from the public, to identify major issues

and potential impacts.

7. Develop work schedules and budgets that will allow you to address
the major issues and impacts in a timely and efficient manner.

8. Where issues are contentious, try alternative process formats (e.g.,

workshops, review committees, independent moderators, etc.) to
resolve conflicts.

In general, EA is aimed at determining the direction, magnitude, and

significance of impacts that may arise from the development of a project.
"Direction" means positive or negative impacts, both of which should be
included in the analysis. For example, a dam project may have positive
impacts on (would reduce) flood losses, but negative impacts on fish
habitat (through inundation and excessive sedimentation). "Magnitude"
means how big the effect will be-for instance, the areal extent of an
impact, or the difference between preproject conditions and postproject
conditions. "Significance" means the importance of the effect to the
community. For instance, a major flood may have a large magnitude (cover
a large area, result in much higher than normal stream flows) but low
significance if the inundated land is not inhabited, contains no valuable
structures or machinery, and the inundation will not affect resources
contained in that land. On the other hand, a flood of the same magnitude,
but in a major city, may have enormous significance. Most jurisdictions
also evaluate "residual" impacts-the effects that are expected to remain
after mitigative measures have been put in place.
 Methods of assessing impact are by no means perfectly precise or
accurate. These issues have been addressed elsewhere in this book,
particularly in Chapters 3, 6, and 7. Chapter 8 presented methods of
incorporating risk and uncertainty into planning that may also have
application in EA. Typically, an EIS contains the following elements:

• A description of the proposed action or project

• A statement of the purpose of, and need for, the project

• A description of the environment likely to be affected by the project

• A discussion of prevailing land use plans, policies, controls, and

applicable legislation in the affected area

• A list of alternatives to the project (for instance, "do nothing")

• A list of alternative ways of carrying out the project

• A discussion of the anticipated positive and negative effects on the

environment of the preferred project and each feasible alternative to it

• A discussion of the steps that could be taken to mitigate those effects,

for the preferred project and each feasible alternative

• A discussion of' the relationship between short-term uses of the

environment and long-term productivity

• Any irreversible or irretrievable commitments of resources that would

be involved in the proposed action if it were implemented

Table 10.I shows a suggested outline for preliminary or detailed EIS

documents, based on CEQ guidelines and other sources:

Applicable guidelines for EA vary considerably from jurisdiction to

jurisdiction. Those engaged in the preparation of EIS documents or in
planning EA processes should consult the appropriate federal, state, or
provincial agencies for advice on the guidelines for their region and type
of project.

10.3.2 Evaluation of Impact Significance

Many basic EA texts discuss the problem of evaluating the significance of

impacts. Possible approaches include those described in Chapters 3, 5, 6,
and 7, for instance, identification of acceptable target values and
evaluation of the likelihood of postproject effects resulting in values in
excess of those targets. "Best professional judgment" and other qualitative
approaches are also cony monly used. Some authors Hilden 1996)
recommend using quantitative approaches when effects can be predicted
with some degree of accuracy (or perhaps when the degree of accuracy is
known) and negotiation with stakeholders when available information is
sparse, uncertain, or contradictory. The approaches described in Chapters
3, 5, 6, and 7 of this hook work well in the former case; the techniques
described in Chapter 4 may be useful in the latter.
 The Canadian Environmental Assessment Agency (1995) has proposed
the following tests for environmental impacts.

An effect is considered "adverse" if it causes one or more of the

following:

• Effects on biota health

• Effects on rare or endangered species

• Reductions in species diversity

• Habitat loss

• Transformation of natural landscapes

• Effects on human health

• Effects on current use of lands and resources for traditional purposes

by aboriginal persons

• Foreclosure of future resource use or production

An effect is considered "significant" if it falls into one or more of the

following categories:

• Large

• Geographically widespread

• Of long duration or high frequency

• Irreversible

• In a sensitive ecological context

A significant effect may be considered "likely" if it has either:

 • A high probability of occurrence, or

• Significant scientific consensus about its occurrence.

Sadler (1996) recommends four steps in the evaluation of significance:

1. Assessment of the nature and extent of impacts (e.g., type, duration)

Impacts are more likely to be significant if they are extensive

over space or time.

Impacts are more likely to be significant if they are intensive in

concentration relative to assimilative capacity, or if they exceed
allowable environmental standards or thresholds.

2. Assessment of the likely effects on the receiving environment (e.g.,

sensitive area, land use, community traditions)

Impacts are more likely to be significant if they adversely and

seriously affect ecologically sensitive areas, heritage resources,
other land uses, community life-style, and/or indigenous
peoples' traditions and values.

3. Assessment of the magnitude of impacts (e.g., low, moderate, high)

4. Identification of options for impact mitigation (e.g., reduction,

avoidance)

Sadler also recommends incorporation of tests of significance at various

stages in the EA process, with focus on the following questions:

1. How adverse are the predicted effects (e.g., change, loss,

foreclosure)?

2. How do these vary in scope and intensity (e.g., in their effect on

ecological and resource values)'?
 3. How significant or serious are the impacts (e.g., irreversible versus
inconsequential)?

4. How probable is it that they will occur (e.g., high risk versus low
risk)?

In a sense, therefore, environmental assessment is a risk assessment of a

project before it is undertaken. The term risk assessment is used in the
sense of formal risk analysis, in which both the effect (for instance,
toxicity) and probability (risk) are taken into account in decision making.
Environmental assessment seeks to identify the possible significant
impacts of a project in the context of their likelihood. Section 10.4
discusses a range of methodologies commonly used for the assessment of
impacts.

10.4 ENVIRONMENTAL ASSESSMENT

METHODOLOGIES

Numerous authors (e.g., Jain and Urban 1975; Henderson 1982; Goodman
1984) have described methodologies available for EA. The following
discussion provides an overview of the main methods currently in use,
with comments about their suitability (or lack of suitability) in certain
situations.

10.4.1 Matrix Analysis

Matrix methods may be the most familiar and easily understood of all EA
methodologies. One of the earliest, and certainly the best known, of these
is that proposed by Leopold et al. (1971) for the U.S Geological Survey.
The matrix developed by these authors lists 88 environmental
characteristics on the vertical axis and 100 actions with potential to affect
the environment on the horizontal axis. The cells of the matrix are then
filled in to indicate where actions may affect environmental characteristics.

Leopold and his colleagues proposed a method of scoring that employs

a diagonal line drawn across the cell, from the lower left to the upper right
corner. In the top half of the cell, a number from I to 10 is entered,
indicating the "magnitude" of the impact. In the lower half of the cell,
another score, also from I to 10, shows the "importance" of the impact.
Higher numbers indicate greater magnitude or importance, as the case may
be. A plus sign (+) can be used to indicate a positive impact. Attention can
be drawn to cells with particularly high scores by circling or highlighting
them so they stand out from the matrix. Figure 10.3 shows a portion of
such a matrix, adapted from the matrix developed by Leopold et al. (1971).

This matrix is intended to be more or less universal. Its large size makes
it impractical to use in toto, however, so its authors suggest that users
develop a reduced matrix incorporating only rows and columns applicable
to the project under study. The authors do not suggest summation of
scores, such as might be used in other multiattribute decision-making
methods (see, for example, Section 5.5.4). Indeed, they point out that cells
cannot be compared directly; in other words, a score of 10 in importance in
terns of "airports" may not equate directly to an importance value of 10 for
railroads or canals. The values in individual cells do, however, provide a
basis of comparison between various project alternatives. That is, it is
possible to compare the magnitude and importance scores for a cell formed
by the intersection of "erosion" and "canals" among preproject conditions,
preferred project conditions, and the conditions that would obtain under
various alternative approaches to the project.

Many variations of the Leopold matrix are possible. Although broadly

applicable, the matrix can, for example, be made more specific for a
certain kind of activity, such as a landfill site or an irrigation project. In
these projects. it may be possible to itemize individual actions and to
separate similar actions in different places. For example, while this matrix
simply cites "paving" as one of the land transformation actions, a more
detailed and specific matrix could be created by listing "paving of
roadways," "paving of administration complex parking lot," and "paving of
airstrip." Alternatively, the matrix could specify steps in the paving
process: "laying of gravel sublayer," "compaction of gravel layer,"
"application of asphalt layer," and so on. Similarly, the environmental
characteristic "water quality" could he broken down into individual water
quality parameters, such as the indicators described in Chapters 3 and 5.
These may include, for instance, fecal coliform bacteria, total phosphorus,
total suspended solids, and total lead as indicators of a range of water
quality concerns. Environmental components or characteristics can be
added as appropriate to the climate and culture.

In short, the Leopold matrix provides an excellent starting place for EA,
and one that is both easily understood and easily customized for the
purposes of the project at hand. It serves as a visual summary of analyses
that appear in narrative form in the EIS document. Figure 10.4 illustrates
the use of a Leopold-type matrix in the Canadian federal-provincial (Nova
Scotia) EA of the proposed Halifax-Dartmouth Metropolitan Sewage
Treatment Facility.
10.4.2 Overlays and Geographic Information Systems

Overlays and transparencies have been used for decades to evaluate the
intersection of activities and environmental components. This technique
has much to recommend it in terms of simplicity and ease of use. The basic
approach is to take a base map, usually a topographic map, of some
appropriate scale, and to place over that base one or more transparencies
on which are drawn information about environmental characteristics
and/or the expected impacts of the proposed action.

For example, a base map of a subwatershed area could be overlaid with

a transparency showing the location of bathing beaches, another showing
historical and archeological structures, a third showing sensitive habitat,
and so on. These three overlays alone will indicate the conjunction of uses
of different types (heavy traffic at bathing beaches could, for instance,
have adverse effects on nearby sensitive habitat). Once sensitive areas
have been mapped in this way, new overlays can be prepared showing the
anticipated effects of the proposed project. For instance, one transparency
might show the expected range of noise effects during construction.
Overlaying this on the beach/habitat maps would quickly reveal whether
those uses are likely to he impacted by noise during construction.

Color can be a useful adjunct in overlay mapping. Single-color overlays

can, for instance, employ a range of color intensity to indicate degree of
concern (or degree of protection needed). A single-color overlay of
sensitive habitat could show areas of local significance in 10% shading and
highly-valued areas of national significance (perhaps endangered species
habitat) in 30% or 50% shading. Similarly, impact overlays could employ
graded color, so areas of minor noise impact would be shown with 10%
shading, moderate impact by 30% shading, and intense impact by 50%
shading. When overlaid, the maps of sensitive habitat and noise impacts
will show intense colors in areas where the beneficial use is highly valued
and the impact is likely to be intense: "significant" impact.

Table 10.2 lists a range of attributes that have been mapped using
overlay techniques.

The advent of Geographic Information Systems (GIS) in the past decade

has vastly increased the power of the overlay method in EA. GIS systems
are essentially computerized maps containing georeferenced data of almost
any kind. The term georet rented implies that each piece of information-for
instance, soil type-has map coordinates associated with it. Most GIS
systems (and there are many different ones available) incorporate basic
mapping capability, a relational database, and a suite of interpretive tools,
such as statistical analyses. Using a GIS, it is possible to store a wide range
of information that would formerly have been held only in map form, and
then to extract that information in different combinations as necessary for a
given analytical purpose.

Many, perhaps most, public agencies are now using GIS to store and
display data. The technique lends itself well to visualization, and some
agencies are now taking advantage of this feature to increase public access
to data and collect public feedback on different planning options. A
computer display terminal may, for instance. be set up in a public library or
shopping center, linked to the main agency database. A member of the
public can view regional and local maps on the computer. "point and click"
to select an area of particular interest, zoom the display to obtain a close-
up view, and then request different types of display: which of the area
bathing beaches were closed to swimming in the past calendar year? which
agricultural areas contain soils with a high risk of erosion'? which of these
are close to streams'?-and so on.

GIS techniques formed the basis of a recent major study of agricultural

land use and environmental impacts in the Great Lakes Basin (Great Lakes
Commission 1996). The study aimed to develop an overview of
agricultural practice and the impacts of agriculture on the environment of
the Great Lakes Basin. A GIS was chosen as the framework for data
storage and analysis, in part because GIS systems were already in use in
the U.S. Department of Agriculture and Agriculture Canada. the two
federal agencies that would he central in the study.
 In practice, the GIS proved difficult to use. This difficulty stemmed
from differences in data collection practices and GIS technology used in
the two agencies. As noted earlier, there are many GIS systems available.
Although the data can usually he "exported" from one and "imported" into
another, these data transfers are seldom straightforward or easy. More
difficult in the Great Lakes Commission's study was the problem of scale
of data collection. In the United States, most data were collected at a
county scale. In Canada, the watershed was the more common unit. As a
result, data were difficult to overlay accurately and large areas lacked data
altogether.
 Finally, where one data set did not exactly match another in areal extent,
orphan polygons were created in the GIS maps. Whereas in a manual
overlay system, the analyst's eye may make a judgment that this little piece
could be lumped in with this or that area (perhaps because the analyst has
direct knowledge of the area involved), in a GIS this is much more
difficult. A GIS employs computer-based computation to analyze data (for
instance, to compile summary statistics for an area) and cannot
automatically distinguish between an intended map unit (e.g., a county)
and an orphan polygon created by default when two overlays do not match
exactly. If the matching of one GIS layer to another is generally poor,
many such orphan polygons will be created and the necessary
computations can increase exponentially, with associated increases in
computing time and costs. Techniques are available to deal with orphan
polygons-for instance, by using a system of rules to merge orphan
polygons with adjacent larger map units. To date, however, these
techniques have been cumbersome to use. In the Great Lakes Commission
work, the most detailed data was available for the U.S. Great Lakes states,
and in some cases lack of data prevented analysis of the Canadian side of
the basin altogether.

The technique is, nevertheless, a powerful one, primarily because of its

flexibility in data display. Figures 10.5 and 10.6 show examples of GIS
results from the Great Lakes Commission study.

Increasingly, GIS systems are being linked to predictive water quality

models, so that predictions about in-stream impacts can he continually
updated as better land use, water quality, soils, and other data become
available. GIS-based agency databases therefore serve two purposes: as a
data storage system and as input to predictive computer simulation models.
It should be emphasized that the linkage between GIS and computer
simulation models remains awkward in most cases, often requiring several
intermediate "translation" steps. The technology is continually improving,
however, and both GIS systems and computer models are now being
developed with the goal of linkage in mind.

10.4.3 Checklist Methods

A checklist approach can he used to itemize impacts with and without a
given project. Sometimes these methods are informal, and sometimes they
involve complex scoring and weighting schemes. In the simplest form, a
list of envi ronmental components is made on the vertical axis, while the
horizontal axis contains the various project options. Table 10.3 illustrates
such an approach.

The list of environmental components can be tailored to suit the

particular environment of the project. Similarly, the "projects" along the
horizontal axis can be broken down into specific activities. Separate
checklists can be used for categories of activities. For instance, the impacts
of accidental events can be evaluated with a checklist that itemizes
environmental components down the vertical axis and individual types of
accident across the horizontal axis. Table 10.4 illustrates the type of
checklist that was used in the federal-provincial EA for the Halifax-
Dartmouth Metropolitan Sewage Treatment Facility.

The checklist method can be adapted to include some information about

the magnitude and direction of impact-for instance, by extending the range
of symbols used, by replacing symbols with positive or negative numbers,
or by the use of footnotes, bold type, or flags to indicate highly significant
effects.
Figure 10.5 Population density change in the Great Lakes Basin: Change in
individuals per sq km of county area 1981/ 1982 to 1991/ 1992 (based on
1991 Canadian Census of Agriculture, 1992 United States Census) (Great
Lakes Commission 1996).
Figure 10.6 Change in herbicide usage in the Great Lakes Basin: Change in
area receiving herbicide 1981/ 1982 to 1991/ 1992 (based on 1991/ 1981
Canadian Census of Agriculture, 1992/ 1982 United States Census) (Great
Lakes Commission 1996).

10.4.4 Combination Methods

There are probably as many methods for EA as there have been

environmental assessments. In part, this is because analysts discover
shortcomings in existing methods and modify them to suit their own needs.
The main challenge in EA methodology is not so much in data collection
(although that too has its difficulties) as in data presentation. The goal is to
present data on the proposed undertaking, as well as a range of alternatives
to it, and to list for each the impacts on a range of environmental
components and the magnitude, direction, and significance of those
impacts. Ideally, the presentation should be simple so that it is accessible
by any member of the community. This is a tall order indeed.

As one example of a "combination" method loosely based on the

Leopold matrix and checklist methods, Jacques Whitford Environment
Ltd. (1992) developed a scheme for cataloging the impacts of a proposed
regional sewage treatment facility on the busy Halifax (Nova Scotia)
Harbor. The harbor has a thriving commercial fishery and complex
hydrodynamics that would likely be altered by discharges from a large-
capacity sewage treatment plant. Jacques Whitford's scheme employs
several elements:

• A symbol to indicate direction and magnitude of impact

• A numerical value for geographical extent of the impact

• A numerical value for the duration of the impact

• A numerical value for the frequency of impact

• A number indicating the basis of the ratings (i.e., professional

judgment only, available data and professional judgment, abundant
data supported by professional judgment)

An example of the Jacques Whitford scheme appears as Figure 10.7.

The method is not easily understood without frequent reference to a
legend. but it does capture several different considerations in a single
matrix cell and thus extends the Leopold matrix approach significantly. In
an effort to reduce the complexity of impact-assessment matrices,
however, Jacques Whitford has simplified the list of environmental
components to the major elements listed in Table 10.4. This approach may
underestimate the impacts on specific environmental components (for
example, a particular water quality parameter) while overemphasizing the
importance of "valued environmental components" such as the great blue
heron and osprey that were selected for inclusion.
Note: This combination methodology uses a rectangle containing
a symbol indicating probable impact (see Figure 10.4) and four
numbers providing additional information on the geographic
extent, duration, and frequency of the impact, and the basis for
these ratings.

Figure 10.7 Example "combination" EA rating methodology (after Jacques

Whitford Environment Ltd. 1992).

Goodman (1984) offers a helpful review of various EA applications,

showing how basic methodologies have been modified to suit the needs of
particular EA processes.
10.5 SOCIAL IMPACT ASSESSMENT

Social impact assessment is now an integral part of most water resources

planning exercises and, indeed, is required by law for many. Such an
assessment provides a way for government agencies and other proponents
of major projects to anticipate impact on people and communities,
especially when the benefits and costs of a project are not equally
distributed. It is often the case that major multiobjective public projects
have benefits distributed over a wide geographic area, but impacts that are
concentrated in the local community. The same is true of many major
private projects such as mine developments, pulp and paper mills, and
similar undertakings. While these projects may have enormous economic,
and therefore social, benefits, those benefits seldom accrue entirely to the
community directly affected by the project. Social impact assessment
attempts to describe the effect on the community of such projects.

As described earlier in this chapter, impact assessment is based on

values: what is a significant impact to one group may be considered a
negligible impact, or even ignored, by another. Many groups bring values
to the planning table. The affected community usually has at least one
group, and may have several different groups, whose perspectives are
informed by their fundamental views of what is right and wrong. Agency
staff and other public officials have perspectives that differ again, perhaps
because they incorporate consideration of a wider geographic region or an
emphasis on service provision over other considerations. Corporate,
commercial, and institutional interests can all be defined in some
processes. Each of these groups may see an effect as "good" or "bad,"
depending on its perspective.

Chapter 4 described a range of techniques suited to public involvement

and public consultation in water resources planning. Many of these also
have value in formal social impact assessment. Goodman (1984) suggests
that the 12 social science tools most frequently used in this context are:

1. Institutional analysis
 2. Policy profiling

3. Value mapping

4. Social profiling

5. Content analysis

6. Small-group process techniques

7. Human-cost accounting

8. Community impact assessment

9. Ethnographic field analysis

10. Questionnaire and nonparametric statistical analysis

1 I. Population projections

12. Trend and cross-impact analysis

The first three of these methods were discussed in Chapter 4 and

represent (along with (6) and (10)) methods of determining community
response to, and concerns about, a proposed project. Chapter 3 discussed
methods for compiling population projections.

Social profiling and content analysis attempt to map the range and depth
of public concerns in a systematic fashion. (Use of the public involvement
techniques described in Chapter 4 can produce a vast quantity of'
information that is difficult to interpret and understand without a content-
analysis framework.) Social profiling is discussed in more detail by Flynn
and Schmidt (1977); Dunning (1982) provides a review of content analysis
techniques.

Human-cost accounting attempts to measure the psychological and

behavioral impacts of a project. Affected individuals are positioned on a
trauma scale, which is then related to standard benefits scales paid by
agencies like the U.S. Veterans Administration for similar disabilities.
Other human health impacts can be translated into costs using hospital-
bed-days and estimates of the costs of economic disruption. Community
impact assessment examines demographic changes in the community (for
instance, worker influx) as a result of project activities. As described in
Chapter 3, worker influxes (as one example) can he translated into demand
for goods and services of various kinds, and thus can be used to forecast
other kinds of community impacts.

Ethnographic field analysis, a method of on-site observation of

community structure and culture, usually requires the assistance of an
expert.

Some of these techniques (e.g., social profiling, small-group processes,

questionnaires) are well suited to defining problems. Others (e.g., small-
group processes, policy profiling, and values mapping) are useful in
generating alternatives (and see Chapter 5 of this hook). Most of' the
techniques are useful in compiling an inventory of the society that will be
affected by the project. The last nine of the list have particular value in
analysis and comparison of plans and in selection of a preferred
alternative.

As with environmental impact analysis, impacts that may be considered

include both "direct" (immediate, obvious) impacts, such as those related
to population relocation, major demographic changes, and economic
trends, and "indirect" (delayed, less obvious) impacts, such as impacts on
the health care and education systems, on community attitudes and values,
and on "quality of life." Both kinds of impacts may be assessed over few
years (short term) or many (long term).

Table 10.5 lists the social science variables commonly considered in

social impact assessment.

Social impact occurs at the level of the individual, the family, the
neighborhood (or community group), the community, and the region. At
the individual or• family level, Maslow (1954) suggests a hierarchy of
needs, beginning with basic survival and physiological needs, followed by
security and self-sufficiency needs, then by social interaction needs, then
by esteem, and, finally, growth or self-actualisation. At the co, nnuniIv
level, concerns center on community cohesion and the structures and
values that promote that cohesion. These may include considerations such
as community stability, opportunities for public participation in the
community, and the presence or absence of community structures to
enhance cohesion. At the regional le vet analysis generally focuses on
patterns of economic growth, employment, infrastructure, and
demographic changes.

Any of the basic methodologies described in Section 10.3 can be used in

social impact assessment, with the "environmental characteristics" or
"components" replaced with social characteristics or components-for
instance, those in Table 10.5. The applicability of matrix and checklist
methods will be easily seen, but overlay and GIS-assisted analysis can also
be used, with overlays showing average income level, population density,
land use, and other features.

In summary, social impact assessment seeks to define the social systems

that currently exist in a watershed or other geographic area and to
determine how management actions will alter those social systems. In
some jurisdictions (for instance, Ontario) social and commercial systems
are included in the definition of "environment" under environmental
assessment legislation. In other jurisdictions, separate legislation
governing social impact assessment may apply.
10.6 STRATEGIC ENVIRONMENTAL ASSESSMENT

Strategic environmental assessment (SEA) is an emerging area with

significant potential to affect water resources management. As discussed in
Section 10.2, SEA is a process of prior examination and appraisal of
policies, plans, and programs and other hi(,her-level or preproject
initiatives. According to Sadler (1996), there is still considerable debate as
to the scope of SEA and its potential application to environmental policy,
as compared with plan or program levels. Sadler makes the case that SEA
of policies, plans, and programs is necessary and important because it:

• Builds environmental considerations into all levels of decision

making, not just project approvals

• Helps to determine the need for the feasibility of government

initiatives and proposals

• Avoids the foreclosures of options and opportunities that arise when

assessment occurs only at the project stage

• Addresses environmental issues and impacts that are best dealt with or
can he considered only at the policy or program level (for example,
those that cannot readily he divided into projects)

• Establishes an appropriate context for project EIA, including pre-

identification of issues and impacts that are likely to warrant detailed
examination

In Canada, the federal government established a process of policy and

program assessment as part of a Cabinet directive of 1990: a parallel
review process is not in place at the provincial level. This federal process
requires all agencies submitting policy and program proposals to document
the environmental impacts of' those policies or programs. The actual level
of compliance with this policy is difficult to determine, but it is clear that
not all federal agencies carry out the required policy EA to the level
intended in the original Cabinet directive.
 In the European Union, internal policies call for EA of new legislation
and actions likely to affect the environment. Member states are
increasingly incorporating SEA-type reviews of policies, plans, and
programs.

In the United Kingdom, there are no formal policy requirements for

SEA, but the Department of the Environment has issued "good practice"
guidelines that require EA of policies drafted by government agencies.
Policy assessment incorporates benefit-cost analysis, scoping,
identification of issues, and development of a policy impact matrix.

In the United States, the National Environmental Policy Act of 1969

(NEPA) in theory applies to all legislation and other major federal actions,
but in practice only program-level EIS reports (PEISs) are prepared. These
cover a wide range of resource management, waste disposal, and other
types of environmentrelated programs and are increasingly used in long-
term planning, dealing with cumulative effects, tiering actions requiring
project EIA, and evading costly litigation (Sadler 1996). Sadler also notes
that, to date, NEPA requirements have not been applied to broad
government policies. Other than in California, SEA is not actively
practiced at the state level.

SEA is hampered in many jurisdictions by a lack of political will and a

limited societal support base. There is, however, considerable support for
the notion among EA professionals. A typical SEA process contains the
following elements:

1. Accountability by agencies who initiate policies, plans, or programs

for assessing the environmental effects of those initiatives

2. Assessment of effects early in policy, plan, or program design

3. Scope of the SEA sufficient to include all potential impacts or

consequences of the policy, plan, or program

4. Clear objectives and terms of reference for the SEA

 5. Inclusion of a full range of effects, including biophysical, economic,
and social, as appropriate for the policy, plan, or program

6. Provision for public involvement

7. Public reporting of the SEA and decisions

8. Explicit inclusion of environmental factors in policy-making

9. Tiering to bring SEA to the level of project EIA where possible

10. Monitoring and follow-up of measures

11. Independent oversight of process implementation, agency

compliance, and government-wide performance

Sadler (1996) suggests the following methods for SEA:

1. Literature search-to determine the state of knowledge and the

probable linkages between policy actions and environmental effects.
State-ofthe-environment reports and environmental policy plans may
be helpful adjuncts.

2. Expert judgment-using Delphi techniques, workshops, or other

smallgroup techniques (see Chapter 4).

3. Analytical techniques-including scenario development, checklists and

matrices, GIS, benefit-cost analysis, multicriteria analysis, or indicator
variables to determine probable areas of impact.

4. Consultative tools-such as interviews, workshops, key informant

interviews, sectoral discussions, multi stakeholder roundtables, and
similar techniques.

10.7 MONITORING AND FOLLOW-UP

No matter how careful or comprehensive the EA, it can only be guesswork

until the project is actually in place. Postproject monitoring and follow-up
is an essential component of the overall EA, both to determine compliance
with legal requirements imposed as a result of the EA and to identify any
new or unforeseen effects. Postproject monitoring is also important to
determine the effectiveness of mitigation measures that may have been
simulated but not actually tested in the field.

Ideally, EA is an iterative process through which impacts are assessed,

and later reassessed when water management changes are again
contemplated. Postproject EA is essential to this iteration. Without it, an
EIS becomes a costly one-time effort, not the basis for ongoing guidance
for future planning initiatives.

Several types of monitoring may be required:

1. Compliance with legal obligations, permits, policies, and similar

requirements

2. Monitoring of rare or endangered species habitat, sensitive or highly

valued habitat (e.g., wetlands, estuaries), wilderness areas, historical
sites, and sites of particular spiritual value, to determine whether
impacts are occurring at or in excess of anticipated levels

3. Monitoring of impacts that were and continue to he highly

controversial, to ensure that adequate mitigation is in place and to
provide additional data to inform the scientific or public debate

4. Monitoring to confirm the effectiveness of mitigation measures

5. Remapping of altered environmental conditions following project

implementation; i.e., as the basis for a future EA or to inform the
design of future monitoring or mitigation programs

Normally, requirements for monitoring and follow-up are specified at

the conclusion of the EA process and are designed around the particular
issues, controversies, and biophysical, social, and economic environments
affected by the project. Where issues have remained unresolved, or where
particularly sensitive species, habitats, or human environments are
affected, monitoring and follow-up programs may be especially detailed in
space, time, and range of indicators. A typical follow-up program includes
several components:

• Surveillance and inspection to ensure that EA terms and conditions are

complied with

• Routine, repetitive monitoring of the affected environment to measure

changes in selected indicator variables

• Compliance monitoring, to ensure that legal requirements (permits and

approvals) are met

• Response to unacceptably altered environmental conditions

• Third-party audit of surveillance and monitoring results

In a sense, then, post-EA monitoring and follow-up provides a quality

control function for the EA process and generates an ongoing source of
information that provides both the basis for response and mitigation and a
foundation for future planning initiatives.

10.8 THE SHORTCOMINGS OF ENVIRONMENTAL

ASSESSMENT

A Lack of Long-Term Data One of the most serious weaknesses of EA is

that it often relies on small data sets. For many assessments, intensive data
collection begins only when the EA process is initiated, ending when the
assessment ends. This means that, at most, one to two years of data are
used as the foundation for assessing "impacts" on systems that may be
highly complex. This weakness is particularly significant in terms of water
resources impacts, which may be seriously underestimated if long-term
flow and quality patterns are not well understood.

Failure to Capture Long-Term Environmental Effects The two weaknesses

of limited databases and simple assessment tools mean that EA cannot
accurately predict long-term environmental effects, even though the
general direction and likelihood of effects may he known. The difficulty
arises in determining whether long-term effects are likely to be
"significant" or "negligible." Indeed, it is this point that often forms the
focus of public opposition to proposed public activities: an agency charged
with responsibility for EA has decided that long-term effects will be
minimal and the project should go ahead, while members of the
community believe that long-term effects will he serious and irreversible,
so the project should be rejected. EA methodology does not currently
provide an objective way to prove one outcome over the other.

Failure to Capture Subtle Environmental Effects Environmental

assessment is, necessarily, a rather crude tool, typically based on a modest
data-collection exercise (because of resource limitations) and employing
largely guesswork about the future impacts of proposed actions. The
further into the future these impacts are assessed, the less accurate the
predictions become.

Even gross effects, such as alterations in sedimentation regime, may he

difficult to predict with accuracy. Subtle effects, such as the altered flavor
of a traditional meat, may he impossible to capture with typical
environmental assessment methods such as overlays and matrices.

Failure to Identify and Incorporate Indirect Effects The vast complexity of

natural ecosystems cannot be understood in a year or two of data collection
and analysis. Consequently, it is simply impossible for the analyst to
understand and predict all possible consequences of a proposed action.
Direct effects-for instance, those related to excavation, blasting, heavy
vehicle traffic, infilling, inundation, and similar activities-may he
understood in a general sense. But indirect effects-for instance, the impacts
on human health of long-term exposure to dust and noise-may be difficult,
if not impossible, to anticipate, even though they may eventually he
devastating in the community.

An Emphasis on Economic Development, Sometimes at the Expense of

Ecological Integrity As Tollan (1993) points out, an EIA seldom questions
economic development as such or considers the long-term goals for
healthy ecosystems. This has led to a widespread emphasis on "mitigation"
of impacts and, thus, a presumption that impacts are necessary if economic
development is to occur. Tollan emphasizes the need to use the
"precautionary principle" in EA. The precautionary principle appears as
Principle 15 of the Rio Declaration on Environment and Development,
signed in Rio de Janeiro in June, 1992, at the United Nations Conference
on Environment and Development:

The precautionary principle stales that where there are threats of serious
or irreversible damage, lack of full scientific certainty should not be
used as a reason for postponing measures to prevent environmental
degradation. The precautionary principle is seen as a basis for
sustainable development.

Too often, EA processes instead rely on postproject monitoring to detect

problems. and on reactive measures to correct them. The precautionary
principle would instead mean that where serious or irreversible damage
may occur, the actions likely to cause that damage should not proceed. In
other words, lack of scientific certainty about the outcome of an action
should not he used as a reason to go ahead with the action.

Failure to Incorporate "Traditional Ecological Knowledge" As suggested in

earlier sections of this chapter, environmental assessment is founded in
values. The decision as to when a "valued resource" has been "impacted"
depends very much on the perspective of the analyst. As many authors
(e.g., Goodman and Edwards 1992: Sadler 1996) have pointed out, modern
environmental assessment methodologies are rooted in the "Western-
knowledge" tradition developed in cultures that are economically
advanced. Typically, this means establishing a hypothesis ("no impact will
occur"), designing a study to test that hypothesis, collecting data in a
systematic fashion, drawing conclusions from the results of this study, and
reevaluating the original hypothesis in light of those conclusions.

This scientific-method tradition is alien to many cultures. Many

cultures, for example, rely on an oral tradition through which knowledge
of the envi ronment is passed from elders to younger members of the
society. "Impacts" as judged in such a society may be tangible and
significant, but different from those assessed by a scientist from the
Western-knowledge tradition. An example of such an impact is altered
taste of caribou flesh following low-level flying in Labrador, Canada
(Larry Innes, Environmental Coordinator, Innu Nation, She- shatshiu,
Labrador, May 15, 1997; personal communication). Innu elders report
subtle changes in the flavor of this meat, which is central to the Innu
culture (as discussed in Section 10.2, the Innu are primarily a hunter-
gatherer society, heavily reliant on caribou). This subtle change would not
likely be apparent to a Western scientist even if attention were drawn to it,
yet it is a serious impact as judged by the Innu.

Such a failure to incorporate "traditional ecological knowledge," as it is

known, is an acknowledged weakness in current environmental assessment
methodologies. Traditional environmental knowledge bases frequently
incorporate many decades, or even centuries, of observation of
environmental phenomena, often passed down through story telling or
other oral traditions. They are thus immensely valuable in extending the
short-term "scientific" observations that may be made during the data-
collection phase of a typical environmental assessment process. Additional
research and guidance from aboriginal communities is needed to improve
the ways that this rich knowledge base can complement Western-
knowledge approaches.

Sadler (1996) documents the failure of contemporary EA methodology

to predict impact on the environment. In his study, which was
commissioned by the Canadian Environmental Assessment Agency and the
International Association for Impact Assessment, Sadler conducted a
survey of EA practitioners throughout the world, concluding that:

Current [environmental impact assessment] practice is unsuccessful or

only marginally successful in making verifiable predictions, in
specifying the significance of residual impacts, and in providing advice
to decision makers on alternatives.

Whatever its weaknesses, however, EA forces us to examine alternatives

and systematically review their advantages and disadvantages. This is a
fundamental change from the project evaluation methodologies in use
three decades ago and goes a long way toward fulfilling the goal stated by
Congress when it passed the National Environmental Policy Act in 1969.

REFERENCES

Baber, W. 1988. Impact assessment and democratic politics. Impact

Assessment Bulletin 6:172-178.

Canadian Environmental Assessment Agency. 1995. A Guide to the

Canadian Environmental Assessment Act. Ottawa: Queen's Printer.

Clement, Daniel. 1990. L'ethnohotanique niontagnai.se de Mingarr. Report

53 (Collection Nordicana). Quebec: Centre d'Etudes Nordiques
Universite Laval.

. 1995. La zoologie des Montagnais. Paris: Editions Peeters.

Dunning, C. M. 1982. Content analysis. In Social Impact Assessments

Training Manual. Fort Bclvoir, Va.: U.S. Army Corps of Engineers,
Institute for Water Resources.

Flynn, Cynthia B.. and Rosemary T. Schmidt. 1977. Sources of

Information for Social Profiling. Fort Bclvoir, Va.: U.S. Army Corps of
Engineers, Institute for Water Resources.

Freeman. Milton, R., and Ludwig W. Carbyn, eds. 1988. Traditional

Knowledge and Renewable Resource Management in Northern Regions.
A joint publication of the IUCN Commission on Ecology and the
Canadian Circumpolar Institute (formerly the Boreal Institute for
Northern Studies). Occasional Paper No. 23. Alberta: University of
Alberta.

Gibson, Robert, and Beth Savan. 1987. Environmental Assessment in

Ontario. Toronto: Canadian Institute for Environmental Law and Policy.

Goodman, Alvin S. 1984. Principles of' Water Resources Planning.

Englewood Cliffs, N.J.: Prentice-Hall.
Goodman, A. S., and K. A. Edwards. 1992. Integrated water resources
planning. Natural Resources Forum 16(1):65-70.

Great Lakes Commission. 1996. An Agricultural Profile of the Great Lakes

Basin: Characteristics and Trends in Production, Land-Use, and
Errironnrertal Impacts. Ann Arbor, Mich.: Great Lakes Commission.

Henderson, James E. 1982. Handbook of Em'ironrnental Quality'

Measurement curd Assessment Methods and Techniques. Publication IR-
E-82-2. Vicksburg, Miss.: Environmental Laboratory, U.S. Army
Engineer Waterways Experiment Station.

Hilden, M. 1996. Evaluation of the Significance of Environmental Impacts:

F.IA Process Strengthening. Canberra: Environment Protection Agency.

Inglis, Julian T., ed. 1993. Traditional Ecological Knowledge: Concepts

and Cases. International Program on Traditional Ecological Knowledge
and International Development Research Council. Ottawa: Canadian
Museum of Nature.

Jacques Whitford Environment Ltd. 1992. Environmental Assessment

Report for the Hali/ox-Dartnrouth Metropolitan Sewage 1 reatment
Facility. Vol. 11, Impacts and Mitigation. Consultant report for Halifax
Harbour Cleanup Inc. Dartmouth. N.S.: Jacques Whitford Environmental
Ltd.

Jain, R. K., and L. V. Urban, 1975. A Review and Analysis of

Envirmimental Impact Assessment Methodologies, Technical Report E-
69. Champaign, Ill.: Construction Engineering Research Laboratory.

Leopold, Luna B., Frank E. Clarke, and Bruce B. Hanshaw. 1971. A

Procedure for Evaluating Enrironniental Impact. U.S. Geological Survey
Circular 645. Washington, D.C.: US Geological Survey.

Maslow, Abraham E1. 1954. Motivation and PersonalitY. New York:

Harper.
Sadler, Barry. 1996. Environmental Assessment in ct Changing World:
Evaluating Practice to Improve Performance: Final Report. International
Study of the Effectiveness of Environmental Assessment. Ottawa:
Canadian Environmental Assessment Agency and the International
Association for Impact Assessment.

Tollan, Anne. 1993. The ecosystem approach to water management. World

Meteorological Organization Bulletin: 28-34.

U.S. Council on Environmental Quality. 1981. NEPA's Fort" Most Asked

Questions. Memorandum to Agencies. Washington, D.C.: U.S. Council
on Environmental Quality, Executive Office of the President.

. 1993. Environmental Quality. Twenty-third Annual Report. Washington,

D.C.: CEQ.

Vanclay, F., and D. Bronstein, eds. 1995. Environment and Social Impact
Assessments. London: John Wiley & Sons.

World Business Council on Sustainable Development. 1995.

Environmental Assessment: A Business Perspective. Geneva: World
Business Council.on Sustainable Development.
11
 Choosing the Best Plan
Among the many decisions that must be made in watershed management,
the choice of the "best" management approach is in many ways probably
the most difficult and controversial. Chapter 10 described the differences
in values that may underlie discussion about water management strategies,
and these differences are not easy to resolve. This chapter examines
approaches to choosing a "best" plan from the alternatives that were
developed (Chapter 5) and tested (Chapters 6 and 7).

11.1 EVALUATING THE EFFECTIVENESS OF OPTIONS

AND STRATEGIES

Chapters 3 and 5 described the need to decide what conditions are desired
in the watershed, and to select indicators of progress toward those goals.
Often, but not always, there is widespread agreement on these goals:
everybody wants the beaches "swimmable" again; everybody wants to be
able to catch fish in the river. It is at the stage of choosing the best
approach to meeting these goals that the need for good indicators becomes
most apparent. What does "everybody" mean by "swimmable," for
instance? Values may underlie even the most basic assumptions about the
plan.

Possibly as difficult is the question of power in the decision-making

process, which includes sensitive issues such as who calls and chairs
meetings, which viewpoints are allowed to he heard, and which are
summarily dismissed. Sometimes, regulatory agencies have the legal
responsibility to make decisions about water management planning (for
example, for navigational put-poses) and may feel that they are weakening
their mandate by allowing different views into the discussion. As Arnstein
(1969) has pointed out, allowing more citizen involvement in the planning
process amounts to delegation of decision-making power. Not all
regulatory agencies are comfortable with this delegation. Yet successful
implementation of the watershed management strategy (Chapter 12) may
depend on the social consensus that underlies policy.

As discussed in Chapter 3, it is therefore essential that the wider

community be involved in decisions about desired outcomes and the
indicators used to track them. The International Joint Commission (IJC)
has recently devoted several years to the social and technical problem of
defining desired outcomes and indicators for the Great Lakes (e.g., IJC
1994a, 1994b) and continues to conduct research on how a system of
indicators can best be implemented.

Assuming that the desired condition of the watershed is known or

guessed, that some set of specific targets to reflect successful attainment of
those conditions has been developed, and that a number of management
options have been identified, the following steps can be used to reduce the
number of possible management options to a workable group. (Refer to
Chapters 5, 6, and 7 for more detailed discussion of the terms used in the
following section.)

Step 1: Apply Planning Constraints Chapter 5 discussed the general

problem of establishing planning constraints, and Chapters 6 and 7,
respectively, provided advice on application of constraints in simple and
detailed planning initiatives. The constraints essentially constitute a set of
limits, preferably supported by the community, beyond which the plan may
not go.

Application of constraints is usually a straightforward matter requiring

little detailed analysis. Table 5.6 illustrates a simple example of such an
analysis. Typically, application of constraints involves ruling out options
that:

• Are inappropriate for the physical conditions (soils, climate,

infrastructure, development, etc.) of the site

• Are too costly

• Take too long to implement

 • Do not meet basic performance requirements (for instance, for
maintenance of desired flow or water levels or for removal of specific
water pollutants)

As important as the actual analysis of options is communication with,

and involvement of, the community in decision making. If a government
agency is undertaking the analysis without active participation by the
public, all decisions should be clearly documented and explained in a
report that is available for public and agency scrutiny and comment.
Sufficient time must be allowed for public and agency review and debate.
(If this requirement seems onerous, decision makers should recall that the
development of a watershed management strategy is, as Bishop (1970)
suggests, a process for creating social change, which must logically be
endorsed by the affected society.) Public and agency commentary can
provide a rich source of feedback and can identify issues or options
overlooked by the decision makers, who may wish to revisit and revise the
decisions made and issue an amended document to record these revisions.

Step 2: Develop a Framework for Application of Decision Criteria The

application of planning constraints should have reduced a large and
confusing list of management options to a shorter list of options that are
clearly feasible. The remaining analysis is directed at determining which of
these best suits the needs of the community, as reflected in the decision
criteria developed by the planning team.

In most cases, decision makers are attempting to make decisions that

will continue to be effective for many years into the future. To do so, they
must decide uhal future they want to plan for and how many intermediate
"futures" should also be assessed. As discussed in Chapters 5, 6, and 7, this
requires decisions about:

• An appropriate planning horizon

• An appropriate discount rate unless risk and uncertainty will he

incorporated in the analysis (see Chapter 8)
 • Any specific planning assumptions to be used in the analysis (for
example, that "best management practices" will be assumed to have a
certain level of acceptance in the community)

There is no single correct approach to these decisions. A longer planning

horizon is useful for major projects and structures, but more distant futures
are less certain in terms of population forecasts, prevailing interest rates,
demand for goods and services, and other key planning factors. Many
detailed watershed planning initiatives therefore incorporate several
planning horizons, including a short-term future (2 to 5 years, highly
predictable), a midrange future (5 to 10 years, less certain but still
foreseeable), and a long-term future (25 to 50 years, very uncertain but
desirable in terms of the expected lives of major structures). Decision
makers must therefore:

• Decide whether performance will be evaluated under one or several

future development conditions

• Define the characteristics of present and future development

conditions, including land use planning forecasts, population densities,
and location of major watershed features and structures

• Identify any other planning assumptions that will he made in

evaluating these conditions-for example, public attitudes toward
resource conservation or economic development

Certainly, future conditions can be chosen more or less arbitrarily: 5

years, 10 years, and 25 years, for example, with an average discount rate of
I2% for all. But planners may also wish to incorporate uncertainty in these
analyses. For this reason, some analyses include future scenarios with low-
medium-high, or best-typical-worst, projections. Table 11.I illustrates how
establishing these scenarios builds the foundation for the evaluation of
management alternatives.

What values should be used to fill in the cells of this matrix'? A single
value would rarely suffice to represent the performance of management
strategy across a range of criteria. Several approaches are possible:
 1. Prepare a matrix like that in Table 11. I for each decision criterion.

2. Prepare a matrix like that in Table 11.1, but use several columns
under each "Plan" heading to display performance on various decision
criteria.

3. Compile an "index" by summing performance on all criteria, using

weights if appropriate.

Step 3: Develop a List of Management Scenarios to Be Modeled Although

the general future cases (Present case, 5 years into the future, etc.) are
easily developed, these are not sufficient for detailed testing of scenarios.
The goal of the analysis is to determine the performance of each individual
management option, and combination of options ("plan" or "strategy"), on
each decision criterion. And each of these cases must be examined under
each of the planning horizons shown in Table 1 1.1. Clearly, scenario
testing can rapidly evolve into a complicated and time-consuming process.
Thus, it is essential that it be carried out in a systematic fashion, with
careful recording of each step of the analysis (for instance, each computer
simulation run).

The analyst must therefore develop a comprehensive list of scenarios

that will be tested under each future condition. These should include:

• "Do nothing" cases for the present and for any separate future
scenarios.
 • Scenarios that examine the impact of a single management option
applied to the greatest possible effect ("best case" for that option). For
example, the analyst could assume 100% adoption of conservation
tillage in cropland within the watershed.

• Scenarios that examine the impact of a single option applied to a

portion of the basin (e.g., a portion of the total stream length: or 50%
of farms; or some similar scenario). This type of scenario might be
used to model various levels of stakeholder (e.(,., farmer) willingness
to implement a given management practice.

Step 4: Test the Performance of Each Individual Management Option One

way to reduce the complexity of scenario testing is to review the
performance of each individual management option first, and then include
only very effective (and very inexpensive but moderately effective)
measures in compiling plans or strategies. It is usually the case that one
option will perform differently on different criteria-in other words, will not
he "best" on all factors. This problem is illustrated in Table 5.8 and
discussed in Section 5.5.3. The decision as to whether to include an option
in subsequent plan development will depend on whether the option
performs well on criteria deemed to be important (heavily weighted) in the
analysis, on cost and availability of funding, and on other factors such as
ease of implementation and degree of public acceptance. The analyst
should eliminate from further consideration those options that

• Do not perform well on important criteria and

• Are expensive or difficult to fund, or

• Are not well supported by the community.

Testing should include both present conditions and the various future
conditions of interest (see the example in Table 1 1.1 ). We could, for
example, extend Table 5.8 to illustrate this concept, as shown in Table
11.2.

If many management options and future conditions are under

consideration, the number of scenarios to he tested can be very large and
there will be a temptation to cut corners and skip to the "obvious"
conclusion. There are several reasons to guard against this tendency:

1. Each decision in the option-screening process must be clearly

documented and transparent in order to ensure full accountability to
the public and public agencies.

2. Cutting corners may lead the analyst to ignore less effective but low-
cost and easily implemented options which may. nevertheless, have
value in the development of management plans.

3. Lack of a systematic approach to option testing can result in whole

categories of options being overlooked altogether.

Decisions to keep or eliminate individual options should be ratified with

regulatory agencies and the community and should he fully documented
and explained in the final planning report.
Step 5: Conduct Final Testing Having decided which individual options are
worth including in a management strategy and which can safely be
eliminated from further consideration, the analyst is faced with the
challenge of developing credible management strategies. To a large extent
this is a subjective process, based on the interests of decision makers and
the community and on considerations of plan implementation. Individual
options can, for example, be grouped into strategies that focus on a
particular pollution source, a particular beneficial use, or some similar
rationale. For example, all agricultural tillage and crop management
measures can be grouped into one scenario, incorporating the most likely
levels of farmer acceptance in the watershed, and their effectiveness
examined under present conditions, 5 years into the future, and 25 years
into the future. The far-distant future might be examined as a worst-case
estimate (e.g., maximum population growth, minimum adoption of
measures by farmers) or some other scenario viewed as likely by the
decision makers. Similarly, all urban stormwater management options can
be grouped into a second scenario that is tested under present and future
conditions, and so on. The advantage of testing like options together is to
evaluate the maximum effect possible with agricultural or urban controls,
respectively. With the results of this testing, it is possible to estimate, for
example, the maximum reduction possible in a given pollutant if a suite of
agricultural measures were implemented. This in turn allows recalculation
of the relative loadings of pollutants to a receiving water under present and
future management scenarios.

When this information has been obtained, "combination" plans can be

compiled, incorporating a range of measures across the watershed. Here
the choice is mainly subjective and will be strengthened by frequent and
careful review by agency and community representatives. (The AEAM
process described in Section 3.4.2 in fact uses scenario development and
testing as the focus of a series of workshops. Copies of the results of
computer simulation are not retained, but instead repetitive scenario testing
forms the framework for a broad dialogue on issues and solutions.
Although this book generally advocates a more structured approach to
problem solving, the AEAM process provides many useful lessons about
the process of community-agency interaction in the development and
testing of management strategies.)

In the author's experience, cost can be a useful framework for plan

development. Several alternative management strategies can be set up,
reflecting different levels of expenditure: a minimum-cost plan, a
moderate-cost plan, and a "bells and whistles" plan, for example. This
approach also has value in providing a range of options for the
consideration of funding agencies, because it clarifies the relationship
between level of expenditure and resulting environmental improvement.

It is particularly important at this final stage of plan development and

screening that the public and interested agencies have an opportunity to
comment on the strategies being tested (including recommending different
strategies for evaluation) and any assumptions about present and future
conditions. The analyst should not regard this involvement as burdensome
or interfering, but rather as an opportunity to build community support for,
and consensus on, effective management strategies. This is particularly
important when the strategies under consideration include those with
potential to displace communities or individuals (for instance, major dam
projects) but which may have benefits in proportion to or greater than their
impacts. Meaningful inclusion of the public through the plan-testing
process can reassure all participants about the costs and benefits of each
strategy under consideration. While such inclusion may not build universal
support for a contentious plan, it may reduce conflict and improve the
success of implementation measures. In part, this is because a good public
involvement program allows disclosure and mapping of the values that
underlie conflict so that alternatives can be developed or modified to
accommodate a range of values.

Step 6: Document the Results of Plan Testing and Recommend One or

More Plans At the conclusion of plan testing, it is recommended that all
results be documented and the rationale for each stage of option screening
and plan development be fully described. Typically, three or four plans will
emerge as better than others. The difficulty, as is often the case in the
testing of individual options, is that one plan may be "best" in terms of
cost, while another is "best" in terms of pollutant reduction and a third is
"best" in the eyes of agency officials because of its administrative
simplicity.

The decision as to which of these plans is implemented is, ultimately, a

political one. Water managers can facilitate the decision-making process
by:

1. Presenting several effective strategies for the consideration of

decision makers

2. Documenting the performance of each plan in terms of the decision

criteria established at the beginning of the planning process

3. Reminding decision makers of the weights they have assigned to the

various decision criteria

4. Providing supporting documentation regarding implementation

considerations for each plan

The Stage 2 Remedial Action Plan report for the Metropolitan Toronto
RAP area (Ontario Ministry of the Environment and Energy 1993)
presented a long list of recommended management actions, each of which
is assigned a priority, an estimated cost, and a recommended
implementation period. The document served as the basis for negotiations
with local and regional municipalities, the provincial government, and the
Canadian federal government as to which agency should or would in fact
fund each measure. Figure 11.1 illustrates a portion of the Toronto RAP
"actions" matrix.

There is no question that funding availability and the presence of an

agency "sponsor" for a measure were significant factors in defining the
final characteristics of the management plan for the Toronto area
watersheds. A major consideration in this debate was the implementability
of different management approaches: how well each plan would be
accepted by the public, how likely each was to succeed over the long term,
and similar considerations. Section 11.2 discusses some of these
considerations. Chapter 12 describes the components of a successful
implementation plan in more detail.

Figure 11.1 Portion of the Metropolitan Toronto Stage 2 Remedial Action

Plan matrix of proposed actions, responsibilities, and timetables for storm
water management.
11.2 IMPLEMENTATION CONSIDERATIONS

Although the cost and effectiveness of a management strategy is of central

interest in integrated watershed management, its "implementability" is also
an essential consideration. In part, this is because a plan that is easily
implemented is usually welcomed by the public and results in support for
the politicians and agencies who have recommended it. In other words, a
strong, popular plan can translate directly into votes. On the other hand, a
plan that is highly effective but onerous to some portion of the population,
very costly, or otherwise disruptive of individuals and communities will
encounter resistance from the public. Resistance means delays, added
costs, public acrimony, and loss of political support.

Ideally, early and meaningful public involvement will have identified

issues of major public concern, will have resulted in decision criteria that
reflect the reasons for those concerns, and will ultimately have prevented
contentious plans from reaching the final phases of screening. In practice,
however, contentious plans are sometimes retained in these final stages,
either because they are highly effective or because of lack of sensitivity on
the part of the analysts. This is one of the reasons to develop several
preferred management strategies for presentation to decision makers.

There may also be obstacles to implementation that are not apparent

either to the analysts or to the wider public. These may include, for
example, statutory prohibitions on certain types of actions. While it is to be
hoped that these prohibitions will be known early in the planning process,
such is not always the case. A legal opinion on the viability of a given
strategy may have to be delayed until the precise details of the plan are
known. In Ontario's MunicipalIndustrial Strategy for Abatement (Ontario's
program to develop Clean Water Act-type regulations for effluent
discharges), for example, it was believed that "permit-by-rule" was a
viable approach for managing certain point-source dischargers. Under a
permit-by-rule system, the law requires only that a specified type of
technology be in place and correctly operated at a facility; no effluent
standards are specified. In final legal review, however, it emerged that
legislative counsel believed there to be insufficient precedent for this
approach, and it was discarded in favor of a more traditional standards-
based approach.

Where obstacles to implementation have been identified, it is, however,

worth exploring with interested agencies and the community whether there
is any way to surmount them-for instance, through financial compensation
or additional structures or technologies. Some obstacles will remain no
matter what mitigation is proposed; for instance, landowners are often
strongly opposed to dam projects because of the potential for displacement
of farms and homes. Spiritual and cultural values may create major
obstacles that are not easily overcome, even if the affected group is a small
minority in the community. But some obstacles (e.g., property value
concerns) may be manageable if values are clearly mapped and conflict
resolution processes are employed to develop outcomes agreeable to all
parties.

In general, it makes sense to eliminate from detailed consideration any

individual options or management strategies, however effective, that will
present significant obstacles to implementation. The underlying logic here
is that it is probably better to have a less effective plan that is fully
implemented, rather than a highly effective plan that remains a paper
exercise.

11.3 CASE STUDY: WATERSHED MANAGEMENT FOR

GANDER LAKE, NEWFOUNDLAND

Gander Lake, Newfoundland (Canada), is a large freshwater lake

approximately 50 km long, 2 km wide, and up to 300 m deep. Cantwell
and Pinhey (1997) describe the development of a watershed management
plan for the lake.

Gander Lake supplies drinking water for the town of Gander and several
other communities. The rivers that supply the lake support a thriving
salmon fishery, said by Cantwell and Pinhey to be among the best in North
America. The area's natural beauty has recently placed it under pressure
from cottage development, while forest-product companies seek to exploit
the timber resources of the watershed.

In the early 1990s, the Newfoundland Department of the Environment

began a water management planning initiative for Gander Lake, targeted at
protection of drinking water resources. The responsible agency was the
Department's water resources management division. The agency intended
that this plan should not only provide for the protection of Gander Lake's
water resources, but also serve as a model for other water management
plans elsewhere in the province.

The process utilized a steering committee with representation from

various provincial government departments, Environment Canada (the
responsible federal agency), and the town of Gander. Public
representatives were not included on the committee. The objectives of the
initiative were to:

1. Identify the primary sources of pollutants to the lake

2. Develop a zoning plan to direct future development (especially

forestry operations and cottage development)

The province placed special emphasis on scientific rigor throughout the

planning process and rejected the notion of generalized planning concepts
such as a uniform 50 in buffer zone all around the lake. Each decision
made by the committee had to be supported by clear and specific evidence.

Gander and its neighboring towns are all relatively new, and their
develop ment is well recorded. Good records were also available for water
quality in Gander Lake and from major point sources of pollution,
including sewage treatment plants and industrial discharges. A review of
these data soon revealed that point sources of pollution did not have a
major influence on the water quality of the lake itself. Instead, nonpoint-
source pollution was found to be the primary source of water pollution.

Review of available data also showed that although Gander Lake is

currently oligotrophic (low in nutrient loadings and therefore not subject to
nuisance algae blooms), it is sensitive to phosphorus loadings. Phosphorus
thus became the focus of subsequent water management planning.

Major sources of phosphorus to the lake include direct precipitation and

delivery via streams and overland runoff. Much of the phosphorus
delivered to the lake was found to he attached to soil particles eroded from
the land and carried by streams and runoff. Thus, erosion was also found to
he an important mechanism in affecting phosphorus concentrations in the
lake. Water managers compiling the plan believed that forest clearance
would be the major source of erosion, and therefore phosphorus, to the
lake.

The analysts for the plan (the consulting company EDM Environmental
Design and Management Ltd., of Halifax, Nova Scotia) chose to employ a
relatively simple computer-assisted watershed analysis, as described in
Chapters 6 and 7 of this hook. They employed a GIS-based predictive
model based on the Universal Soil Loss Equation (LISLE) using a I ha grid
size. Digital maps of land use, development type, soil type, slope, and
other factors were prepared, and a simple USLE-based predictive model
("watershed decision model") was calibrated to existing conditions.

A range of future management scenarios were then developed and tested

using this model, by altering the cartographic models of' suspended solids
and phosphorus input to the lake under different management conditions.
The model produced estimates of total annual phosphorus and suspended
solids loads to the lake, both for existing conditions and for the various
management strategies under consideration.

The results of scenario testing showed that planned clear-cuts in the

forested portions of the watershed and additional urban development
would actually have less impact than was originally expected. By contrast,
cottage development and expansion of the Trans-Canada Highway (a
major traffic route through the area) would he likely to have a greater
impact on water quality than was originally foreseen. In part, this was
because clear-cuts were proposed for areas of the watershed that were
relatively resistant to erosion, while highway expansion activities would
occur in more sensitive areas. The analysts explained: "In this catchment,
location tended to be more critical in terms of lake water impact than
actual land use."

The results of scenario testing were used to prepare a zoning plan to

restrict land use within the watershed. The plan illustrates locations where
development would create a negative impact on water quality. About 7%
of the land in the watershed is categorized as "restricted to development"
because of its sensitivity to erosion and likelihood of affecting water
quality in the lake. The planners note that the finding that a majority of
lake water sediment derives from only a small land area in the catchment is
consistent with findings from other watersheds (cf. Wall et al. 1978). The
results also allowed the identification of allowable activities in the
watershed, based on the potential of those activities to disturb the soil and
create erosion. Controls on the density of development were negotiated
with proponents. These included limits on construction activities,
stormwater management practices, and cottage development guidelines.

The Gander Lake planning initiative was relatively inexpensive but

yielded important insight into watershed management practices for the
community. Although the I ha GIS grid is relatively coarse, the model
provides the basis for more detailed planning at the local level and has
created a basis for dialogue with the community and with industry
regarding future land use planning and water quality protection.

The output of the process was, first and foremost, a map, showing land
use zones and erosion-sensitive areas. The snap is easily understood by
both lay people and technical experts and is an effective means of
communicating the results of complex analysis.

The watershed decision model has now been accepted by the provincial
government as an appropriate tool for decision making in other
watersheds. Cantwell and Pinhey (1997) observe that integrating
appropriate land use planning techniques and an understanding of
landscape ecology will "allow planners to produce `living plans' that
respond to development demands without compromising those parts of our
environment that cannot be compromised."
REFERENCES

Arnstein, Sherry. 1969. A ladder of citizen participation. Journal of the

American Institute of Planners 35:216-224.

Bishop, Bruce. 1970. Public Participation in Planning: A Multi-Media

Course. IWR Report 70-7. Fort Belvoir, Va.: U.S. Army Engineers
Institute for Water Resources.

Cantwell, Margot D., and Jeffrey A. Pinhey. 1997. Watershed management

plan for Gander Lake and its catchment. Plan Canada (March): 27-30.

DeGarmo, E. Paul, William G. Sullivan, James A. Bontadelli, and Elin M.

Wicks. 1997. Engineering Economy. 10th ed. Upper Saddle River, N.J.:
Prentice-Hall.

International Joint Commission. I994a. Bioindicators as a Measure of

Success for Virtual Elimination of Persistent Toxic Substances. Windsor,
Ont.: IJC Great Lakes Regional Office.

. I994b. Indicators to Evaluate Progress Under the Great Lakes Water

Quality Agreement. Report of the Indicators for Evaluation Task Force to
the International Joint Commission. Ottawa and Washington, D.C.: IJC.

Lumsdaine, Edward, and Monika Lumsdaine. 1995. Creative Problem

Solving: Thinking Skills ffrr a Changing World. New York: McGraw-
Hill.

Novotny, V., and H. Olem. 1994. Water Quality: Prevention, ldentifzcation,

and Management of Diffuse Pollution. New York: Van Nostrand
Reinhold.

Ontario Ministry of the Environment. 1987. Technical Guidelines for

Preparing a Pollution Control Plan. Report from Urban Drainage Policy
Implementation Committee, Technical Sub-Committee No. 2. Toronto:
Ontario Ministry of the Environment.
Ontario Ministry of the Environment and Energy. 1993. Metropolitan
Toronto Stage II Remedial Action Plan: Clean Waters, Clear Choices.
Toronto: Ministry of the Environment and Energy, Central Region.

Wall, G. J., L..1. P. van Vliet, and W. T. Dickinson. 1978. Contribution

o/'Sediments to the Great Lakes from Agricultural Activities in Ontario.
Windsor, Ont.: International Joint Commission.
12
 Implementing the Plan
Up to this point, this book has been concerned with developing an
understanding of watershed components and processes, developing
societal consensus as to the ideal condition of those watershed elements,
and developing a plan to move the watershed toward the desired condition.
In short, this book has so far been concerned largely with a paper exercise.

What distinguishes the successful watershed management plan from its

paper counterpart is implementation: putting into action the plans and
programs that planners believe will be effective in improving the state of
the watershed ecosystem. Successful implementation implies a societal
consensus that a plan is acceptable as a means of reaching community
goals for a watershed. As Viessman (1990) observes, however:

Water management institutions evolve from needs identified at some

milestone in time. The problem is that times change, and so do needs.
Unfortunately, institutions seem to march on with entrenched
constituencies, and many in existence today are addressing yesterday's
goals or addressing today's problems with yesterday's practices.

Figure 12.1 (reproducing Figure I.1 for convenience) illustrates the

interplay between social, economic, and biophysical forces in a watershed.
It will be seen from this figure that there is a sort of dynamic tension
between the watershed's water resources, its water users, the broader
watershed ecosystem (including soil, air, vegetation, and animals), and the
various social and economic forces at work in the watershed. Changes in
social and economic forces can trigger changes in water management
practices. Similarly, changes in the watershed ecosystem can indicate a
need for altered water management practices. Supplyoriented or demand-
oriented management action affect the horizontal axis of the figure: the
quantity and quality of water in the watershed and the uses made of' that
water.
 The system is dynamic because, as Viessman points out, neither the
natural watershed ecosystem nor the human social and economic systems
of' the watershed are static. Just as watershed planning must respond to
these changes, so must watershed management, including plan
implementation, be dynamic rather than static.

This chapter addresses the special needs of' the implementation stage,
beginning with key principles of water administration and moving through
the elements of a recommended implementation approach. At the end of
the chapter, the administrative structures and implementation success of
several case studies are examined.

12.1 PRINCIPLES OF WATER RESOURCES

ADMINISTRATION

Integrated watershed planning and management has two principal goals

(Goodman and Edwards 1992):
 Figure 12.1 Forces affecting integrated watershed management (after
Koudstaal et al. 1992).

1. To plan programs and projects that are economically efficient and

socially desirable, and

2. To execute projects that will be sustainable over a Tong period of

time beyond the exodus of domestic or foreign financing, technical
assistance, and the repayment of loans.

Young (1992) suggests that these goals must be fulfilled within certain
constraints, namely, that:

1. Environmental quality (ecosystem functioning, soil, water, and air

quality, and landscape amenity) must be maintained or improved. No
 economic activity should unduly disturb the regenerative capacity, the
assimilative capacity, or the productivity of life-support systems.

2. All resources use must be technically and economically efficient, and

resources must not be wasted.

3. Governments may fail. Planners must recognize and plan for this
contingency by creating incentives for environmental improvement
and plans and programs that encourage political and economic
stability.

4. Future options must be maintained. This is the central challenge of

sustainability: that each future generation will have the same capacity
and option to solve its own problems as currently exists in the society.
The creation or substitution of future options is acceptable only when
they are consistent with this notion. In addition, there may be benefit
in giving preference to institutional arrangements that encourage
ecological, social, and economic diversity, which are more capable of
responding quickly, positively, and flexibly to change.

5. Population growth should be controlled, so that future generations

will have equivalent opportunities for solving their own problems and
will have equal access to resources. At least primary education, and
preferably secondary education, are central to this constraint.

6. Nations must conserve and enhance the value of their natural capital
through controls on pricing, harvesting, and regeneration rates.

7. Where additional resources are required, nonrenewable capital

(resources) should be depleted in preference to renewable resources.
The logic behind this is that a sustainable future must include the life-
support systems and ecological functions that underlie renewable
resources. Nonrenewable resource depletion should occur at a rate that
maintains the aggregate value of renewable and nonrenewable
resources.
 8. Wealth should be redistributed to poorer countries, because poorer
countries cannot he expected to manage their resources sustainably
without assistance from wealthier nations. Young suggests that this
could be done by liberalizing trade arrangements to allow global
markets to take greater account of environmental conditions.

Caponera (1985) reviews the work of the Institut de Droit International

(IDI), the International Law Association (ILA), and the International Law
Commission (ILC) of the United Nations, regarding the general principles
of international water law; many of these apply also at the national and
regional levels. The principles Caponera lists are not legally binding and
do not impose any obligations on water management institutions. They are,
however, valuable in determining appropriate implementation
mechanisms. Essentially, the principles reviewed by Caponera are based
on two fundamental propositions, both of which echo principles of English
common law and other unwritten behavioral codes:

1. Common water resources are to be shared equitably between the

states entitled to use them, and

2. States are responsible for substantial transboundary injury originating

in their respective territories.

Where a watershed crosses state or national boundaries, the watershed

unit can be taken as encompassing not only a biophysical system, but a
"community of interests" related to water in the drainage basin. Successful
implementation of watershed management plans must therefore be based
on certain principles, as discussed in the following paragraphs.

12.1.1 The Principle of Reasonable and Equitable Sharing of Resources

According to the Helsinki Rules of the ILA, what is 'reasonable and

equitable" is to be determined in light of all the relevant factors operating
in the case, including but not limited to basin geography, basin hydrology,
climate, past and present water uses, economic and social needs of water
users, the population dependent on the waters of the basin, the comparative
costs of alternative means of satisfying the social and economic needs of
each water user, the availability of other resources, the avoidance of
unnecessary waste in water use, the practicability of compensation as a
means of adjusting conflicts between users, and the degree to which the
needs of one state (user) may be satisfied without causing substantial
injury to a co-basin state (user). Water users (states) are to resolve conflict
through consultation, negotiation, and agreement, all with reference to
specific watershed systems.

12.1.2 The Principle of Limited Sovereignty

The principles of reasonableness and equity obviate a system whereby a

state has full rights to its water resources because of territorial sovereignty.
As Caponera (1985) puts it, "confronted with a factual situation, nothing
prevents the independence of sovereignty from giving way to
interdependence suggested by the principles of reasonableness and equity."
So although a jurisdiction may claim territorial sovereignty over an area, it
may nevertheless be expected to manage its water resources in a fashion
consistent with the needs of downstream or adjacent water users.

12.1.3 The Duty to Cooperate in Development

If a watershed represents a community of interests, and in view of the

principle of reasonableness, it seems appropriate that one jurisdiction be
expected to cooperate in watershed development-for instance, by delaying
new works or water uses until the needs and interests of neighboring water
users can be satisfied. Caponera points out that implementing this principle
is not easy. Parochial interests may prevail over "equitable utilization,"
especially where choices must be made among conflicting water uses and
where water supplies are limited.

12.1.4 The Duty to Cooperate in Protection

According to Caponera, principles of international water law are emerging

as states contend with the protection of transboundary water resources,
especially in regard to pollution of water with toxic chemicals and
waterborne diseases. Several elements are included here:
 • Individual states are expected to have adequate legal and
administrative frameworks for the protection of water within their
jurisdictions.

• States are also expected to keep abreast of changing technologies for

the use, recycling, and purification of waters within their jurisdictions.

• States are expected to prevent water pollution that may harm or

threaten the interests of neighboring states.

Consistent with this last element is the next principle, responsibility for
injury across frontiers.

12.1.5 Principle of Responsibility for Injury Across Frontiers

In general, states are responsible under general international law for acts
and omissions concerning activities within their jurisdictions. In terms of
water management, the relative harmfulness of pollution or water
extractions may depend on the beneficial uses affected by those acts or
omissions. And since interpretation of "impact" on "beneficial" use is
highly context-specific, harm must be evaluated relative to each individual
stream and watershed situation.

Caponera's principles of international water law are equally valid in

terms of watershed cooperation and management and, indeed, underlie
successful watershed management initiatives. Whether national, state,
regional, or local governments are involved, integrated watershed
management implies a reasonable and equitable sharing of the water
resources of the basin among interested users, responsibility for harm
caused, and a duty to protect the shared resources within the bounds of
technical and economic feasibility.

In a sense, then, implementation means getting individuals and

organizations to:

• Agree on a desired plan of action

 • Agree on indicators of' progress toward the shared goal

• Allocate tasks among the various water users in the basin

• Allocate the costs of water management among the various water users
in the basin

• Agree on periodic review of, and revisions to, water management

activities under way in the basin

This sequence of tasks means that individuals and organizations must

set aside their "territorial sovereignty," with its narrow goals and
potentially exploitive actions, in favor of the equitable sharing of water
resources and the responsibility for water resources management. In effect,
it means balancing private interests against public interests. This balancing
act requires agreement among players who may have very different values
and different visions of an ideal watershed condition.

There is no question that the most difficult part of implementation lies

in reaching agreement among the various players. David Crombie, former
mayor of Toronto and chair of the Royal Commission on the Future of the
Toronto Waterfront, once said that the two biggest obstacles to resolving
environmental problems are "turf and ego." There is much to support this
idea in the process of implementing a watershed plan. While everyone
supports the notion of a cleaner watershed and the restoration of beneficial
uses, support may be likely to dwindle in the implementation phase. The
main problem is usually money, although jurisdiction ("turf") can also get
in the way. The difficulty is that most watershed restoration actions are
costly. The Remedial Action Plan for the Metropolitan Toronto Watershed
will cost in excess of $3 billion (Canadian), including capital and operating
costs, if it is fully implemented (planned implementation will extend over
at least 20 years). Individual agency shares of this total amount to tens of
millions of dollars-a hefty price at a time when governments are trying to
cut costs and shrink budgets. To avoid paying more than they need to,
agencies may try to argue that another level of government has
responsibility for the necessary action. At the same time, however, both
agencies and politicians are eager to reap the political benefits of
successful remediation ("ego"), so there may also be a struggle for political
profile in the implementation process. Teclaff and Teclaff ( 1987) affirm
this perspective:

In 1111ulti-state basins, problems are bound to arise over the allocation

of expenses among member governments. Unlike the costs of
development projects. which are often apportioned among states
according to benefits ultimately received, the expenses of pollution
research, monitoring, and surveillance are rarely so equitably divided.
European countries have made much of the so-called "polluterpays"
principle as applied to individual polluters and polluting industries.
They seem less willing to apply it as among states.

12.2 PLANNING FOR SUCCESSFUL IMPLEMENTATION

Chapter I I discussed the development and testing of alternative

management strategies, but recommended that no single plan be put
forward as the preferred approach. Instead, it was recommended that two
or three good strategies be offered to decision makers for consideration.
The Royal Commission on the Future of the Toronto Waterfront (1992)
observed that the decision as to which management approach is preferred
usually lies with an elected body, such as a regional or local municipal
government, or a state or provincial Cabinet. Because few plans will
satisfy all interests in the community, tradeoffs must be made in selecting a
final plan. Ideally, these decisions should be made by elected officials who
represent the larger society. Endorsement by elected officials often carries
the authority necessary to expend resources of time and money and to
actually "get things done." If a plan is recommended and approved only by
a bureaucratic (or technocratic) group, but endorsement-and expenditure-
by elected officials does not follow, the plan will likely remain on the
shelf. It requires political will, encouraged by well-developed public
consensus, to ensure that the plan is put into action.

Successful implementation requires several elements that have not

uniformly been part of previous stages in the planning process:
 1. A single lead agency to act as an advocate and facilitator for the plan
with the community and with political representatives

2. Strong linkages to existing programs, including local and regional

land use planning processes, water quality and flow monitoring
programs, and similar programs, to optimize use of available
information and minimize duplication of' effort

3. Clear designation of responsibilities, timetables, and anticipated costs

for project actions

4. Effective laws, regulations, and policies to provide a framework for

the tasks identified in item 3

5. Ongoing tracking of the degree of implementation of management

actions and of the success of those actions once implemented

6. Ongoing monitoring and reporting of progress, both to assess the

effectiveness of individual actions and to sustain public and political
interest in and enthusiasm for the plan

7. Ongoing public education and communication programs to

consolidate and enhance the social consensus achieved in the planning
process

8. Periodic review and revision of the plan

9. Adequate funding for these activities

The following sections examine each of these requirements in more

detail.

12.2.1 A Single Lead Agency

Studies carried out in numerous watersheds (cf. Teclaff and Teclaff 1987)
have demonstrated the importance of a simplified administrative structure
for the plan. In particular, plan implementation should not require the
creation of additional layers of administration, although change may be
needed in existing administrative structures to accommodate the ongoing
needs of the watershed plan.

For these reasons, groups like the Royal Commission on the Future of
the Toronto Waterfront (now the Waterfront Regeneration Trust) have
found it helpful to have a third-party sponsor/advocate for the plan. This
third party should be nonpartisan and objective, yet able to work with all
groups to encourage participation in elements of the plan. Often, it requires
an individual who is widely respected to fulfill this need. In the case of the
Royal Commission, David Crombie himself, a convert to and passionate
advocate of watershed management, is that person. Speaking of his work
with the Trust, Mr. Crombie says that his job is simply to get everyone
sitting at the table talking to each other about who is going to do what.
This, in itself, is a challenging task.

McLelland (1987) observes that the structure of the sponsoring agency

may differ, depending on the level of government involved. Local agencies
are primarily concerned with community water resources needs-for
instance, water supply or wastewater treatment. Local governments are
more likely to undertake water resources planning on a case-by-case basis
as the need arises, rather than as a more forward-looking, integrated
strategy. She notes the examples of local water management in China,
where many rural villages have assumed responsibility for designing,
constructing, and operating their own water systems and in doing so have
achieved results superior to those obtained in some developing countries
that have instead chosen a strongly centralized organizational structure.
The success of water resources planning at the local level (in terms of
watershed, we could perhaps say the subwatershed level) requires well-
developed public participation. If problems are severe and the local system
complex (as, for instance, in the case of Mexico City), regional or national
planning may be more appropriate than local planning.

Regional planning is probably most compatible with the concept of

watershed management, but the administration of water management plans
at the regional level poses challenges simply because a region may
encompass several or many individual jurisdictions. It is seldom clear,
therefore. as to which administrative structure would best serve the needs
of the water management initiative.

Regional planning was the focus of much discussion at the United

Nations Mar del Plata conference of 1977 and, as discussed in Chapter 1,
has been the subject of increasing interest and international effort in the 20
years since that conference. At the time of Mar del Plata, about 40% of the
world's population was estimated to live within international river basins,
and a large proportion of these people therefore relied on water "imported"
from upstream countries (McLelland 1987). McLelland, like Caponera
(1985), emphasizes the barriers to planning and management that may
arise when different jurisdictions within the watershed have different
interests in, and uses of, basin water resources. In France, basin authorities
have been established for the management of multijurisdictional river
basins, and the "river catchment area is recognized] as a natural entity"
(Dubose 1992). Multijurisdictional agencies are also employed in the
management of the Rhine River basin (Wilkes 1975; Kiss 1985) and the
Danube River basin (Linnerooth 1990). The International Joint
Commission (IJC), a Canada-United States treaty-based organization
charged with oversight of the "boundary waters," provides a framework for
binational management of the Great Lakes Basin and other major
international waters. The U.S.- Mexico Border Environment Cooperation
Commission (BECC), established in 1993, coordinates transboundary
environmental infrastructure projects under the North American Free Trade
Agreement (NAFTA). Other countries with long-range plans and
administrative structures for the management of domestic and international
watersheds include Argentina, Egypt, Mexico, and Romania (McLelland
1987).

Planning at the national level is consistent with one of the Mar del Plata
Action Plan recommendations, which calls for the development of strong
national programs. However, as Lee (1992) points out, centralized national
planning and administration of water resources has failed, as it has failed
in the administration of social and economic programs, simply because it
cannot provide adequate recognition of local and regional needs and
differences. McLelland (1987) echoes this sentiment, observing that
although most countries in the Asia-South Pacific region have central
water agencies and well-defined national water policies, only about half
have been able to develop effective master plans for water management
and most of these cannot implement them because of structural
weaknesses in their central agencies. In less developed countries, these
weaknesses typically include lack of staff, lack of financial support, and a
judicial system unprepared to enforce the laws. McLelland also raises the
interesting case of the Ogun-Oshem river basin in Nigeria, where two
conflicting water management plans were developed, one at the federal
level and the other by state authorities. Each plan responded to a different
perception of need, the federal plan focusing on agricultural development
and the state plan on provision of adequate water supply in urban centers.
The two plans were not coordinated, so conflicts and redundancies
occurred and neither plan was fully successful.

12.2.2 Strong Linkages to Existing Programs

Implementation of a comprehensive water management plan can be an

expensive business; monitoring its effectiveness after implementation will
add to that expense. Implementing agencies can reduce the costs of actions
and of postimplementation monitoring by making strong linkages to
existing programs inside and outside government agencies. Among these
programs may be routine monitoring conducted by environmental and
health-protection agencies, independent monitoring carried out by public
interest groups or community organizations, data collection and analysis
by university researchers and their students, and so on. Even elementary
school students can he helpful in monitoring watershed status-for instance,
in tracking the timing of manure application on agricultural fields.

Efficient utilization of these varied data sources usually requires a staff'

person to act as liaison with the community and other agencies. This
person may also be, but is usually not, the same person charged with
maintenance of a centralized database, ideally accessible by all parties.
This data coordinator can be a knowledgeable volunteer but should be able
to provide continuity over time and across data contributors. Each
contributing agency or program would normally retain a copy of its data
and file a copy with the central database.

Utilization of existing programs can also help to enhance the profile of,
and community support for, the plan. Individual actions can he delegated
to existing groups, either for execution or for oversight, allowing the
sponsoring agency to retain an advisory role at lower cost and time
commitment. Linkages to existing programs can increase media exposure
of the plan, improve public awareness of issues and actions at the
household level, and increase political support for ongoing programs and
policies. Periodic review of the watershed plan will also he stronger and
more effective in a community where previous planning initiatives have
received extensive coverage and where a variety of other groups are
clearly interested in, and involved with, the planning process.

Forging strong linkages to existing programs was a central feature of the

Metropolitan Toronto (Ontario) Remedial Action Plan (RAP) (Ontario
Ministry of the Environment and Energy 1993). A previous major planning
effort, conducted in the early 1980s, had failed to produce an
implementable plan despite the expenditure of millions of dollars.
Gradually, private and public interest group initiatives had sprung up in the
river basins to take the place of the failed government plan. These
initiatives were often highly successful and enthusiastically supported in
the community, but were usually local in scope and limited by slim
financial resources. The Metro Toronto RAP therefore sought to extend the
work of the previous government plan while building a much stronger base
of public support in the community. Key to this strategy was a well-
publicized effort to recognize and link up with local community projects.
This was a delicate business, because the RAP team did not want to give
the appearance of taking over those local projects and, indeed, had no wish
to alter their course or to assume an oversight role. Rather, the RAP team
wanted to ensure that all efforts were recognized and coordinated so as to
minimize duplication of effort and unnecessary wastage of resources.

12.2.3 Clear Designation of Responsibilities, Timetables, and Anticipated Costs

To a large extent, implementation means program delivery: concrete,
observable actions, paid for by a public or private agency. Program
delivery in turn implies an administrative structure for program
development, delivery, and oversight. Actual plan implementation
therefore begins with assignment of responsibility and timetables for
specific actions. An obvious place to begin in making these assignments is
in jurisdictional responsibilities, such as those defined under national or
state/provincial constitutions. If a jurisdiction has the authority to enact
legislation, and has in fact done so, then that jurisdiction can legitimately
be expected to carry out activities consistent with that legislation and its
enforcement.

Sometimes the authors of the plan review jurisdictional responsibilities

and propose an assignment of tasks. This proposal is then debated in a
suitable forum, often in political councils or Cabinets, and the details of
responsibilities gradually hammered out. Partnership agreements and cost-
sharing arrangements may be useful mechanisms for achieving agreement
about responsibilities. Similarly, technical factors may dictate the most
probable costs and timetables for major project actions, but these may later
he amended in discussion with political bodies, either to defer or accelerate
actions to accommodate political and economic needs, or to share or trade
off responsibilities as appropriate for the implementing agencies. Here
again, as in so many stages of the watershed planning process, consensus
is essential. A consensus built on a decision to delay a costly project may
he a good approach if all parties are in agreement and the project is
actually (if eventually) built. A prescriptive arrangement assigning
responsibility to an agency without its concurrence may result in repeated
delays and, ultimately, in failure to build the project at all.

12.2.4 Effective Water Management Laws, Regulations, and Policies

As an extension of its constitutional responsibilities, a jurisdiction usually

has a number of tools available for water management. Often, these will
have been employed in the past and can be extended or reapplied for
current management purposes. Figure 12.1 shows how demand
management tools can be used to reduce pressure on water resources
quantity and quality (as opposed to supply management tools that would
increase supply to meet unrestricted demand). Appropriate demand
management tools include a range of legal, paralegal, and nonlegal
measures, such as:

• Legal instruments, including extraction permits or licenses, ambient

water quality standards, effluent discharge standards, and similar
instruments, and the fines and other sanctions associated with
noncompliance

• Economic instruments, including effluent charges, grants and

subsidies for pollution abatement activities, taxes, and regulations that
create markets affecting water and emission rights

• Voluntary instruments, such as nonbinding memoranda of

understanding between a government and an industrial or industrial
association, to reduce the volume or improve the quality of waste
emissions, to avoid or eliminate the use of certain chemicals, or to
effect similar actions

There are advantages and disadvantages to each of these instruments.

Some generate revenues that can be used for water management purposes.
Others, like voluntary programs, generate no revenues but are inexpensive
to administer and therefore may be cost-effective even if less prescriptive
than legislative tools. In the past, regulatory agencies have been reluctant
to adopt economic instruments, whether because of a fear that they will be
seen as licenses to pollute or otherwise misuse water, or because they are
cumbersome, or perhaps just different, to administer, as compared with
traditional legislative instruments like permits. (Economic instruments are,
in fact, usually based in statutes or regulations; that is, a legal instrument is
used to create a framework within which economic instruments like taxes
and effluent charges can operate.)

Probably the most common implementation obstacle in terms of laws

and policies is enforcement, especially in less developed countries.
Implementation of programs that involve or depend on legal instruments
requires that those affected by the legislation be monitored regularly by the
responsible agency to determine whether they are in compliance with legal
requirements. And there must be sufficient political will and resources
directed to enforcing regulations, including prosecution or other sanctions
against offenders, in the event of noncompliance. McLelland (1987) and
other authors have observed that most countries now have in place a
general legislative framework for environmental protection, but
enforcement of that framework is rarely satisfactory. Problems that may
arise in enforcement include:

• Insufficient staff for compliance monitoring

• Inadequate staff training

• Exceptions and variances to legal requirements granted where

improvements would be costly or difficult or would adversely affect
an export market

• Graft and corruption (e.g., bribery) to avoid compliance with legal

requirements

• Lack of public understanding of the issue and of cause-effect

relationships

• High costs of monitoring (e.g., costs of labor, laboratory analysis of

wastewater samples, sample and data storage and management, etc.)

• Lack of incentives to comply with legislation

12.2.5 Ongoing Tracking of the Degree and Success of Implementation

Despite the best intentions of all concerned, some intended projects will
not be implemented at the scope or within the time period originally
envisioned. It is important that implementation progress be tracked
regularly and consistently, both to reassure political bodies that progress is
being made and to encourage continued participation by all concerned.
Implementation tracking also allows screening of individual projects as
they are put into place, to make sure that their design and performance are
consistent with the goals of the watershed plan.

Regular reports on the progress of implementation also allow the plan to

be updated to take advantage of new ideas and emerging technologies that
may not have been apparent when the plan was originally completed.

Finally, regular monitoring of project implementation is important to

determine the condition and operability of structures and devices that have
already been installed. It is not uncommon, especially in less developed
countries where other infrastructure may be lacking, community awareness
may be low, and parts difficult to obtain, that installed equipment quickly
falls into disrepair and may even be stripped for other uses. McLelland
(1987) reports a survey of 589 villages in Maharashtra, India, where 36%
of hand pumps were found to have failed, either from lack of proper
installation or because they had not been adequately maintained. Schramm
(1980) cites the example of the state of Zacatecas, Mexico, where in 1978,
50% of all rural water supply works were not functioning and a further
20% were only partially operational. Only 30% of all rural water supply
works were working properly. In the same state, only 60% of the
nominally irrigated land actually produced a crop in 1976.

In these cultures it may especially important to employ what is

sometimes termed "appropriate technology"-that is, technology that is
known and understood in the community and that has a high likelihood of
continued success over many years.

12.2.6 Ongoing Monitoring and Reporting of Progress

The goal of the watershed plan is, ultimately, to move the watershed
toward a more desirable condition. This progress will cost money,
probably much of it from public coffers, and the implementing agency
must therefore be accountable to the public for its expenditures. Part of this
accountability is being able to demonstrate progress toward the goals of
the plan.
 Regular reporting of progress allows the effectiveness of individual
management actions to be evaluated. Recall that, prior to implementation,
water man agers must forecast the performance of each option and select
those that appear most promising. Postimplementation monitoring allows
each action's true performance to he evaluated and compared against
predictions. If a given action is not found to be as effective as anticipated,
it can be replaced or altered as appropriate.

Routine surveillance and monitoring should build on the baseline data

collected during the planning stage and should allow water managers to:

• Evaluate changes in watershed conditions because of, or independent

of, plan-related actions

• Assess compliance with statutory, regulatory, and voluntary

requirements

• Reveal areas where continued or enhanced remediation is required

As a general rule, all monitoring data should he available to the public,

either in annual reports or on file in a central but easily accessible location.
Some jurisdictions are now employing computer technology to make data
available to the public. As one example, Geographic Information Systems
(GIS) databases can he displayed using simpler display software so that a
user can just "point and click" on a map to bring up more detailed tables,
graphs, or maps. In another instance, a kiosk with weatherproof protection
housed a computer terminal on a busy street corner. Passersby were able to
view monitoring data from continuous sensors in an effluent discharge
stream. As discussed in Section 12.2.2, it is essential that all project data he
adequately stored and maintained so that they can he retrieved and
manipulated as necessary for project analysis. In multijurisdictional and
international basins this can he a challenging task, because data collection
and laboratory analysis practices may differ from jurisdiction to
jurisdiction. Section 10.4.2 described the problems encountered in trying
to match Canadian and U.S. GIS records on agricultural practices, slope,
soil type, and similar factors in the Great Lakes Basin.
12.2.7 Ongoing Public Education and Communication Programs

Ideally, the watershed plan represents a social consensus about the

condition of the watershed. Some management actions, such as control of
household hazardous wastes and responsible use of' storm drains, require
participation throughout the community. Other, more costly, actions, such
as sewage treatment plant upgrading or flood control structures. may
involve tax increases or other economic impacts on the community.
Ongoing public education and communication (which, as noted earlier, can
be linked to other external programs) can consolidate and enhance the
social consensus achieved in the planning process.

McLelland (1987) notes that national papers on developing countries,

prepared for the United Nations, list the following topics in decreasing>
order of importance: industry, agriculture, health. energy, natural
resources, t ansporta- tion, human settlements, communications, and
environment. Yet, as she points out, public awareness of, and response to,
environmental issues is a well-documented phenomenon, stimulated by
increasing media coverage and by local and regional environmental crises
that have focused public attention.

In less developed countries, literacy rates are often lower than in

economically advanced countries, and this may pose a challenge to public
education. In Uttar Pradesh, India, and in many other nations, appropriate
sewage disposal and safe-birthing information has been successfully
distributed, using pictorial representations such as sketches of mothers
washing children's hands following defecation. More difficult in some
developing countries is the widespread belief that governments are
omnipotent and will be able to identify and correct any problems that may
arise.

Sometimes indirect public education results from unrelated activities,

with benefits that are not anticipated when the primary project is designed.
Jacobson and Robles (1992) report that careful tour guide training has been
an important vehicle for public education about environmental protection
in Costa Rica. As ecotourism continues to increase in that country, it has
become especially important to ensure that ecological reserves can support
the educational needs of visitors while protecting the reserves' natural
resources. This goal can be fulfilled by careful tour guide training, taking
into account resource management requirements, visitor needs, and local
economic conditions. Jacobson and Robles have demonstrated that a good
tour guide training program, developed in collaboration with local
communities, scientists, park managers, and the hotel industry, can help
mitigate negative tourism impacts and provide environmental education to
adults and children in the local community, in addition to enhancing visitor
experience and providing local economic benefits.

12.2.8 Periodic Review and Revision of the Plan

To be an effective framework for watershed management, the plan must be

reviewed and evaluated periodically, preferably at predetermined intervals.
There are many reasons for this. Probably the most important of them is
that the watershed itself is dynamic: its population is growing or shrinking,
its water resources are improving or deteriorating, land uses and resource
utilization patterns are changing, and societal views of these changes are
shifting as they are influenced by internal and external factors. The plan
must be reviewed and revised as information about the watershed is
updated.

Periodic review also allows water managers to take advantage of newer

water control and pollution abatement technologies, which may offer
advantages of cost or performance over older approaches. Twenty years
ago, for example, channel improvements such as dredging, bank
stabilization with riprap or gabion baskets, and similar measurements were
common; today there is a trend toward natural channels and vegetative
rather than structural measures.

Computer simulation methods are also continually improved, especially

as computer technology changes. When the Grand River Basin Water
Management Plan was published in Ontario in the early 1980s, the plan
was based on 20year computer projections run on an IBM mainframe
computer at considerable expense. Fifteen years later the Grand River
Simulation Model, a purpose-built Fortran model of the river system,
could be run on any basic desktop computer, faster and more efficiently
than was possible in the early years. Now, as the plan is under revision, the
Grand River Conservation Authority is taking advantage of improvements
in computing technology and recent research advances to update portions
of the Grand River Simulation Model, to improve its simulation of algal
and rooted aquatic plant growth, and to link it to a regional groundwater
model.

Some planning exercises have remained static, one-time efforts. But a

regular schedule of plan review encourages water managers to use the
plan, and perhaps the option-testing framework on which the plan is based,
as a foundation for regular dialogue about water resources within the river
basin. Indeed, as the Royal Commission on the Future of the Toronto
Waterfront (1992) puts it, a good implementation framework, including
periodic review, will truly integrate environmental matters, provide a fair
and consistent process, and ensure that information, evaluation, and
decision making are shared and accessible by all parties in the watershed.
In the long run, it believes, this will lead to greater planning efficiency and
may shorten the time required for subsequent studies and agency
approvals.

12.2.9 Adequate Funding

It almost goes without saying that implementation will fail if there is not
enough money to pay for projects, infrastructure, and an appropriate
administrative system for the plan. As McLelland (1987) has pointed out,
the risks of failure because of economic constraints, and the implications
of failure for a fragile economy, are more extreme in less developed
countries than in economically advanced countries. Teclaff and Teclaff
(1987) observe that control of transme- dia, transboundary pollution hinges
on the availability of money and manpower. These authors further note
that, in fact, the existence of an administrative structure (they point to
binational water management commissions and other similar institutions)
may obscure the fact that insufficient money is available to actually carry
out projects. Even large and powerful organizations may run short of
funds. Teclaff and Teclaff cite the example of the Commission of the
European Economic Community, which completed a major study in 1979
on the underground water resources of the Community but was unable to
print and disseminate the study's 10 country reports and 152 maps because
of insufficient funds.

The Canada-United States International Joint Commission (IJC)

regularly reports on how the scarcity of research funding and manpower
limits the two countries' ability to manage water resources effectively in
the Great Lakes Basin. While funds may be available within each country
for short-term local projects, funding for oversight agencies like the IJC is
seldom adequate for the management, or even oversight, of an entire
watershed.

12.3 WHY IMPLEMENTATION SOMETIMES FAILS

Goodman and Edwards (1992) review the history of integrated water

resources planning in developed and less developed countries and provide
some insight into why implementation sometimes fails. They note that a
single framework for effective integrated water resources planning is not
possible for all countries, or even for all regions within a single country,
because these entities differ as to:

• Natural resources

• Population distribution and styles of living

• Economy

• Political, institutional, and legal structures

Planned programs may therefore fail if they are not successful in

matching projects to watershed conditions, particularly if they:

• Incorporate projects that are too ambitious to be achieved with

available financial and other resources, especially in less developed
countries
 • Incorporate projects that do not meet the expressed goals of the plan

• Fail to guide development properly

• Underestimate or ignore environmental impacts of the plan or

individual projects

• Underestimate or ignore social impacts, such as disruption of

community cohesion, displacement of groups or individuals, or loss of
key social structures that provide a focus for community activities

• Overestimate the efficiency or availability of existing institutional

frameworks to support implementation, or fail to create an
organization with adequate staffing and responsibilities to ensure that
projects are sustained beyond construction and early operation stages

• Lack good leadership

• Lack adequate staff training

• Disregard or underestimate the legal implications of the plan

• Lack an adequate infrastructure of facilities and services such as roads,

marketing organizations, and other components

• Fail to recognize national, regional, local, and individual costs and

benefits of the project in formulating and analyzing management
options and selecting priorities for action

Goodman and Edwards (1992) emphasize the special needs of less

developed countries in integrated watershed planning. In those countries,
sophisticated methodologies, such as the computer simulation methods
described in Chapter 7 of this book, and the considerations of risk and
uncertainty described in Chapter 8, require data resources that are simply
not available and will be too costly to acquire and/or require interpretation
by an experienced analyst who understands the applications and limitations
of both data and methodology. These requirements may be beyond the
capability of planners in less developed countries. The simpler techniques
described in Chapter 6 of this hook may be helpful in such cases, but even
access to necessary literature may he limited in some countries. In
particular, the problems of uncertainty described in Chapter 8 have
received increasing attention in water resources planning in economically
advanced countries. Less developed countries are particularly ill equipped
to forecast the outcomes of projects under risk and uncertainty, or to deal
with the economic implications of failure of a development project.
because of scarce capital and trained labor and because the infrastructure
for basic services may be incomplete. If a development project fails, the
consequence may not he simple inconvenience, as would be the case in an
economically advanced country, but rather a failure to provide basic needs
and services to communities.

Wilkes (1975) describes implementation issues in the management of

the Rhine River basin; many of these issues will also arise in developing
countries. A number of obstacles are identified:

1. The need to go outside the basin boundaries for water, and the
paucity of international arrangements to ensure adequate water supply
for all users.

2. The failure to integrate pollution control with control of water quality,

including sufficient water purification to ensure that water flowing
beyond political boundaries is of adequate quality for consumption in
neighboring jurisdictions. This problem becomes more acute when
protection of' future water supplies and future populations are
considered.

3. The danger of diverting attention from supraregional problems as a

result of decent alization of water management responsibilities. By
this, Wilkes means that although there are many benefits to regional
management of water resources, there may also be a danger in that
only problems that are exclusively regional or communal in dimension
may he "managed," while supraregional problems, such as
transboundary water supply, may fall outside a regional agency's
mandate.
 4. The lack of automatic links between planning for "growth" and for
water supply, including the coordination of land use permitting with
permits for water withdrawals. When water suppliers have no real say
in where development will go, shortages will occur and water
management problems will increase.

5. Failure to give officials duties to ensure lowered use of water,

especially tools that allow officials to obtain leverage over existing
users to force efficient water use, and tools that permit controls on
new and additional water extractions.

6. The risk of ignoring solutions developed outside the river basin

through excessive focus on management within the basin. Solutions
developed by nations and regions outside the watershed may have
value in the watershed but may be overlooked or discarded in favor of
those that are based on direct experience within the basin.

7. Failure to keep political influence out of technical judgments,

especially in relation to increased water use, future development, and
political good faith in avoiding unwarranted or irresponsible water
use.

8. Lack of planning for out-of-basin needs and demands, including

trends in international development and other forces.

Although Wilkes was writing more than 20 years ago, his observations
remain important and current today. For example, Ontario's Planning Act,
which governs land use and zoning, contains no explicit or implicit
requirement that development proposals must consider environmental
impacts (cf. Wilkes' obstacle 4 in the preceding list). Several years ago a
land-use planning commission was established in Ontario, headed by
former Toronto mayor John Sewell, and charged with reexamining the
Planning Act to improve its responsiveness to environmental concerns. The
commission duly reported its recommendations, which included a general
statement of principle that land use planning must be based on sound and
sustainable environmental practices, and suitable amendments to the Act
were drafted. With a change of provincial government in 1995, however,
the New Democratic Party was replaced by the Conservative Party and the
proposed changes to the Planning Act were swiftly discarded. There is
currently no requirement in Ontario to incorporate environmental
considerations in land use planning or zoning approvals.

According to Schramm (1980), one common error in making the jump

from planning to implementation is to assume that sound economic criteria
applied in planning analysis will guarantee economic feasibility in the real
world. As discussed in Chapter 8, the choice of a discount rate, or
minimum acceptable rate of return, is essentially an arbitrary decision
made by the planner. Depending on whether public or private investments
are involved (or a mixture of the two), different economic criteria and
different market rates will apply. Schramm notes that:

Many private investments and activities, confidently predicted on the

basis of the original benefit-cost analysis, simply do not materialize.
The overall projects turn out to be costly failures, unless government
takes the further step to subsidize the private investments as well.

This idea-that a planning prediction is simply that, and not a fact-again

seems so obvious as to be trivial. And yet, time and again, planners
produce economic forecasts, computer simulations, and population
projections that, although essentially guesses, arc interpreted and used as
fact. As Schramm notes, when the underlying assumptions for those
guesses are incorrect or inaccurate, grave errors can be made in planning
estimates. This may he the particular value of the Adaptive Environmental
Assessment and Management approach, as described by Holling (1978)
and Grayson et al. (1994) (see Section 3.4.2): the product of' the planning
exercise is not a set of "facts," but rather a community consensus as to the
desired plan of action and its probable effectiveness.

12.4 CASE STUDIES

This hook has presented many examples of watershed management

techniques in use around the world. In this section, several case studies are
examined and compared in an attempt to determine the factors that
encouraged success or caused failure.

12.4.1 The Tennessee Valley Authority, United States The Tennessee

Valley Authority (TVA) is often cited as the best-known attempt at
integrated watershed management (McDonald and Kay 1988). Indeed,
through its sheer size and duration, it is believed to have influenced water
management strategies throughout the world.

The TVA, originally envisioned as an electric power development

project, was begun in the early 1930s. It grew out of an earlier
development for a nitrate plant to supply war munitions. This plant was
begun in 1917 but decommissioned in 1919 before it was fully operational.
The plant was to have required significant quantities of power, to be
provided by two thermal generating plants and, later, by hydroelectric
power from the Wilson Darn development. It is sometimes referred to as
the Muscle Shoals facility, in reference to a nearby river feature.
Throughout the 1920s and early 1930s, the facility was the subject of hot
debate. Some people wanted to use it to produce nitrate fertilizer, but the
plant was suited only to the outdated cynamid process, which required
large quantities of electric power. If the plant had been converted to this
use, it would have demanded all the power supplied by the hydroelectric
plant, denying domestic customers access to that power. A variety of other
public and private interests were also involved in this debate, which
lingered into the beginning of the Depression. President Franklin D.
Roosevelt saw the Muscle Shoals plant as an opportunity for wider
economic development of the basin, not just as a local project. He
envisioned an electric power development at the center of a suite of
projects, including navigational improvements and flood controls in the
Tennessee River, as well as soil erosion control, reforestation,
improvements in agricultural land use, and increased and diversified
industrial development in the watershed.

A central feature of the TVA, consistent with Roosevelt's aims for the
project, was the "Force Account," which allowed construction in the river
and valley to be carried out by TVA itself, employing local laborers, rather
 than by construction companies from outside the basin. The TVA provided
welfare and education to its employees and their families and was
therefore fundamental in the economic recovery of the area.

TVA activities continued well into the 1940s with the development of
additional electric power generation capacity, using a series of thermal
plants. Construction projects were funded by bond sales and revenues from
the sale of power and other sources. Federal subsidies are still received for
environmental protection, agricultural extension in the valley, and various
other purposes, but these funds are small relative to the gross revenues
from TVA projects. Today the combination of inexpensive electricity,
easily available, trained workers, and a pleasant environment contribute to
the area's appeal as a site for industrial development. The project remains a
landmark of successful public ownership and multiproject development.

The Danube River Basin

The Danube has been called one of the most international rivers in the
world (Linnerooth 1990), with 10 European countries bordering the 2,850
km-long river. The river has more than 300 tributaries, and its basin
supports a population of more than 70 million people. The various
countries bordering on the Danube collaborate on management of the
shared resource through the nonbinding Danube Declaration, which
requires cooperation (Linnerooth 1990):

• Between eight countries spanning Eastern and Western Europe,

• In the absence of effective and enforceable international legal rules,

• In the absence of a basinwide planning or decision agency,

• Between numerous national and international authorities with diverse,

conflicting interests,

• On problems for which the geopolitics of the "upstream" and

"downstream" countries creates disincentives for cooperative
behavior,
 • On issues characterized by serious scientific gaps and uncertainties,
and

• In an atmosphere of increasing concern about the long-term effects of

toxic pollutants and acute awareness that pollutants cross national
boundaries.

Today the most pressing issues in the Danube are deteriorating water
quality and conflicting water use demands, particularly those related to the
generation of electric power. Several proposals for hydroelectric
development and the construction of barrages are under discussion, but
agreement has been difficult to achieve because of the inequitable
distribution of project costs and benefits (most opportunities for
hydroelectric development exist in the headwaters of the river, in Western
Europe, but the environmental costs of development will be felt throughout
the river basin). Specific issues now under discussion in the Danube basin
include:

• Maintenance of river flow for electric power generation and for waste
disposal.

• Maintenance and expansion of navigable waterways (the river

experiences heavy winter ice and high spring floods and has large
shallow stretches and hazardous rapids).

• Maintenance of an adequate and secure water supply for irrigation,

industrial uses, and other purposes.

• Protection of water quality for potable water, irrigation, fishing,

recreation, tourism, and nature preservation (water quality has
traditionally been measured using dissolved oxygen, pH, bacteria
counts. temperature, and hardness; more recently toxic substances
have been added to the list).

• Flood prevention.
 • Preservation of the river and its watershed for recreation, tourism, and
nature preservation.

Linnerooth (1990) emphasizes the importance of cooperation through

bilateral, stepwise negotiations. In view of the weak framework for
international water law and institutions, co-basin nations may cooperate
through building up a reservoir of goodwill upon which to draw in future
negotiations. Compensation arrangements ("side payments") may assist in
reaching agreement, as may straightforward payment by one country to
have another country clean up pollution.

Progress in the Danube is hampered by the lack of an existing

mechanism for multilateral, integrated decision making. A Danube
Commission exists for the resolution of navigation issues on the river but
has not, to date, been involved in wider management debates. Although
this Commission might have been expanded to incorporate a wider
mandate, this move was blocked by regional interests that preferred to
limit the powers of the central agency. Nevertheless, the countries
signatory to the Danube Declaration have expressed their willingness to
work toward improvement using narrowly focused agreements between
pairs or small clusters of countries. This means that, rather than employing
integrated decision making at the watershed scale, water management in
the Danube will proceed through a series of locally bounded and bilateral
bargaining sessions. As a first step, the signatories to the Danube
Declaration are wrestling with the problem of what constitutes adequate
water quality and what testing methods are to be preferred in collecting
water quality data. Linnerooth argues that as discussions move into
consideration of management measures, there may be value in the use of
computer simulation techniques as a means of communicating and
displaying complex information and as a way of providing "neutral"
evidence. Linnerooth terms this approach "mutual learning" and
emphasizes the value of shared problem definition, information collection,
and data analysis in the promotion of a shared goal for the watershed.

The Rhine River Basin

Like the Danube, the Rhine is one of Europe's largest rivers, its watershed
supporting a population of more than 50 million people. The Rhine
catchment area encompasses nine European nations and one of the most
important industrialized regions in the world. Other important water uses
on the Rhine include fishing, navigation, waste disposal, and a variety of
recreational uses. As a result of these activities, the Rhine has become
heavily polluted with mercury, arsenic, cadmium, lead, copper, zinc,
chromium, and chlorides (Kiss 1985).

Two organizations are expressly charged with integrated management of

the Rhine River and its basin. The Central Commission for the Navigation
of the Rhine dates from 1815 and has as its main goal the provision of free
navigation on the river and equal treatment for all ships. The Commission,
through its navigation-related activities, has some responsibility for
pollution control on the Rhine, including the transportation of dangerous
goods, but its main function is to provide good navigational facilities
throughout the river. The International Commission for the Protection of
the Rhine Against Pollution was created in 1963. Its primary task is to
prepare and carry out research on the nature, importance, and origin of
pollution on the Rhine, on behalf of several member nations. The
Commission conducts a range of such activities and reports its findings
annually to member nations. It also collaborates and cooperates with other
agencies interested in the protection of natural waters.

Studies by the International Commission have clearly shown that most

of the pollution on the Rhine originates from industrial sources and from
discharges from major urban centers. A Rhine Convention on Chemical
Pollution emphasizes integrated management of river water quality and
reminds member nations that Rhine River water is used for human
consumption, for livestock watering, and for the support of wild animal
populations. The Convention also stresses the need to support other
beneficial water uses such as fishing, recreation, irrigation, and industrial
water supply. Notwithstanding these provisions, surface water discharges
are subject primarily to approval by individual governments, but not by
any multilateral agency. Discharges are supposed to comply with emission
standards established by the International Commission. The Convention
 also includes provisions for notification of chemical spills and for the
resolution of water use conflicts (arbitration is the preferred approach).

Recent legal decisions in the Rhine basin have shown that pollution
control will require international cooperation. Kiss (1985) notes that
progress on water quality management will not occur solely through the
application of liability rules, which do not work well at the international
level because concerned governments did not wish to raise the issue of
international responsibility of industrial polluters. Despite slow progress
on water management and reluctance to acknowledge international
liability, Kiss states that the drafting of international rules and the constant
institutional cooperation between concerned nations offer promising
approaches to effective management of international basins like the Rhine.
Intergovernmental institutions are necessary to foster permanent, or at least
stable, cooperation, but their resources must be extended by linkages to
individual member states and sharing of resources across nations. It may
require political action to ensure that the necessary resources are diverted
and that shared environmental resources are protected.

The Upper Wye Catchment, United Kingdom

The River Wye is one of the United Kingdom's most important rivers,
draining the south-central part of Wales and portions of western England.
The basin has a total area of more than 4,000 km2. but for planning
purposes the total area has been divided into two subbasins: the Upper
Wye catchment (the River Wye and its lakes and tributaries down to the
city of Hay on Wye) and the Lower Wye catchment (the rest of the basin
down to the mouth of the river at its confluence with the River Severn,
near the Bristol Channel). The Wye passes through varied country,
including hilly upland areas, rolling agricultural lands, and industrial
towns. Its complex land uses and the international extent of its basin make
it essential to plan future uses of the basin carefully. In June 1993, the
National Rivers Authority (NRA) began a comprehensive catchment
planning exercise for the upper portion of the River Wye basin to ensure
that natural waters are protected and, where possible, improved for the
benefit of future generations. The NRA approached the plan in several
steps (NRA 1993a, 1993b):

1. Identifying catchment uses

2. Setting appropriate water use and water quality targets

3. Determining the current "state of the catchment"

4. Identifying specific issues and management options

5. Developing a process for public consultation

Catchment uses in the Upper Wye are diverse. Agriculture is

predominant, including sheep and dairy farming, crop production, and
moorlands. There is a small amount of mining activity and a larger forestry
industry in upland areas. A good network of roads and a major railway line
serve the larger centers of Builth Wells (population 2,040), Llandrindod
Wells (population 4,943), Llan- wrtyd Wells, Rhayader, Talgarth, and a
number of smaller villages. Several of the urban centers are tourist
attractions (a result of their historical development as spa towns), so there
is a seasonal influx of population during the summer months. Normal
population density is about 18 people per square kilometer. The area
supports a wide variety of plants, mammals, fish, and birds. Some areas of
the catchment have been designated as Environmentally Sensitive Areas,
and the River Wye itself is a "Site of Special Scientific Interest" under
national law. The area also contains part of a national park. Water supply
and sewage treatment in the area is provided by two companies, Welsh
Water (Dwr Cymru) and Severn Trent Water. The activities of these
companies, including the volume of their water abstractions and the quality
of their finished water, are overseen by Her Majesty's Inspectorate of
Pollution.

The Upper Wye is a major spawning and nursery area for salmon and
may be, according to the NRA, the best salmon fishery in England and
Wales. During the spawning season illegal fishing of salmon occurs, with
the effect of reducing fish stocks directly, through taking of adult fish, and
indirectly, by reducing the spawning population. There is a lively market
for salmon from the river, whether caught legally or illegally, and this
market may tend to encourage poaching (illegal fishing). In recent years
the salmon catch has declined in the Upper Wye, suggesting that the
numbers of salmon in the stream have also declined. If real, this decline
may be a result of various factors, including illegal fishing, legal fishing,
acidification of waters causing depleted phytoplankton and zooplankton
stocks, forestry practices, avian predators, physical barriers to migration,
and weather conditions. Other important fish species include brown trout
and rainbow trout (Both of which are stocked), eels, shad, and a variety of
"coarse" (as compared to sport) fish, including carp, chub, dace, pike, and
grayling. Commercial fishing is limited to an eel trap maintained by the
NRA on a tributary of the Wye.

Water abstractions are regulated by the NRA under a permit system.

Small abstractions, such as residential wells, do not require a permit. The
area's groundwater resources are an important resource for Dwr Cymru,
the local water company. Abstracted water is mostly returned to the
catchment via discharges of sewage treatment plant effluent to rivers and
streams. Surface water is also abstracted for drinking water supply,
primarily from reservoirs, some of which are used for local electricity
generation (possible only when the reservoirs are full). Some increase in
potable water demand is expected as the area develops further, but this
increase is expected to be small. Reservoir water is also released to the
river during low flow periods under an agreement between Dwr Cymru
and the NRA. Extensive "drawdown" in the reservoirs has been a matter of
concern to ecologists because of its potential to affect aquatic biota.
Agricultural abstractions occur throughout the watershed for general
agricultural use (including livestock watering), spray irrigation, and fish
farms. About 60% of general-use water is returned to the river after use,
almost all of the fish-farm water (which can be large in volume), but
virtually none of the sprayirrigation water. Industrial water use is limited in
the Upper Wye because of the area's largely rural character. Two
abstraction licenses have been granted for sand- and gravel-washing
operations and related activities. A third license allows groundwater
abstraction for industrial use. Future growth in demand is hard to predict
but will likely be less than I% per year.

There are 39 sewage treatment works operating in the Upper Wye

catchment, most of them owned by Dwr Cymru. All are monitored by the
NRA. In urban areas, there are several stormwater overflows (drainage
from streets and roofs), but the impact of these is thought to be negligible.
There are very few industrial effluent dischargers in the basin. Filter
backwash is discharged from Dwr Cymru's water treatment plants, and
there is a small amount of discharge of effluent from fish farms, quarries,
and sawmills. Spillage of chemicals including chlorine (used in water and
wastewater disinfection) is always a possibility. Three landfill sites are
located in the basin, but none is located close to a stream and so they are
not thought to pose any threat to water resources.

The appearance of the river is important to residents and others who

visit the area. Recreational activities in the basin include walking, boating
(primarily white-water canoeing), sailing and water skiing on the lakes,
and a very limited amount of swimming (which is discouraged by the
NRA).

The water management issues identified by NRA, and the options for
their resolution, were as follows:

1. Acidification, resulting from the deposition of acid precipitation

(caused by the burning of fossil fuels) on acid-sensitive soils and
"soft" (low-hardness) waters. Acidification impairs the survival of fish
eggs and young fish, reduces aquatic species diversity, and thus may
affect carnivorous species like otters and fish-eating birds. It may also
increase the need for treatment of drinking water. Options ►nay
include restricting emissions of acid-causing gases from smokestacks,
choosing species other than conifers for reforestation, and possibly
adding lime (calcium carbonate) to raise pH in natural waters.

2. Impaired .fisheries, resulting from low dissolved oxygen in some

areas. The sources of-and thus the solutions for-this issue are not
 presently clear. The problem may be a result of pollution from
industrial and/or agricultural activities.

3. Blue-green algae in lakes and ponds. Algae blooms create water

quality problems by using large quantities of oxygen at night, through
respiration, and through the oxygen-demanding decay of dead plant
tissue. Here again, there is little information available on the causes of
the problem, although they are likely related to enrichment of natural
waters with nutrients, especially phosphorus and nitrogen, which may
enter water through a variety of human activities in the form of
sewage discharge, agricultural runoff, and industrial effluents.

4. Low flows in summer months. Portions of the river exhibit very low
flows during the summer, the time when the demand for spray
irrigation is greatest. Water used in spray irrigation is not returned to
the river, so this type of abstraction can have a significant negative
effect on river flows. Solutions include storage of more water in
reservoirs through the winter months so that more is available for
release in summer. Farmers can also construct on-farm reservoirs or
grow crops that are less dependent on spray irrigation.

5. Lack of knowledge about the environmental impacts of abstraction.

Current understanding of the impacts of water abstractions on
groundwater supplies and surface water systems is very weak.
Additional research in this area is needed.

6. Lack of knowledge about the impacts of development of base flows.

Some people believe that land drainage and land use changes have
altered the natural flow regime in the river, making it "flashier"-faster
to respond to rainfall-and lower in its base flows than in the natural
condition. Again, more research into these possible impacts is needed
to clarify the extent of hydrologic changes following development.

7. Need to protect and enhance the wildlife resource. Human activities

affect wildlife habitat in many ways. Protective measures may include
consideration of wildlife requirements in approving abstraction
licenses, land drainage permits, discharge permits, and land use
 planning applications. River-edge vegetation should be protected
where sheep and cattle trample banks as they enter the river to drink.
The impact of recreational activities on wildlife should be evaluated
carefully, with a view to limiting certain activities when and where
necessary.

8. Decline in salmon and brown trout stocks. Over the past 20 years, the
NRA has observed a decline in the number of spring salmon and
brown trout caught on the Wye. The reasons for this decline are not
clear but are likely to be complex, including acidification, habitat
degradation, physical barriers to migration, and overfishing. Controls
or improvements in all these areas are possible and should result in
increased fish stocks. The impact of avian predators can be evaluated
through separate research.

9. Flooding at Builth Wells and Llanelwedd. Flooding of a main road

through these communities occurs in floods with a return period of
greater than I in 4 years. Thirteen homes and 20 commercial
properties are affected by this flooding. Flood defenses have been
considered but were rejected as too costly and environmentally
undesirable.

The preceding list identifies some areas where existing uses match or
conflict with existing quality in the Upper Wye catchment. It also gives
some indication of where problems are most urgent, where additional
information is required, and where concerns are probably minor. It does
not, however, give a good picture of future land uses and pressures on
catchment water resources.

For the Upper Wye, the NRA chose to focus the planning process on a
draft catchment management plan. This plan, which was prepared by the
NRA alone, then became the basis for consultation between the NRA and
all those with interests in the catchment. In the Upper Wye process,
participants were encouraged to comment on the issues and options
identified in the plan, suggest alternative options for resolving identified
issues, and raise additional issues not identified in the plan. Following the
consultation period, the NRA considered comments for incorporation into
 a revised final plan that will then form the basis for the NRA's actions
within the catchment and for the agency's interaction with other
organizations.

No information is currently available about the success of this planning

initiative, although the NRA expects that it will influence not only the
agency's own action plans and statements of policy but also those of
developers, planning authorities, and others involved in day-to-day
management of the catch ment. Individual project actions are still being
identified, and implementation will follow over the next decade or so.

The Grand River Basin Water Management Study, Canada

The Grand River is one of the largest rivers in the Province of Ontario,
with a huge river basin encompassing mixed residential, urban, and large
areas of intensive agricultural land use. The watershed contains live major
cities, which together account for at least 10% of the province's population.
The river flows from the Dundalk Highlands, the highest point in Ontario,
through rolling countryside and glacial outwash plains, into Lake Erie.
Until the last few decades the river supported a cold-water trout fishery,
but in recent years only less desirable warm-water species have been
present. Swimming, boating, and a variety of other recreational uses are
popular in the river basin, which includes a number of park and
conservation areas. River water is used for potable water supplies, for
irrigation and livestock watering, for industrial process and cooling waters,
and for many other purposes.

Although the river experiences inputs of a variety of contaminants,

including oxygen-demanding materials, suspended solids, heavy metals,
and industrial organic compounds, the major focus of public concern has,
in recent years, been the growth of rooted aquatic plants and algae
(nuisance species) resulting from excess nutrient loadings to the river. This
plant growth is also a major obstacle to the restoration of the fishery
because it uses large quantities of dissolved oxygen at night, when
photosynthesis is suppressed. Some reaches regularly exhibit nighttime
dissolved oxygen levels below those necessary to support cold-water fish
species. The main cause of this problem is thought to be phosphorus,
which enters the stream in sewage treatment plant discharges and in
agricultural runoff.

The impetus for watershed planning in the Grand River Basin was, and
continues to be, urban development pressure. The basin supports many
industries, and population is growing quickly as industrial development
proceeds. When the original Grand River Basin Water Management Study
was conducted, the Ontario Ministry of the Environment placed limits on
further development in the basin pending expansion and improvement of
sewage treatment facilities. A separate set of concerns centered on
persistent flooding and high flood damages in some communities along the
river.

The Grand River Basin Water Management Study was initiated by the
Grand River Conservation Authority (GRCA), a basin management agency
that operates at the provincial level with the cooperation of local
municipalities. Funding for the study was obtained from the Ontario
Ministry of the Environment, which formally cosponsored many of the
plan activities and all of its reports. As discussed in Chapter 3, the plan
employed it two-tiered advisory committee structure. The senior advisory
committee consisted of local decision makers, including local politicians
and senior government bureaucrats. A second tier of technical committees
provided specialized advice on a range of topics from hydrology to
economics. Both levels included community representatives. The advisory
committees were charged with developing a scope of work for the basin
management plan, beginning with problem definition. Technical
committees and subcommittees discussed issues such as data acquisition
and data quality, nature and adequacy of computer simulation efforts, and
similar matters.

The Grand River study employed a comprehensive continuous computer

simulation model of the lower river basin (see Figure 7.7), where most
development was occurring. The model was developed by Ministry of the
Environment engineers and scientists, in collaboration with GRCA,
specifically for the planning initiative and was not adapted from any
existing model. It incorporated separate subroutines to generate point- and
nonpoint-source flows, simulate the growth of several species of aquatic
plants and algae, and predict water levels, dissolved oxygen concentration,
and the concentration of several other water quality constituents, including
phosphorus.

The model proved cumbersome to calibrate and validate (see Chapter

7), and the complex output from the simulation runs were difficult to
translate into lay language. Eventually the planning team developed an
index of water quality that incorporated data on several key parameters,
and this index was used as one decision criterion. Other criteria included
flood frequency and magnitude, and percentage of time in which dissolved
oxygen levels dipped below target levels.

The study team developed and tested a wide range of urban,

agricultural, and in-stream management measures, including major
structures such as a dam and channel improvements. A range of sewage
treatment plant options were also considered.

The planners employed a series of regional consultations and direct

involvement of community members in decision making to evaluate and
screen management strategies. The process was time-consuming and
costly, requiring about six years and more than $4 million, in addition to
"in kind" contributions such as staff time and computing facilities
contributed by the provincial government. Several good plans were
proposed, each with clear advantages and disadvantages. In the end, the
committees chose a medium-cost alternative that included sewage works
upgrades at several locations and a variety of other channel improvements,
but no dam. These improvements were all in place by the mid1980s, and
local municipalities continue to collaborate with GRCA as the plan, and
simulation model, are continually updated to reflect changing conditions
and technology.

12.5 LESSONS LEARNED

Section 1.2 presented some elements of a successful integrated watershed
management effort. In summary, these are:

1. Adequate expertise for multiple-objective planning and evaluation

procedures, especially in economic, social, and environmental areas

2. Adequate resources of time and money for planning and

implementation

3. Consideration of a wide range of alternatives to solve observed

problems

4. A flexible, adaptable plan. reviewed and amended at regular intervals

5. Representation of all parties affected by the plan and its

implementation

6. Sufficient authority to enforce conformity of execution with

construction and operating plans

Examination of the case studies presented earlier in this chapter and

throughout this book reveals the truth of these requirements. The
development and implementation of workable solutions is not a process
that has a single correct outcome. Rather, it is an ongoing process of
dialogue with the community, learning about needs, teaching about
options, and building consensus about an ideal watershed condition and the
best way to get there.

The central feature of this process is choice: humans use and affect
water resources and have many choices available as to how and when that
use occurs. Integrated watershed planning means developing a social
consensus about best choices. Figures 12.2 and 12.3 illustrate mass flows
of energy, raw materials, and waste under two different operating
philosophies, the first a high-use highwaste system, and the second a
conservative, low-waste approach.
 Sustainable management of water resources requires water users to
make conscious choices that may sometimes reduce personal benefits in
favor of community or intergenerational benefits. The continuing theme
throughout this book is that those choices cannot he made solely on the
basis of scientific evidence. Indeed, the science is not very clear, or may be
contradictory, on key questions affecting water management. In addition to
scientific evidence, sustainable watershed management requires
community understanding and support, which in turn will generate
political will and, thus, economic and human resources to make changes.
Although those resources are now scarcer than they have been for many
decades, public awareness of and concern about water resources issues are
now higher than ever before. Integrated watershed management will
therefore depend on the formation of partnerships between governments
and the public, across disciplines and international borders, and among
water users with different interests and values. This is a huge challenge,
but one for which the payoff will be sustainable management of water
resources for our own and future generations.
Figure 12.2 Forces affecting water quantity and water quality under high-
use, highwaste management systems.

REFERENCES

Caponera, Dante A. 1985. Patterns of cooperation in international water

law: Principles and institutions. Natural Resources Journal 25:563-587.

Dubosc, A. 1992. The French Water Agencies. Paris, France: Ministere de

1'En- vironnement.

Goodman, A. S., and K. A. Edwards. 1992. Integrated water resources

planning. Natural Resources Forum 16(1):65-70.

Grand River Conservation Authority. 1982. The Grand River- Basin Water
Management Study. Summary Report. Cambridge, Ont.: Grand River
Conservation Authority, in conjunction with the Ontario Ministry of the
Environment.

Note: Dashed lines indicate reduced use, reduced impact, or reuse/recycle.

 Figure 12.3 Forces affecting water quantity and water quality under
conservative, lowwaste management systems.

Grayson, R. B., J. M. Dooland, and T. Blake. 1994. Application of AEAM

(Adaptive Environmental Assessment and Management) to water quality
in the Latrobe River catchment. .I. Emir Management 41:245-258.

Holling, C. S. 1978. Adaptive Environment Assessment and Management.

Chichester, United Kingdom: John Wiley & Sons.

Jacobson, S., and R. Robles. 1992. Profile: Ecotourism, sustainable

development, and conservation education: Development of a tour guide
training program in Tortuguero, Costa Rica. Environmental Management
I6(6):70I-713.

Kiss, Alexandre. 1985. The protection of the Rhine against pollution.

Natural Resources Journal 25:613-637.

Koudstaal, Rob, Frank R. Rijsherman, and Hubert Savenije. 1992. Water

and sustainable development. Natural Resources Forma 16(4):277-290.

Lee, Terence. 1992. Water management since the adoption of the Mar del
Plata Action Plan: Lessons for the 1990s. Natural Resources Forum
16(3):202-21 1.

Linnerooth, Joanne. 1990. The Danube River Basin: Negotiating

settlements to transboundary environmental issues. Natural Resources
Journal 30:629-660.

McDonald, Adrian T. and David Kay. 1988. Water Resources Issues and
Strategies. Harlow: Longman Scientific and Technical.

McLelland, Nina. 1987. Improved efficiency in the management of water

quality. Natural Resources Forum 11(l):49-57.

National Rivers Authority. 1993a. National Rivers Authority Strategy

(eight-part series encompassing water quality, water resources, flood
 defense, fisheries. conservation, recreation, navigation, research, and
development). Bristol. United Kingdom: National Rivers Authority
Corporate Planning Branch.

. I993b. Upper Wre Catchment Management Plan Consultation Report.

Cardiff. Wales: National Rivers Authority, Welsh Region.

Ontario Ministry of the Environment and Energy. 1993. Metropolitan

Toronto Stage 11 Remedial Action Plan: Clean Waters, Clear Choices.
Toronto: Ministry of the Environment and Energy, Central Region.

Royal Commission on the Future of the Toronto Waterfront. 1992.

Regeneration: Toronto's Waterfront and the Sustainable City: Final
Report. Toronto: Queen's Printer.

Schramm, Gunter. 1980. Integrated river basin planning in a holistic

universe. Natural Resources Journal 20: 787-805.

Teclaff, Ludwik A., and Eileen Teclaff. 1987. International control of

cross-media pollution: An ecosystem approach. Natural Resources
Journal 27:21-53.

Viessman, Warren Jr. 1990. Water management issues for the nineties.
Water Resources Bulletin 26(6):883-891.

Wilkes, Daniel. 1975. Water supply regulation. In Regional Management o/

the Rhine, edited by Chatham House Study Group. London: Chatham
House.

Young, Mike. 1992. Sustainable investment: The economic challenge. In

Environment and Development, edited by UNESCO. Publication No.
166, Vol. 42(2). London: Taylor and Francis.
Index